Manual on
MFI interest rate statistics
January / 2017
Manual on MFI interest rate statistics − Contents
1
Contents
1 Aim of this Manual 7
2 Scope and uses of MFI interest rate statistics 9
3 Basic definitions 12
4 Types of interest rate 14
4.1 Nominal versus effective interest rates 14
4.2 Annualised agreed rate and narrowly defined effective rate 15
4.2.1 Annualised agreed rate 15
4.2.2 Narrowly defined effective rate 17
4.2.3 Clarification of the variable n in the annualised agreed
rate formula 20
4.2.4 Treatment of exceptional repayments of principal 22
4.2.5 The annualised agreed rate formula for indefinite loans 22
4.2.6 The annualised agreed rate formula applied to revolving
loans and overdrafts and extended credit card credit 22
4.2.7 The annualised agreed rate formula for one-off deposits 23
4.2.8 Treatment of disagio 23
4.2.9 Treatment of agio 24
4.2.10 Annualising variable interest rates 25
4.3 Treatment of taxes, subsidies and regulatory arrangements 26
4.3.1 Taxes, subsidies and favourable rates 26
4.3.2 Special national practices including regulatory
arrangements 29
4.3.3 The annualised agreed rate formula for subsidised
loans 29
4.4 Annual percentage rate of charge 30
4.4.1 Definition of and link to the Consumer Credit Directive
and the Mortgage Credit Directive 30
4.4.2 Indicator for other loan charges 33
Manual on MFI interest rate statistics − Contents
2
4.4.3 Charges to be taken into account at national level 34
4.4.4 Period of fixation in the calculation of the APRC 35
4.4.5 Treatment of subsidies in the APRC 36
4.4.6 Treatment of non-profit institutions serving households
in the APRC 37
5 Business coverage 39
5.1 Bad loans and loans for debt restructuring below market
conditions 39
5.2 Interest rates on outstanding amounts 40
5.3 Interest rates on overnight deposits, deposits redeemable at
notice, convenience and extended credit card credit, and
revolving loans and overdrafts 41
5.3.1 The balance at the time reference point as an indicator
for new business 41
5.3.2 Determining the interest rate on an overnight deposit 42
5.3.3 Combined deposit and loan facilities 44
5.3.4 Regular savings on a deposit redeemable at notice 45
5.4 Interest rates on new business in instrument categories other
than overnight deposits, deposits redeemable at notice,
revolving loans and overdrafts, and convenience and extended
credit card credit 46
5.4.1 Definition of new business 46
5.4.2 Definition of renegotiation 47
5.4.3 New business on deposits with agreed maturity 48
5.4.4 Matured deposit with agreed maturity 50
5.4.5 Regular savings on a deposit with agreed maturity 51
5.4.6 New lending with a fixed interest rate and with initial rate
fixation 53
5.4.7 To p -up loans 54
5.4.8 Conversion of an overdraft into another type of loan 55
5.4.9 Loan taken out in tranches 55
5.4.10 The definition of new business for variable interest rates 57
Manual on MFI interest rate statistics − Contents
3
5.4.11 Choice of money market index 58
5.4.12 Money market index with a floor and a ceiling 59
5.4.13 Change in the value of a currency as an external index 59
5.4.14 Timing differences 60
5.4.15 “Cooling-off” period 61
5.4.16 Loan offer and preliminary offer 61
5.4.17 Moratorium on a loan 62
5.4.18 Documentary credit 62
5.5 Renegotiated loans: further issues 63
5.5.1 Renegotiations within a month 63
5.5.2 Renegotiations including an increase in the loan amount
and renegotiations involving a partial redemption of the
loan 64
5.5.3 Renegotiations of loans not yet fully drawn (loans taken
out in tranches) 65
5.5.4 Loan transfers, sales of loans and reorganisation of
loans 65
5.5.5 Calculation of the APRC for renegotiated loans 67
6 Time reference point 68
6.1 Time reference point for interest rates on outstanding amounts
of overnight deposits, deposits redeemable at notice, revolving
loans and overdrafts, and convenience and extended credit card
credit 68
6.2 Time reference point for interest rates on new business 70
7 Instrument categories 71
7.1 Summary tables of indicators 71
7.2 General provisions 71
7.3 Foreign-currency deposits and loans 73
7.4 Breakdown by sector 73
7.5 Breakdown by type of instrument 76
7.5.1 Types of deposits 76
Manual on MFI interest rate statistics − Contents
4
7.5.2 Types of loans 77
7.6 Breakdown by amount category 82
7.7 Breakdown by original and residual maturity, notice and interest
rate reset period or initial rate fixation 84
7.7.1 Time bands 84
7.7.2 Original and residual maturity and period of notice 84
7.7.3 Period of maturity for a loan taken out in tranches 85
7.7.4 Initial period of fixation of the interest rate 86
7.8 Breakdown by secured loans with collateral and/or guarantees 89
8 Specific instruments and national products 91
8.1 Step-up (step-down) deposits and loans 91
8.2 Revolving loans and lines of credit 92
8.3 Umbrella contracts 94
8.4 Savings plans for housing loans 94
8.5 Savings plans with a fidelity and/or growth premium 95
8.6 Interest rate on zero coupon bond-like savings bonds 96
8.7 Splitting of loans into two parts 97
8.8 Option to convert a deposit into equity shares 97
8.9 Interest rate linked to a share price 98
8.10 Treatment of a deposit comprising two components 99
8.11 Pension savings accounts 102
8.12 Purchase of mortgage loans by a credit institution 102
8.13 Securitisation of mortgage loans by a credit institution 103
8.14 Payday loans 104
8.15 Restricted deposits 105
8.16 Cash pooling 106
8.17 Treatment of factoring and the calculation of the AAR/NDER for
very short-term loans 107
8.18 Treatment of student loans without a definite maturity 108
Manual on MFI interest rate statistics − Contents
5
8.19 Treatment of convertible deposits 108
8.20 Leasing contracts 109
8.21 Recording of syndicated loans 109
9 Aggregation of the data and reporting obligations 111
9.1 Overview 111
9.2 Statistical information at the level of the reporting agents 112
9.3 National weighted average interest rates and national total
business volumes 113
9.4 Aggregated results for the euro area 114
10 Validation rules 115
11 Revision policy 121
11.1 Principles of the revision policy 121
11.2 The revision criteria 122
12 Selection of the reporting agents 126
12.1 Selection of the actual reporting population 126
12.2 Census versus sampling approach 127
12.3 Stratification of the reference reporting population 129
12.4 Allocation of the sample across strata 133
12.5 Minimum national sample size 135
12.6 Special provisions in the case of group reporting 137
12.7 Estimation of total new business volume 138
12.8 Maintenance of the sample 141
12.9 Further sampling issues 144
12.10 Description of the euro area sample 146
13 Derived indicators 148
13.1 Reason for including this chapter in the Manual 148
13.2 Bennet binary indices for rates and weights 148
13.3 Bennet chain indices 150
Manual on MFI interest rate statistics − Contents
6
13.4 Coefficients of cross-country variation 152
13.5 Cost-of-borrowing indicators 153
13.5.1 Cost-of-borrowing indicator for households for house
purchase 155
13.5.2 Cost-of-borrowing indicator for non-financial
corporations 155
13.5.3 Cost-of-borrowing indicator for short-term loans to
households and non-financial corporations 157
13.5.4 Cost-of-borrowing indicator for long-term loans to
households and non-financial corporations 157
Appendix 158
Index of terms 163
Manual on MFI interest rate statistics − Aim of this Manual
7
1 Aim of this Manual
On 20 December 2001, the Governing Council of the European Central Bank (ECB)
adopted Regulation ECB/2001/18 concerning statistics on interest rates applied by
monetary financial institutions to deposits and loans vis-à-vis households and non-
financial corporations. It was published in the Official Journal of the European Union
on 12 January 2002
1
and came into force on 31 January 2002. This Regulation was
amended three times
2
and was then recast by Regulation ECB/2013/34
3
(hereafter
referred to as “the Regulation”). Together with Guideline ECB/2014/15 on monetary
and financial statistics
4
(hereafter “the Guideline”), the Regulation defines the
statistical standards according to which MFI interest rate statistics
5
are collected and
compiled in the European Union (EU). The Regulation is addressed to the reporting
agents, i.e. monetary financial institutions (MFIs) except central banks and money
market funds
6
, while the Guideline is addressed to the national central banks
(NCBs), containing additional reporting requirements and instructing NCBs on data
to be transmitted to the ECB, by when, in what format, etc. Unlike the Regulation, the
Guideline leaves some discretion as to the choice of source for the data, provided
that certain quality standards are met.
The Regulation and the Guideline are binding on the Member States participating in
the Monetary Union (hereafter referred to as the “euro area Member States”).
However, for non-participating Member States (hereafter the “non-euro area Member
States”), Regulation (EC) No 2533/98 concerning the collection of statistical
information by the European Central Bank implies an obligation to design and
implement at national level all the measures that they consider appropriate to fulfil
the ECB’s statistical reporting requirements in order for them to become participating
Member States. Thus, non-euro area Member States are encouraged to follow the
reporting requirements laid down by Regulation ECB/2013/34.
This Manual, by contrast, contains no additional requirements to those included in
the Regulation and the Guideline and has no binding legal status. It aims to further
clarify and illustrate these requirements, mainly through the use of extended
definitions, explanations of the underlying concepts and examples. It also brings
together detailed transcriptions from other legal acts referred to in the Regulation. A
clear and consistent understanding of the statistical requirements contained in the
Regulation and the Guideline by the statisticians in the NCBs of the European
System of Central Banks (ESCB) and also in the accession countries is essential for
1
OJ L 10, 12.1.2002, p. 24.
2
Regulations ECB/2004/21, ECB/2009/7 and ECB/2010/7.
3
OJ L 297, 7.11.2013, p. 51.
4
OJ L 340, 26.11.2014, pp. 1-209.
5
To avoid making a potentially misleading distinction between “retail” and “wholesale” interest rates –
expressions that can carry different meanings – the statistics developed under the above-mentioned
Regulations are referred to as “MFI interest rate statistics”.
6
In this Manual, when reference is made to an MFI or to MFIs, this excludes central banks and money
market funds.
Manual on MFI interest rate statistics − Aim of this Manual
8
the production of harmonised MFI interest rate (MIR) statistics. The information in
this Manual may also interest reporting agents and users of the statistics.
The Manual is composed of 13 chapters. Chapter 2 sets out the scope of MFI
interest rate statistics with special emphasis on the use of these statistics for
monetary policy purposes. Chapter 3 defines the main terms contained in the
Regulation. Chapter 4 discusses the types of interest rates compiled under the
Regulation and the Guideline. Chapter 5 describes the business coverage of these
statistics both in terms of new business and outstanding amounts and Chapter 6
explains the time reference point for these two statistical concepts. Chapter 7
provides an overview of the indicators available at the euro area and at the national
level. Chapter 8 gives guidance on the treatment of specific deposit and loan
products. Chapter 9 summarises the steps needed to aggregate the individual data
to obtain euro area results. Chapter 10 defines the validation rules. Chapter 11
describes the principles of the revision policy and the revision criteria. Chapter 12
sets out the method for selecting the reporting agents for MFI interest rate statistics.
It tackles a full range of sampling issues, including the stratification procedure, the
definition of the minimum sample size, the way of allocating the sample across
strata, and the maintenance of the sample over time. Finally, Chapter 13 illustrates
the calculation and usage of derived indicators based on the MFI interest rate
statistics, such as the Bennet binary indices for rates and weights, the Bennet chain
indices, the coefficients of cross-country variation and the cost-of-borrowing
indicators.
Manual on MFI interest rate statistics − Scope and uses of MFI interest rate statistics
9
2 Scope and uses of MFI interest rate
statistics
The scope of euro area MFI interest rate statistics is all interest rates that MFIs
resident in the euro area apply to euro-denominated deposits and loans vis-à-vis
non-financial sectors (other than government) resident in the euro area, i.e. vis-à-vis
households and non-financial corporations of any size. In practice, mainly credit
institutions (as defined in Article 4(1) of Regulation (EU) No 575/2013) need to report
MFI interest rate statistics.
7
The statistics are compiled for the euro area as a whole
and individually for each euro area Member State in order to give information about
the level and development of interest rates both at euro area and at national level.
Additionally, for each non-euro area Member State, the scope of MFI interest rate
statistics is all interest rates that MFIs resident in that Member State apply to
deposits and loans in euro and national currency vis-à-vis non-financial sectors
(other than government) resident in the Member State, i.e. vis-à-vis households and
non-financial corporations of any size.
The MFI interest rate statistics are broken down by original and residual maturity,
notice and interest rate reset period, or initial period of interest rate fixation. Loans to
households are further broken down by the purpose of the loan, while loans to non-
financial corporations are divided into categories according to the size of the loan. In
addition, information on loans with collateral and/or guarantees is collected. Data on
renegotiated loans are also provided. All this information is organised into 117
indicators, of which 91 refer to new business and 26 to outstanding amounts. New
business is defined as any new agreement between the customer and the MFI.
Outstanding amounts are defined as the stock of all deposits placed by customers,
i.e. households and non-financial corporations, with MFIs, and the stock of all loans
granted by MFIs to customers.
8
MFI interest rate statistics are collected and compiled at monthly frequency. The first
data were compiled for the reference month January 2003. Additional historical
series on some interest rates on new business are available for euro area Member
States starting with data referring to January 2000. These interest rates have been
estimated either by the respective NCB or by the ECB in agreement with the NCB.
The MFI interest rate statistics have four main uses:
• To analyse the monetary policy transmission mechanism, as monetary policy is
transmitted through the economy via changes in interest rates. First, the
statistics enable the studying of the pass-through of changes in official rates
and market interest rates to lending and deposit interest rates faced by
households and non-financial corporations. Information about the speed and
7
More detailed definitions and further explanations are given in Chapter 3.
8
See Chapter 5 for a more detailed description of new business and outstanding amounts.
Manual on MFI interest rate statistics − Scope and uses of MFI interest rate statistics
10
extent of the pass-through is essential to understand the effect of monetary
policy on the demand for consumption and investment. Second, changes in MFI
interest rates affect the cost of capital and so influence investment decisions
and substitution between current and future consumption. MFI interest rate
statistics are, therefore, vital for any economic analysis over time. Third, the
statistics allow income effects to be analysed, as changes in MFI interest rates
affect the interest paid or received by households and non-financial
corporations and hence the disposable income of these sectors. Finally, MFI
interest rate statistics enable users to analyse the credit channel of monetary
policy, in particular the cost spread between self-financing and credit, or the so-
called external finance premium.
• To enhance the monetary analysis in the euro area. Ideally, prices and
quantities are analysed together. Information on the remuneration of M3 and its
components is one essential factor to assess portfolio shifts between monetary
and non-monetary assets as well as among monetary instruments. For
instance, a steepening of the yield curve may determine an increase in the
remuneration of longer-term deposits which results in an increased volume of
these deposits placed with MFIs (either inside or outside M3). As deposits
constitute the largest component of M3, monitoring their remuneration is
essential to explain the dynamics of monetary growth and its effects on price
stability. Similarly, detailed MFI interest rates allow developments in loans to the
private sector to be analysed.
• To monitor structural developments in the banking and financial system and to
analyse financial stability issues. Users may study the development of banks’
interest rate margins and changes in their profitability, and potentially adverse
developments that may damage financial stability, such as how quickly banks’
interest rate margins react to external developments or how the interest burden
changes for households and non-financial corporations.
• To analyse financial integration. Harmonised interest rate statistics provide
information on the convergence of interest rates across euro area Member
States and shed light on the degree of integration of the retail banking markets
in the euro area.
In conclusion, MFI interest rate statistics are essential to well-founded monetary
policy decision-making. Monetary authorities need to be frequently and rapidly
informed of the changes in these interest rates, so as to assess the reach, scope or
delayed effect of monetary decisions and their change over time. Furthermore, these
statistics are a very valuable tool for the analysis of financial stability as they support
the assessment of bank profitability and competition. MFI interest rate statistics are
designed to meet all these needs.
Manual on MFI interest rate statistics − Scope and uses of MFI interest rate statistics
11
Table 1
Main uses of the reported indicators
Scope of analysis
Monetary transmission
Monetary analysis
Financial
stability
Financial
Integration Interest rate channel
Credit
channel
Pass-
through of
interest
rates
Cost of
capital
substitution
effect
Income
effect
External
finance
premium
Money
demand
Credit
demand
Bank
competition,
bank
profitability
Convergence of
retail interest
rates and of cost
of borrowing
Interest rates
on
outstanding
amounts
9
X X X X X
Interest rates
on new
business
X X X X X X X
9
Where outstanding amounts consist of a significant part of variable rate business, the related rates may
also provide information on the pass-through of interest rates.
Manual on MFI interest rate statistics − Basic definitions
12
3 Basic definitions
In order to ensure the comparability of MFI interest rate statistics with other
macroeconomic statistics produced at European level, the Regulation relies, to the
extent possible, on existing frameworks such as the European System of National
and Regional Accounts (ESA 2010)
10
and the ECB’s MFI balance sheet statistics
11
.
The Regulation therefore uses a number of expressions that are common to
European and, in particular, to euro area statistics, but also terminology that is
specific to MFI interest rates. The main terms are defined in Article 1 of the
Regulation. In the following, these definitions are explained in more detail.
A euro area Member State is a country that has adopted the single currency in
accordance with the Treaty establishing the European Community.
12
An entity is regarded as a resident
13
of a Member State (Article 1(1)) when it has a
centre of economic interest in the territory of that Member State, i.e. when it has
engaged for a year or more in economic activity in that territory, or when it has
registered or indicated an intention to operate permanently in that territory. For euro
area Member States’ MFI interest rate statistics, the interest rates and weights refer
to deposits by and loans to customers resident in the euro area.
14
No distinction is
made for customers between domestic residents and residents of the other euro
area Member States. For example, the Bank of Greece collects data on interest
rates that MFIs resident in Greece apply for customers resident in Greece and in the
other euro area Member States, but not for customers resident outside the euro
area. The main reason for excluding residents outside the euro area is that the
monetary authority of the euro area needs to monitor how the monetary policy
actions are being transmitted to savers and investors of the euro area, since these
two kinds of agents are the ones that determine the demand for consumption and
investment within the euro area. The demand for consumption and investment, in
turn, are relevant variables to achieve price stability. It is outside the scope of euro
area monetary policy to influence the demand for consumption and investment of
those residents outside the euro area.
The reporting scheme defined in the Regulation applies only to MFIs other than
central banks and money market funds included in the “list of MFIs”
15
. E-money
institutions are MFIs, so they are covered by MFI interest rate statistics. In principle,
10
Regulation (EU) No 549/2013 of the European Parliament and of the Council of 21 May 2013 (OJ L
174, 26.6.2013).
11
Regulation ECB/2013/33 (OJ L 297, 7.11.2013, p. 1).
12
This also includes the associated territories or countries that have been authorised to adopt the euro as
their legal currency and in which the single monetary policy of the ECB is conducted. For instance, this
is the case of Monaco and the French overseas territories of Saint-Pierre-et-Miquelon and Mayotte.
13
Defined in Article 1 of Regulation (EC) No 2533/98.
14
For non-euro area Member States, this should read here and in the rest of the document “resident in
the same Member State as the reporting MFI”.
15
A comprehensive list of all MFIs in the European Union is produced and published by the ECB. Further
information and the list are available on the ECB’s website.
Manual on MFI interest rate statistics − Basic definitions
13
money market funds should be covered also; however, as their business is not to
receive deposits or grant loans, money market funds do not pay interest on their
liabilities or receive interest on their assets in the same way as other MFIs.
MFI interest rate statistics (Article 1(5)) cover interest rates that resident MFIs apply
to euro-denominated
16
deposits and loans vis-à-vis households and non-financial
corporations (Article 1(2 and 3)). The sector classification follows the principles
established in Chapter 2 of the ESA 2010
17
, except for non-profit institutions serving
households (NPISHs) which, as in Regulation ECB/2013/33, are included in the
household sector, as NPISHs are usually not as important as the other two sectors
with respect to the amount of loans and deposits on their balance sheet.
18
The reference reporting population (Article 1(7)) follows from the scope of MFI
interest rate statistics. In each Member State, the reference reporting population
comprises all resident MFIs except central banks and money market funds which
take euro-denominated deposits from and/or grant euro-denominated loans to
households and/or non-financial corporations resident in the euro area Member
States. This means that the customers may be resident anywhere in the euro area,
not necessarily in the same Member State as the reporting MFI. As explained above,
the Regulation does not require a distinction between domestic residents and
residents of the other euro area Member States and does not apply to central banks
and money market funds.
NCBs select the reporting agents (Article 1(1)) for MFI interest rate statistics from the
reference reporting population which, for each Member State, comprises only
resident entities. All reporting agents together constitute the actual reporting
population (Article 2)
19
. Reporting agents are the legal and natural persons that are
subject to the ECB’s statistical reporting requirements. They include the entities that,
according to the national law of their Member State of residence, are neither a legal
person nor a collection of natural persons, but can be subject to rights and
obligations. The persons legally representing these entities must fulfil their reporting
obligations.
16
For non-euro area Member States, this should read here and in the rest of the document “euro and
national currency”.
17
See also Money and Banking Statistics – Sector Manual – Guidance for the statistical classification of
customers, ECB, third edition, March 2007. The sector breakdown is further discussed in Section 7.4.
For the treatment of NPISHs, see also Section 4.4.6.
18
In a few Member States, loans to NPISHs form, however, a non-negligible part of loans to households
and NPISHs.
19
Further discussed in Section 12.1.
Manual on MFI interest rate statistics − Types of interest rate
14
4 Types of interest rate
20
4.1 Nominal versus effective interest rates
The terms “nominal interest rate” and “effective interest rate” have a range of
different meanings depending on the Member State and the context. Also, there is a
specific terminology in the banking business related to interest rates, e.g. advertised
nominal rates and prime rates. Advertised nominal rates are interest rates that are
displayed by banks as the headline rates in the windows of their branches, their
leaflets, advertisements, newspapers, other journals, on their websites, etc. These
rates give customers an indication of the current interest rate level for different
banking products. The advertised nominal rates might be applied to highly
standardised deposits and loan products, but are not necessarily the rates that the
bank actually pays or charges its customers. The advertised nominal rates might
also be prime rates that the banks offer to their most creditworthy customers. In this
case, the rate actually applied to a deposit or loan might be less favourable than the
advertised nominal rate.
Given this diversity in the terms applied and the fact that the same term does not
always carry the same meaning or follow the same methodology in different
countries, it is potentially misleading to use ambiguous terms in the context of MFI
interest rates. Therefore, it might be more useful to briefly describe the main features
of these rates and to leave it to the individual user of the statistics to decide what to
call them.
• MFI interest rates are agreed rates: The data collected refer to the interest rate
that is individually agreed between an MFI and its customer. MFI interest rates
are hence distinct from advertised nominal rates, because households and non-
financial corporations might be able to negotiate better terms and conditions
21
than those advertised. However, customers may also be offered worse terms
and conditions due to e.g. a lack of creditworthiness.
• MFI interest rates are annualised: They are converted to an annual basis and
quoted in percentages per annum. This means that MFI interest rates take into
account the frequency of interest payments. Ceteris paribus, the more frequent
the interest payments, the higher the MFI interest rate recorded in the
statistics.
22
Two possibilities exist for annualising interest rates: either an
algebraic formula leading to the annualised agreed rate (AAR) or successive
approximation resulting in the narrowly defined effective rate (NDER), both of
which are discussed in Section 4.2.
20
This chapter refers mainly to Part 1 of Annex I to the Regulation.
21
The meaning of the expression “terms and conditions” is specified in Section 5.4.1.
22
See also Equations 2 and 3 below.
Manual on MFI interest rate statistics − Types of interest rate
15
• 115 out of the 117 MFI interest rate categories exclude charges: For deposits,
MFIs pay interest to the customer but might also charge fees. Analogously, for
loans the customer has to pay an amount comprising an interest rate
component and a component made up of other related charges. The annual
percentage rate of charge (APRC) is an effective lending rate that covers the
total costs of the credit to the consumer, i.e. the interest payments as well as all
other related charges. The concept of “total costs for the consumer” was
designed for the purpose of consumer protection. The compilation of the APRC
is defined in Directives 2008/48/EC and 2014/17/EU and further explained in
Section 4.4. The two exceptions are the additional series collected for loans to
households for consumption and for house purchases, where in addition to a
rate without charges the APRC is also required.
4.2 Annualised agreed rate and narrowly defined
effective rate
23
4.2.1 Annualised agreed rate
The annualised agreed rate (AAR) is defined in paragraph 1 of Annex I to the
Regulation as “the interest rate that is individually agreed between the reporting
agent and the household or non-financial corporation for a deposit or loan, converted
to an annual basis and quoted in percentages per annum”. The AAR covers all
interest payments on deposits and loans, but no other charges that may apply.
Disagio (or discount in common language)
24
, defined as the difference between the
nominal amount of the loan and the amount received by the customer, is considered
as an interest payment at the start of the contract (time t
0
) and is therefore reflected
in the AAR.
An annualised agreed rate reflects the creditworthiness and other qualities of the
customer (in respect of loans) and the solvency and other qualities of the credit
institution as determined by the customer (in respect of deposits). The AAR is
influenced by the budget, capital or other constraints faced by the credit institution in
granting loans and taking deposits, including competition with other types of financial
institutions and products. It is a result of the demand and supply conditions in the
deposit and loan markets.
Paragraph 2 of Annex I to the Regulation provides the formula for annualising the
agreed interest rate, i.e. for converting interest payments that are due at regular
intervals within a year to a yearly basis. It is applied in cases where the interest
payments that are agreed between the MFIs and the customer are capitalised at
regular intervals within a year, for example per month or quarter, rather than per
annum:
23
See also paragraphs 1, 2 and 3 of Annex I to the Regulation.
24
See also Section 4.2.8.
Manual on MFI interest rate statistics − Types of interest rate
16
= 1 +
1 [1]
where:
• x is the annualised agreed rate
• r
ag
is the interest rate per annum that is agreed between the reporting agents
and the household or non-financial corporation for a deposit or loan where the
dates of the interest capitalisation of the deposit and all the payments and
repayments of the loan are at regular intervals in the year
• n is the number of interest capitalisation periods for the deposit and
(re)payment periods for
25
the loan per year, i.e. 1 for yearly payments, 2 for
semi-annual payments, 4 for quarterly payments, and 12 for monthly payments
For example, a customer and a credit institution agree on a five-year loan at 10% per
annum (p.a.) for the entire maturity, where the interest is paid at the end of each
quarter and the principal repaid at the end of the fifth year. The annualised agreed
rate for this loan is then 10.3813% p.a. and it is calculated as follows:
= 1 +
1 = 1 +
.
1 = 0.10381289 [2]
If, in the same example, the interest payments were at monthly frequency, then the
AAR would be slightly higher at 10.4713% p.a. and it is calculated as follows:
= 1 +
1 = 1 +
.
1 = 0.10471307 [3]
In the case of daily interest capitalisation, n = 365 should be used in Equation 1
following the convention of a standard year of 365 days as specified in paragraph 12
of Annex I to the Regulation.
26
In the above example for a loan, the AAR for monthly
interest payments would be 10.5156% p.a. and it is calculated as follows:
= 1 +
1 = 1 +
.
1 = 0.105156 [4]
Equation 1 may also be used to derive the AAR, for example in the case of a deposit
of EUR 10,000 that is placed for two years where EUR 11,000 is paid out to the
customer at maturity. During the two years, the customer earns 10%. The AAR is
4.8809% and calculated as follows:
= 1 +
1 = 1 +
.
1 = 0.048809 [5]
25
The text in strikethrough is included in the Regulation. However, the application of the formula is clearer
without it. See also Section 4.2.3.
26
If 360 days were used in the formula as a standard year instead of the convention of 365 days, a
different result would be achieved. The size of the difference depends on the level of the agreed
interest rate r
ag
.
Manual on MFI interest rate statistics − Types of interest rate
17
4.2.2 Narrowly defined effective rate
Instead of the annualised agreed rate, NCBs may require their reporting agents to
implement the narrowly defined effective rate (NDER) for all or some deposit or loan
instruments referring to new business and outstanding amounts. The NDER refers to
an annual basis and is defined as the interest rate that equalises the present value of
all commitments other than charges (deposits or loans, payments or repayments and
interest payments), future or existing, agreed by the MFI and the household or non-
financial corporation. The NDER is equivalent to the interest rate component of the
APRC
27
, i.e. it does not take into account the component of other charges. Hence,
the basic formula for the APRC in Annex I of Directive 2008/48/EC and in Annex I of
Directive 2014/17/EU applies, which is given as Equation 7 in Section 4.4, but
without the references to other charges. This Equation 7 is equivalent to the formula
proposed by the International Capital Market Association
28
for the exponential
interest rate calculation for all maturities. Hence, in the case when years are
considered to have 365 days and the amount of the deposit or loan is placed or paid
out in one amount, the following applies:
=
(
)
=
(
1 +
)
[6]
where:
• i is the interest rate (NDER)
• CF
n
is the cash flow n, from the perspective of the investor in the case
of deposits and from the point of view of the credit institution in the case of
loans
• N is the number of cash flows associated with the financial instrument
• A is the present value of the total paid-out amount
• D
n
is the timing of the cash flow n, expressed in days after the first cash flow (in
general, the date of investment of the deposit or valuation of the loan)
One of the differences between the NDER and the AAR is the underlying method for
annualising interest payments. The NDER uses successive approximation and can
be applied to any type of deposit or loan. The AAR uses the algebraic formula in
Equation 1 and is only applicable to deposits and loans where the dates of interest
capitalisation are at regular intervals and the interest payments are as frequent as, or
more frequent than, the repayments of the principal.
Both types of rates, the NDER and the AAR, may be reported for the purpose of MFI
interest rate statistics. The reason is that for products with regular capitalisation
periods, where interest payments occur more frequently or equally frequently than
the repayments of the principal, including all cases where the principal is repaid in
27
Further discussed in Section 4.4.
28
Formerly the International Securities Market Association.
Manual on MFI interest rate statistics − Types of interest rate
18
full at the end of the contract, the AAR and the NDER coincide.
29
One formula can
be derived from the other. This applies also to products with irregular or exceptional
repayments of the principal as long as these do not occur more frequently than the
interest payments.
30
Hence, for the majority of retail products the NDER and the
AAR lead to the same result.
It should be noted, however, that for products with complex cash flows only the
NDER gives the mathematically correct result and the AAR an approximation.
For example, Equation 2 gives an AAR of 10.3813% p.a. for a five-year loan at 10%
where the interest is paid at the end of each quarter and the principal repaid at
maturity. The same result is achieved with Equation 6 for the NDER, if the calculation
refers to a standard year of 365 days (see Table 2).
Table 2
Five-year loan, quarterly interest rate payments, repayment of principal at the end of
year 5: standard year of 365 days
t
Outstanding
loan
Interest
rate p.a.
Interest
payments
Repayments
of principal
Cash
flow
Discount
factor =
(1+NDER)^
(-t/365)
Present
value of
cash flow NDER
(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)
0 10,000 -10,000
1YQ1
91.25 10,000 10% 250 0 250 0.976 243.9 10.3813%
1YQ2
182.5 10,000 10% 250 0 250 0.952 237.95 10.3813%
1YQ3
273.75 10,000 10% 250 0 250 0.929 232.15 10.3813%
1YQ4
365 10,000 10% 250 0 250 0.906 226.49 10.3813%
2YQ1
456.25 10,000 10% 250 0 250 0.884 220.96 10.3813%
2YQ2
547.5 10,000 10% 250 0 250 0.862 215.57 10.3813%
2YQ3
638.75 10,000 10% 250 0 250 0.841 210.32 10.3813%
2YQ4
730 10,000 10% 250 0 250 0.821 205.19 10.3813%
3YQ1
821.25 10,000 10% 250 0 250 0.801 200.18 10.3813%
3YQ2
912.5 10,000 10% 250 0 250 0.781 195.3 10.3813%
3YQ3
1,003.75 10,000 10% 250 0 250 0.762 190.54 10.3813%
3YQ4
1,095 10,000 10% 250 0 250 0.744 185.89 10.3813%
4YQ1
1,186.25 10,000 10% 250 0 250 0.725 181.35 10.3813%
4YQ2
1,277.5 10,000 10% 250 0 250 0.708 176.93 10.3813%
4YQ3
1,368.75 10,000 10% 250 0 250 0.69 172.62 10.3813%
4YQ4
1,460 10,000 10% 250 0 250 0.674 168.41 10.3813%
5YQ1
1,551.25 10,000 10% 250 0 250 0.657 164.3 10.3813%
5YQ2
1,642.5 10,000 10% 250 0 250 0.641 160.29 10.3813%
5YQ3
1,733.75 10,000 10% 250 0 250 0.626 156.38 10.3813%
5YQ4
1,825 10,000 10% 250 10,000 10,250 0.61 6,255.27 10.3813%
Sum 5,000 10,000 RHS: 10,000
29
See also Section 4.2.3.
30
See also Section 4.2.4.
Manual on MFI interest rate statistics − Types of interest rate
19
The interest rate would be slightly lower in this example, at 10.3758% p.a., if it was
recognised that, for example, year 4 is a leap year with 366 days. This is shown in
Table 3.
Table 3
Five-year loan, quarterly interest rate payments, repayment of principal at the end of
year 5, year 4 is leap year
t
Outstanding
loan
Interest
rate p.a.
Interest
payments
Repayments
of principal
Cash
flow
Discount
factor =
(1+NDER)^
(-t/365)
Present
value of
cash flow NDER
(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)
0 10,000 -10,000
1YQ1
91.25 10,000 10% 250 0 250 0.976 243.9 10.3758%
1YQ2
182.5 10,000 10% 250 0 250 0.952 237.95 10.3758%
1YQ3
273.75 10,000 10% 250 0 250 0.929 232.15 10.3758%
1YQ4
365 10,000 10% 250 0 250 0.906 226.49 10.3758%
2YQ1
456.25 10,000 10% 250 0 250 0.884 220.96 10.3758%
2YQ2
547.5 10,000 10% 250 0 250 0.862 215.57 10.3758%
2YQ3
638.75 10,000 10% 250 0 250 0.841 210.32 10.3758%
2YQ4
730 10,000 10% 250 0 250 0.821 205.19 10.3758%
3YQ1
821.25 10,000 10% 250 0 250 0.801 200.18 10.3758%
3YQ2
912.5 10,000 10% 250 0 250 0.781 195.3 10.3758%
3YQ3
1,003.75 10,000 10% 250 0 250 0.762 190.54 10.3758%
3YQ4
1,095 10,000 10% 250 0 250 0.744 185.89 10.3758%
4YQ1
1,187.25 10,000 10% 250 0 250 0.725 181.34 10.3758%
4YQ2
1,278.5 10,000 10% 250 0 250 0.708 176.91 10.3758%
4YQ3
1,369.75 10,000 10% 250 0 250 0.69 172.6 10.3758%
4YQ4
1,461 10,000 10% 250 0 250 0.674 168.39 10.3758%
5YQ1
1,552.25 10,000 10% 250 0 250 0.657 164.29 10.3758%
5YQ2
1,643.5 10,000 10% 250 0 250 0.641 160.28 10.3758%
5YQ3
1,734.75 10,000 10% 250 0 250 0.626 156.38 10.3758%
5YQ4
1,826 10,000 10% 250 10,000 10,250 0.61 6,255.15 10.3758%
Sum 5,000 10,000 RHS: 1,0000
Equation 5 gives an AAR of 4.8809% p.a. for a deposit of EUR 10,000 that is placed
for two years, where the customer receives EUR 11,000 from the credit institution at
maturity. The NDER for this example leads to the same result; the calculation is
shown in Table 4.
Manual on MFI interest rate statistics − Types of interest rate
20
Table 4
Deposit with an agreed maturity of two years, interest payment at the end of the
second year
t
Outstanding
loan
Interest
rate p.a.
Interest
payments
Repayments
of principal Cash flow
Discount factor =
(1+NDER)^
(-t/365)
Present
value of
cash flow NDER
(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)
0 10,000 -10,000
1Y
365 10,000 0% 0 0 0 0.953 0 4.8809%
2Y
730 0 10% 1,000 10,000 11,000 0.909 10,000 4.8809%
Sum 1,000 10,000 RHS: 10,000
4.2.3 Clarification of the variable n in the annualised agreed rate formula
Question:
Which value should variable n take in the AAR formula (Equation 1) when the
interest payments and the repayment of a loan occur at different but regular
intervals?
Case A: A customer and a credit institution agree on a two-year loan at 10% p.a. with
monthly interest payments. The principal is repaid at the end of the second year.
Should variable n in Equation 1 be:
1. equal to 12 (based on the frequency of interest payments), or
2. equal to ½ (based on the repayment frequency)?
Case B: A customer and a credit institution agree on a five-year loan at 10% p.a. for
the entire maturity, where the interest should be paid at the end of each quarter and
the principal should be paid back in tranches on a monthly basis. Should n in
Equation 1 be:
1. equal to 4 to only reflect quarterly interest payments, or
2. equal to 12 to also include the repayments?
Answer:
As a general rule, for MFI interest rate statistics the value of variable n in Equation 1
is determined by the frequency of the interest payments and not by the repayment
periods of the principal. If this rule is followed, the AAR coincides with the NDER
31
whenever the interest payments are more frequent than or as frequent as the
repayments of the principal. This includes all cases where the principal is repaid at
the end of the contract. An overview of possible combinations of interest payment
and repayment frequencies is given in the following matrix.
31
Further discussed in Section 4.2.2.
Manual on MFI interest rate statistics − Types of interest rate
21
Table 5
Possible combinations of interest payment and repayment frequencies
Frequency of interest payments
Frequency of repayment of
principal
Monthly (.M) Quarterly (.Q) Yearly (.Y)
Monthly (M.) M.M M.Q M.Y
Quarterly (Q.) Q.M Q.Q Q.Y
Yearly (Y.) Y. M Y. Q Y. Y
End of contract (E.) E.M E.Q E.Y
It is expected that, in general, the frequency of the interest payments is higher than
or equal to the repayment frequency of the principal (light grey boxes in Table 5) and
that hence the AAR and the NDER coincide. An example is Case A, where variable n
in Equation 1 is equal to 12 reflecting the number of interest payments per year. As
the interest payments are at regular monthly intervals and the principal is repaid in
full at the end of the contract, i.e. situation E.M in Table 5, Equation 1 provides for
n = 12 the same result as the NDER, i.e. an interest rate of 10.4713%.
In the remaining cases, where the frequency of the interest payments is lower than
that of the repayments of the principal (dark grey boxes in Table 5), the AAR deviates
from the NDER. The NDER then represents the mathematically correct calculation
and the AAR only an approximation. An example is Case B, where variable n in
Equation 1 is equal to 4, reflecting the quarterly interest payments. Since the
repayments of the principal are at monthly frequency, the AAR differs from the
NDER, i.e. situation M.Q in Table 5. The AAR is 10.3813% and the NDER is
10.3795%.
The AAR and the NDER for a five-year loan at 10% p.a., with the combinations of
interest payment and repayment frequencies as given in Table 5are summarised in
Table 6.
Table 6
Five-year loan at 10% p.a., with the combinations of interest payment and repayment
frequencies
Frequency of interest payments
Monthly (.M) Quarterly (.Q) Yearly (.Y)
AAR NDER AAR NDER AAR NDER
Frequency of
repayment of
principal
Monthly (M.) 10.4713 10.3813 10.3795 10.0000 9.0022
Quarterly (Q.) 10.4713 10.3813 10.0000 9.0331
Yearly (Y.) 10.4713 10.3813 10.0000
End of
contract (E.)
10.4713
10.3813
10.0000
Manual on MFI interest rate statistics − Types of interest rate
22
4.2.4 Treatment of exceptional repayments of principal
Question:
Is it in line with the Regulation to apply the AAR formula (Equation 1) to financial
contracts with regular interest capitalisation, for example quarterly interest payments,
which provide the option to the customer to have exceptional repayments?
Answer:
Repayments of the principal whether regular or irregular only influence the level of
the AAR and NDER if the repayments are more frequent than the interest payments.
This is analogous to the issue discussed in Section 4.2.3. For example, in the case
of monthly interest payments and exceptional repayments of the principal that are
not more frequent than monthly, the AAR and the NDER coincide, with n = 12 in
Equation 1 reflecting the frequency of interest payments per year. Likewise, if the
interest is paid quarterly and the possibility of exceptional repayments of the principal
occurs not more frequently than quarterly, the AAR and the NDER coincide, with
n = 4 in Equation 1.
It is assumed that, in general, the interest is paid more frequently than exceptional
repayments take place. Hence, the variable n in Equation 1 is determined by the
frequency of the interest payments and the possibility of exceptional repayments can
be ignored. If exceptional repayments occur more frequently than the interest
payments, then the AAR deviates from the NDER. In that case, the NDER
represents the mathematically correct calculation and the AAR an approximation.
4.2.5 The annualised agreed rate formula for indefinite loans
Question:
Can the AAR formula (Equation 1) also be applied to loans that have been granted to
customers indefinitely but still have regular interest payments?
Answer:
The variable n in Equation 1 refers not to the maturity of the loan but to the
frequency of the interest payments. Hence, Equation 1 can be applied.
4.2.6 The annualised agreed rate formula applied to revolving loans and
overdrafts and extended credit card credit
Question:
What should n be in the AAR formula (Equation 1) for revolving loans and overdrafts
and extended credit card credit
32
that are characterised by irregular, rather than
predefined, periods of utilisation and/or repayment, and interest payments that are
based on the daily outstanding amount? Should n be equal to 365?
32
Defined in Chapter 7.
Manual on MFI interest rate statistics − Types of interest rate
23
Answer:
The use of n = 365 in Equation 1 assumes that the interest is paid on a daily basis
as in Equation 4. In this example, however, the interest is computed (ex post) by
using the daily outstanding amount, but the interest is paid at the end of the month.
In the case of monthly interest payments, variable n is equal to 12; in the case of
quarterly interest payments it is equal to 4.
4.2.7 The annualised agreed rate formula for one-off deposits
Question:
Should the AAR formula (Equation 1) be applied in the case of a one-off deposit with
an agreed maturity of three months?
Answer:
For a one-off deposit with an agreed maturity of three months, two cases can be
distinguished. If the interest is paid at the end of each month, variable n in
Equation 1 is equal to 12. If the interest is paid at the end of the three months, then n
is equal to 4. For example if 10% is agreed for the three-month deposit, then the
AAR for n = 12 is 10.47% and for n = 4 it yields 10.38%.
4.2.8 Treatment of disagio
33
Question:
The last sentence in paragraph 1 of Annex I to the Regulation states: “Disagio,
defined as the difference between the nominal amount of the loan and the amount
received by the customer, shall be considered as an interest payment at the start of
the contract (time t
0
) and shall therefore be reflected in the AAR.” How should the
following cases of disagio be treated in the AAR?
Case A: Disagio in the sense that the agreed monthly interest payments are made at
the end of the previous period.
Case B: Disagio in the sense that the agreed monthly interest payments are made at
the end of the previous period, with a different interest rate for the first period.
Case C: Disagio in the sense of a payment at the beginning of the contract that has
no link with the subsequent interest payments, which are based on the outstanding
amount of the loan in the previous period.
Answer:
In Cases A, B and C, the correct interest rate can be given by the NDER. In Cases A
and B, it might be possible to reflect the advance payment for the first period as
disagio in Equation 1. However, this formula ignores for all other periods that the
33
“Disagio” is usually known as “discount”.
Manual on MFI interest rate statistics − Types of interest rate
24
interest is paid in advance and not at the end of the period. For such a complex
product, only the NDER will yield the correct result.
4.2.9 Treatment of agio
Question:
How should the AAR be determined when the financial contract includes an agio
defined as the “inverse” of a disagio,
34
i.e. the price being higher than the nominal
amount of the loan or non-negotiable debt security? Should the agio be incorporated
at all? An example is the following purchase of a (borrowers’) note loan
(Schuldscheindarlehen), where the agio is included in the purchase price of the note
loan:
• Purchase of note loan on 6 March 2014, nominal: EUR 5,112,918.81
• Value date: 10 March 2014
• Interest rate: 7.60%
• Original maturity: 25 March 2012 – 25 March 2015
• Purchase price: EUR 5,403,332.60
• Agio (included in purchase price): EUR 5,403,332.60 – EUR 5,112,918.81 =
EUR 290,413.79
Answer:
The (secondary) purchase of this note loan on 6 March 2014 is not considered to be
new business
35
because no new funds are being granted to the borrower and no
new terms and conditions
36
are negotiated for the borrower. The (secondary)
purchase is simply a change in ownership of the note loan. New business would only
occur in the case of a (primary) issue of the note loan by an MFI, where at the time
of issue the borrower receives money through the issuance of a financial instrument,
which according to the ECB’s MFI balance sheet statistics is classified as a loan
rather than a debt security. If this primary issue involves an agio, it would be
reflected in the MFI interest rate.
The agio is reflected in the MFI interest rate statistics on outstanding amounts,
37
irrespective of whether the note loan or non-negotiable debt security is initially issued
or subsequently acquired by the MFI. The agio results in a reduction of the interest
rate. The rate was 7.6% on EUR 5,112,918.81, which results in an interest payment
of EUR 388,581.83. This amount applied to the now higher purchase price of EUR
5,403,332.60 is equivalent to an interest rate of 7.1915%. The latter is, from 6 March
2014 onwards, reflected in the MFI interest rate statistics on outstanding amounts.
34
See also Section 4.2.8.
35
See also Section 5.4.1.
36
The meaning of this expression is specified in Section 5.4.1.
37
See also Section 5.2.
Manual on MFI interest rate statistics − Types of interest rate
25
The same holds true for the (primary) issue and (secondary) purchase of any other
non-negotiable debt security, which according to MFI balance sheet statistics is to be
classified as a loan.
4.2.10 Annualising variable interest rates
Question:
In March 2014 a customer and a credit institution agree on a two-year loan with a
variable interest rate. The contract specifies the lending rate as a certain spread over
an underlying index. The value of the interest rate for the first month, i.e. March
2014, is 10%. In March 2014 there is no information regarding the future
development of the interest rate for the loan as it will rise and fall based on the
movement of the underlying index. In which way would the reporting agent calculate
the AAR or the NDER in March 2014? Should it be assumed that the interest rate is
10% for the remaining periods?
Answer:
For the statistics on new business,
38
the interest rate that is taken into account in the
calculation of the AAR or the NDER is 10%. In general, for variable rates it is
assumed that the interest rate remains constant at the level on the date of
agreement on the contract, in this example at 10%. This is the same as assuming
that the future unknown development of the interest rate is not taken into account.
Hence, if the variable rate is agreed and paid per annum, the AAR or the NDER to
be recorded under new business in March 2014 is 10%. If the variable rate is to be
paid per month or quarter, the interest rate needs to be annualised. The AAR and the
NDER would then be 10.4713% and 10.3813% respectively.
The interest rate that is recorded for new business is also reflected in the statistics
on outstanding amounts
39
as the first reporting in March 2014, assuming a total or
partial withdrawal of the loan on this date. The subsequent reporting on outstanding
amounts reflects the interest rate applied by the reporting institution at the time of the
calculation of MFI interest rates and hence shows the variability of the interest rate
over time.
38
See also Section 5.4.1.
39
See also Section 5.2.
Manual on MFI interest rate statistics − Types of interest rate
26
4.3 Treatment of taxes, subsidies and regulatory
arrangements
40
4.3.1 Taxes, subsidies and favourable rates
The AAR and the NDER reflect what the reporting agent pays on deposits and
receives for loans. If the amount paid by one party and received by the other differs,
the point of view of the reporting agent determines the interest payment covered by
MFI interest rate statistics. Following this principle, MFI interest rates:
• are recorded on a gross basis before tax, since the pre-tax interest rates reflect
what reporting agents pay on deposits and receive for loans;
• do not take into account subsidies granted to households or non-financial
corporations by third parties such as government, because these subsidies are
not paid or received by the reporting agent. These subsidies can be received
either directly by the households or non-financial corporations or, indirectly,
through the reporting agent;
• reflect favourable rates that reporting agents apply to their employees.
Favourable rates do not include a subsidy granted by a third party but are
actually applied by the reporting agent.
Therefore, in the case of deposits, MFI interest rate statistics capture what the MFI
pays but not what the household or non-financial corporation receives in terms of
interest payments. For example, if a customer receives 5% p.a. on a deposit where
3% is actually paid by the reporting agent and the other 2% is a subsidy by a third
party, which is transferred to the customer via the reporting agent, then the 3% p.a.
is covered by MFI interest rate statistics. Further examples are given in Figure 1.The
deposit interest rates covered by MFI interest rate statistics are shaded, i.e. the 3%
on the deposits of customers 1, 2 and 3 and the 5% for the employee.
40
See also paragraphs 4 to 8 of Annex I to the Regulation.
Manual on MFI interest rate statistics − Types of interest rate
27
Figure 1
Subsidies and favourable rates on deposits
Analogously, in the case of loans, MFI interest rate statistics capture what the MFI
charges in terms of interest rates but not what the household or non-financial
corporation pays. For example, if a customer pays 6% p.a. for a loan where 10% is
actually charged by the reporting agent but a third party deducts a 4% subsidy from
this interest rate, and this is transferred to the customer via the reporting agent, the
10% p.a. is covered by MFI interest rate statistics. Further examples are given in
Figure 2. The lending interest rates reflected in MFI interest rate statistics are
shaded, i.e. the 10% for the loans of customers 1, 2 and 3 and the 6% for the loan of
the employee.
Figure 2
Subsidies and favourable rates on loans
It should be stressed that as the MFI perspective is the relevant one to determine the
interest rates that have to be recorded in MFI interest rate statistics, any subsidy
received by the MFI that is linked to a favourable interest rate applied to the
household or non-financial corporation should be considered as a subsidy to
correctly compute MFI interest rate statistics. The following example aims at
clarifying further this issue.
Pays 3%
Transfer s subsidy of 2%
Cr edit instituti on
Government
Customer 3
Customer 2
Customer 1
Employee
Pays 3%
Pays 3%
Pays 3%
5
%
Subsidy of 2% for customer 3
Subsidy of 2%
Char g es 10%
Transfer s subsidy of 4%
Cr edit instituti on
Government
Customer 3
Customer 2
Customer 1
Employee
Char g es 10%
Char g es 10%
Char g es 6%
6%
Subsidy of 4% for customer 3
Subsidy of 4%
Manual on MFI interest rate statistics − Types of interest rate
28
Question:
In order to foster growth in credit to small and medium-sized enterprises (SMEs), in
particular because this type of enterprise was especially affected by the global
financial crisis, the NCBs could launch special programmes, through the credit
institutions, which would consist of offering new loans to SMEs with an interest rate
ceiling significantly below the prevailing market rate for similar loans (let us suppose
that the interest rate applied by the credit institution is 2%, whereas the market rate
for such loans is 6%). The credit institutions participating in the programme would
receive the necessary funds for this particular activity from the NCB through loans at
a rate of 0%. Let us suppose that the normal current rate on the credit institutions’
funding from the NCB is 4%.
Which interest rate should be recorded in MFI interest rate statistics for these loans
to SMEs?
Answer:
The fact that the interest rate charged by the credit institutions participating in the
programme is below market conditions means that these loans are subsidised loans.
In this case, the SMEs do not receive the subsidy directly but indirectly through the
credit institutions. Therefore, the interest rate that has to be computed, in both new
business and outstanding amounts, is the interest rate that the credit institution
actually receives. This is in line with the Regulation which states that, in determining
the interest rate of instruments with subsidies, the perspective of the reporting
agents is the one that needs to be taken into account.
In the example described above, the credit institutions involved in the special
programme receive an interest rate for the loans granted of 2% from the SMEs plus
4% from the NCB for their own financing. Therefore, the interest rate to be computed
in MFI interest rate statistics is 6%. Not taking into account the subsidy being
granted to the credit institutions for their own financing, at a 0% rate for these types
of loans, would artificially bias downwards the interest rates for the non-financial
corporation sector as a whole.
At this point, it could be of interest to clarify the wording in paragraph 6 of Annex I to
the Regulation: “Furthermore, subsidies granted to households or non-financial
corporations by third parties are not to be taken into account when determining the
interest payments, because the subsidies are not paid or received by the reporting
agent.” The paragraph refers to subsidies received directly by households and non-
financial corporations. In the case considered in this example, the subsidy is
received indirectly by SMEs through the credit institutions.
The case under analysis can be considered in a different “formal” way with the same
results. Let us suppose that the credit institutions have to borrow from the NCB at
the normal rate of 4%. As the credit institutions want to obtain a margin of 2%, they
would grant the loans to the SMEs at 6% (4%+2%) and then the government gives a
subsidy “directly” to the SMEs of 4%. The result for both the credit institutions and
the SMEs is the same but, in this second presentation, paragraph 6 of Annex I of the
Regulation can be more directly followed.
Manual on MFI interest rate statistics − Types of interest rate
29
Special programmes similar to the programme explained above can be launched by
institutions such as the European Investment Fund (EIF), the European Bank for
Reconstruction and Development (EBRD), the European Investment Bank (EIB),
etc., under which MFIs are funded at lower interest rates while being under the
obligation to extend loans to different groups of economic agents (SMEs, farmers,
etc.) at rates lower than what they would be able to get if no such programmes
existed. In these cases, the treatment in MFI interest rate statistics is the same as
the one explained above. The MFIs have to calculate the amount of the subsidy as
the difference between their “normal” funding rate and the lower rate obtained linked
to the special programme. The amount of the subsidy plus the rate charged to the
customers benefiting from the programme is the rate to be reported for both new
business and outstanding amounts.
4.3.2 Special national practices including regulatory arrangements
Where national regulatory arrangements affect interest payments, for example direct
subsidies on certain types of instruments or interest rate ceilings, these are reflected
in MFI interest rate statistics. Any change to the regulatory arrangements, for
example an increase (decrease) in the level of administered interest rates, is shown
in the statistics as an increase (decrease) of the respective MFI interest rate.
In addition to national regulatory arrangements, the level and development of MFI
interest rates may also be influenced by special national practices that are not legally
binding. These can include national conventions, institutional arrangements and
specific deposit or lending products offered at national level:
• National conventions cover usual banking practices which give rise to lower or
higher than usual interest rates, which are not necessarily due to legal acts, for
example a general but not legally enforced 0% remuneration on overnight
deposits.
• Institutional arrangements are similar to national conventions but affect only a
specific group of institutions. The legal status of these arrangements is
irrelevant for the purpose of analysing the MFI interest rate statistics.
• Specific national products may, as a result of their special features, carry
unusually high or low interest rates compared with other products falling into the
same instrument category. The treatment of some specific products is included
in Chapter 8 of this Manual.
4.3.3 The annualised agreed rate formula for subsidised loans
Question:
How should the AAR be calculated for a subsidised loan, where the credit institution
receives interest payments from its customer and the subsidy from the government
at different frequencies? For example, the credit institution receives 6% interest for a
subsidised loan, 4% directly from the customer at a monthly frequency and 2% from
Manual on MFI interest rate statistics − Types of interest rate
30
the government. The interest payments by the customer take place monthly,
whereas those by the government are made only twice a year. Hence, no single
value exists for the variable n in the AAR formula (Equation 1): n would be 12 for the
interest paid by the customer and 2 for the interest paid by the government.
Answer:
In this example of a subsidised loan at 6% p.a. where the customer pays 4% p.a. at
monthly frequency and the government pays the subsidy of 2% p.a. every six
months, the rate recorded in MFI interest rate statistics is 6.1416% calculated as the
NDER.
Equation 1 cannot derive this result. This simple formula allows either the use of n =
12, which gives 6.17%, assuming that both the customer and government pay
monthly or, alternatively, it permits the division of the loan into two parts: one at 4%
with n = 12 and one at 2% with n = 2, leading to 6.08% in total. Both results are
incorrect. Equation 1 can be applied as long as there is only one frequency for all the
payments, but does not deliver correct results for more complex products. In these
cases, NCBs should ask the reporting agents to calculate and report the NDER.
4.4 Annual percentage rate of charge
41
4.4.1 Definition of and link to the Consumer Credit Directive and the
Mortgage Credit Directive
The first Consumer Credit Directive was enacted in 1986 with the purpose of setting
rules to protect consumers through the harmonisation of terms, information and
methodology relating to the main concepts and instruments of credit markets. Twenty
years later, an analysis of the national laws transposing the Directive showed that,
due to differences in legal or economic situations at national level, in addition to the
Directive, the Member States used a variety of consumer protection mechanisms.
These differences were creating significant distortions in competition among
creditors in the European Union and, in turn, obstacles to the internal market. On
these grounds and in response to this evidence, Directive 2008/48/EC
42
on credit
agreements for consumers (hereafter the “Consumer Credit Directive”) was enacted.
However, credit agreements covering the granting of credit secured by real estate
were excluded from the scope of that Directive, owing to the special nature of this
type of credit. Likewise, credit agreements, the purpose of which is to finance the
acquisition or retention of property rights in land or in an existing or projected
building were also excluded from the Directive. These two types of credit
agreements have been regulated more recently, through the Directive 2014/17/EU
41
See also paragraphs 9 to 11 of Annex I to the Regulation.
42
Directive 2008/48/EC of 23 April 2008 on credit agreements for consumers and repealing Council
Directive 87/102/EEC (OJ L 133, 22.5.2008, p. 66), as amended by Directive 2011/90/EU of 14
November 2011 (OJ L 296, 15.11.2011, p. 35).
Manual on MFI interest rate statistics − Types of interest rate
31
on credit agreements for consumers relating to residential immovable property
43
(hereafter the “Mortgage Credit Directive”).
One of the key concepts regulated by both Directives is the annual percentage rate
of charge (APRC), defined as the total cost of the credit to the consumer expressed
as an annual percentage of the total amount of credit. The APRC equals, on an
annual basis, the present value of all drawdowns and the present value of all
repayments and charges agreed between the creditor and the consumer, and should
be calculated in accordance with a precise mathematical formula, which is
represented below as Equation 7.
()
=
()
’
[7]
where:
• X is the APRC
• m is the number of the last drawdown
• k is the number of a drawdown, thus 1≤ k ≤ m
• C
k
is the amount of drawdown k
• t
k
is the interval, expressed in years and fractions of a year, between the date
of the first drawdown and the date of each subsequent drawdown, thus, t_
1
t
1
=
0
• m’ is the number of the last repayment or payment of charges
• l is the number of a repayment or payment of charges
• D
l
is the amount of a repayment or payment of charges
• s
l
is the interval expressed in years and fractions of a year, between the date
of the first drawdown and the date of each repayment or payment of charges
Remarks:
(a) The amounts paid by both parties at different times shall not necessarily
be equal and shall not necessarily be paid at equal intervals.
(b) The starting date shall be that of the first drawdown.
(c) Intervals between dates used in the calculation shall be expressed in
years or in fractions of a year. A year is presumed to have 365 days (or
366 days for leap years), 52 weeks or 12 equal months. An equal month is
presumed to have 30.41666 days (i.e. 365/12) regardless of whether or
not it is a leap year.
43
Directive 2014/17/EU of 4 February 2014 on credit agreements for consumers relating to residential
immovable property and amending Directives 2008/48/EC and 2013/36/EU and Regulation (EU) No
1093/2010 (OJ L 60, 28.2.2014, p. 34).
Manual on MFI interest rate statistics − Types of interest rate
32
(d) The result of the calculation shall be expressed with an accuracy of at
least one decimal place
44
. If the figure at the following decimal place is
greater than or equal to 5, the figure at this particular decimal place shall
be increased by one.
On the costs that have to be included, the Directives mention expressly the following:
• interest, commissions, taxes and any other kind of fees which the consumer is
required to pay in connection with the credit agreement and which are known to
the creditor, except for notarial costs;
• costs in respect of ancillary services relating to the credit agreement, in
particular insurance premiums, are also included if, in addition, the conclusion
of a service contract is compulsory in order to obtain the credit or to obtain it on
the terms and conditions marketed;
• the cost of valuation of property where such valuation is necessary to obtain the
credit, but excluding registration fees for the transfer of ownership of the
immovable property.
With regard to the costs excluded, apart from what has already been mentioned, the
Directives also exclude the following:
• charges payable by the consumer for non-compliance with any of his/her
commitments laid down in the credit agreement;
• charges other than the purchase price which, in the purchase of goods and
services, the consumer is obliged to pay whether the transaction is paid in cash
or by credit;
• the costs of maintaining an account recording both payment transactions and
drawdowns, the costs of using a means of payment for both payment
transactions and drawdowns, and other costs relating to payment transactions
included in the total cost of credit to the customer unless the opening of the
account is optional and the costs of the account have been clearly and
separately shown in the credit agreement.
Even though the scope covered by both Directives is very broad, the following credit
agreements are not covered:
• In the Consumer Credit Directive, credit agreements involving a total amount of
credits less than EUR 200 or more than EUR 75,000; however, unsecured
credit agreements, the purpose of which is the renovation of a residential
44
This is a direct quote from the Directives and only included here to illustrate the requirements of the
Directives. Paragraph 42 of Part 14 of Annex II to the Guideline of the ECB of 4 April 2014 on monetary
and financial statistics (recast) (ECB/2014/15) specifies that “NCBs provide the MFI interest rates on
outstanding amounts and on new business to the ECB to four decimal places. This is without prejudice
to any decision taken by the NCBs on the level of precision they wish to apply in collecting the data.
The published results do not contain more than two decimal places.” Hence it is up to the NCB to
define the level of detail at which the reporting agents submit the data to the NCB.
Manual on MFI interest rate statistics − Types of interest rate
33
immovable property involving a total amount of credit above EUR 75,000 are
covered;
• credit agreements in the form of an overdraft facility and where the credit has to
be repaid within one month;
• credit agreements where the credit is granted free of interest and without any
other charges and credit agreements under the terms of which the credit has to
be repaid within three months and only insignificant charges are payable;
• credit agreements where the credit is granted by an employer to employees as
a secondary activity free of interest or at annual percentage rates of charge
lower than those prevailing on the market and which are not offered to the
public generally.
Both Directives include clauses on harmonisation and sanctions.
With regard to harmonisation, the Directives state that Member States may not
maintain or introduce in their national law provisions diverging from those laid down
in the Directives, and that Member States shall take the necessary measures to
ensure that consumers do not lose the protection granted by the Directives by
means of any sort of circumvention. On sanctions, both Directives state that Member
States shall lay down the rules on sanctions applicable to infringements of the
national provisions adopted pursuant to the Directives and shall take all measures
necessary to ensure that they are implemented. The sanctions shall be effective,
proportionate and dissuasive.
Finally, on transposition, the Consumer Credit Directive established 12 May 2010 as
the deadline for transposition, while the Mortgage Credit Directive established
21 March 2016 as the deadline.
Regarding the MFI interest rate statistics, although significant charges might be
applied in all lending categories, the Regulation requires the compilation of an APRC
only for loans to households for consumption and for house purchases (indicators 30
and 31 of Table 5 in the Appendix).
4.4.2 Indicator for other loan charges
In general, MFIs pay interest on deposits to their customers but also charge fees.
Analogously, MFIs usually receive payments for the loans granted from their
customers comprising an interest rate component and a component of other related
charges. Both components may have an impact on the monetary transmission
mechanism.
The AARs and the NDERs collected for the purpose of MFI interest rate statistics
provide extensive information about the interest paid on deposits and charged for
loans by MFIs. The APRCs collected for consumer credit and loans to households
for house purchases incorporate indistinguishably information about the interest rate
and related fees charged by the same institutions. Hence, the APRC can change
Manual on MFI interest rate statistics − Types of interest rate
34
from one month to another as a result of changes in the interest rate component or
the component of other charges or both.
Since there are no methodological differences between the AAR and the NDER, on
the one hand, and the APRC, on the other, the APRC should not be lower than the
AAR (or the NDER) because the APRC includes charges in addition to the interest
rate. The mathematical difference between the APRC and the AAR (or the NDER)
therefore represents the component of other charges applied to loans. Indeed, the
main aim of collecting the APRC for the purpose of MFI interest rate statistics is to
construct an indicator of other loan charges applied to loans to households for
consumption and for house purchases, which allows the monitoring of developments
in loan charges over time.
Significant charges might be applied in all lending categories. However, in order to
limit the reporting burden on MFIs, the Regulation requires no information related to
charges for revolving loans and overdrafts, extended credit card credit, loans to
households for other purposes and loans to non-financial corporations. For deposits,
charges are assumed to be less significant than for loans, hence no data are
collected.
4.4.3 Charges to be taken into account at national level
Question:
The transposition of the Consumer Credit Directive and the Mortgage Credit
Directive into national law may require the calculation of the APRC with respect to
loans to households for house purchases based on assumptions that are not
identical to those in the Directives. In particular, the national legislation in one
Member State requires the exclusion of mortgage protection insurance as a relevant
charge from the APRC. In this Member State, mortgage protection insurance is
compulsory for the vast majority of housing loans. However, the MFIs do not always
know the amount of the mortgage protection premium applicable to the individual
borrower, as the insurance policy may be arranged through an insurance
intermediary other than the lender. Accordingly, under national law the costs of such
insurance are not included in the APRC for housing loans. Should the national
definition of the APRC be followed in this case for the purpose of MFI interest rate
statistics?
Answer:
Paragraph 11 of Annex I to the Regulation acknowledges that the composition of the
fees to be taken into account in the APRC may differ across countries, because the
Consumer Credit Directive is differently applied and because national financial
systems and the procedure for securing credits differ. Nevertheless, it is expected
that the clauses for harmonisation that the Consumer Credit Directive contains and
the fact that the date for its transposition into national legislation was 12 May 2010
will have reduced the differences in the application of the APRC across countries.
But as the Mortgage Credit Directive had to be transposed by 21 March 2016, it is
very probable that for credit agreements for consumers relating to residential
Manual on MFI interest rate statistics − Types of interest rate
35
immovable property, such differences are still significant. Hence, if national
legislation on the APRC provides that mortgage protection insurance should not be
considered as a relevant charge in the calculation of the APRC, the reporting agents
in that Member State may exclude such mortgage protection from the calculation of
the MFI interest rate statistics.
4.4.4 Period of fixation in the calculation of the APRC
Question:
The current national law in a Member State may require the calculation of the APRC
with respect to loans to households for house purchases based on assumptions that
are not identical with those in the Mortgage Credit Directive. In particular, in one
Member State the legislator took the view that undue emphasis should not be placed
on the initial interest rate, which applies only for a short period relative to the full term
of the loan. Therefore, the national legislation requires the calculation of the APRC
based on the assumption that the fixed interest rate at the beginning of the contract
applies only for the period specified and that the interest rates applicable to other
periods of the contract are the current variable rates. Should the national definition of
the APRC be followed in this case for the purpose of MFI interest rate statistics?
Answer:
The initial period of fixation is discussed in detail in Section 7.7.4. In general, MFI
interest rates on new business only reflect the interest rate that is agreed for the
initial period of fixation at the start of the contract or after renegotiations of the loan.
If, after this initial period of fixation, the interest rate automatically changes to a
variable interest rate, which might be at a very different level, this is not reflected in
new business statistics but in the statistics on outstanding amounts. For example, if
a ten-year loan is granted where it is agreed at time t
0
that for the first 12 months the
interest is fixed at 10% and then automatically adjusted to EURIBOR plus x basis
points, then the MFI interest rate on new business captures the 10%. The changes in
the interest rate over time, i.e. the change from 10% to EURIBOR plus x basis points
and then the movements of EURIBOR over time, are only captured in the MFI
interest rates on outstanding amounts.
In general, this approach should be followed for all contracts with initial rate fixation.
However, if according to the national legislation the calculation of the APRC warrants
a different treatment for certain products with initial rate fixation, then the same
treatment should also be applied for these products to calculate the interest rate
without charges. The reason for using the same method of calculation for the interest
rate with and without charges is that the difference between them should serve as an
indicator for loan charges as discussed in Section 4.4.2. In this example, the national
legislation requires that the APRC for loans to households for house purchases
should be based on the assumption that the fixed rate applies for only the period
specified and that the interest rates applicable to other periods of the contract are the
current variable rates. If the reporting agents follow this approach for calculating the
APRC for the purpose of MFI interest rate statistics, this is reflected in the new
business indicator 31 of Table 5 in the Appendix referring to loans to households for
Manual on MFI interest rate statistics − Types of interest rate
36
house purchases. The difference between indicator 31 and the interest rate for loans
to households for house purchases without charges, i.e. the weighted average of
new business indicators 16 to 19 in Table 2, should only be the charges. Hence, the
assumption for the initial fixed rate that is applied to indicator 31 also needs to be
applied for new business indicators 16 to 19 referring to loans to households for
house purchases with a variable rate and up to one year initial rate fixation, over one
and up to five years initial rate fixation, over five and up to ten years initial rate
fixation, and over ten years initial rate fixation. As a consequence, the NDER needs
to be compiled for new business indicators 16 to 19 in this Member State, because
the AAR could only take into account the fixed rate at the beginning of the contract.
The alternative is that the MFIs ignore the national legislation regarding the APRC
and report for the purpose of MFI interest rate statistics the fixed rate including
charges as the APRC and the fixed rate without charges as the AAR.
4.4.5 Treatment of subsidies in the APRC
Question:
The transposition of the Consumer Credit Directive and the Mortgage Credit
Directive into national law may require that the APRC reflects the total costs for a
loan from the point of view of the customer. For example, a credit institution grants a
loan at 10% p.a., where the customer pays 6% and the government transfers the
remaining 4% as a subsidy directly to the MFI. In this case, the APRC calculated
according to national legislation would reflect the 6% p.a. paid by the customer plus
any charges. Should the national definition of the APRC be followed in this case for
the purpose of MFI interest rate statistics?
Answer:
The treatment of subsidies is discussed in detail in Section 4.3. In general, MFI
interest rate statistics take the point of view of the reporting agent and not of the
customer. This means that in the above example, MFI interest rate statistics would
reflect what the MFI charges in terms of interest rate, i.e. the 10%, but not the 6%
interest that the customer pays.
As explained in Section 4.4.2, there should be no methodological differences
between the APRC and the AAR (or the NDER). The APRC should not be lower than
the AAR (or the NDER) and the difference between them should be solely
attributable to the existence of other charges applied to loans. It is therefore
essential that subsidies are treated in the same way in the AAR (or the NDER) and
the APRC, as otherwise the indicator of other loan charges cannot be interpreted.
45
As a consequence, when calculating the APRC for the purpose of MFI interest rate
statistics the point of view of the reporting agent needs to be taken and the national
legislation regarding the treatment of subsidies ignored. Therefore, in the above
example, both the AAR (or the NDER) and the APRC need to reflect the 10% interest
rate that is charged by the MFI.
45
This is analogous to the argumentation for the treatment of the initial rate fixation in Section 4.4.4.
Manual on MFI interest rate statistics − Types of interest rate
37
In theory, it would also be possible to apply in this Member State, for all types of
rates, the treatment of subsidies as specified in the national legislation, i.e. to take
the point of view of the consumer. However, this would not be in line with paragraph
4 of Annex I to the Regulation and would moreover compromise the comparability of
MFI interest rate statistics across the euro area.
4.4.6 Treatment of non-profit institutions serving households in the APRC
As deposits from and loans to non-profit institutions serving households (NPISHs)
are usually quantitatively less important than deposits from and loans to households
and non-financial corporations, the household sector includes NPISHs for the
purpose of MFI interest rate statistics in general.
46
However, for the APRC on loans
to households for consumption and for house purchases, (indicators 30 and 31 in
Table 5 of the Appendix) NCBs may grant a derogation to reporting agents regarding
such loans to NPISHs.
47
The reason is that certain loans to NPISHs may not be
covered by the national legislation referring to the calculation of the APRC. It is
expected that loans to NPISHs for consumption or house purchase are negligible. If
they were not to be negligible, however, the exclusion of loans to NPISHs from the
APRC for consumer credit and loans to households for house purchases and their
inclusion in the AAR or the NDER would lead to a distortion of the indicator for other
loan charges.
48
Question:
If an NCB grants the derogation to its reporting agents, this implies that all loans to
NPISHs are excluded from the calculation of the APRC. Therefore, the APRC does
not cover the interest rates on new loans for consumption and new loans for house
purchases granted to NPISHs. The question is whether, in the calculation of the
APRC, NCBs should also exclude the amount of new loans granted to NPISHs?
Answer:
In order to calculate the APRC as a weighted average rate, the Regulation assumes
that it is possible to use the new business volumes that are provided for the
calculation of the AAR (or the NDER).
49
However, these new business volumes
include the amount of new business in respect of loans to households and NPISHs,
whereas if the derogation is applied, the weighted interest rate to be calculated
refers only to households, i.e. excluding NPISHs. Hence, if the derogation is applied,
reporting agents and NCBs require additional data to calculate the APRC.
As the reporting agents apply the derogation, it can be assumed that they are in a
position to distinguish between households and NPISHs. Therefore, NCBs should
ask reporting agents to calculate the APRC with weights covering only households,
i.e. excluding NPISHs, and to supply the NCB with additional series for the amount
46
See also Chapter 3 and Section 7.4.
47
See footnote 1 to paragraph 9 of Annex I to the Regulation.
48
See also Section 4.4.2.
49
Further discussed in Sections 9.3 and 9.4.
Manual on MFI interest rate statistics − Types of interest rate
38
of such loans to households. The additional series on the amount of loans to
households for consumption excluding NPISHs, i.e. indicator 30, and loans to
households for house purchases excluding NPISHs, i.e. indicator 31, need to be
transmitted to the ECB. The amounts might not be significant but this is not
necessarily known for certain in advance.
Manual on MFI interest rate statistics − Business coverage
39
5 Business coverage
50
5.1 Bad loans and loans for debt restructuring below market
conditions
Interest rates on bad loans and on loans for debt restructuring below market
conditions are not collected for the purpose of MFI interest rate statistics. The reason
is that the interest rate agreed on a loan for debt restructuring is not the result of the
general demand and supply conditions in the loan market at the time of the
agreement, but rather of what the indebted customer is able to pay and the
agreement that the customer is able to reach with creditor MFIs. Hence, interest
rates on loans for debt restructuring at rates below market conditions are, like other
bad loans, not the type of interest rate that is supposed to be covered as an agreed
rate by MFI interest rate statistics.
Bad loans are defined in accordance with Annex II to Regulation (EU) No 1071/2013
(ECB/2013/33) which, in turn, relates to Article 178 of Regulation (EU) No
575/2013.
51
As for loans for debt restructuring, i.e. restructuring in relation to
financially distressed debtors, they should be defined in accordance with existing
national definitions.
A key question that would have to be clarified is when an interest rate applied to a
loan can be considered to be below market conditions. It would be desirable to set
precise rules to try to harmonise this concept across countries; however, the
differences that still exist in the financial systems of the Member States and in the
level of retail interest rates, exacerbated by the economic and financial crisis, make it
very difficult to arrive at a harmonised position on this issue. Therefore, at the
moment, it seems more convenient to leave the assessment of when to qualify the
rate applied to loans as below market conditions to the reporting agents, as this
concept varies depending on the conditions prevailing in the different national
markets, the different instruments and even the different kinds of customers.
Undoubtedly, the reporting agents are in the best position to assess whether they
have given special treatment to a customer in terms of the interest rate applied to a
loan and the reasons for that. As not all the loans granted at interest rates below
market conditions have to be excluded from MFI interest rate statistics, only those for
the purpose of debt restructuring and also all bad loans, the reporting agents know
exactly if a loan is classified as a bad loan and if a loan to which they are applying
the lower than usual interest rate corresponds to restructured debt.
Another aspect that should be clarified is that although the interest rates on bad
loans and loans for debt restructuring below market conditions are excluded from
50
This chapter refers mainly to Part 2 of Annex I to the Regulation.
51
Regulation (EU) No 575/2013 of the European Parliament and of the Council of 27 June 2013 on
prudential requirements for credit institutions and investment firms and amending Regulation (EU) No
648/2012 (OJ L 176, 26.6.2013, p. 1).
Manual on MFI interest rate statistics − Business coverage
40
MFI interest rate statistics, their amounts are included in the weighting information
used for aggregating MFI interest rates on outstanding amounts at the national and
euro area levels. This exception aims to alleviate the burden on MFIs which do not
have to report the outstanding amounts of the different instruments to the NCBs for
them to calculate the interest rates at national level. Instead, the amounts
outstanding are taken from MFI balance sheet statistics. However, the amounts in
the balance sheet cover bad loans and loans for debt restructuring, but as these
loans enter into the numerator (the weights) and the denominator (the total
outstanding amounts), the incurred error is not considered relevant.
5.2 Interest rates on outstanding amounts
52
MFI interest rate statistics on outstanding amounts give information about the
interest paid and received by households and non-financial corporations, which
allows the analysis of any changes in the disposable income of these sectors and
their interest burden. From the point of view of the MFIs, the statistics also refer to
the interest paid or received, which allows the analysis of changes in interest rate
margins and banks’ profitability. Interest rates on outstanding amounts are
furthermore needed to calculate the own rate of return on M3 and its components. A
more exhaustive list of uses is given in Chapter 2.
Outstanding amounts are defined as the stock of all deposits placed by customers,
i.e. households and non-financial corporations, with MFIs, and the stock of all loans
granted by these MFIs to their customers. An interest rate on outstanding amounts
reflects the weighted average
53
interest rate level applied to the stock of deposits or
loans in the relevant instrument category as at the time reference point:
• Interest rates on outstanding deposits cover all deposits placed and not yet
withdrawn by customers in all the periods up to and including the reference
period.
• Interest rates on outstanding loans cover all loans withdrawn and not yet repaid
by customers in all the periods up to and including the reference period; this
excludes bad loans and loans for debt restructuring at rates below market
conditions.
MFI interest rate statistics on outstanding amounts therefore include the interest
rates actually applied to the stock of all deposits and loans.
52
See also paragraphs 14 and 15 of Annex I to the Regulation.
53
Aggregations are discussed in Chapter 9.
Manual on MFI interest rate statistics − Business coverage
41
5.3 Interest rates on overnight deposits, deposits redeemable
at notice, convenience and extended credit card credit,
and revolving loans and overdrafts
54
5.3.1 The balance at the time reference point as an indicator for new
business
The general concept of interest rates on outstanding amounts is subject to
Section 5.2. Interest rates on new business are explained in Section 5.4. Five
instrument categories, i.e. (1) overnight deposits, (2) deposits redeemable at notice,
(3) revolving loans and overdrafts, convenience and extended credit card credit, (4)
revolving loans and overdrafts and (5) extended credit card credit,
55
form a separate
group for which the interest rates on outstanding amounts and new business
coincide. They are included as new business indicators 1, 5, 6, 7, 12, 23, 32 and 36
in Table 2 of the Appendix and indicators 86 and 87 in Table 8 of the Appendix. The
method for compiling these indicators is, however, the same as for interest rates on
outstanding amounts. Overnight deposits, deposits redeemable at notice (in
particular non-transferable sight savings deposits
56
), revolving loans and overdrafts,
and convenience and extended credit card credit form a separate group, because
they experience a large number of inflows and outflows throughout the month. The
increases and decreases in the amount on these accounts arise from receipts and
payments related to the customer’s economic activity, and are therefore related to
transactions rather than to the autonomous investment decisions of the customer.
Moreover, the bulk of these deposits or revolving loans and overdrafts and credit
card debt
57
is usually turned over during the period. The balance at the time
reference point
58
is considered to be the most appropriate indicator for new
business:
• For overnight deposits and deposits redeemable at notice, the balance at the
time reference point reflects the amount of money the customer has chosen to
put into this type of deposit instead of placing the money elsewhere.
• For convenience credit, the balance at the time reference point reflects the
amount of money spent from the dedicated card account but not yet repaid in
the period between the use of the card and the relevant billing date. Extended
credit card credit is the remaining debt on the account after the relevant billing
date at the end of the reporting period.
54
See also paragraphs 16 to 18 of Annex I to the Regulation.
55
All defined in Section 7.5.
56
Non-transferable sight savings deposits, which – although legally redeemable on demand – are subject
to significant penalties, have features that are very close to overnight deposits and are offered by credit
institutions in some Member States. According to Regulation ECB/2013/33, they are classified as
deposits redeemable at up to three months’ notice.
57
Credit card debt, as defined in Part 2 of Annex II to Regulation ECB/2013/33, comprises both
convenience and extended credit card credit.
58
The time reference point is further discussed in Section 6.1.
Manual on MFI interest rate statistics − Business coverage
42
• For revolving loans and overdrafts, the balance at the time reference point
reflects the amount of money the customer has chosen to draw on a loan facility
or to leave as a debit balance on a current or cheque account instead of
borrowing the money elsewhere.
By leaving a “net” (debit/credit) balance on the overnight deposit
59
or deposit
redeemable at notice, the customer has implicitly agreed to the terms and conditions
of the account. The customer adjusts this balance as part of his/her portfolio
management. The balance at the time reference point is, in fact, the outstanding
amount at the time reference point, which means, in other words, that the concept of
new business is extended to the whole stock in the case of overnight deposits,
deposits redeemable at notice, revolving loans and overdrafts, and convenience and
extended credit card credit.
In addition to these conceptual considerations, there are also several practical
reasons for using the balance at the time reference point as an indicator for new
business in this set of instruments. If instead the definition of new business were to
be based on increases in the amount on existing deposit or loan accounts, this would
lead to a heavy reporting burden on reporting agents due to the large number of
inflows and outflows throughout the month. Also, the weight based on all inflows and
outflows would overestimate the amount of new business on these accounts.
Analogous to the interest rates on outstanding amounts discussed in Section 5.2, the
MFI interest rates for overnight deposits, deposits redeemable at notice, revolving
loans and overdrafts, and convenience and extended credit card credit reflect the
weighted average interest rate level applied to the stock of deposits or loans in the
relevant instrument category as at the time reference point:
• The interest rates for overnight deposits and deposits redeemable at notice
cover all amounts placed and not yet withdrawn by customers in all the periods
up to and including the reporting date.
• The interest rates on revolving loans and overdrafts, and convenience and
extended credit card credit cover all amounts withdrawn and not yet repaid by
customers in all the periods up to and including the reporting date.
• The interest rate on convenience credit is not reported separately as it is by
definition 0%. The outstanding amount of convenience credit granted is
however included in the new business volumes, and correspondingly in the
weights, of indicators 86 and 87 in Table 8.
5.3.2 Determining the interest rate on an overnight deposit
Chart 1 shows an example of an overnight deposit yielding 2% p.a. for an amount up
to EUR 2,500 and 3% p.a. for any amount exceeding EUR 2,500. The same figure
could also represent a non-transferable sight savings deposit, a revolving loan, an
59
The overnight deposit becomes a bank overdraft in the case of a debit balance; see also Section 7.5.2.
Manual on MFI interest rate statistics − Business coverage
43
overdraft or an extended credit card credit. In the latter three cases, the account
would not yield, but would incur interest.
In this Manual, MIR (OA) indicates the MFI interest rate on outstanding amounts and
MIR (NB) the MFI interest rate on new business. Weight (OA) and weight (NB) show
the weighting information that would be applied to the interest rates on outstanding
amounts and new business respectively in aggregations with values for other
comparable accounts in this or other reporting agents.
60
As a result of extending the
definition of new business for overnight deposits to the whole stock, the interest rates
and weights for new business and outstanding amounts coincide in Chart 1.
Chart 1
Overnight deposit
t0 t1 t2 t3 t4 t5 t6 t7 t8 t9
MIR (OA) = MIR (NB) - 2.00% 2.00% 2.17% 2.29% 2.00% 2.00% - 2.00% 2.38%
Weight (OA) = Weight (NB) -
1,500 2,500 3,000 3,500 1,000 1,000
-
2,000 4,000
- yielding 2% -
1,500 2,500 2,500 2,500 1,000 1,000
-
2,000 2,500
- yielding 3% -
0 0 500 1,000 0 0
-
0 1,500
Each of the periods t
0
to t
9
represent one month. In this example, the MFI interest
rates are calculated as a snapshot of end-month observations.
61
Hence, the amount
and interest rate observations in Chart 1 are end-month values. For example, at the
end of the first month t
1
an amount of EUR 1,500 is available on the overnight
deposit yielding 2% interest and, therefore, the MFI interest rate on this account at
time t
1
is 2.0%. At time t
3
, EUR 3,000 is available on this deposit, where EUR 2,500
yields 2% and EUR 500 yields 3% interest. Hence, the MFI interest rate for t
3
is
calculated as (0.02 * 2500 + 0.03 * 500) / 3000 = 2.17%. At time t9, the MFI interest
rate is calculated as (0.02 * 2500 + 0.03 * 1500) / 4000 = 2.38%. These interest rates
reflect a snapshot of the deposit at the time of data collection. They may of course
60
Aggregations are discussed in Chapter 9.
61
Further discussed in Section 6.1. An alternative calculation based on implicit rates is possible but is not
be demonstrated here.
t0 t1 t2 t3 t4 t5 t6 t7 t8 t9
Overnight deposit
0 1,500 2,500 3,000
3,500 1,000
1,000 0 2,000 4,000
Threshold
2,500 2,500 2,500 2,500 2,500 2,500 2,500 2,500 2,500 2,500
0
500
1,000
1,500
2,000
2,500
3,000
3,500
4,000
4,500
← 2% 3% →
Manual on MFI interest rate statistics − Business coverage
44
differ from the actual amount of interest earned during the month, the latter being
covered by implicit rates.
5.3.3 Combined deposit and loan facilities
Question:
How should the AAR or the NDER be calculated for financial products on accounts
that, depending on their balance, can either be a deposit or a loan, such as for
example for overnight deposits and overdrafts? Reporting agents do not know in
advance whether the account will be a deposit or turn into a loan in the coming
period.
Answer:
MFI interest rates are determined ex post, i.e. for the month prior to the reporting
date. To calculate the MFI interest rate for a combined deposit and loan facility,
periods where the account was a (positive) overnight deposit have to be
distinguished from periods where the account was an (negative) overdraft. For
analytical reasons, it would not be appropriate to compute an average interest rate
combining (low) overnight deposit rates and (high) overdraft rates. An example for a
combined deposit and loan facility is given in Chart 2. For combined deposit and loan
facilities, two compilation methods need to be distinguished:
62
• If the interest rate is compiled as a snapshot of end-month observations, only
one balance during the month is taken into account to decide whether the
account in the reference month is an overnight deposit or an overdraft. This
balance is a snapshot at a certain point in time on the last day of the month.
• If the interest rate is calculated as an implicit rate referring to the average of the
month, for each daily balance the reporting agent needs to assess whether the
account is a deposit or a loan. The reporting agent then calculates an average
of the daily credit balances and the daily debit balances to derive the average
monthly stocks for the denominator of the implicit rate. Also, for the flows in the
numerator the accrued interest payable on deposits and receivable on loans
needs to be distinguished.
62
The time reference point and hence the two compilation procedures are further discussed in Section
6.1.
Manual on MFI interest rate statistics − Business coverage
45
Chart 2
Combined deposit and loan facility
If in each of the periods t
0
to t
9
represents one month, then t
1
, t
2
, t
3
and t
6
are
recorded as overnight deposits and t
4
, t
5
, t
7
and t
9
as overdrafts. If each of the
periods t
0
to t
9
represents one day and t
9
the reporting day, then an overdraft is
reported if the method of end-period observations is chosen, because only the
situation at the time of data collection is relevant.
5.3.4 Regular savings on a deposit redeemable at notice
63
Question:
A deposit redeemable at notice is subject to regular savings of EUR 200 per month
for five years. What should be reflected in the statistics on new business and
outstanding amounts?
Answer:
For deposits redeemable at notice, the concept of new business is extended to the
whole stock. Hence, the credit balance, i.e. the amount outstanding at the time
reference point, is used as an indicator for new business. In this example, the initial
savings of EUR 200 and all subsequent savings of EUR 200 are reflected in the
amount outstanding on the account and hence in MFI interest rate statistics. An
example for the first year is given in Chart 3.
63
See also regular savings on deposits with agreed maturity in Section 5.4.5.
t0
t1 t2 t3 t4 t5 t6 t7 t8 t9
Amount
0 2,000 3,000 1,000 -1,000 -2,000 2,000 -300 0 -300
-3,000
-2,000
-1,000
0
1,000
2,000
3,000
4,000
← 2% 3% →
Manual on MFI interest rate statistics − Business coverage
46
Chart 3
Deposit redeemable at notice
y-axis: Outstanding amount (left-hand scale), Interest rate (right-hand scale)
t0 t1 t2 t3 t4 t5 t6 t7 t8 t9 t10 t11 t12
MIR (OA) =
- 2.50% 2.50% 2.70% 2.70% 2.65% 2.60% 2.55% 2.50% 2.45% 2.45% 2.45% 2.45%
MIR (NB)
Weight (OA) =
- 200 400 600 800 1,000 1,200 1,400 1,600 1,800 2,000 2,200 2,400
Weight (NB)
5.4 Interest rates on new business in instrument categories
other than overnight deposits, deposits redeemable at
notice, revolving loans and overdrafts, and convenience
and extended credit card credit
64
5.4.1 Definition of new business
MFI interest rate statistics on new business are statistics on the interest rates agreed
in new agreements. They reflect the demand and supply conditions in the deposit
and loan markets at the time of the agreement, including competition with other
types of financial institution and product. These statistics are needed to analyse the
pass-through of changes in official rates and market interest rates to lending and
deposit interest rates faced by households and non-financial corporations. They
provide information about the cost of capital and the cost spread between self-
financing and credit. MFI interest rates on new business allow the study of prices
and quantities together which helps for example to explain portfolio shifts. They also
show how quickly banks’ interest rate margins react to external developments.
64
See also paragraphs 19 to 27 of Annex I to the Regulation.
t0 t1 t2 t3 t4 t5 t6 t7 t8 t9 t10 t11 t12
Deposit
0 200 400 600 800 1,000 1,200 1,400 1,600 1,800 2,000 2,200 2,400
Interest rate
2.50% 2.50% 2.50% 2.70% 2.70% 2.65% 2.60% 2.55% 2.50% 2.45% 2.45% 2.45% 2.45%
2.30%
2.35%
2.40%
2.45%
2.50%
2.55%
2.60%
2.65%
2.70%
2.75%
0
500
1,000
1,500
2,000
2,500
3,000
Manual on MFI interest rate statistics − Business coverage
47
New business in these instrument categories is defined as any new agreement
between the customer and the MFI. New agreements are:
• all financial contracts, terms and conditions that specify for the first time the
interest rate of the deposit or loan, and
• all renegotiations of existing deposit and loan contracts.
65
For this definition of new business, the expression terms and conditions needs to be
specified since it appears many times in the Manual as a key element. The terms
and conditions of a loan or deposit contract consist of the agreed interest rate or the
spread over a relevant reference rate, the size of the loan or the deposit, the access
conditions, and other terms and conditions in the form of non-interest rate charges
(e.g. fees), collateral or guarantees which the customer of an MFI needs to provide
(including compensation balances), loan or deposit covenants and the agreed
maturity.
An interest rate on new business reflects the weighted average
66
interest rate level
that has been agreed for all new deposits or loans in the relevant instrument
category during the reference month. Interest rates on new business cover all new
agreements made during the whole month irrespective of the point in time a deposit
is placed or a loan is withdrawn
67
. MFI interest rate statistics on new business are
distinct from MFI interest rate statistics on outstanding amounts in that, as explained
in Section 5.2, the latter reflect the interest rates actually applied to the stock of
deposits and loans. In an extreme case, an interest rate laid down in a new
agreement may never actually be applied to any deposit or loan. For example, a
customer and a credit institution might agree on an interest rate for a certain amount
of money, but the customer might in the end choose not to place any deposit with
this institution or decide not to withdraw any of the money granted as a loan. As a
consequence, both the agreed interest rate and the amount would be reflected in the
MFI interest rate statistics on new business at the time of agreement, but never
appear in the MFI interest rate statistics on outstanding amounts.
5.4.2 Definition of renegotiation
68
As defined in paragraph 21 of Annex I to the Regulation, renegotiation refers to the
active involvement of the household or non-financial corporation in adjusting the
terms and conditions of an existing loan or deposit contract, including the interest
rate. Prolongations of existing contracts that occur automatically, i.e. without any
active involvement of the customer, and do not involve any renegotiation of the terms
and conditions of the contract, including the interest rate, are not new business.
65
The definition of renegotiations is included in Section 5.4.2.
66
Aggregations are discussed in Chapter 9.
67
Further discussed in Section 6.2.
68
Renegotiated loans are further discussed in Section 5.5.
Manual on MFI interest rate statistics − Business coverage
48
The Regulation requires the compilation of the amounts of renegotiated loans
separately from the new business amounts for the instrument categories included in
Table 6 of the Appendix. The purpose is to have a measure of “pure new loans” in
the sense of gross “fresh money” arriving on the credit market, distinguishing these
from renegotiated loans where, by definition, no new money is arriving on the credit
market.
For this definition of renegotiation, it should be clarified that contract changes that
have an impact on the terms and conditions (e.g. the maturity or amount granted)
are considered renegotiations, and not only changes in interest rates. However,
changes in formal elements of the contract, such as the address of the credit
institution or that of the customer, must not be considered as renegotiations.
An additional aspect to be clarified is the expression “active involvement”; that is to
say, how active does the involvement of the customer have to be to consider a
change in the prevailing conditions of a loan or a deposit a renegotiation? Let us
consider the case of a deposit contract which stipulates that, at maturity, the deposit
will be renewed automatically with the same terms and conditions as the previous
ones unless the customer or the bank decides not to renew the deposit or decides to
change the terms. In one of these renewal periods, the bank informs the customer
about its willingness to renew the deposit but at a different interest rate and gives the
customer a period in which to cancel the deposit in case of disagreement. The
customer does not respond and the bank applies the new conditions. Then, the
question is: could the passive consent of the customer be considered as active
involvement so that this change in the conditions of the deposit is recorded as a
renegotiation? The question is very relevant, because as the conditions of the
deposit have changed and these new conditions were not agreed beforehand when
the contract was signed, a new business operation has to be recorded, but there is
no new money coming onto the deposit market. This operation seems to be a
renegotiation, but the active involvement of the customer could be called into
question. The answer to this case is that tacit agreement by the customer by not
opposing the new proposal of the bank is, in fact, an active involvement because the
customer could have opted for other alternatives; therefore, new business and a
renegotiation should be recorded in MFI interest rate statistics.
It needs to be clarified, however, that tacit agreements should be reported as
renegotiations only if contract changes have an impact on the contract’s terms and
conditions (e.g. the interest rate or the maturity). Therefore, changes in formal
elements of the contract, such as the address of the credit institution or that of the
customer, must not be considered as renegotiations.
5.4.3 New business on deposits with agreed maturity
For most deposits with agreed maturity, i.e. classic time deposits where a fixed sum
is placed for a predefined period of time, new business only arises when a new
account is opened for the first time, at which point the deposit amount and the
interest rate are agreed. Normally, for deposits with agreed maturity no further new
Manual on MFI interest rate statistics − Business coverage
49
business occurs until maturity. An exception to this rule is discussed in Section 5.4.5.
The treatment of maturing deposits is subject to Section 5.4.4.
Chart 4 illustrates the calculation of MFI interest rates on outstanding amounts and
new business in an example involving three separate deposits with agreed maturity:
1. at time t
1
EUR 2,000 is placed at 4% p.a. with an agreed maturity of two years;
2. at time t
2
EUR 2,000 is deposited at 5% p.a. also with an agreed maturity of two
years; and
3. at time t
4
EUR 3,000 is paid in at 6% p.a. with an agreed maturity of three
years.
In this example, at time t1, an amount of EUR 2,000 is placed as a new deposit. At
that time, the MFI interest rate on new business MIR (NB) and on outstanding
amounts MIR (OA) is 4.0%, which is indicated in Chart 4. At time t
2
, EUR 2,000 is still
in the first deposit yielding 4% and EUR 2,000 is newly paid into the second deposit
yielding 5%. Hence, at time t
2
the MFI interest rate on outstanding amounts
comprising both deposits is calculated as (0.04 * 2000 + 0.05 * 2000) / 4000 = 0.045
= 4.5%, while the MFI interest rate on new business at time t
2
is 5% referring solely
to the new second deposit. At time t
3
, the first deposit of EUR 2,000 reached
maturity; therefore, there is an outstanding amount of EUR 2,000 referring to the
second deposit yielding 5%. At that time, there is no new business. At time t
4
, the
second deposit of EUR 2,000 reached maturity and EUR 3,000 is newly paid into the
third deposit yielding 6%. Hence, at that time, the MIR (NB) and the MIR (OA) is 6%.
At t
5
and t
6
, there is no new business to be reported and EUR 3,000 is to be reported
for outstanding amounts referring to the third deposit yielding 6%, as indicated in
Chart 4.
Manual on MFI interest rate statistics − Business coverage
50
Chart 4
Deposit with agreed maturity
(y-axis: Outstanding amount)
t0 t1 t2 t3 t4 t5 t6
MIR (OA) - 4.00% 4.50% 5.00% 6.00% 6.00% 6.00%
Weight (OA)
- 2,000 4,000 2,000 3,000 3,000 3,000
MIR (NB) - 4.00% 5.00% - 6.00% - -
Weight (NB)
- 2,000 2,000 - 3,000 - -
5.4.4 Matured deposit with agreed maturity
Question:
A customer’s deposit with agreed maturity matures. Following the standard
procedure, the reporting agent informs the customer and credits the corresponding
amount to another account of the customer, or alternatively, the customer could have
two other options, both stipulated in the contract.
Option 1: cancel the deposit and inform the credit institution that he or she wants to
make a new deposit with similar or different terms and conditions than the matured
deposit, or
Option 2: do nothing, in which case the credit institution, as stipulated in the contract,
will automatically renew the deposit on identical terms.
Would options 1 and 2 constitute new business?
Answer:
The task is to distinguish between new business and an (automatic) prolongation of
an existing contract. The key issue is the active involvement of the customer and the
alternatives laid down in the existing contract for when the deposit reaches maturity:
an (automatic) prolongation without the active involvement of the customer is not
new business, whereas a prolongation with active involvement of the customer is
new business.
t0 t1 t2 t3 t4 t5 t6
Two-year maturity, 4% interest
2,000 2,000
Two-year maturity, 5% interest
2,000 2,000
Three-year maturity, 6% interest
3,000 3,000 3,000
0
500
1,000
1,500
2,000
2,500
3,000
3,500
Manual on MFI interest rate statistics − Business coverage
51
Option 1 is new business because the customer is involved in renegotiations with the
credit institution. Whether the new deposit is on terms and conditions similar or even
identical to the old deposit or on new terms and conditions is irrelevant for
determining whether there is new business.
Option 2 is not new business because the automatic renewal of the deposit was laid
down in the existing contract in the event that the customer did not take any action.
In general, the contract for a deposit with agreed maturity defines what happens with
the money after maturity, if it is not withdrawn. For example, the contract might
determine that the deposit is originally placed for one year at EURIBOR plus 50
basis points. If the funds are not withdrawn at maturity, then they are reinvested at
the same terms, i.e. at EURIBOR plus 50 basis points. In this case, if the customer
says nothing, it is not new business but an (automatic) prolongation of the contract. If
the customer decides to change the external index or the spread, the customer is
actively involved and the reinvestment constitutes a renegotiation and, as a
consequence, it is new business.
It is also possible that the customer on placing a deposit with one-month maturity
instructs the credit institution to roll over the account each month provided the terms
and conditions remain unchanged from one month to the next. The customer
stipulates that he or she wishes to be notified if there is a change in the terms and
conditions attached to the account. The automatic rolling over is not new business.
However, if the terms and conditions change, and the customer is contacted and
engages in new negotiations, this is recorded as new business, irrespective of
whether the negotiations result in the old or new terms and conditions. If the credit
institution informs the customer about the new terms and conditions of the deposit,
giving him or her a period to object, and the customer does not respond, this tacit
agreement is considered new business and a renegotiation.
When the terms and conditions of a maturing deposit with agreed maturity are
renegotiated and, as a consequence, the deposit is classified as new business, the
maturity of that (new) deposit is counted as commencing at the point of the new
business classification.
5.4.5 Regular savings on a deposit with agreed maturity
69
Question:
A company savings plan is linked to a deposit with agreed maturity. In the contract it
is agreed that the company makes regular monthly deposits of EUR 2,000 for a
period of two years. The first payment is made a few weeks after the contract date.
1. What is to be considered as new business, the first deposit of EUR 2,000 or all
payments during the lifetime of the contract?
69
See also regular savings on deposits redeemable at notice in Section 5.3.4.
Manual on MFI interest rate statistics − Business coverage
52
2. What is to be considered as the starting date, the contract date (in which case
there has not yet been a payment and there is no new business amount) or the
first payment date?
Answer:
Ceteris paribus an MFI will offer a different interest rate on a deposit that is:
• progressively increasing over time from EUR 2,000 in t
0
to EUR 48,000 in t
24
,
• fixed at EUR 2,000 for two years, and
• fixed at EUR 48,000 for two years.
The company’s regular monthly savings in this example lead to a progressive
increase in the outstanding amount on the deposit with agreed maturity. This
increase is reflected month after month in the MFI interest rate statistics on
outstanding amounts. By contrast, the MFI interest rate statistics on new business
capture the deposit only once at the time of agreement. Hence, the new business
statistics can either mirror the initially placed EUR 2,000 or the maximum amount of
EUR 48,000, but cannot reflect the progressive increase of the amount on the
deposit. The treatment of regular savings on deposits with agreed maturity in new
business statistics cannot be defined in general, but depends on the agreements laid
down in the individual contract:
Case 1: The customer and the MFI agree that EUR 2,000 has to be placed each
month on the account until maturity. The customer’s obligation can be enforced by
the MFI in such a way that it is sure ex ante that at time t
24
EUR 48,000 will be
accumulated on the account. In this case, the full amount of EUR 48,000 is reflected
as new business in MFI interest rate statistics at the time of agreement on the
contract. Most likely the level of the interest is linked to the commitment on the part
of the customer. For variable interest rates, the value of the rate is determined at the
time of the agreement.
70
Case 2: The customer and the MFI agree on a flexible savings plan, which states
that EUR 2,000 should be placed each month on the deposit until maturity. The
payments can be lower or higher with (or without) a targeted maximum saving of
EUR 48,000 at time t
24
. In this case, the amount accumulated on the account at
maturity is certain only ex post. Therefore, the MFI reports as new business the
amount of EUR 2,000 when it is initially placed. For variable interest rates, the value
of the rate is determined at the time the deposit is placed.
71
In the absence of any knowledge to the contrary, Case 2 should be assumed, i.e.
that the amount accumulated on the account at maturity is certain only ex post.
Hence the default option is that the EUR 2,000 is reported as new business when it
70
This treatment is similar to that of a loan in tranches discussed in Section 5.4.9, where it is also the
agreement that determines the amount and the time of recording as new business. However, the
difference between the savings plan and the loan in tranches is that for the savings plan the interest
rate is linked to the increase in the amount and for the loan to the full amount stated in the contract.
71
This treatment is similar to that of savings plans for housing loans described in Section 8.4. See also
Section 8.1.
Manual on MFI interest rate statistics − Business coverage
53
is initially placed. In the case of variable interest rates, the value of the rate is
determined at the time the deposit is placed.
5.4.6 New lending with a fixed interest rate and with initial rate fixation
Chart 5 shows the example of a loan that is granted for ten years. At time t
0
it is
agreed between the customer and the MFI that the interest is fixed at 10% p.a. for
the first four years (up to t
3
) and that after this initial period of fixation a new interest
rate level will be agreed for the remaining maturity of the loan.
72
This new interest
rate might be fixed for another period or variable, but is in any case unknown at time
t
0
. As an example, in Chart 5 the result of the renegotiations at time t
4
is a fixed rate
of 8% p.a. for the remaining six years of the loan. The MFI interest rate statistics on
new business capture at time t
0
the interest rate of 10% agreed for the first four years
and at time t
4
the interest rate which is the result of the renegotiations.
Chart 5
Te n -year loan renegotiated after four years
(y-axis: Outstanding amount (left-hand scale); interest rate (right-hand scale))
t0 t1 t2 t3 t4 t5 t6 t7 t8 t9
MIR (OA) 10% 10% 10% 10% 8% 8% 8% 8% 8% 8%
Weight (OA)
10,000 10,000 10,000 10,000 10,000 10,000 10,000 10,000 10,000 10,000
MIR (NB) 10% - - - 8% - - - - -
Weight (NB)
10,000 - - - 10,000 - - - - -
Chart 6 provides another example of a ten-year loan. In this example, at time t
0
the
customer and the MFI agree that the interest is fixed at 9% p.a. for the first 12
months, and that after this initial period of fixation the interest rate automatically
72
The original maturity of this loan is ten years with an initial period of fixation of four years. See also
Section 7.7.4.
t0 t1 t2 t3 t4 t5 t6 t7 t8 t9
Loan
10,000 10,000 10,000 10,000 10,000 10,000 10,000 10,000 10,000 10,000
Interest rate
10.00% 10.00%
10.00%
10.00% 8.00% 8.00% 8.00% 8.00%
8.00% 8.00%
0%
2%
4%
6%
8%
10%
12%
0
2,000
4,000
6,000
8,000
10,000
12,000
Manual on MFI interest rate statistics − Business coverage
54
adjusts to EURIBOR plus x basis points.
73
This rate is then applied for the next 12
months, after which it will again automatically adjust to EURIBOR plus x basis points.
Only the interest rate of 9% for the first year is considered as new business at time
t
0
. Neither the switch to variable rates nor the associated automatic adjustments are
reflected in the statistics on new business. They are not new agreements but part of
the terms and conditions of the loan laid down at time t
0
. In contrast to the previous
example, in this example the customer and the MFI agreed on the variability
mechanism for the entire maturity of the loan at t
0
, i.e. the day of the signature of the
loan contract. A change from fixed to variable interest rates or vice versa during the
course of the contract, which has been agreed at the start of the contract (time t
0
), is
not a new agreement but part of the terms and conditions of the loan laid down at
time t
0
. These changes in the interest rate over time are only captured in the MFI
interest rates on outstanding amounts.
Chart 6
Te n -year loan with a fixed interest rate for the first year
(y-axis: Outstanding amount left-hand scale)
T0 t1 t2 t3 t4 t5 t6 t7 t8 t9
MIR (OA) 9.00% 8.02% 8.31% 6.59% 5.88% 4.76% 4.36% 5.57% 7.16% 7.75%
Weight (OA)
10,000 10,000 10,000 10,000 10,000 10,000 10,000 10,000 10,000 10,000
MIR (NB) 9.00% - - - - - - - - -
Weight (NB)
10,000 - - - - - - - - -
5.4.7 Top-up loans
Question:
A customer has an outstanding consumer credit of EUR 10,000 at 9% and asks the
73
The original maturity of the loan is ten years with an initial period of fixation of one year. This example is
different to the case where the initial period of fixation is very short compared with the whole maturity of
the loan and the interest rate offered during this period is significantly below market conditions. See
also Section 7.7.4.
t0 t1 t2 t3 t4 t5 t6 t7 t8 t9
Loan
10,000 10,000 10,000 10,000 10,000 10,000 10,000 10,000 10,000 10,000
Interest rate
9.00% 8.02% 8.31% 6.59% 5.88% 4.76% 4.36% 5.57% 7.16% 7.75%
0%
1%
2%
3%
4%
5%
6%
7%
8%
9%
10%
0
2,000
4,000
6,000
8,000
10,000
12,000
Manual on MFI interest rate statistics − Business coverage
55
MFI to lend a further EUR 5,000. How should this top-up loan be treated in MFI
interest rate statistics?
Answer:
If an agreement is reached for the incremental amount, then only this agreement is
new business, i.e. the new loan of EUR 5,000. The additional loan could be at the
same interest rate of 9% as the outstanding loan of EUR 10,000, but could also be at
a higher or lower rate.
If the new operation includes also the previous outstanding amount (EUR 10,000) in
such a way that the new terms and conditions affect the whole loan, i.e. the EUR
15,000, then the amount of EUR 15,000 is new business. Also, a renegotiated
amount of EUR 10,000 has to be computed. The interest rate on the loan recorded
under new business is then the one agreed in the new negotiations between the
customer and the MFI, which could be 9% as for the old loan, or higher or lower.
5.4.8 Conversion of an overdraft into another type of loan
Question:
A customer incurred an overdraft of EUR 10,000 on which the interest rate is 15%
p.a. In agreement with the MFI, the customer transforms this overdraft into a
consumer credit on which the interest rate is 10% p.a. Does the conversion
represent new business?
Answer:
The conversion of an overdraft into a consumer credit constitutes new business. It
requires a new agreement between the MFI and the customer. Both parties are
actively involved in new negotiations. This operation has to be included also as a
renegotiation.
5.4.9 Loan taken out in tranches
74
A household or non-financial corporation is normally expected to take out a loan –
other than a revolving loan or an overdraft – in full at the start of the contract (time
t
0
). In some cases, however, it may be agreed that the customer takes out the loan in
tranches at times t
1
, t
2
, t
3
, etc. instead of withdrawing the full amount at time t
0
.
Chart 7 shows the first year of a ten-year loan in tranches. In this example, a loan of
EUR 10,000 is granted at time t
0
for ten years fixed at 8% p.a. The customer takes
out the first tranche of EUR 1,000 at time t
1
and then further tranches of EUR 1,000
in each of the following nine months.
74
Loans in tranches that fall into the instrument categories consumer credit, loans to households for
house purchases and loans to non-financial corporations are covered. Revolving loans and overdrafts,
which follow the treatment explained in Section 5.3.1, are not covered. Umbrella contracts, which are
further discussed in Section 8.3, are also excluded.
Manual on MFI interest rate statistics − Business coverage
56
A household or non-financial corporation is normally expected to take out a loan –
other than a revolving loan or an overdraft – in full at the start of the contract (time
t
0
). In some cases, however, it may be agreed that the customer takes out the loan in
tranches at times t
1
, t
2
, t
3
, etc. instead of withdrawing the full amount at time t
0
.
Chart 7 shows the first year of a ten-year loan in tranches. In this example, a loan of
EUR 10,000 is granted at time t
0
for ten years fixed at 8% p.a. The customer takes
out the first tranche of EUR 1,000 at time t
1
and then further tranches of EUR 1,000
in each of the following nine months.
Chart 7
Loan in tranches
(y-axis: Outstanding amount (left-hand scale); interest rate (right-hand scale))
t0 t1 t2 t3 t4 t5 t6 t7 t8 t9 t10 t11 t12
MIR (OA) - 8% 8% 8% 8% 8% 8% 8% 8% 8% 8% 8% 8%
Weight (OA)
- 1,000 2,000 3,000 4,000 5,000 6,000 7,000 8,000 9,000 10,000 10,000 10,000
MIR (NB) 8% - - - - - - - - - - - -
Weight (NB)
10,000 - - - - - - - - - - - -
For MFI interest rate statistics on new business, the fact that the loan is taken out in
tranches is irrelevant. These statistics capture the agreement between the customer
and the MFI at time t
0
, i.e. the full amount of the loan granted at time t
0
and the
interest rate agreed in relation to this full amount. New business is therefore the
amount of EUR 10,000 at an interest rate of 8% at time t
0
. For fixed and variable
rates, the risk premium included in the interest rate depends on the loan amount and
economic conditions. These facts can only be captured consistently by recording the
interest rate and the full amount of the loan at time t
0
in MFI interest rates on new
business. As explained in Section 5.4.1, MFI interest rate statistics on new business
cover agreements, not the actually withdrawn credit.
By contrast, MFI interest rate statistics on outstanding amounts cover the actual
withdrawn credit at any point in time. They reflect the interest rates applied and the
t0 t1 t2 t3 t4 t5 t6 t7 t8 t9 t10 t11 t12
Loan
0
1,000 2,000 3,000 4,000 5,000 6,000 7,000 8,000 9,000 10,000 10,000 10,000
Interest rate
8.00% 8.00% 8.00%
8.00%
8.00% 8.00% 8.00% 8.00% 8.00%
8.00%
8.00% 8.00% 8.00%
0%
1%
2%
3%
4%
5%
6%
7%
8%
9%
0
2,000
4,000
6,000
8,000
10,000
12,000
Manual on MFI interest rate statistics − Business coverage
57
amount of the loan actually outstanding
75
at the time of data collection, i.e. the
interest rate is 8% for an amount of EUR 1,000 at t
1
, 8% for an amount of EUR 2,000
at t
2
, and finally 8% for an amount of EUR 10,000 at t
10
until maturity. No interest rate
on outstanding amounts is recorded at t
0
as no money has yet been withdrawn.
The reporting agents granting the loan in tranches, for example for financing the
building of a house, may charge interest on the amount granted but not yet
withdrawn.
76
In this example of an agreed lending rate of 8% for the loan in tranches,
the credit institution could ask for an additional 3% on the amount it has agreed to
lend but the borrower has not yet withdrawn. The 3% is not covered by MFI interest
rate statistics, neither as new business nor as outstanding amounts. The 3% is not
considered part of the lending interest rate: it is a related charge that might be
covered in the APRC according to national conventions, but not in the AAR or the
NDER.
5.4.10 The definition of new business for variable interest rates
A rise or fall of a variable interest rate in the sense of an automatic adjustment of the
interest rate performed by MFIs is not a new agreement and therefore not new
business. Such changes or movements in variable interest rates over time are not
captured by MFI interest rates on new business but by MFI interest rates on
outstanding amounts.
A change from a fixed to a variable interest rate or vice versa during the course of
the contract is not a new agreement, if the possibility of such a change is an agreed
part of the terms and conditions of the deposit or loan. When the change occurs this
is not new business. However, a change from a fixed to a variable interest rate or
vice versa is new business, if the possibility of the change was not laid down in the
initial contract. The change is therefore the result of a renegotiation between the
customer and the MFI and constitutes new business and, as it refers to a
renegotiation of an existing agreement, it also constitutes a renegotiation and should
be reported separately in the case of loans.
77
MFI interest rate statistics on outstanding amounts and new business lead to
different results not only for deposits and loans with initially fixed or fully fixed interest
rates, but also for deposits and loans to which fully variable interest rates apply.
Since MFI interest rate statistics reflect the interest rates that are actually agreed
between the customer and the MFI, interest rates on outstanding amounts and new
business only coincide for that part of the interest rate that reflects the external
index, i.e. in this example EURIBOR. The spread over this external index might differ
across customers but also for the same customer over time.
75
Derived from MFI balance sheet statistics.
76
In Belgium this is referred to as commission de non utilisation, in Germany as Bereitstellungszinsen.
77
See also Section 5.4.2.
Manual on MFI interest rate statistics − Business coverage
58
For example, in Chart 8 at time t
0
a four-year loan at
EURIBOR plus x basis points is agreed. The
creditworthiness of the customer, his or her scope for
negotiating the interest rate and market conditions vary
over time. Hence, at time t
2
the same customer might
only reach agreement on a four-year loan at EURIBOR
plus y basis points, but achieve EURIBOR plus z basis
points at time t
5
, with z < x < y.
Contracts might be agreed between customers and
reporting agents that contain very flexible terms and
conditions including the level that the variable interest
rate may take. In some cases, certain conventions are
needed in order to determine the level of the variable
interest rate that should be recorded under new
business in MFI interest rate statistics. As not all
possibilities can be covered in this Manual, the
examples in Sections 5.4.11 to 5.4.14 are intended to
provide general guidance.
5.4.11 Choice of money market index
Question:
A loan of EUR 10 million is agreed, which is taken out in tranches
78
. The contract
specifies that at the time the withdrawals are being made the customer may choose
which of the following interest rates applies: one-month, three-month, six-month or
12-month LIBOR plus 50 basis points. According to paragraph 27 of Annex I to the
Regulation, this loan in tranches is new business at the time the contract is agreed
for the total amount of EUR 10 million. At that time, however, it is unknown which
interest rate the customer will choose for the first and each of the following
withdrawals. Which interest rate should be reported as new business?
Answer:
For a loan contract that provides the customer with a choice of money market rates
as an external index, a convention is needed for determining the value of the new
business rate. By convention, MFI interest rate statistics on new business record as
the value of the variable interest rate either one-month, three-month, six-month or
12-month LIBOR plus 50 basis points, whatever gives the lowest value at the time
the contract is agreed. The rationale for this convention is that ceteris paribus the
lowest rate is what the customer would most likely have chosen, if he or she
withdrew the whole or a part of the amount on the day of agreement.
For determining the annualised value of the variable rate, it is assumed that the
interest rate stays the same for the life of the contract
79
. For example, if at the time of
78
See also Section 5.4.9.
79
This is in line with Section 4.2.10.
Chart 8
Variable interest rates
0%
2%
4%
6%
8%
10 %
12 %
14 %
t0 t1 t2 t3 t4 t5 t6 t7 t8 t9
EURIBOR + x
EURIBOR + y
EURIBOR + z
EURIBOR
Manual on MFI interest rate statistics − Business coverage
59
agreement on the contract one-month LIBOR plus 50 basis points is 4% p.a. and as
such the lowest interest rate, then the AAR or the NDER is determined based on the
4% and taking into account the frequency of interest rate payments.
The interest rates on outstanding amounts reflect the actual interest rate the
customer chooses at each withdrawal, the development of the variable rate over time
and also the actual amount of the loan which has been withdrawn.
5.4.12 Money market index with a floor and a ceiling
Question:
A customer agrees to place a deposit with an MFI that has an agreed maturity of one
year receiving interest of 12-month LIBOR plus 40 basis points with a floor of 2%
and a ceiling of 6%. The interest rate floor and ceiling are not derivative contracts,
which may or may not be exercised by the customer or the MFI as described in
paragraph 6 of Part 15 of Annex II to the Guideline. Instead, the floor and ceiling
constitute a corridor for the possible interest rates. What interest rate should be
reported as new business?
Answer:
The treatment is in principle the same as for any other variable interest rate but
taking into account the floor and ceiling for the possible value. Hence, new business
is the value of 12-month LIBOR plus 40 basis points at the time of agreement on the
contract taking into account that the interest rate cannot fall below 2% and cannot
exceed 6%. For example, if at the time of agreement 12-month LIBOR is 1.5%, then
2% is recorded under new business, because 1.5 plus 0.4 is below the agreed floor.
If at the time of agreement 12-month LIBOR is 4.5%, then 4.5 plus 0.4 = 4.9% is the
new business rate. If at the time of agreement 12-month LIBOR is 5.9%, then 6% is
captured, because 5.9 plus 0.4 is above the agreed ceiling.
The AAR or the NDER on outstanding amounts covers the interest rate applied by
the reporting agent at the time of the calculation of MFI interest rates, i.e. it is based
on the value of 12-month LIBOR at the time of data collection. The floor and ceiling
are taken into account in the same way as for new business.
5.4.13 Change in the value of a currency as an external index
Question:
A customer agrees with an MFI on a deposit with agreed maturity of one year where
the interest rate is linked to the percentage change in the value of a currency with a
floor of 2% and a ceiling of 6%, i.e. the interest rate is always positive. The
percentage change in the value of the currency is not known at the time of
agreement t
0
, but only at the time of maturity. What interest rate should be reported
as new business?
Manual on MFI interest rate statistics − Business coverage
60
Answer:
In contrast to the example in Section 5.4.12, where a value can be given for 12-
month LIBOR at the time of the agreement on the contract, the percentage change in
the price of a currency is only known ex post at the time of maturity. Predictions for
the value of the currency at maturity could be made to determine a value for the
(potential) percentage change of the currency at the time of agreement. However,
this is difficult and not consistent with the method for determining the value of a
variable rate that is linked to a money market or bond market index, where no
predicted future value, but rather the current value of the index at the time of
agreement, is used. Hence, the only interest rate that can be included in the
statistics on new business is the agreed floor, i.e. 2%, as this is known with certainty
by the customer and the reporting agent at time t
0
80
.
The AAR or the NDER on outstanding amounts covers the interest rate applied by
the reporting agent at the time of the calculation of MFI interest rates. Hence, it is
based on the percentage change in the price of the currency at the time of data
collection. The agreed floor and ceiling are taken into account in the same way as in
Section 5.4.12.
5.4.14 Timing differences
Question:
A customer and an MFI agree on a housing loan with a variable interest rate. The
interest rate may be explicitly referenced to an external index, e.g. EURIBOR, or it
may be subject to change at the discretion of the lender. As the contractual rate is
variable, the value of this rate may vary (within the terms of the contract) at any point
from the receipt of the letter of offer through to acceptance or withdrawal of the loan
and then during the life of the contract. For contracts with variable rates that have
been agreed, but where funds have not yet been withdrawn, some institutions record
in their IT system the prevailing rate for that product type rather than the interest rate
for the specific contract. It is the view of these institutions that reporting the interest
rate for the product type is both more accurate and more efficient as this obviates the
need for an additional significant data capture exercise. Furthermore, reporting the
interest rate for the product type will allow institutions to provide data on contracts
that have been signed by the borrower, but for which the institution has not yet
received the completed forms, which are in transit via post, etc.
Answer:
The prevailing rate for the product type and the rate quoted in the individual contract
might differ because the customers have different creditworthiness. MFI interest rate
statistics on new business cover the actual rate in the individually agreed contract.
80
If no floor is agreed and no minimum return is guaranteed, then a convention is needed for determining
the level of the interest rate on new business. In this case, by convention the AAR or the NDER on new
business captures an interest rate of 0% for the deposit whenever the contract guarantees, at least, the
principal of the deposit (i.e. no incurrence of capital losses). If, however, the principal of the deposit is
not guaranteed, this operation is not recorded in MFI interest rate statistics. See also Sections 8.9 and
8.10.
Manual on MFI interest rate statistics − Business coverage
61
The value of the variable rate is determined based on the value of the external index
at the time of agreement on the contract. If the customer signs first and then the MFI,
agreement is reached when the MFI signs. If the MFI signs first and then the
customer, which is unusual, then the contract is accepted when the customer signs.
There might be a transmission time for the contract back to the MFI and the new
business might therefore be recorded with a delay of some days. As MFI interest rate
statistics on new business capture the average interest rate over the month, this
delay will only be apparent at the end of the month, in the sense that new business
on 30 March might only be recorded on 1 April because of the transmission delay.
This is accepted for MFI interest rate statistics.
5.4.15 “Cooling-off” period
Question:
Laws for consumer protection might require a “cooling-off” period, thus necessitating
for example a period of at least ten days between the conclusion of the agreement
and the withdrawal of the funds. In practice, greater delays can occur. What is the
new business?
Answer:
The “cooling-off” period and hence the possibility that a customer steps back from
the contract has no influence on MFI interest rate statistics on new business. MFI
interest rates on new business reflect the conditions, i.e. the interest rate and the
amount of the loan, as laid down in the contract at the time of agreement. The actual
day of the withdrawal of the funds is irrelevant for interest rates on new business and
is shown only in the statistics on outstanding amounts
81
.
5.4.16 Loan offer and preliminary offer
Question:
Before looking for a new house, a customer may have already acquired a loan offer
from an MFI. In the loan offer, the MFI promises to grant a loan to the customer at
specified, agreed terms in the event that the customer finds a house. The advantage
for the customer is that he or she can make a bid for a house without having to
consult the credit institution first. The final loan contract is signed once the customer
has found the house. In some Member States, lenders must issue preliminary offers
before granting loans. The preliminary offer specifies the terms and conditions of the
loan, which cannot be changed by the lender if the customer finally agrees them.
The offer stands for a given period of time, the length of which is established by
national law. During this time, the customer will generally look for competing offers
from other lenders. Do loan offers constitute new business?
81
This treatment is similar to that of the loan in tranches, where it is also the agreement that determines
the amount and the time of recording as new business. The withdrawal of the first and all following
tranches is reflected only in the statistics on outstanding amounts. See also Section 5.4.9.
Manual on MFI interest rate statistics − Business coverage
62
Answer:
A loan offer is not new business. People may attempt to buy houses that are worth
less than the MFI is willing to accept. In this case, an MFI may either withdraw the
offer or lower the amount it is willing to lend. New business therefore only arises
when on the basis of the loan offer the final loan contract has formally been signed.
Only when the customer signs the legally binding contract does it constitute new
business.
5.4.17 Moratorium on a loan
Question:
If a customer temporarily stops the repayments on a loan and starts again sometime
later, does the restart constitute new business? Does the treatment depend on
whether the bank had started to classify this loan as a bad debt? Does it depend on
whether the customer signs a new agreement with the bank?
Answer:
New business occurs only if the customer signs a new agreement. The restart of
loan repayments after a moratorium is per se not new business. If the loan is
considered a bad loan during the moratorium, it is taken out of the statistics on
interest rates on outstanding amounts, but is reincluded when the customer restarts
paying the interest
82
.
5.4.18 Documentary credit
A documentary credit, primarily used in international trade, is a commitment made by
a bank (the issuing bank) upon the request of an importer that ensures payment to
an exporter once the duties of the latter have been fulfilled. The document containing
the commitment by the bank serves as a guarantee to the exporter that it will be paid
regardless of whether the importer ultimately manages to pay. The bank of the
exporter (the advisory bank) may additionally confirm the documentary credit to
improve the guarantee.
Question:
Should the documentary credit be recorded in MFI interest rate statistics?
Answer:
A documentary credit is not a credit per se but a guarantee, and therefore, like other
types of guarantee, it should not be recorded in MFI interest rate statistics. New
business and outstanding amounts should be recorded only when the guarantee is
triggered and the corresponding bank (the issuing bank and/or the advisory bank,
depending on the terms and conditions agreed for the guarantee) pays the exporter.
At this point, the bank will record a loan granted to the importer. The terms and
conditions of this new loan, including the interest rate, will have been agreed at the
82
See also Section 5.1.
Manual on MFI interest rate statistics − Business coverage
63
time the documentary credit was signed, but at that moment no loan exists, only a
guarantee.
5.5 Renegotiated loans: further issues
As already mentioned in Section 5.4.2, the purpose of collecting the amount of
renegotiated loans is to distinguish “pure new loans” in the sense of gross “fresh
money” arriving on the credit market from renegotiated loans where no new money is
arriving on the credit market. Both types of loans are new business, but a clear
distinction needs to be drawn between the two within the MFI interest rate statistics.
This distinction is also very important for the users, since the data on pure new loans
– calculated as the difference between the total new business volumes and the
amount of renegotiations – offer a measure of the gross flow of new credit which
serves to assess the MFIs’ credit market stance.
As indicated in Section 5.4.2, the Regulation defines renegotiated loans in paragraph
21 of Annex I. Additionally, paragraphs 22 and 37 of the Annex complement the
definition by establishing the following:
“For the separate reporting of new business volumes of renegotiated loans to
households and non-financial corporations in MFI interest rate statistics,
renegotiated loans comprise all new business loans, other than revolving loans and
overdrafts and credit card debt, which have been granted but not yet repaid at the
time they are renegotiated.” (paragraph 22)
“All renegotiations of existing deposits and loan contracts should be taken into
account, even if the same contract is renegotiated more than once during the
reference month.” (paragraph 37)
However, the Guideline states that: “In respect of loans transferred from another
institution, renegotiation refers to new business loans that were granted by the
institution selling or handing over the loan.” (Annex II, Part 13, paragraph 4)
Notwithstanding all these provisions in the Regulation and Guideline, some issues
need to be further clarified, notably renegotiations that take place within a month,
renegotiations that include an increase or decrease in the loan amount,
renegotiations of loans that are not yet fully drawn (loans taken out in tranches), and
loan transfers, sales of loans and reorganisation of loans.
5.5.1 Renegotiations within a month
Question:
A loan of EUR 1,000 was granted on 3 February 2014. This loan was renegotiated
on 10 February 2014 and on 17 February 2014. What should be reported in
February under new business and under renegotiations?
Manual on MFI interest rate statistics − Business coverage
64
Answer:
In February this loan was granted for the first time and then renegotiated twice
during the same month. As renegotiations are also new business, EUR 3,000 should
be reported in February as new business and EUR 2,000 as renegotiations.
Therefore, the new money arriving on the credit market in February is EUR 1,000,
which is the difference between the gross new business (EUR 3,000) and the
amount of renegotiated loans (EUR 2,000).
5.5.2 Renegotiations including an increase in the loan amount and
renegotiations involving a partial redemption of the loan
If the renegotiation of a contract involves an increase in the outstanding loan
amount, both the agreed amount from the original contract not yet repaid and the
renegotiated increase in the loan amount are considered new business, while only
the original amount granted and not yet repaid by the time the renegotiation takes
place is to be reported under renegotiated loans. If the renegotiation involves or is
preceded by a partial redemption of the loan, the amount amortised as a result of the
renegotiation is excluded from the amount of both new business and renegotiated
loans.
Question:
A loan of EUR 2,000 is granted on 3 February 2014. The loan is renegotiated on 17
February 2014 and an additional amount of EUR 1,000 is granted to the customer.
What should be reported in February under new business and under renegotiations?
Answer:
As the original loan (EUR 2,000) and the renegotiated original amount (EUR 2,000)
plus the increase in the loan amount (EUR 1,000) are all included in new business,
EUR 5,000 is reported as new business for February. The renegotiated original
amount of EUR 2,000 is recorded under renegotiated loans. Thus, the new money
arriving on the market in February is EUR 3,000.
Question:
A loan of EUR 5,000 is granted on 1 March 2014. An amount of EUR 1,000 is repaid
on 31 March 2014. On 15 April 2014 the remaining maturity of the loan is
renegotiated and at the same time an amount of EUR 1,000 is repaid. What should
be reported in March and April under new business and under renegotiations?
Answer:
Since no renegotiation takes place in March, the whole loan amount of EUR 5,000
reported under new business is new money arriving on the market. In April the loan
is renegotiated and the EUR 3,000 granted and not yet repaid is reported under both
new business and renegotiated loans.
Manual on MFI interest rate statistics − Business coverage
65
5.5.3 Renegotiations of loans not yet fully drawn (loans taken out in
tranches)
Question:
Let us consider again the example in Section 5.4.9, where a loan of EUR 10,000 is
granted at time t
0
(January 2014) for ten years at a fixed rate of 8%. The customer
takes out the first tranche of EUR 1,000 in February and then further tranches of
EUR 1,000 in each of the following nine months but, before the first tranche is taken,
a renegotiation between the MFI and the customer takes place and the interest rate
is then fixed at 7.5%. How should this loan be reported in MFI interest rate statistics
in January and February?
Answer:
As the contract is signed in January 2014, in this month a new business amount of
EUR 10,000 at 8% will be recorded. In February 2014, as a renegotiation takes
place, a new business amount will be recorded for EUR 10,000 at 7.5% and the
same amount will be recorded also as a renegotiation. In the subsequent nine
months, nothing will be recorded either as new business or as a renegotiation. In
terms of interest rates on outstanding amounts, the reporting agent will record
nothing in January, EUR 1,000 at 7.5% in February, and the same amount and
interest rate in the following nine months.
5.5.4 Loan transfers, sales of loans and reorganisation of loans
Several issues have to be considered here.
a) Sales or transfers of loans among MFIs, without the active involvement of
the customer
If an MFI sells or transfers loans to another MFI, which happens frequently as a
consequence of a corporate restructuring or, in general, as part of companies’
strategies, no new business should be reported, as these types of operations are
excluded from new business in MFI interest rate statistics and there is no active
involvement of the customer. The corresponding interest rates on outstanding
amounts are, however, reported.
b) Transfer of loans from one MFI to another, with the active involvement of the
debtor
If a household or non-financial corporation decides or agrees to transfer a loan
granted to it from one MFI to another, the transfer is reflected in new business as a
renegotiated loan, assuming that a negotiation of the terms and conditions of the
contract takes place. While the principle is clear, it is acknowledged that, in practice,
the new lender might not be aware of the earlier loan in another institution (and
hence may not be able to report the loan as a renegotiation). The recommendation
of this Manual, in these cases, is that reporting agents should report such
renegotiated loans coming from another MFI on a best-efforts basis; nevertheless,
Manual on MFI interest rate statistics − Business coverage
66
this solution is considered as provisional until the collection of data allows further
assessment.
c) Transfer of loans to an MFI from another institution outside the MFI sector
If the loan transferred to an MFI with the agreement of the customer was originated
by another institution outside the MFI sector, it is even more difficult for the reporting
agents to report the loan as a renegotiation. The provisional recommendation of this
Manual is to exclude these operations from renegotiations. Therefore, the MFI
receiving the transferred loan should report it as new business and not as a
renegotiation. Such recording would be appropriate for the analysis of the credit
market stance within the MFI sector, since such a transfer of loans would represent
“true loans” coming to the MFI sector. Nevertheless, if the analysis of the credit
market stance relates to the whole economy, not recording such operations as
renegotiations would be a significant drawback, should the amount of these
operations be material.
d) Change of the debtor of a loan
Question:
When a non-financial corporation steps in to replace another non-financial
corporation as a debtor and, subsequently, the MFI and the new debtor agree to
change some of the conditions of the loan, e.g. the collateral, should this change be
reported under renegotiated loans and new business or only under new business?
Answer:
To properly answer this question, it has to be kept in mind that a change in the terms
and conditions of the loan has occurred, and that there has been an active
involvement of a non-financial corporation (not the original customer, but the one
which has taken over the loan). Therefore, this change in the terms and conditions of
the loan has to be reported as both new business and a renegotiation.
e) Consolidation and splitting of loans
Loan consolidation in the sense that several loans granted to the same customer are
combined into one loan is considered new business, and the combined amounts
should be included in renegotiated loans. This applies also to the splitting of existing
loans into several separate loans.
f) Currency conversions
Currency conversions of existing loans are considered renegotiations if both the
reporting institution and the household or non-financial corporation have to agree on
the change in currency. If the conditions for conversion are, however, agreed when
the original loan is granted and the conversion takes place automatically, following a
pre-specified event, no renegotiation or new business is recorded.
Manual on MFI interest rate statistics − Business coverage
67
5.5.5 Calculation of the APRC for renegotiated loans
Question:
A household agrees with a bank on a consumer loan being granted with an original
maturity of 13 months and a fee of EUR 200 paid in advance. Three months later,
the customer and the bank renegotiate the maturity of the loan, shortening it to six
months from the date of the original contract. Since this renegotiation is reported as
new business, the APRC of this loan has to be reported as well. How should the
APRC be calculated in this case?
Answer:
Since the APRC indicates the total cost of a loan paid by a customer, the total fee
should be used in the calculation of the APRC of the renegotiated loan to reflect,
precisely, the increase in the cost arising from the only element of the loan that has
changed, i.e. the term, which is now shorter.
Manual on MFI interest rate statistics − Time reference point
68
6 Time reference point
83
6.1 Time reference point for interest rates on outstanding
amounts of overnight deposits, deposits redeemable at
notice, revolving loans and overdrafts, and convenience
and extended credit card credit
84
MFI interest rates on outstanding amounts,
85
i.e. indicators 1 to 26 in Table 1 of the
Appendix, may be compiled as a snapshot of end-month observations or as implicit
rates referring to the average over the month:
• As a snapshot of end-month observations, they are calculated as weighted
averages
86
of the interest rates applied to the stock of deposits and loans at a
certain point in time on the last day of the month. At that point in time, the
reporting agents collect the interest rates and the amounts involved for all
outstanding contracts relevant for MFI interest rate statistics. As end-month
observations, the MFI interest rates on outstanding amounts have the same
time reference point as MFI balance sheet statistics. Hence, the amounts
needed for weighting the interest rates may be taken from MFI balance sheet
statistics. By means of the collected interest rates and amounts, the reporting
agents then compile a weighted average interest rate for each instrument
category.
• As implicit rates referring to the average of the month, interest rates on
outstanding amounts are calculated as quotients. The numerator is the
accumulated flow of interest during the reference month, i.e. the accrued
interest payable on deposits and receivable on loans, and the denominator is
the average monthly stock. At the end of the reference month, for each
instrument category, the reporting agent needs to report the accrued interest
payable or receivable during the month and the stock of deposits and loans on
average during the same month. The NCB then calculates from these data the
monthly implicit interest rate per instrument category. As averages of the month,
these implicit rates on outstanding amounts have the same time reference point
as the MFI interest rates on new business, which are covered in Section 6.2.
As explained in Section 5.3.1, overnight deposits, deposits redeemable at notice,
convenience and extended credit card credit, and revolving loans and overdrafts
experience a large number of inflows and outflows throughout the month. Therefore,
the balance at the time reference point is taken as an indicator for the amount the
customer has chosen to leave on the overnight deposit or the deposit redeemable at
83
This chapter refers mainly to Part 3 of Annex I to the Regulation.
84
See also paragraphs 29 to 34 of Annex I to the Regulation.
85
Outstanding amounts are defined in Section 5.2.
86
See also Chapter 9.
Manual on MFI interest rate statistics − Time reference point
69
notice, or has chosen to take out as convenience or extended credit card credit or a
revolving loan or overdraft instead of placing or borrowing the money elsewhere.
Hence, for indicators 1, 5, 6, 7, 12, 23, 32 and 36 in Table 2 of the Appendix, and for
indicators 86 and 87 in Table 8 of the Appendix, the compilation procedure and the
time reference point are exactly the same as for the interest rates on outstanding
amounts. The choice between a snapshot of end-month observations and an implicit
rate referring to the average of the month has the following implications:
• In a snapshot of end-month observations, only one balance, i.e. the balance at
a certain point in time on the last day of the month, is taken into account when
computing the MFI interest rate. In the case of combined deposit and loan
facilities,
87
it is sufficient to look at this balance to decide whether the account in
the reference month is an overnight deposit or an overdraft.
• The denominator in an implicit rate referring to the average of the month is
based on the average of the daily balances on the overnight deposit, deposit
redeemable at notice, extended credit card credit or revolving loan and
overdraft. Therefore, in the case of combined deposit and loan facilities, each
day the reporting agent needs to assess whether the account is a deposit or a
loan. The reporting agent then calculates an average of the daily credit
balances and the daily debit balances to derive the average monthly stocks for
the denominator of the implicit rate. Also, for the flows in the numerator of the
implicit rate, accrued interest payable on deposits and receivable on loans need
to be distinguished. Reporting agents should not report weighted average
interest rates combining (low) overnight deposit rates and (high) bank overdraft
rates.
It is up to each NCB to decide which of the two methods, i.e. a snapshot of end-
period observations or implicit rates referring to period averages, is more appropriate
in the national context. The ECB strongly recommends not mixing both approaches
at national level, as this makes the interpretation of the national data more difficult.
However, if there are very good reasons to mix both approaches because the type of
institutions and the products offered require a different treatment, then this is
accepted.
Calculating the interest rates and the weights based only on the amounts
outstanding on the last day of the month, as in the case of a snapshot, is less
accurate than referring to the average stock during the month, as required for implicit
rates. For example, day-of-the-month effects might distort the end-month results.
However, these distortions are expected to be small as a result of the number of
accounts covered. Indeed, MFI interest rates on outstanding amounts cover all
deposits placed and not yet withdrawn or all loans withdrawn and not yet repaid by
customers in all the periods up to and including the reporting date. A snapshot of
end-month observations only covers those contracts that are still outstanding at the
time of data collection. By contrast, an implicit rate referring to the average of the
87
Further discussed in Section 5.3.3.
Manual on MFI interest rate statistics − Time reference point
70
month also includes contracts that were outstanding at some time during the month,
but are not outstanding at the end of the month.
The main difference between the two methods is, however, the way in which the data
are collected. In the case of a snapshot, interest rates are directly collected, whereas
in the case of implicit rates, the interest rates are derived as a quotient. Assuming
that the profit and loss accounts provide enough detail, the numerator in the quotient
derived from these accounts should be of sufficiently high quality. The quality of the
resulting implicit interest rates therefore depends on the denominator. Minimum
standards for compiling the average monthly stock per instrument category are laid
down in the Regulation. The ideal is the average of daily stocks over the month, as it
ensures a close link between the flow in the numerator and the reference stock in the
denominator. If instead the denominator refers to the end-month stock, then the flow
in the numerator may include payments for contracts that have already expired at the
end of the month, resulting in inaccurate implicit rates. Nevertheless, the ECB
accepts approximations of the average of daily stocks if they fulfil the following
minimum standards:
• For volatile instrument categories, i.e. at least overnight deposits, deposits
redeemable at notice, extended credit card credit and revolving loans and
overdrafts, the average monthly stock needs to be derived from daily balances.
• For all other instrument categories, the average monthly stock needs to be
derived from weekly or more frequent balances.
6.2 Time reference point for interest rates on new business
88
As defined in Section 5.4.1, MFI interest rates on new business reflect the average
interest rate level applied to deposits and loans in new agreements that have been
agreed between customers and MFIs during the reference month. The time
reference point for new business rates is therefore the average of the month. This
holds true for all indicators in Table 2 of the Appendix, except for indicators 1, 5, 6, 7,
12, 23, 32 and 36, which refer to overnight deposits, deposits redeemable at notice,
extended credit card credit and revolving loans and overdrafts, and for indicators in
Tables 3, 4, 5, 7 and 9.
For each instrument category, reporting agents calculate the new business interest
rate as a weighted average of all interest rates on new business operations in the
instrument category during the reference month. They then transmit these new
business interest rates referring to the average of the month to the NCB of the euro
area Member State in which they are resident. The transmission includes weighting
information on the amount of new business conducted during the reporting month for
each instrument category. Reporting agents need to take into account the new
business operations conducted during the entire month, rather than just a selected
period during the month.
88
See also paragraphs 35 to 37 of Annex I to the Regulation.
Manual on MFI interest rate statistics − Instrument categories
71
7 Instrument categories
89
7.1 Summary tables of indicators
MFI interest rate statistics provide detailed monthly information on 117 indicators, of
which approximately three-quarters refer to new business and slightly less than one-
quarter to outstanding amounts. All of the essential instrument categories for euro-
denominated deposits and loans of households and non-financial corporations are
covered. All indicators are, in principle, required at both euro area level and national
level, but national deviations in the number of indicators exist, which is further
discussed in Section 7.2. This set of indicators is organised in nine tables in
Appendix 1, in accordance with the Regulation and the Guideline.
Table 1 shows the 26 instrument categories for which MFI interest rates on
outstanding amounts are collected at euro area level. Depending on the choice of
NCBs, the interest rates are either AARs or NDERs and are compiled either as a
snapshot of end-month observations or as implicit rates referring to the average of
the month.
Tables 2 to 9 list the 91 MFI interest rates on new business that are collected at euro
area level. Indicators in Tables 2, 3, 4, 7, 8 and 9 are, depending on the NCBs’
choice, either AARs or NDERs and as such reflect only the interest rate paid or
charged by MFIs, excluding any other related fees. Indicators in Table 5 are APRCs
and therefore comprise an interest rate component and a component of other
charges. All new business rates other than indicators 1, 5, 6, 7, 12, 23, 32, 36, 86
and 87 are compiled as weighted average interest rates referring to the whole
month. For indicators 1, 5, 6, 7, 12, 23, 32, 36, 86 and 87, new business interest
rates are calculated in the same way as interest rates on outstanding amounts, as a
snapshot of end-month observations or as implicit rates referring to the average of
the month. The brackets around these indicators, included in Table 2 and in Table 8,
illustrate their special status. The indicators in Table 6 are business amounts
referring to the whole month.
With the 117 indicators in Tables 1 to 9, a balance is struck between the policy and
analytical needs of the users and the reporting burden on MFIs. The set of indicators
required by the Guideline (indicators in Tables 7, 8 and 9 of the Appendix) is
submitted to the ECB by the NCBs on a best-efforts basis.
7.2 General provisions
90
MFI interest rate statistics for the euro area refer to instrument categories rather than
to individual products as is the case with the national statistics on interest rates in
89
This chapter refers mainly to Part 4, Appendix 1 and Appendix 2 of Annex I to the Regulation.
90
See also paragraphs 38 to 41 of Annex I to the Regulation.
Manual on MFI interest rate statistics − Instrument categories
72
many countries. For example, MFI interest rates refer to consumer credit with “up to
x years” maturity or interest rate fixation, whereas, for instance, in the United States
interest rates are collected for a 48-month new car loan as a typical retail product. It
is recognised that interest rates referring to typical retail products might be easier to
interpret than interest rates referring to instrument categories and might provide a
better comparison with capital market interest rates. In the euro area, however, there
are large national differences in the banking business of MFIs. It is therefore not
possible to define a sufficient number of typical retail products that are
representative, or even available, in each Member State.
Also as a result of the national differences in products, deviations exist in the number
of indicators compiled per Member State. In some Member States, MFIs do not take
deposits or offer loans in some of the instrument categories listed in Tables 1 to 9
from/to households and non-financial corporations resident in the euro area. Where
inapplicable, the NCBs in these Member States ignore the instrument categories in
the data collection from reporting agents and collect less than the required 117
indicators at national level
91
. An instrument category is inapplicable at national level
only if resident MFIs do not at all offer products belonging to this category to
households and non-financial corporations resident in the euro area. Data have to be
provided if some business exists, however limited it is.
92
NCBs need to cover each instrument category listed in Tables 1 to 9 that exists in the
banking business of resident MFIs with euro area households and non-financial
corporations, but not each product offered at national level. Covering all products
could lead to a near-census situation in some countries with a wide variety of
products. However, NCBs cannot exclude a whole instrument category on the
grounds that the amounts involved are very small. So if an instrument category is
only offered by one institution, then this institution needs to be a reporting institution.
Furthermore, once an MFI is selected as a reporting agent, it has to cover for each
instrument category all interest rates applied to all the products that come under this
category. This principle implies that it is not up to NCBs to define a set of national
products within each instrument category on which reporting agents collect data.
The exception to the rule of covering for each instrument category all interest rates
applied to all the products are interest rates on bad loans and loans for debt
restructuring at rates below market conditions. As already explained in Section 5.1,
these loans are not covered by MFI interest rate statistics.
If a new product belonging to an existing instrument category is created, the
reporting institutions cover it with the next reporting, as all reporting agents have to
91
It is possible that some NCBs collect more than 117 indicators at national level, because the statistical
reporting requirements of the ECB are part of a broader statistical reporting framework which the NCB
establishes under its own responsibility.
92
Council Regulation (EC) No 2533/98 concerning the collection of statistical information by the
European Central Bank specifies in Article 8 that the ECB and NCBs may collect confidential statistical
information for the tasks of the ESCB. The ECB and NCBs have to take all regulatory, administrative,
technical and organisational measures to ensure the physical protection and logical security of
confidential statistical information. The ECB has the responsibility to define common rules and
implement minimum standards to prevent unlawful disclosure and unauthorised use of such
information.
Manual on MFI interest rate statistics − Instrument categories
73
report all their products. If an instrument category did not exist in a Member State at
the time the sample was first drawn, but one institution subsequently creates a new
product belonging to this category, this institution needs to be introduced into the
sample at the time of the next representativity check
93
. The reason for this is that
with the creation of this new product the instrument category is no longer
inapplicable at national level.
7.3 Foreign-currency deposits and loans
94
MFI interest rate statistics cover the interest rates applied by resident MFIs (except
central banks and money market funds) to euro-denominated deposits and loans vis-
à-vis households and non-financial corporations resident in the euro area. Dual
currency loans exist where the currency of the interest payments differs from the
currency in which the loan is granted. The treatment of dual currency loans follows
two principles:
1. either both the amount lent and the interest rate are covered by MFI interest
rate statistics or both are excluded; and
2. one main use of MFI interest rate statistics is to monitor the transmission of the
ECB’s monetary policy.
Hence, if the lending rate agreed between the customer and the MFI is based on
considerations relating to the monetary policy as carried out by the ESCB, then the
interest rate and the amount of the loan should be covered by MFI interest rate
statistics. If the loan is in the local currency, then it should be covered by the MFI
interest statistics independently of the fact that interest payments are made in local
currency or not.
7.4 Breakdown by sector
95
MFI interest rate statistics distinguish between interest rates applied in the banking
business vis-à-vis households (including NPISHs) and vis-à-vis non-financial
corporations. The ESA provides the standard for this sector classification:
1. The non-financial corporation sector consists of institutional units which are
independent legal entities and market producers, and whose principal activity is
the production of goods and non-financial services (ESA 2010, paragraph 2.45).
The institutional units covered are the following:
(a) private and public corporations which are market producers principally
engaged in the production of goods and non-financial services;
93
Further discussed in Sections 12.8 and 12.9.
94
See also paragraph 42 of Annex I to the Regulation.
95
See also paragraphs 43 to 45 of Annex I to the Regulation.
Manual on MFI interest rate statistics − Instrument categories
74
(b) cooperatives and partnerships recognised as independent legal entities
which are market producers principally engaged in the production of goods
and non-financial services;
(c) public producers which are recognised as independent legal entities and
which are market producers principally engaged in the production of goods
and non-financial services;
(d) non-profit institutions or associations serving non-financial corporations,
which are recognised as independent legal entities and which are market
producers principally engaged in the production of goods and non-financial
services;
(e) head offices controlling a group of corporations which are market
producers, where the preponderant type of activity of the group of
corporations as a whole – measured on the basis of value added – is the
production of goods and non-financial services;
(f) special-purpose entities (SPEs), the principal activity of which is the
provision of goods or non-financial services; and
(g) private and public quasi-corporations which are market producers
principally engaged in the production of goods and non-financial services.
2. The household sector consists of individuals or groups of individuals as
consumers and as entrepreneurs producing market goods and non-financial
and financial services (market producers) provided that the production of goods
and services is not by separate entities treated as quasi-corporations. It also
includes individuals or groups of individuals as producers of goods and non-
financial services for exclusively own final use (ESA 2010, paragraph 2.118).
The household sector includes (ESA 2010, paragraph 2.119):
(a) individuals or groups of individuals whose principal function is
consumption;
(b) persons living permanently in institutions who have little or no autonomy of
action or decision in economic matters (e.g. members of religious orders
living in monasteries, long-term patients in hospitals, prisoners serving
long sentences, old persons living permanently in retirement homes). Such
people are treated as a single institutional unit, i.e. a single household;
(c) individuals or groups of individuals whose principal function is
consumption and that produce goods and non-financial services for
exclusively own final use; only two categories of services produced for
own final consumption are included within the system: services of owner-
occupied dwellings and domestic services produced by paid employees;
(d) sole proprietorships and partnerships without legal status, other than those
treated as quasi-corporations, and which are market producers; and
Manual on MFI interest rate statistics − Instrument categories
75
(e) non-profit institutions serving households which do not have independent
legal status, or those which do but which are of only minor importance.
Sole proprietorships and partnerships which are not recognised as independent legal
entities and are market producers are classified as follows (ESA 2010, paragraph
2.141):
1. If they are quasi-corporations:
(a) those principally engaged in the production of goods and non-financial
services: in sector S.11 (non-financial corporations);
(b) those principally engaged in financial intermediation and financial auxiliary
activities: in sector S.12 (financial corporations)
2. If they are not quasi-corporations, they are classified in sector S.14
(households).
For the purpose of MFI interest rate statistics, the household sector includes also the
non-profit institutions serving households (NPISH) sector, defined in ESA 2010
paragraph 2.129 as non-profit institutions which are separate legal entities, which
serve households and which are private non-market producers. Their principal
resources come from voluntary contributions in cash or in kind from households in
their capacity as consumers, payments made by general government and property
income.
The NPISH sector includes the following main kinds of NPISHs that provide non-
market goods and services to households (ESA 2010, paragraph 2.130):
1. trade unions, professional or learned societies, consumers’ associations,
political parties, churches or religious societies (including those financed but not
controlled by governments), and social, cultural, recreational and sports clubs;
and
2. charities or relief or aid agencies financed by voluntary transfers in cash or in
kind from other institutional units.
This sector also includes charities or relief or aid agencies serving non-resident units
and excludes entities where membership gives a right to a predetermined set of
goods and services.
Indicator 5 in Table 1 and indicator 11 in Table 2 refer to repos. At the euro area level,
no sector breakdown for repos is required, because their remuneration is often,
although not in all Member States, independent of the holding sector. Thus, for the
purpose of MFI interest rate statistics on repos, the two sectors, households and
non-financial corporations, are merged. Indicator 5 in Table 1 and indicator 11 in
Table 2 therefore indistinguishably refer to both sectors instead of either households
or non-financial corporations. In line with the 5th recital and Article 3(2) of the
Regulation, NCBs may ask for a sector breakdown for repos at national level.
Manual on MFI interest rate statistics − Instrument categories
76
Indicators 5 and 6 in Table 2 refer to deposits redeemable at notice held by
households. In the euro area, deposits redeemable at notice are overwhelmingly
owned by households, i.e. currently about 95% of all deposits redeemable at notice
held with MFIs. Non-financial corporations hold about 5% of these deposits. Since it
was found to be more cost-efficient to include households and non-financial
corporations indistinguishably in one sector, MFI interest rates are collected for
deposits redeemable at notice vis-à-vis both sectors and then entirely allocated to
deposits redeemable at notice held by households.
7.5 Breakdown by type of instrument
96
7.5.1 Types of deposits
The instrument breakdown by type of deposit for MFI interest rate statistics and the
definitions used follow MFI balance sheet statistics and hence
Regulation ECB/2013/33. It defines overnight deposits
97
as deposits which are
convertible into currency and/or which are transferable on demand by cheque,
banker’s order, debit entry or similar means, without significant delay, restriction or
penalty. This item includes:
• balances (interest bearing or not) which are immediately convertible into
currency on demand or by close of business on the day following that on which
the deposit was made, without any significant penalty or restriction, but which
are not transferable;
• balances (interest bearing or not) representing prepaid amounts in the context
of “hardware-based” or “software-based” e-money (e.g. prepaid cards).
MFI interest rates on overnight deposits, i.e. indicators 1 and 7 in Table 2, cover all
overnight deposits, whether or not they are interest bearing. Zero-interest overnight
deposits influence the transmission mechanism of monetary conditions, the own rate
of return on M3, the interest burden on households and non-financial corporations,
etc., and are therefore captured by MFI interest rate statistics.
According to Regulation ECB/2013/33 deposits with agreed maturity are “non-
transferable deposits which cannot be converted into currency before an agreed
fixed term or that can only be converted into currency before that agreed term
provided the holder is charged some kind of penalty. This item also includes
administratively regulated savings deposits where the maturity related criterion is not
relevant (classified in the maturity band “over two years”). Financial products with
roll-over provisions must be classified according to the earliest maturity. Although
deposits with agreed maturity may feature the possibility of earlier redemption after
prior notification, or may be redeemable on demand subject to certain penalties,
96
See also paragraphs 46 to 55 of Annex I to the Regulation.
97
This and the following definitions can be found in Part 2 of Regulation ECB/2013/33.
Manual on MFI interest rate statistics − Instrument categories
77
these features are not considered to be relevant for classification purposes.”
Indicators 1 to 4 in Table 1 and indicators 2 to 4 and 8 to 10 in Table 2 follow this
definition.
The same Regulation defines deposits redeemable at notice as “non-transferable
deposits without any agreed maturity which cannot be converted into currency
without a period of prior notice; before the expiry the conversion into currency is not
possible or possible only with a penalty. They include deposits which, although
perhaps legally withdrawable on demand, would be subject to penalties and
restrictions according to national practice (classified in the maturity band “up to and
including three months”), and investment accounts without period of notice or agreed
maturity, but which contain restrictive drawing provisions (classified in the maturity
band “over three months”).” This definition applies to indicators 5 and 6 in Table 2.
Also, Regulation ECB/2013/33 defines repos as “counterpart of cash received in
exchange for securities sold by reporting agents at a given price under a firm
commitment to repurchase the same (or similar) securities at a fixed price on a
specified future date. Amounts received by reporting agents in exchange for
securities transferred to a third party (“temporary acquirer”) are to be classified under
“repurchase agreements” where there is a firm commitment to reverse the operation
and not merely an option to do so. This implies that reporting agents retain all the
risks and rewards of the underlying securities during the operation.”
7.5.2 Types of loans
The instrument breakdown by type of loan for MFI interest rate statistics and the
definitions used follow as far as possible MFI balance sheet statistics (Regulation
ECB/2013/33). For the purpose of MFI interest rate statistics, revolving loans and
overdrafts (i.e. indicators 12 and 23 in Table 2) are defined as follows:
Revolving loans are loans that have all of the following features: (1) the borrower
may use or withdraw funds up to a pre-approved limit without giving prior notice to
the lender; (2) the amount of available credit can increase and decrease as funds
are borrowed and repaid; (3) the credit may be used repeatedly; and (4) there is no
obligation to regularly repay the funds.
Revolving loans include the amount obtained through a line of credit and not yet
repaid (outstanding amounts). A line of credit is an agreement between a lender and
a borrower that allows a borrower to take advances, during a defined period and up
to a certain limit, and repay the advances at his/her discretion before a defined date.
Amounts available through a line of credit that have not been withdrawn or have
already been repaid should not be considered under any loan category.
Overdrafts are debit balances on current accounts.
Both revolving loans and overdrafts exclude loans provided through credit cards. The
total amount owed by the borrower has to be reported, irrespective of whether it is
Manual on MFI interest rate statistics − Instrument categories
78
within or beyond any limit agreed beforehand between the lender and the borrower
with regard to the size and/or maximum period of the loan.
The interest rate on overdrafts refers to the rate charged when an overnight deposit
becomes negative, i.e. the overnight deposit and the overdraft are linked to the same
account. In contrast to loans to non-financial corporations up to one year and
consumer credit and other loans to households up to one year, overdrafts are without a
defined maturity and, in general, are authorised but taken without giving prior notice to
the bank. Usually, the MFI defines an upper limit on the size and a maximum period of
the overdraft. All overdrafts are covered, whether they are within or beyond the limit
agreed between the reporting agent and the household or non-financial corporation.
Typically, penalties apply if the overdraft is extended beyond the agreed limit. The
penalties may be charged as an interest rate component, a component of other
charges, or a combination of both. Penalties on overdrafts applied as a component of
other charges, i.e. in the form of special fees, are not covered by the AAR or the
NDER, because these rates only cover the interest rate component of loans. If
penalties on overdrafts are applied as an interest rate component, i.e. a higher interest
rate, this higher interest rate is reflected in MFI interest rate statistics.
For the purpose of MFI interest rate statistics, credit card debt has the same
meaning as defined in Part 2 of Annex II to Regulation ECB/2013/33. Credit cards
allow households and non-financial corporations to postpone the payment of a
transaction. Depending on the conditions of this postponement, two kinds of cards
can be distinguished: delayed debit cards and credit cards, which give the
cardholder access to two different types of credit:
1. Convenience credit is defined as the credit granted at an interest rate of 0% in
the period between the payment transaction(s) made with the card during one
billing cycle and the date at which the debit balances from this specific billing
cycle become due.
Through these cards and also through credit cards, the cardholder makes
purchases during one period, generally one month, paying with the card. The
receipts of the purchases arrive successively at the credit institution which
issued the card. The credit card contract with the cardholder specifies that the
receipts that arrive during a month will be debited, all together, for example, on
the 5th day of the following month, from the cardholder’s current account or
loan account. No interest is charged to the cardholder for this deferral of
payment, which lasts at most one month.
2. Extended credit
98
comprises credit granted to the cardholder after the due
date(s) of the previous billing cycle(s) has/have passed, i.e. debit amounts on
the card account that have not been settled when this was first possible, for
which an interest rate or tiered interest rates usually greater than 0% are
charged. Often, minimum instalments per month have to be made, to at least
partially repay extended credit.
98
See also paragraph 49 of Annex I to the Regulation.
Manual on MFI interest rate statistics − Instrument categories
79
According to these two definitions of credit granted through credit cards, a delayed
debit card can give access to both types of credit: convenience credit for the
transactions made during one month until the due date, and extended credit for the
balance not paid by the due date. This would be the case if the current or loan account
where the total bills of the month should be settled has insufficient funds. The terms
and conditions agreed between the MFI issuing the card and the cardholder include
the rate of interest that the MFI will charge the cardholder for the balance not paid by
the due date. Therefore, the amount not paid by the due date has to be classified as
“extended credit”. Nevertheless, normally “extended credit” through a credit card
comes from the transactions made by the cardholder with a credit card, the terms and
conditions of which stipulate the deferred payment of such transactions and the
interest rate that the MFI will charge on the amount spent and not yet repaid.
Data on the interest rate are collected separately only in respect of extended credit
card credit in indicators 32 and 36 of Table 2.The interest rate on convenience
credit card credit is not reported separately, as it is by definition 0%. However, the
outstanding convenience credit card credit is included in MFI interest rate statistics
on outstanding loan amounts, together with the outstanding extended credit card
credit. In addition, convenience credit is included in indicators 86 and 87 of Table 8.
Yet, a question still needs to be answered: given that extended credit card credit
allows the cardholder to postpone the payment of the transactions until a certain
date, should this be recorded as new business at the time the credit card contract is
signed or should only the amounts drawn be recorded?
Answer: The recording of new business figures for extended credit card credit is one
of the cases (together with overnight deposits, deposits redeemable at notice and
revolving loans and overdrafts) where the interest rates on new business and on
outstanding amounts coincide, as both are calculated as the rate of interest applied
to the outstanding amount on the last day of the corresponding reference month
(“snapshot” option) or as the average of the stock throughout the month.
The following example aims to clarify the recording of convenience and extended
credit in MFI interest rate statistics.
A customer has a credit card (not linked to any deposit or loan account) with which
he or she withdraws cash and makes purchases of goods and services up to a
maximum amount of EUR 5,000. The credit card receipts are sent to the MFI that
issued the card, and every month-end, the credit institution sends the cardholder an
invoice for the value of the credit card receipts. The due date agreed to pay the bill is
the 26th of each following month, which means that the cardholder is granted a
maximum interest-free period of 56 days, during which he/she has to pay a minimum
amount of 3% of the balance. If the customer does not pay this minimum amount, in
addition to the interest payments, additional charges will be imposed. Interest is only
charged after this interest-free period has passed on the outstanding balance, and it
is 16% per annum.
The following transactions were carried out during September and October 2013:
Manual on MFI interest rate statistics − Instrument categories
80
Table 7
Credit card transactions
Credit card
operation date
Operation Amount Observations
Total debt:
Convenience
credit
Repayments
Total debt:
Extended
credit
10-Sep-13
Purchase of
good 1
€ 1,750 € 1,750
16-Sep-13
Withdraw € 500 € 2,250
30-Sep-13
End of billing
cycle
€ 2,250
12-Oct-13
Payment € 1,500 € 750
26-Oct-13
End of the interest
free period
€ 1,500
28-Oct-13
Purchase of
good 2
€ 500 € 500 € 1,500
31-Oct-13
End of billing
cycle
€ 500 € 1,500
Question:
1. What amounts and interest rates should be recorded in MFI interest rate
statistics for October 2013, as outstanding amounts and new business for the
different instruments involved and, specifically, does the maximum amount set
by the credit institution (EUR 5,000) constitute new business?
2. How would penalties due to late payments affect MIR statistics?
Answer:
1. The recording in October 2013 of the different transactions undertaken is as
follows:
Table 8
The recording (October 2013) of the different transactions
Table Loan / deposit Sector Type of instrument
Original maturity /
period of notice /
initial rate fixation
Rate Volume
Outstanding
amounts
L HH
Loans for consumption
and other purposes
[decided at national
level]
12% (2,000 BSI)
(indicators 9 to 11 in
Table 1 of the Appendix)
New business
L HH
Extended credit
16% (1,500 BSI)
(indicator 32 in Table 2
of the Appendix)
New business
L HH
Revolving loans and
overdrafts, convenience
and extended credit
12% 2,000
(indicator 86 in Table 8
of the Appendix)
The 12% is calculated as a weighted average of the 0% and the 16%:
(16%*1500+0%*500)/2000=12%
Manual on MFI interest rate statistics − Instrument categories
81
The amount of new business is the money that has actually been drawn. The
maximum amount set by the credit institution does not constitute new business. The
repayments of the outstanding balances during the month are captured by
decreases in the outstanding amounts. If the cardholder pays off the balance in full
at the counter or by bank transfer before the due date and thereby incurs no interest
charges, only the interest-free period is reflected for overdrafts in the form of a 0%
interest rate on the outstanding amount.
Regarding the penalties, if a minimum amount of 3% of the balance is not repaid
within 56 days and a penalty is charged in the form of higher interest rates, this
penalty is also reflected in MFI interest rate statistics. If the penalty is applied in the
form of fees or other non-interest components, it is not covered.
Regulation ECB/2013/33 defines loans as “holdings of financial assets created when
creditors lend funds to debtors which are not evidenced by documents or are
evidenced by non-negotiable documents”.
Loans to non-financial corporations comprise all loans to non-financial
corporations regardless of their size.
99
In the case of outstanding amounts, i.e. for
indicators 12 to 14 and 21 to 26 in Table 1, loans to non-financial corporations cover
revolving loans and overdrafts, and convenience and extended credit card credit. By
contrast, new (other)
100
loans in indicators 37 to 54 in Table 2, indicators 62 to 79 in
Table 3, indicators 80 to 85 in Table 4, indicator 91 in Table 6, indicators 24 to 29 in
Table 7 and indicator 91 in Table 9 exclude revolving loans and overdrafts and credit
card debt for the purpose of MFI interest rate statistics. As explained above,
revolving loans and overdrafts and extended credit card credit to non-financial
corporations constitute, as indicators 23 and 36, separate instrument categories in
Table 2. Revolving loans and overdrafts, and convenience and extended credit card
credit to non-financial corporations constitute in turn a separate indicator 87 in Table
8. The interest rate on convenience credit card credit is not reported separately, as it
is by definition 0%.
According to Regulation ECB/2013/33, loans for consumption granted to
households are “loans granted for the purpose of mainly personal use in the
consumption of goods and services”. Credit for consumption granted to sole
proprietors and partnerships without legal status is included in this category if the
reporting MFI knows that the loan is predominantly used for personal consumption
purposes. In the case of outstanding amounts, i.e. for indicators 9 to 11 in Table 1,
consumer credit covers revolving loans and overdrafts and credit card debt. By
contrast, new loans to households for consumption, i.e. indicators 13 to 15 in Table
2, indicators 55 to 57 in Table 3, indicator 30 in Table 5, indicator 88 in Table 6 and
indicator 88 in Table 9, exclude revolving loans and overdrafts and credit card debt
for the purpose of MFI interest rate statistics. These instruments are included in
Table 2 as separate instrument categories, i.e. indicators 12 and 32.
99
Further discussed in Section 7.6.
100
“Other” refers to “other than revolving loans and overdrafts and credit card debt”.
Manual on MFI interest rate statistics − Instrument categories
82
Regulation ECB/2013/33 defines loans to households for house purchases
101
as
“credit extended for the purpose of investing in houses for own use or rental,
including building and refurbishments”. It comprises loans secured on residential
property that are used for the purpose of house purchase and other loans for house
purchases made on a personal basis or secured against other forms of assets.
Housing loans granted to sole proprietors and partnerships without legal status are
included in this category unless the reporting agent knows that the house is
predominantly used for business-related purposes, in which case it is reported as
“for other purposes, of which: Sole proprietors”. MFI interest rate statistics in Tables
1 and 2 cover indistinguishably secured and unsecured loans to households for
house purchases, while indicators 58 to 61 in Table 3 refer only to loans with
collateral and/or guarantees. In the case of outstanding amounts (i.e. for indicators 6
to 8 in Table 1) loans to households for house purchases include revolving loans and
overdrafts and credit card debt, while in the case of new business (i.e. for indicators
16 to 19 in Table 2, indicators 58 to 61 in Table 3, indicator 31 in Table 5, indicator 89
in Table 6 and indicator 89 in Table 9) revolving loans and overdrafts and credit card
debt are excluded. Instead, revolving loans and overdrafts and extended credit card
credit to households are covered in Table 2 as part of separate instrument
categories, i.e. indicators 12 and 32.
Other loans to households are defined in Regulation ECB/2013/33 as loans
granted for purposes other than consumption and house purchase such as business,
debt consolidation, education, etc. In the case of outstanding amounts (i.e. for
indicators 9 to 11 in Table 1, other loans to households include revolving loans and
overdrafts and credit card debt. In the case of new business (i.e. for indicators 20 to
22 in Table 2, indicator 90 in Table 6 and indicator 90 in Table 9), new loans to
households for other purposes exclude revolving loans and overdrafts and extended
credit card credit, because they are covered in Table 2 as part of separate instrument
categories (i.e. indicators 12 and 32).
For MFI interest rates on outstanding amounts, consumer credit, loans to
households for house purchases and other loans to households together cover all
loans granted to households by resident MFIs. For MFI interest rates on new
business, the possible loan types for households are revolving loans and overdrafts,
convenience and extended credit card credit, credit for consumption, loans to
households for house purchases and other loans.
7.6 Breakdown by amount category
102
Large non-financial corporations have considerable economic significance and
account for a substantial share of banking business. The loans granted to large firms
will, due to the amounts involved, dominate any weighted average interest rate
referring to the non-financial corporation sector as a whole. Small firms, however,
101
ECB/2013/33 refers to “lending for house purchases”.
102
See also paragraph 56 of Annex I to the Regulation.
Manual on MFI interest rate statistics − Instrument categories
83
may play a special role in the transmission of monetary policy, since they have
limited access to capital markets and are therefore more vulnerable to changes in
the lending rates offered by MFIs. In general, larger firms are more able to negotiate
the interest rate and conditions than small firms so that large non-financial
corporations have available to them less standardised products with interest rates
that are closer to (or identical with) market interest rates.
The size of non-financial corporations can be defined on the basis of various
measures such as the number of employees or the turnover. Instead of the size of
the non-financial corporation, the size of the loan is considered to be more relevant
for distinguishing banking conditions in broad terms for the purpose of MFI interest
rate statistics.
103
Indeed, most Member States confirmed a link between the amount
of the loan granted and the interest rate. It should, however, be borne in mind that
the amount of a loan is only one of many criteria that are taken into account when
negotiating the interest rate. The sum of all criteria may best be summarised by
“credit risk”.
For new (other)
104
loans to non-financial corporations, i.e. indicators 37 to 54 in Table
2, 62 to 79 in Table 3 and 80 to 85 in Table 4, three size categories for the amount of
the loan granted are distinguished, i.e. “up to and including EUR 0.25 million”; “over
EUR 0.25 and up to and including EUR 1 million”; and “over EUR 1 million”. The
amount refers to the single loan transaction considered as new business. It does not
cover all business between the non-financial corporation and the reporting agent.
Additionally, the NCBs have to calculate and submit to the ECB a complementary
size category for indicators 24 to 29 in Table 7 up to and including EUR 1 million, and
over an amount of EUR 1 million, in addition to the other size categories mentioned
above.
Amount categories apply only to new business rates vis-à-vis non-financial
corporations and only to the instrument category of other loans. No breakdown by
amount category is required for revolving loans and overdrafts and credit card debt
or for deposits of non-financial corporations. Also, no breakdown by amount is
requested for deposits and loans from/to households or for interest rates on
outstanding amounts.
The breakdown by amount category for other loans to non-financial corporations
increases the comparability of the interest rates on such loans since the size of the
individual loans covered varies widely. As already mentioned at the beginning of this
section, without amount categories a weighted average interest rate on loans to non-
financial corporations would be dominated by the interest rates on the largest loans
during the reporting month. This weighted average lending rate might not be
representative of most loans to non-financial corporations. The indicators for (other)
loans to non-financial corporations of an amount up to and including EUR 0.25
million are assumed to reflect the interest rates faced by smaller corporations, which
are “price takers” rather than “price makers”.
103
This is also partly due to the fact that up-to-date information on the size of the company is not always
available.
104
“Other” refers to “other than revolving loans and overdrafts and credit card debt”.
Manual on MFI interest rate statistics − Instrument categories
84
7.7 Breakdown by original and residual maturity, notice and
interest rate reset period or initial rate fixation
105
7.7.1 Time bands
Depending on the type of instrument and whether interest rates on outstanding
amounts or new business are being looked at, MFI interest rate statistics provide a
breakdown by original and residual maturity, notice and interest rate reset period, or
initial period of fixation of the interest rate. These breakdowns refer to time bands or
ranges. For example, an interest rate on a deposit with an agreed maturity of up to
two years refers to an average interest rate across all deposits with an agreed
original maturity between one day and a maximum of two years. Time bands or
ranges are also used in the MFI balance sheet statistics.
The alternative would be to define exact points in time, i.e. interest rates on
instruments with a maturity or a notice, interest rate reset or initial rate fixation period
of x years/months. The advantages and drawbacks of basing statistics on exact
products rather than instrument categories are discussed in Section 7.2.
7.7.2 Original and residual maturity and period of notice
A breakdown by original maturity is applied in Table 2 to new business in deposits
with agreed maturity. In Table 1, a breakdown by original maturity is required for all
lending rates and all deposit rates on outstanding amounts with the exception of
repos. A breakdown by original maturity in combination with residual maturity and the
next interest rate reset is required for indicators 15 to 20 and 21 to 26 in Table 1. For
lending rates on new business, the corresponding breakdowns are expressed in
terms of initial period of interest rate fixation which, according to the Regulation, is
defined as a predetermined period of time at the start of the contract during which
the value of the interest rate will not change
106
. Separate data on loans to non-
financial corporations with an initial rate fixation period of up to one year in
combination with an original maturity of above one year are also reported (indicators
80 to 85 in Table 4).
For repos (i.e. indicator 5 in Table 1 and indicator 11 in Table 2), no maturity
breakdown is required at the euro area level, as repos are assumed to be
predominantly very short-term. At the national level, NCBs may ask for a maturity
breakdown for repos, which would be in line with the 5th recital and Article 3(2) of the
Regulation.
The definition of (and the breakdown by) original maturity follows
Regulation ECB/2013/33, which states that “maturity at issue (original maturity)
105
See also paragraphs 57 to 61 of Annex I to the Regulation.
106
Further details can be found in Section 7.7.4.
Manual on MFI interest rate statistics − Instrument categories
85
refers to the fixed period of life of a financial instrument before which it cannot be
redeemed (e.g. debt securities) or before which it can be redeemed only with some
kind of penalty (e.g. some types of deposits)”.
Question:
Revolving loans and overdrafts, and convenience and extended credit typically have
no predefined maturity date. In which maturity band should these loans be included
in Table 1 of the Regulation?
Answer:
If no specific agreement is made regarding the repayment schedule, revolving loans
and overdrafts, and convenience and extended credit should be considered short-
term, i.e. these loans should be reported in outstanding amounts of loans with an
agreed maturity of up to one year.
The same Regulation defines further that the period of notice corresponds to the
time between the moment the holder gives notice of an intention to redeem the
instrument and the date on which the holder is allowed to convert it into cash without
incurring a penalty. Financial instruments are classified according to the period of
notice only when there is no agreed maturity. For MFI interest rates, a breakdown by
period of notice is applied in Table 2 to new business in deposits redeemable at
notice.
7.7.3 Period of maturity for a loan taken out in tranches
107
Question:
A loan is agreed between a credit institution and a customer at t
0
, but the customer
takes out the funds only two months later at t
1
. Should the original maturity of this
loan be defined as:
• the date of the contract (t
0
) until the end of the contract (t
n
), or
• the first value date (t
1
) until the end of the contract (t
n
)?
Answer:
A household or non-financial corporation will normally take out a loan (other than an
overdraft) in full at the start of the contract. Hence, in general, the contract date t
0
and the first value date t
1
coincide. It may happen that a household or non-financial
corporation decides to take out the loan in tranches at times t
1
, t
2
, t
3
, etc. instead of
withdrawing the full amount at the start of the contract (time t
0
). Whatever the pattern
of withdrawals, for new business the original maturity of the loan always refers to the
agreed original loan period according to the loan contract, i.e. in this example to the
period from t
0
to t
n
.
Nevertheless, the loan actually appears in the MFI balance sheet statistics only at
time t
1
. Therefore, at that time, it appears for the first time in MFI interest rate
107
Defined in Section 5.4.9.
Manual on MFI interest rate statistics − Instrument categories
86
statistics on outstanding amounts, but the original maturity of this loan in these
statistics is the same as in new business. The same applies to the original maturity
of outstanding amounts should successive loan tranches be withdrawn.
7.7.4 Initial period of fixation of the interest rate
Users have always expressed a strong need for a distinction between variable and
fixed interest rates on loans. This is an unquestionable need since the effects of
monetary policy decisions are more quickly transmitted through variable interest
rates than through fixed interest rates. In order to implement such a breakdown in a
harmonised way across all Member States, variable and fixed need to be clearly
defined. The definition needs to be unambiguous for all Member States and able to
cope with financial innovations, which tend to cause difficulties at the borderline.
However, large national differences exist in the retail banking business in the EU in
that there is no common view on the period of fixation that classifies a lending
interest rate as fixed or variable. An interest rate which is fixed for one year and then
variable might be considered as fixed in one Member State and as variable in
another depending on national lending practices.
To overcome this problem, the initial period of rate fixation was introduced as the
basis for a breakdown of new lending business in MFI interest rate statistics. It gives
an indication of the variability of interest rates at euro area level, without prejudging
whether a loan with a specific fixation period is considered as fixed or variable in the
national context. The initial period of fixation is a predetermined period of time at the
start of a contract during which the value of the interest rate cannot change. The
value of the interest rate is only considered to be unchangeable if it is defined as:
• an exact level, for example as 10%, or
• a spread over an external index at a certain time, for example as six-month
EURIBOR plus two percentage points, which is equivalent to an exact interest
rate level.
If at the start of the contract the customer and the reporting agent agree on a
procedure for calculating the lending rate, this is not considered to be an initial rate
fixation. For example, it might be agreed that six-month EURIBOR plus two
percentage points is the interest rate for the loan. Here two cases should be
distinguished: (1) whether the variable rate used, here the six-month EURIBOR plus
two percentage points, is changing on a frequent basis, for instance every day or at
an unknown or irregular frequency; or (2) whether the rate is fixed for the first six
months and then updated every six months using the six-month EURIBOR as a
base. In the first case, there is no initial rate fixation. In the second case, the initial
period of interest rate fixation is six months. This application of a rate changing at a
frequency linked to the external index is common in many countries. It could be the
case, however, that the initial period of rate fixation is shorter or longer than the
external index reference period. In these cases and as a general rule, the Manual
recommends recording as the initial rate fixation period the actual period defined at
the start of the contract during which the level of the interest rate applied to the
Manual on MFI interest rate statistics − Instrument categories
87
deposit or loan does not change, independently of the external index reference
period. For new loans granted to non-financial corporations, six periods of initial rate
fixation are distinguished (indicators 37 to 54 in Table 2 and indicators 62 to 79 in
Table 3). This detail, together with the five
108
loan amount categories, allow a
detailed picture to be obtained of how monetary policy is affecting the structure of the
loans for each kind of enterprise.
Three periods of initial rate fixation are defined for new loans to households for
consumption and for other purposes, i.e. indicators 13 to 15, 20 to 22, and 33 to 35
in Table 2, and indicators 55 to 57 in Table 3, loans without any interest rate fixation
are included as loans with floating rates that have up to a one-year period of initial
rate fixation:
• floating rate and up to (and including) 1 year initial rate fixation;
• over 1 and up to (and including) 5 years initial rate fixation; and
• over 5 years initial rate fixation.
Considering that 95% of loans to households for house purchases are granted with
an original maturity of over five years, and in some countries with an interest rate
fixation of ten years or more, for this type of loan four periods of initial rate fixation
are distinguished; again, loans without any interest rate fixation are included as loans
with floating rates that have up to a one-year period of initial rate fixation:
• floating rate and up to (and including) 1 year initial rate fixation;
• over 1 and up to (and including) 5 years initial rate fixation;
• over 5 and up to (and including) 10 years initial rate fixation; and
• over 10 years initial rate fixation.
Normally, the initial period of rate fixation is shorter than or equal to the original
maturity of the loan.
109
The initial period of rate fixation might be short and the
interest rate agreed between the customer and the reporting agent for this initial
period of fixation might not be representative of the entire maturity of the loan. New
business statistics only reflect the interest rate that is agreed for the initial period of
fixation at the start of a contract or after renegotiation of the loan. If, after this initial
period of fixation, the interest rate automatically changes to a variable rate, which
might be at a very different level, this is not reflected in the MFI interest rates on new
business, but rather in those on outstanding amounts. Hence, both sets of statistics
are needed to capture the interest rate level and developments in the euro area and
at national level. The two following examples further illustrate the interdependence of
MFI interest rate statistics on new business and on outstanding amounts.
108
Three required by the Regulation and two, additionally, by the Guideline.
109
In exceptional cases, the initial period of rate fixation may be longer than the maturity of the loan. For
example, if a loan is agreed to which a fixed interest rate applies for the first year and this loan may be
repaid with one month’s notice, then the shortest possible maturity is one month and the period of initial
rate fixation is one year.
Manual on MFI interest rate statistics − Instrument categories
88
A ten-year consumer credit is granted. At time t
0
, the customer and the reporting
agent agree that the interest rate is fixed at 10% p.a. for the first four years and that
a new interest rate level is negotiated at time t
4
.
110
The result of the renegotiation at
time t
4
is a loan fixed at 8% p.a. for the remaining maturity. At time t
0
, the maturity of
this loan is ten years with an initial period of fixation of four years. At time t
4
, the
original maturity of the loan is still ten years and the classification of the loan in the
statistics on outstanding amounts is therefore independent of the new negotiation.
111
.
Hence, the MFI interest rate statistics on outstanding amounts would reflect at time t
4
the decrease from 10% to 8% p.a. in the category maturity over 5 years. The results
of the renegotiation are also captured as a new agreement in MFI interest rate
statistics on new business; the 8% p.a. is recorded for loans with an initial rate
fixation of six years in the category over 5 years initial rate fixation.
Another ten-year consumer credit is granted. This time the customer and the
reporting agent agree at time t
0
that the interest is fixed at 9% p.a. for the first 12
months and that it then automatically adjusts to EURIBOR plus x basis points.
112
This rate is then applied for the next 12 months, after which it will again automatically
adjust to EURIBOR plus z basis points. At time t
0
, the maturity of the loan is ten
years with an initial period of rate fixation of one year. Only the interest rate of 9% for
the first year is considered as new business at time t
0
. The movements in the interest
rate over time are captured in the MFI interest rates on outstanding amounts, but no
further new business occurs.
The situation is different if the initial period of fixation is very short compared with the
whole maturity of the loan and the interest rate offered during this period is
significantly below market conditions. For example, a ten-year loan is granted where
the interest rate is fixed at 2% for the first six months and then automatically adjusts
to either a fixed or a variable interest rate at a level reflecting market conditions. To
qualify for the treatment described here, the 2% must be an introductory offer and
“eye-catcher” to attract new customers, in the sense that the interest rate is
significantly lower than current market conditions, i.e. at least 200 basis points, and
applies for a very short period of the loan. The interest rate that applies after this
initial period of fixation has to be already laid down in the contract agreed at time t
0
. It
can be fixed or variable but in any case it should be at a level reflecting market
conditions. The treatment of loans comprising such introductory offers to attract
customers is the same as that for step-up loans explained in Section 8.1. This
means that the new business statistics cover the loan at time t
0
. The Manual
recommends that in these cases, the interest rate on new business is computed as
an NDER comprising the introductory offer and the interest rate agreed for after the
initial period of fixation. A geometric average of the factors “i+interest rate” could also
be computed, but this would only be an approximation. If a variable interest rate is
110
See also Chart 5.
111
This treatment is in line with MFI balance sheet statistics. This situation of a loan that is renegotiated
before it reaches maturity is different from the deposit with agreed maturity that is renegotiated at
maturity described in Section 5.4.4. In the latter case, the maturing deposit with agreed maturity is
renegotiated and therefore classified as new business, and the maturity of the (new) deposit is
regarded as commencing at the point of the “new business” classification.
112
See also Chart 6.
Manual on MFI interest rate statistics − Instrument categories
89
agreed to apply after the initial period of fixation, for example EURIBOR plus x basis
points, as a convention the value of that variable rate is, for the purpose of MFI
interest rates on new business, determined as at time t
0
. The statistics on
outstanding amounts always reflect the interest rate actually applied by the reporting
agent at the time of calculation of the MFI interest rate, i.e. the 2% during the first
months and then the agreed interest rate at a level reflecting market conditions.
For all step-up and step-down loans, the initial period of fixation is equal to the
maturity of the loan, because a fixed interest rate is agreed for the whole maturity of
the loan at time t
0
when the contract is signed.
7.8 Breakdown by secured loans with collateral and/or
guarantees
113
The provision or the absence of collateral and/or a guarantee can lead to different
levels of interest rates being charged on loans. A breakdown by secured loans
granted to households and non-financial corporations should allow interest rates to
be broken down into more homogeneous risk groups and, as a consequence, into
more homogeneous interest rate categories, appropriately reflecting the associated
credit risk. In addition, the data allow more in-depth analyses of how credit risk
influences the way in which lenders structure their terms and conditions, thus also
allowing conclusions to be drawn about borrowers’ financing costs.
Therefore, indicators 55 to 79 in Table 3 referring to loans to households and non-
financial corporations secured with collateral and/or guarantees are reported
separately for all MIR statistics new business categories except for credit card debt,
revolving loans and overdrafts, and lending for other purposes.
Furthermore, no breakdown by collateral/guarantees is required for the indicators
referring to new business in renegotiated loans.
For the purpose of MFI interest rate statistics, the breakdown of loans according to
collateral/guarantees includes the total amount of new business loans which is
collateralised using the “funded credit protection” technique and/or guaranteed using
the “unfunded credit protection” technique, in such a way that the value of the
collateral and/or guarantee is higher than or equal to the total amount of the loan.
• “Funded credit protection”, as defined in Article 4(1)(58) and Articles 197 to 200
of Regulation (EU) No 575/2013 of the European Parliament and of the Council
of 26 June 2013 on prudential requirements for credit institutions and
investment firms, is a credit risk mitigation technique where the reduction of the
credit risk on the exposure of an institution derives from the right of that
institution, in the event of the default of the counterparty or on the occurrence of
other specified credit events relating to the counterparty, to liquidate, or to
obtain transfer or appropriation of, or to retain certain assets or amounts, or to
113
See also paragraphs 63 and 64 of Annex I to the Regulation.
Manual on MFI interest rate statistics − Instrument categories
90
reduce the amount of the exposure to, or to replace it with, the amount of the
difference between the amount of the exposure and the amount of a claim on
the institution.
• “Unfunded credit protection”, as defined in Article 4(1)(59) and Articles 201, 202
and 203 of Regulation (EU) No 575/2013, is a credit risk mitigation technique
where the reduction of the credit risk on the exposure of an institution derives
from the obligation of a third party to pay an amount in the event of the default
of the borrower or the occurrence of other specified credit events.
If an MFI applies an approach different from the “Standardised Approach” as defined
in Regulation (EU) No 575/2013 for supervisory purposes, it may also apply the
same approach to the reporting of loans included under this breakdown.
Manual on MFI interest rate statistics − Specific instruments and national products
91
8 Specific instruments and national
products
Not all specific national products can be covered by this Manual. Hence, in the same
way as in Part 15 of Annex II to the Guideline, the treatment of the selected products
defined in this chapter should be used as a reference for other national products with
similar characteristics.
8.1 Step-up (step-down) deposits and loans
A step-up (or step-down) deposit or loan is “a deposit or loan with a fixed maturity to
which an interest rate is applied that increases (decreases) from year to year by a
pre-fixed number of percentage points”. As mentioned in Section 7.7.4, the initial
period of rate fixation for a step-up or step-down loan is equal to the maturity of the
loan, because a fixed interest rate is agreed for the whole maturity at time t
0
when
the contract is signed.
Chart 9
Step-up loan
(y-axis: Outstanding amount (left-hand scale); Interest rate (right-hand scale))
t0 t1 t2 t3
MIR (OA) 5% 7% 9% 13%
Weight (OA) 1,000 1,000 1,000 1,000
MIR (NB) 8.24% (8.46)
- - -
Weight (NB) 1,000 - - -
An example of a step-up loan is given in Chart 9, where a four-year consumer credit
is granted, for which the credit institution charges 5% interest in the first year, 7% in
the second, 9% in the third and 13% in the fourth. The new business statistics cover
the step-up loan at time t
0
in the category over 1 and up to 5 years initial rate fixation.
t0 t1 t2 t3
Loan
1,000 1,000 1,000 1,000
Interest rate
5.00% 7.00% 9.00% 13.00%
0%
2%
4%
6%
8%
10%
12%
14%
0
200
400
600
800
1,000
1,200
Manual on MFI interest rate statistics − Specific instruments and national products
92
The interest rate on new business can either be computed as a geometric average of
the factors “1 + interest rate” or as an NDER. The NDER for this example is 8.2446%
and gives the mathematically correct result. The geometric average of the factors “1
+ interest rate” provides an approximation. It is 8.46% for the example and
calculated as follows:
(
)
=
(
1 + 0.05
)(
1 + 0.07
)
+ (1 + 0.09)(1 + 0.13)
1 = 0.0845998 [8]
Unlike the NDER, the geometric average does not take into account the timing of the
interest rate payments. This means that the interest rates computed as a geometric
average are the same for a step-up loan with 5% interest at time t
0
and up to 13% at
t
3
and for a step-down loan with 13% at t
0
and down to 5% at t
3
. Therefore, the
NDER is the method recommended in this and similar cases.
The statistics on outstanding amounts cover from time t
0
to t
3
the interest rates
actually applied by the reporting agent at the time of calculation of the MFI interest
rate, i.e. 5% at time t
0
, 7% at t
1
, 9% at t
2
and 13% at t
3
. These interest rates appear
in the instrument category over 1 and up to 5 years maturity.
8.2 Revolving loans and lines of credit
As stated in Section 7.5.2, revolving loans also include lines of credit, given the
similarities between both instruments. As defined in Regulation ECB/2013/33, a line
of credit is an agreement between a lender and borrower that allows a borrower to
take advances, during a defined period and up to a certain limit, and repay the
advances, usually at his/her discretion, before a defined date. The following two
questions are deemed relevant:
Case 1:
Question:
Considering a line of credit agreed between a customer and an MFI, which amount
should be covered by the MIR statistics as new business in revolving loans and
overdrafts? The maximum amount that the customer can take as advances or only
the amount that has already been withdrawn by the customer and not yet repaid?
Answer:
Only outstanding amounts, i.e. amounts withdrawn and not yet repaid in the context
of a credit line, are covered as new business in revolving loans and, therefore,
reflected in the MFI interest rate statistics in accordance with the Section 5.2
description. Amounts available through a line of credit that have not been withdrawn
or have already been repaid should not be considered in MFI interest rate statistics,
either as new business or as outstanding amounts.
The same treatment is applied to extended credit card credit.
Note that the treatment of these instruments is different to that for loans in tranches
where the whole amount of the loan granted, regardless of the way in which the
customer draws upon it, has to be recorded as new business. The reason for this is
Manual on MFI interest rate statistics − Specific instruments and national products
93
that in a loan in tranches, the whole amount of the loan would normally be drawn, but
this is not the case for revolving loans, lines of credit or extended credit card credit,
where there is a limit that may or may not be reached.
Case 2:
There are lines of credit under which some of the drawdowns are repaid not at the
discretion of the borrower but according to a repayment plan, agreed at the time the
drawdown takes place. Hence the maturity of such a drawdown is known.
Sometimes, these drawdowns have to be used for a specific purpose (e.g. financing
the purchase of certain business equipment) agreed upon between the borrower and
the lender.
By contrast, there are some loans that have all of the features of revolving loans
except that:
• the borrower has to give notice before withdrawing funds and/or prove to the
lender that the funds will be used for a specific purpose agreed upon in the
contract (this requirement may apply to some but not all drawdowns);
• the drawdowns are repaid according to a repayment plan.
Question:
Should the different drawdowns be reported separately in such a way that those that
do not follow the normal rules for a line of credit or a revolving loan are reported
separately under an instrument category different from revolving loans and
overdrafts and all other withdrawals taking place without notice or without any
agreed time of redemption are reported under the category revolving loans and
overdrafts?
Answer:
From a conceptual perspective, it is doubtful that if the contract has the
characteristics of a line of credit or a revolving loan it could be split up and
considered as several loan instruments. There is probably only a single interest rate,
and only the conditions for making some withdrawals envisage several commitments
for the customer. Also, for practical reasons, it needs to be taken into account that
the classification of lines of credit as revolving loans and overdrafts is a special case
in the sense that the calculations of the interest rates on new business and on
outstanding amounts are the same. Therefore, splitting these kinds of loans would
make the calculation of the interest rates very complicated. The case where, as a
consequence of the different agreements on withdrawals, the operation has to be
divided into two, three or more parts is still possible. Therefore, the Manual
recommends the classification of the whole operation in only one instrument
category: revolving loans and overdrafts.
Manual on MFI interest rate statistics − Specific instruments and national products
94
8.3 Umbrella contracts
An “umbrella contract” allows the customer to draw loans on several types of loan
accounts up to a certain maximum amount applying to all loan accounts together. At
the time of the agreement on an umbrella contract, the form the loan will take, the
date at which the loan will be drawn and/or the interest rate are not specified, but a
range of possibilities may be agreed. These characteristics make umbrella contracts
different from the case described in the last part of the previous section.
Question:
Should the initial amount agreed under the umbrella contract at t
0
be reflected in MIR
statistics as new business and/or outstanding amounts?
Answer:
An umbrella contract, at the time it is agreed, is not covered by MFI interest rate
statistics. However, as soon as a loan agreed under an umbrella contract is drawn, it
should be included in the corresponding item in MFI interest rate statistics, both in
new business and outstanding amounts.
8.4 Savings plans for housing loans
Question:
Savings plans for housing loans are long-term low return savings schemes that, after
a certain period of saving, give the saver the right to a housing loan at a discounted
rate. In MFI balance sheet statistics, these savings plans are classified under
deposits with agreed maturity over two years. When the savings plan is transformed
into a loan, this loan is classified as loan to households for house purchases. For the
purpose of MFI interest rate statistics, reporting agents report as new deposit
business the interest rate that is agreed at the time the initial deposit is placed. The
corresponding amount of new business is the amount of money that has been
placed. The increase of the amount on the deposit over time is only covered by the
outstanding amounts. At the time when the deposit is transformed into a loan, this
new loan is recorded as new lending business. The interest rate is the discounted
rate that is being offered by the reporting agent. The weight is the total amount of the
loan that is being granted to the household.
1. Which amount should be covered as new deposit business? Only the first
tranche paid into the account, or the total amount of the deposit which has been
agreed and which will be paid into the account in tranches over time?
2. Which amount on the deposit should be covered by outstanding amounts? The
single tranche, which has been paid into the account in the reference month, or
the cumulated amount of all tranches which have been paid into the account so
far?
Answer:
Analogous to Section 5.4.5, the treatment in new business statistics depends on
whether or not it is known for certain ex ante which amount will be held on the
Manual on MFI interest rate statistics − Specific instruments and national products
95
deposit account at maturity. In general, savings plans for house purchases are highly
flexible products and it is only known ex post what amount has been accumulated on
the account and can be transformed into a loan or paid out to the customer. Hence,
in general, the amount of new business is the initial amount of money that has been
placed, i.e. the first tranche paid into the account. The corresponding interest rate for
new business is the deposit interest rate that has been agreed for when the initial
deposit is placed.
If it is known ex ante for certain what exact amount will be held on the account at
maturity, then the total amount on the deposit should be recorded as new business.
In both cases, the outstanding amount is the stock of all deposits placed by the
customer on the account. At the time of reporting, the MFI interest rate statistics on
outstanding amounts therefore reflect the accumulated amount of all tranches that
have been paid into the account so far and the corresponding interest rate.
8.5 Savings plans with a fidelity and/or growth premium
Savings deposits with a basic interest rate plus a fidelity and/or growth premium may
exist. At the time the deposit is placed, it is not certain whether or not the premium
will be paid. The payment depends on the future unknown saving attitude of the
household or non-financial corporation.
Question:
Let us consider a five-year savings deposit with regular
savings allowed and yearly interest payments. The
minimum interest rate paid on this deposit is 3% and
the premium depends on the total deposit amount, as
presented in Chart 10.
The initial deposit amounts to EUR 2,500.
What interest rate should be considered in the MIR
statistics on new business?
Answer:
The new business statistics cover the deposit at time t
0
in the category deposits with agreed maturity over 2
years. The convention is that such fidelity or growth
premiums, as they depend on the future unknown
saving attitude of the customer, are not included in the
AAR on new business. Therefore, the interest rate on
new business is 3.30%, calculated as
3.00% ×
1,000
2,500
+ 3.50% ×
1,500
2,500
The AAR on outstanding amounts always covers the
Chart 10
Saving deposit with a growth premium
Outstanding amount
(left-hand scale: interest rate premium (percentage), right-hand scale: deposit amount
(EUR)
Deposit amount Interest rate premium
Up to EUR 1,000
0%
Over EUR 1,000 and up to EUR 5,000
+0.50%
Over EUR 5,000 and up to EUR 10,000
+1.00%
Over EUR 10,000
+1.50%
0
2,000
4,000
6,000
8,000
10,000
12,000
14,000
0.0%
0.5%
1.0%
1.5%
2.0%
2.5%
3.0%
3.5%
4.0%
4.5%
5.0%
1 2 3 4
loan
interest rate
Manual on MFI interest rate statistics − Specific instruments and national products
96
rates applied by the reporting agent at the time of calculation of MFI interest rates.
Hence, if such a growth premium is granted by the reporting agent, this is reflected in
the statistics on outstanding amounts.
8.6 Interest rate on zero coupon bond-like savings bonds
Question:
In some countries, certain non-marketable savings bonds are classified as deposits
in MFI balance sheet statistics. These savings bonds are discounted zero coupon
bond-like products involving only two cash flows: the initial placement of the deposit
at a discounted value and the final redemption payment at nominal value which
includes the interest. Which amounts should be considered as new business and
outstanding amounts for the purpose of calculating interest rates on savings bonds?
Should the initial value of the deposit placed in t
0
be reflected both as new business
and outstanding amounts, with any accrued interest over the lifetime of the savings
bonds being excluded? This treatment would be in line with the MFI balance sheet
statistics, which requires accrued interest payable on deposits to be classified
separately as remaining liabilities.
Answer:
If, for example, the discounted value of the savings bond is EUR 80 and the nominal
value at which it is redeemed after two years is EUR 100, then at time t
0
the amount
of EUR 80 is recorded as new business. The interest rate on new business is then
11.8034%. It may be calculated as an AAR like in Equation 9 or as NDER shown in
Table 9.
Table 9
Savings bond with agreed maturity of two years, interest rate payment at end of
second year
t
Outstanding
deposit
Interest
rate p.a.
Interest
payments
Repayments
of principal
Cash
flow
Discount
factor =
(1+NDER)^
(-t/365)
Present
value of
cash flow NDER
(1)
(2) (3) (4) (5) (6) (7) (8) (9) (10)
0 80 -80
1Y
365 80 0% 0 0 0 0.894 0 11.8034%
2Y
730 0 25% 20 80 100 0.8 80 11.8034%
Sum 20 80 RHS: 80
=
1 +
.
.
1 = 1 +
.
.
.
1 = 0.118034 [9]
The amount of EUR 80 is also the outstanding amount from time t
0
until maturity,
including the day of maturity where the bond is redeemed at EUR 100. The reason is
Manual on MFI interest rate statistics − Specific instruments and national products
97
that for MFI interest rate statistics, the EUR 20 difference between the issuance and
redemption value is interpreted as the interest payment for the two years and not as
principal. The interest of EUR 20 is known ex ante and interpreted as accrued over
the maturity of the bond, in this example over two years. Hence, 11.8034% is
reflected as the interest rate on outstanding amounts for the whole period. In this
way, the savings bond is treated just like a deposit with an agreed maturity of two
years where the interest is fixed at 12.5% p.a. and paid at maturity.
8.7 Splitting of loans into two parts
Question:
A customer takes out a mortgage loan of EUR 50,000 for 20 years at a variable
interest rate. It is agreed with the credit institution and laid down in the contract that
at any point in time the customer may split the mortgage. For example, after three
years the customer may choose to split it into two parts: EUR 20,000 at a fixed rate
and EUR 30,000 at a variable rate. How would this split be treated in MFI interest
rate statistics?
114
Answer:
The possibility of the split is already agreed in the contract for the mortgage loan of
EUR 50,000. The split does not constitute new business if the same contract already
determines the terms and conditions of the two loans after a possible split, i.e. the
amount of money that would be lent at variable and at fixed interest rate as well as
the interest rates applied to these amounts. In this case, the customer is free to
choose the timing of the split without engaging in renegotiation of the contract. The
change in the interest rates from fully variable to partly fixed rate after the split is
then only reflected in the statistics on outstanding amounts.
If the split of the mortgage loan into two parts implies new negotiation of the terms
and conditions of the contract, for example the amount lent at variable and at fixed
rate or the level of these interest rates, then each of the loans after the split, i.e. the
mortgage loan of EUR 20,000 and of EUR 30,000, constitutes new business.
8.8 Option to convert a deposit into equity shares
Question:
A deposit with agreed maturity of one year is placed at a higher than market interest
rate, currently for example at 5%, where at maturity the credit institution (not the
customer) has the option of converting the deposit into a number of shares of a
specific company. If at maturity the share price is high, the credit institution will not
exercise the option, but pay the 5% interest and return the deposit. If at maturity the
share price is low, then the credit institution will also pay the 5% interest and
exercise the option. In this case, at the time of the conversion into shares, the
114
If part of the loan split is connected with debt restructuring (with rates below market conditions) that
part is excluded from the MIR statistics.
Manual on MFI interest rate statistics − Specific instruments and national products
98
customer will lose part of the deposit because the value of the shares has fallen. The
interest rate is high because it includes a risk premium for the possibility of losing
part of the capital. This instrument is also referred to as reverse convertible. How
should this instrument be treated in MFI interest rate statistics?
Answer:
As there is no capital certainty for the deposit, it is considered that this product is
closer to a financial instrument like securities than to a deposit. Therefore, it is not
covered by MFI interest rate statistics on new business. The AAR or the NDER on
outstanding amounts always covers the rates applied on the deposits by the
reporting agent at the time of the calculation of MFI interest rates. Hence, if at
maturity, the credit institution opts not to convert the deposit into shares and, as a
consequence, pays the 5% agreed on the deposit, this interest rate paid will be
reported in MFI interest rate statistics, whenever the method followed by the
reporting agent is the implicit rate referring to the average of the month to calculate
the interest rate on outstanding amounts. Note that the recording of the interest rate
on outstanding amounts takes place only when the operation matures, not before,
because before maturity no interest is paid. If, however, the credit institution
exercises the option and, at maturity, converts the deposit into shares, the amount
converted is no longer a deposit and hence is not included in the MFI interest rate
statistics. Specifically, the interest rates on new business allow the monitoring of the
effect of monetary policy decisions on the rates applied by MFIs to loans to
households and non-financial corporations; therefore, the more the new business
interest rates on loans and deposits reflect “true” market conditions, the better. In the
example above, neither the 5% interest rate (where the option is not exercised) nor
any other agreement at the time the contract is signed regarding the probable yield
of the operation would reflect the market conditions for the deposit.
Another question relates to the different recording of such a deposit at the time it is
agreed in MIR statistics (where there is no recording) and in balance sheet item
(BSI) statistics (which do record the deposit). Two considerations are of interest
here: (i) although it is advisable that the recording of instruments in the BSI statistics
and in MIR statistics coincide as much as possible, achieving this aim should not go
beyond a reasonable limit, taking into account what MIR statistics try to capture; and
(ii) when a credit institution records this kind of deposit on its balance sheet, it has to
record any derivative at its market price, according to International Financial
Reporting Standards (IFRS). The Regulation does not currently include any such a
provision regarding the calculation of interest rates.
8.9 Interest rate linked to a share price
Question:
A deposit is agreed where the credit institution does not guarantee any interest rate
but the return on the deposit is entirely linked to the stock market price of shares in a
specific company. If, for example, the share price rises by 10%, the credit institution
will pay 20% interest on the deposit to the customer. However, if the share price falls
Manual on MFI interest rate statistics − Specific instruments and national products
99
by 30%, then the customer loses 30% of the money placed as a deposit. How should
this instrument be treated in MFI interest rate statistics?
Answer:
As in Section 8.8, this instrument is not covered in new business in MFI interest rate
statistics, because there is no capital certainty for the deposit.
115
The AAR or the
NDER on outstanding amounts covers the interest rate applied by the reporting
agent at the time of the calculation of MFI interest rates, i.e. it should be based on
the stock market price of the shares of the specific company at the time of
calculation.
8.10 Treatment of a deposit comprising two components
116
Question:
A customer and a credit institution agree on a deposit of EUR 10,000 comprising two
parts. These two parts are inextricably linked and cannot be placed separately. The
first part of EUR 6,000 is placed with an agreed maturity of six months at a (higher
than market) fixed interest rate of 15% paid at monthly frequency. The second part of
EUR 4,000 is placed with an agreed maturity of three years and the return is linked
to a stock exchange index with a guaranteed minimum return of 0%. How should this
product be treated in MFI interest rate statistics? Would the treatment change if the
0% interest, i.e. the minimum return, on the second part of the deposit were not
guaranteed, implying the possibility of a capital loss for the investor?
Answer:
Paragraph 7 of Part 12d of the Guideline reads as follows: “Deposits may be offered
comprising two components: a deposit with agreed maturity to which a fixed interest
rate is being applied and an embedded derivative with a return that is linked to the
performance of a defined stock exchange index or a bilateral exchange rate, subject
to a minimum guaranteed return of 0%. The maturity of both components may be the
same or may differ. The AAR on new business covers the interest rate for the deposit
with agreed maturity, as it reflects the agreement between the depositor and the
reporting agent and it is known when the money is being placed. The return on the
other component of the deposit linked to the performance of a stock exchange index
or a bilateral exchange rate is only known ex post when the product matures and
therefore cannot be covered by the new business rate. Hence, only the guaranteed
minimum return (usually 0%) should be captured. The AAR on outstanding amounts
always covers the interest rate applied by the reporting agent at the time of the
calculation of MFI interest rates. Until the day of maturity, the rate on the deposit with
115
This is different to Section 5.4.13 where a floor of 2% is agreed as a minimum return and to
Section 8.10 where a minimum return of 0% is agreed for the second part of the deposit.
116
This section does not cover financial products where the first component is a deposit and the second
component an investment in mutual fund shares/units or capitalisation products offered by insurance
companies. For these financial products only the amount and the interest rate relating to the first
component, i.e. the deposit, is covered by MFI interest rate statistics. The amount and the interest rate
of the second component are outside the scope of MFI balance sheet statistics and are not covered by
MFI interest rate statistics. In general, for these financial products the interest rate offered for the first
component, i.e. the deposit, is close to market interest rates.
Manual on MFI interest rate statistics − Specific instruments and national products
100
agreed maturity is captured as well as the guaranteed minimum return on the deposit
containing the embedded derivative. Only at maturity do the MFI interest rates on
outstanding amounts reflect the AAR that is paid by the reporting agent.”
Hence, according to the Regulation, both parts of the deposit are covered by MFI
interest rate statistics, i.e. the first part of EUR 6,000 at 15% for a term of six months,
and the second part of EUR 4,000 at the guaranteed 0% for three years. However,
the Regulation is silent about the precise treatment of the two parts, i.e. if they
should be treated as two separate deposits or as a single deposit for the purpose of
calculating MFI interest rates. If both parts were at the same maturity and would
hence fit into the same instrument category for MFI interest rate statistics, no
question would arise, as it would lead to the same result whether both parts were
treated separately or together.
If the two parts have different maturities, as in the example, the two parts of the
deposit are classified under two different instrument categories. The two parts
should, however, be looked at together, because together they form one contract and
could not be placed independently of each other, and hence the interest rates on the
two parts need to be seen as a “package”. For such products, either the NDER is
calculated for the two components or, as an approximation, a weighted average of
the AAR is calculated, although the application of the NDER is more suitable for this
kind of product.
For example, the calculation of the AAR is given in Equation 10 and the calculation
of the NDER is shown in Table 10.
1 +
.
1
+ 0
= 3.2151% [10]
Manual on MFI interest rate statistics − Specific instruments and national products
101
Table 10
Deposit with two components: EUR 6,000 at 15% for 6 months and EUR 4,000 at 0% for 36 months
At the time of agreement on the contract, the compiled average interest rate is
reflected both as new business at six months’ maturity for an amount of EUR 6,000
t
Outstanding
deposit
Interest
rate p.a.
Outstanding
deposit
Interest
rate p.a.
Interest
payments
Repayments of
principal Cash flow
Discount factor
= (1+NDER)^
(-t/365)
Present
value of
cash flow NDER
(1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12)
0 6,000 4,000 -10,000
1YM1
30.42 6,000 15% 4,000 0% 75 75 0.997 74.81 3.1370%
1YM2
60.83 6,000 15% 4,000 0% 75 75 0.995 74.61 3.1370%
1YM3
91.25 6,000 15% 4,000 0% 75 75 0.992 74.42 3.1370%
1YM4
121.67 6,000 15% 4,000 0% 75 75 0.990 74.23 3.1370%
1YM5
152.08 6,000 15% 4,000 0% 75 75 0.987 74.04 3.1370%
1YM6
182.50 0 15% 4,000 0% 75 6,000 6,075 0.985 5,981.9 3.1370%
1YM7
212.92 0 4,000 0% 0 0 0.982 0 3.1370%
1YM8
243.33 0 4,000 0% 0 0 0.980 0 3.1370%
1YM9
273.75 0 4,000 0% 0 0 0.977 0 3.1370%
1YM10
304.17 0 4,000 0% 0 0 0.975 0 3.1370%
1YM11
334.58 0 4,000 0% 0 0 0.972 0 3.1370%
1YM12
365 0 4,000 0% 0 0 0.970 0 3.1370%
2YM1
395.42 0 4,000 0% 0 0 0.967 0 3.1370%
2YM2
425.83 0 4,000 0% 0 0 0.965 0 3.1370%
2YM3
456.25 0 4,000 0% 0 0 0.962 0 3.1370%
2YM4
486.67 0 4,000 0% 0 0 0.960 0 3.1370%
2YM5
517.08 0 4,000 0% 0 0 0.957 0 3.1370%
2YM6
547.50 0 4,000 0% 0 0 0.955 0 3.1370%
2YM7
577.92 0 4,000 0% 0 0 0.952 0 3.1370%
2YM8
608.33 0 4,000 0% 0 0 0.950 0 3.1370%
2YM9
638.75 0 4,000 0% 0 0 0.947 0 3.1370%
2YM10
669.17 0 4,000 0% 0 0 0.945 0 3.1370%
2YM11
699.58 0 4,000 0% 0 0 0.943 0 3.1370%
2YM12
730 0 4,000 0% 0 0 0.940 0 3.1370%
3YM1
760.42 0 4,000 0% 0 0 0.938 0 3.1370%
3YM2
790.83 0 4,000 0% 0 0 0.935 0 3.1370%
3YM3
821.25 0 4,000 0% 0 0 0.933 0 3.1370%
3YM4
851.67 0 4,000 0% 0 0 0.930 0 3.1370%
3YM5
882.08 0 4,000 0% 0 0 0.928 0 3.1370%
3YM6
912.50 0 4,000 0% 0 0 0.926 0 3.1370%
3YM7
942.92 0 4,000 0% 0 0 0.923 0 3.1370%
3YM8
973.33 0 4,000 0% 0 0 0.921 0 3.1370%
3YM9
1,003.75 0 4,000 0% 0 0 0.919 0 3.1370%
3YM10
1,034.17 0 4,000 0% 0 0 0.916 0 3.1370%
3YM11
1,064.58 0 4,000 0% 0 0 0.914 0 3.1370%
3YM12
1,095 0 0 0% 0 4,000 4,000 0.911 3,645.99 3.1370%
Sum 450 10,000 RHS: 10,000
Manual on MFI interest rate statistics − Specific instruments and national products
102
and as new business at three years’ maturity for an amount of EUR 4,000, with the
interest rate for both instruments being the resulting NDER or the AAR, as calculated
in Equation 10. The MFI interest rate statistics on outstanding amounts, at the time
of the placement of the deposit, also reflect the average interest rate based on the
minimum guaranteed return on the two components of the deposit. Hence, in this
example the interest rate on outstanding amounts both at six months maturity for an
amount of EUR 6,000 and at three years maturity for an amount of EUR 4,000 reflect
the average interest rate as compiled above until the respective part of the deposit
reaches maturity. A higher than 0% interest rate that is paid by the reporting agent at
maturity of the second part of the deposit is only reflected in the interest rate on
outstanding amounts for this instrument category at maturity, whenever the method
applied is the implicit rate referring to the average of the month. As always, the
interest rate needs to be annualised, i.e. it has to be taken into account that the
interest paid at maturity on the EUR 4,000 is an one-off payment after 36 months.
If there is no capital certainty for the second component of the deposit, the treatment
in MFI interest rate statistics for new business and for outstanding amounts is the
same as in Section 8.9.
8.11 Pension savings accounts
Deposits with a maturity of over two years as defined in Part 2 of Annex II to
Regulation ECB/2013/33 may contain pension savings accounts. The main part of
pension savings accounts may be placed in securities and the interest rate on the
accounts then depends on the yield of the underlying securities. The remaining part
of pension savings accounts may be held in cash and the interest rate should be
determined by the MFI in the same way as for other deposits.
At the time when the deposit is placed, the total return to the household from the
pension savings account is not known and could also be negative. In addition, at the
time the deposit is placed, an interest rate is agreed between the household and the
MFI which applies only to the deposit part (not to the part invested in securities).
Hence, only the deposit part that is not invested in securities is covered by MFI
interest rate statistics. The AAR on new business that is reported is the rate agreed
between the household and the reporting agent for the deposit part at the time the
deposit is placed. The AAR on outstanding amounts is the rate applied by the
reporting agent to the deposit part of the pension savings accounts at the time of
calculation of the MFI interest rate.
8.12 Purchase of mortgage loans by a credit institution
Question:
A credit institution acquires the economic but not legal ownership of mortgages that
are granted to households by a (related) life insurance corporation. The mortgages in
question have different starting dates and are transferred normally in small,
sometimes in large, portions from the life insurance corporation to the credit
Manual on MFI interest rate statistics − Specific instruments and national products
103
institution. Originally, the life insurance corporation grants the mortgage loans. It is in
fact selling mortgages through its network to households and these mortgages are
funded or refinanced by the credit institution. The amount paid by the credit
institution to the life insurance corporation is the going market rate at the time of the
purchase of the mortgage loans and includes, among others, a fee for handling the
mortgage administration and collection of payments by the life insurance corporation.
The customer pays the interest for the mortgage loan directly to the insurance
company and not to the credit institution. In MFI balance sheet statistics, the credit
institution records these (purchased) mortgage loans as mortgage loans extended to
households. Is this type of transfer of mortgage loans new business? If it is, which
interest rate should be reported for new business and outstanding amounts?
Answer:
The key for the treatment of such loans in MFI interest rate statistics is their
recording in MFI balance sheet statistics. The mortgage loans are originally granted
by the insurance company and therefore initially appear on the books of the
insurance company and not of the credit institution. There is in fact no contractual
relationship between the household and the credit institution, only between the
household and the insurance company. The latter is beyond the scope of MFI
interest rate statistics on new business, as the insurance company is obviously not a
credit institution and therefore not included in the reference reporting population
117
of
the Regulation.
Once the credit institution purchases the mortgage loans, however, a loan to
households for house purchases is recorded in the books of the credit institution. At
this point in time, the interest rate actually paid by the household for the mortgage
loan is recorded in the MFI interest rate statistics on outstanding amounts. The
purchase of the loans by the credit institution does not constitute new business
because there is no active involvement of the household in the purchase or transfer
of the loans from the insurance company to the credit institution. As these mortgage
loans are covered in MFI balance sheet statistics under loans to households for
house purchases, they are also included in the weighting information which is used
to calculate the euro area interest rate for outstanding housing loans to households.
8.13 Securitisation of mortgage loans by a credit institution
Question:
A credit institution transfers the economic but not legal ownership of mortgage loans
that it has granted to households to a financial vehicle corporation. Analogous to
Section 8.12, the mortgages in question have different starting dates and are
transferred normally in small, sometimes in large portions from the credit institution
to the financial vehicle corporation. Originally, the credit institution grants the
mortgage loans and records them as loans to households for house purchases in
MFI balance sheet statistics. Once the mortgage loans are sold to the financial
117
Defined in Chapter 3 and further discussed in Section 12.1.
Manual on MFI interest rate statistics − Specific instruments and national products
104
vehicle corporation, the loans no longer appear in the MFI balance sheet of the credit
institution. How should this type of securitisation be treated in MFI interest rate
statistics?
Answer:
As in Section 8.12, the key for the treatment of securitisation in MFI interest rate
statistics is the recording in MFI balance sheet statistics. The mortgage loans are
originally granted by the credit institution to the household and constitute new
business at the time of agreement on the contracts. They are hence reflected as
loans to households for house purchases in the MFI interest rate statistics on new
business and also appear at that time in MFI interest rate statistics on outstanding
amounts.
However, once the credit institution sells the mortgage loans to the financial vehicle
corporation, along with all the rights and returns associated with the loans, the loans
disappear from the books of the credit institution. At that point in time, they are
removed from the MFI balance sheet of this institution and also no longer covered in
MFI interest rate statistics on outstanding amounts. The transfer of the loans to the
financial vehicle corporation does not constitute new business for MFI interest rate
statistics.
8.14 Payday loans
Payday loans are very short-term micro loans that credit institutions and other
financial institutions offer to their customers, usually households, with the purpose of
bridging the gap between one salary and another. The credit institutions charge very
high fees. At maturity, the customer pays the fee and repays the loan. A rollover of
the principal is not allowed.
Question:
As the charge that the credit institution applies to these loans is a fee, should they be
included in MFI interest rate statistics? And, if so, how should these loans be
classified? Another question is, assuming that these loans are included in MFI
interest rate statistics, is it necessary to apply special recording rules, given the
sizeable distortion of credit aggregates in some countries caused by such loans?
Answer:
First, the characteristics of the high fee charged need to be analysed. If the fee is
very high compared with the fees that credit institutions charge on other loan
instruments, it is doubtful that it can be considered as such. However, if that charge
is just a fee, the loans would be granted at a 0% interest rate, which may not make
sense. Therefore, in these operations, the fee has to be considered as an interest
rate, unless the underlying “true” fee can be disentangled from the interest rate. This,
together with the other characteristics of payday loans, lead to the conclusion that
they must be recorded both in the NDER/AAR and the APRC figures, with the most
appropriate classification being loans for consumption, given the fact that they are
granted to bridge the gap between one salary and another and that, very probably,
Manual on MFI interest rate statistics − Specific instruments and national products
105
only a very urgent need for these loans, like consumption, would justify paying such
high fees. In addition, the low amounts usually involved in such loans support such a
classification.
As the inclusion of these loans in MFI interest rate statistics could cause a significant
distortion in the aggregates, a possible solution would be to report the payday loans
under a separate category, but this option lacks legal support to oblige the reporting
agents to differentiate this type of loan from the rest. The evidence gathered to date
shows that the distortion is undoubtedly reflected in the statistics of countries where
the weight of payday loans in the total consumption loan amount is very high. In
other countries, such a distortion is not apparent, either because the institutions
granting this type of loan are classified as other financial institutions (OFIs) and are
thus not subject to reporting for MFI interest rate statistics, or because the amounts
involved are very low compared with the rest of loans or because these types of
loans do not exist. As this phenomenon is relatively recent, this Manual recommends
the following provisional solution: in countries where payday loans have a significant
impact on interest rates on consumer credit, they should be excluded from the
current MIR statistical breakdowns and included as a separate category. It would be
the responsibility of the NCBs in these countries to collect the separate information
on these loans from the reporting agents on a best-efforts basis.
8.15 Restricted deposits
Restricted deposits are deposits that may be transferred from current or giro
accounts only for a specific purpose. Most restricted deposits have characteristics of
transaction accounts for specific purposes and an interest rate around 0%. Some
others are similar to deposits with agreed maturity with rates close to the rates
applied to the normal ones. The relative importance of the characteristics of the
restricted deposits depends on the country. The restriction on these deposits is on
the customer, who cannot withdraw them for other purposes than those specified
when they are deposited. In addition, the specific purpose of the restriction varies
from country to country.
Question:
Should the restricted deposit be recorded in MFI interest rate statistics and, if so, in
which instrument category?
Answer:
According to Regulation ECB/2013/33 (Annex I, Part I, Section 1), deposits are
amounts which are owed to creditors by reporting agents and which are not
negotiable instruments and, therefore, not marketable. They combine the features of
transferability, convertibility and certainty (in terms of their nominal value).
118
Furthermore, the Regulation, like the European System of Accounts 2010,
distinguishes between transferable and non-transferable deposits, but nothing is said
specifically about restricted deposits; it is only mentioned that, regarding the
118
See also Section 2.1.2 of the Manual on MFI balance sheet statistics.
Manual on MFI interest rate statistics − Specific instruments and national products
106
category deposits with agreed maturity, “This item also includes administratively
regulated savings deposits where the maturity related criterion is not relevant
(classified in the maturity band “over two years”).” Nonetheless, as mentioned above,
restricted deposits comprise, in some countries, a wide variety of deposits and the
rules for their use are very different across countries. Therefore, when deciding on
the classification of restricted deposits, this Manual recommends taking into account
the following issues:
• Restricted deposits can be classified either as remaining liabilities or as
deposits in BSI statistics. The dividing line depends on whether the credit
institution can invest the funds raised without restrictions or not. If restricted,
they should be classified as remaining liabilities, otherwise as deposits.
• If classified as deposits, it has yet to be decided if their inclusion would be as
deposits with agreed maturity over two years if the deposits qualify as
administratively regulated savings deposits or as non-transferable deposits
within the category overnight deposits as these deposits usually do not have a
predetermined maturity.
• The main problem with the classification of these special deposits as with
agreed maturity over two years is that this would imply their exclusion from M3
in the euro area, should the country be a Member State, or from the monetary
aggregates of the country, if this country follows the same definition of broad
money as the one used for the euro area.
All in all, it seems that the classification of restricted deposits as overnight
non-transferable deposits better matches the role they play in the economy. In any
case, the classification as deposits implies that the credit institution can invest the
funds raised without any restriction; if this is not the case, they should be classified
as remaining liabilities.
8.16 Cash pooling
Cash pooling
119
is a bank service that allows corporates to externalise intragroup
cash management in order to manage their global liquidity effectively and with lower
costs. Recently, it has become increasingly relevant in western and northern
European countries. There are three types of cash pooling:
• A single legal account cash pool consists of transaction accounts, which are
often directly used by the individual entities and the parent company for their
daily operations, and a top account, which consolidates the funds of the group.
Under such a scheme, only the top account constitutes an obligation of the
bank vis-à-vis the beneficiary and consequently the interest rate is calculated
on the top account only.
119
See also “The statistical classification of cash pooling activities”, Statistics Paper Series, No 16,
European Central Bank, July 2016 Statistics Paper Series.
Manual on MFI interest rate statistics − Specific instruments and national products
107
• In a physical cash pool, all accounts that are part of the cash pool represent a
resource or an obligation of the bank vis-à-vis the entities and the interest rate
is calculated on all accounts separately.
• A notional cash pool comprises accounts that represent a resource or an
obligation of the bank vis-à-vis the entities as in a physical cash pool, but under
this scheme no liquidity transfers take place between the individual accounts.
The pooling is done on a notional top account created by the bank that virtually
consolidates the positions of the pool participants, but does not represent itself
a resource or an obligation of the bank. The interest rate is calculated on the
net amount recorded on the notional account and is then distributed (at market
conditions, but without margins) among the participants.
Question:
How should cash pooling be treated in MFI interest rate statistics?
Answer:
The overall reporting in MIR statistics should follow the practices adopted in MFI BSI
statistics, namely the gross recording of positions within a cash pool arrangement.
8.17 Treatment of factoring and the calculation of the
AAR/NDER for very short-term loans
Factoring is a financial operation where a firm (factoring client) sells its invoices to a
factoring company (which might be an MFI). In the case of recourse factoring, the
factoring company purchases the invoices at a price that is less than the face value
of the invoices. The discount is kept as collateral to cover any risks associated with
the operation. After the customers have paid the invoices, the factoring company
transmits the receivables net of the advanced cash (and the applicable fees and
interest) to the factoring client. In the case of non-recourse factoring, the fees and
interest are charged immediately to the factoring client, which receives the full face
value of the invoices net of these charges. In this type of factoring, the factoring
company fully assumes the risk of non-payment by the customer; therefore, the fees
and interest are higher. In particular, if the interest rate is calculated based on the
difference between the face value of the invoices and the price paid by the factoring
company, the annualised rate is very high, which has a visible impact on national
aggregated interest rates. The annualised rate does not however reflect the true cost
of the loan to the company, as these loans are typically of very short maturity (from a
few days to a few weeks). Thus, it could be argued that the annualised rate of these
loans should not be included in the national MIR aggregates.
Question:
How should factoring, and in particular non-recourse factoring, be treated in MIR
statistics and how should the AAR/NDER be recorded for very short-term contracts?
Manual on MFI interest rate statistics − Specific instruments and national products
108
Answer:
The case is very similar to payday loans
120
, which are excluded from MIR
aggregates and reported separately if they have an impact on national aggregates.
Nevertheless, as the impact of factoring and other short-term loans is clearly smaller,
it is recommended to include these loans in the calculation of national aggregated
interest rates and to annualise the interest rates charged using the standard
formulae.
8.18 Treatment of student loans without a definite maturity
In the case of student loans without a definite maturity, the full loan amount is not
known at the start of the contract, but instead the maximum amount per year is
limited to a certain predefined amount. In addition, the number of times the amount
can be withdrawn depends on the length of the studies. The interest rate is fixed for
the whole time. In the first year, the student submits the loan application to the bank
and signs the loan agreement. In the following years, there is no need to reapply if it
has been noted in the loan application that in the subsequent academic years the
student would like the loan amounts to be automatically transferred to his/her
account, assuming that all of the conditions have been met. The student can also
refuse any new loan instalments, in which case the additional amounts are not
automatically paid out. After this, the student needs to submit a new loan application
to the bank (no new agreement is signed) for any new instalments. The loan is
repaid on the basis of the schedule prepared by the bank and starts no later than 12
months after the student has graduated or left the educational institution for any
other reason. The repayment period is twice as long as the duration of the studies
according to the curriculum, but no more than 20 years.
Question:
Should the amount withdrawn each year be reported as new business only or should
it be reported also as a renegotiation? What amount has to be reported as a
renegotiation?
Answer:
The change in the loan amount is clearly a change in the terms and conditions of the
contract. Thus, the amount withdrawn each year should be reported as a
renegotiation.
8.19 Treatment of convertible deposits
At the beginning of the contract, a deposit in euro with an agreed maturity is placed
and an interest rate is agreed to be paid at maturity. However, at maturity, depending
on whether, for example, the EUR/USD exchange rate exceeds a strike rate agreed
in the contract or not, the payback (including the interest) is done in euro or US
120
See also Section 8.14.
Manual on MFI interest rate statistics − Specific instruments and national products
109
dollars. Alternatively, the interest rate might be linked to an external index, such as
the EURIBOR.
Question:
How should convertible deposits be recorded in MFI interest rate statistics?
Answer:
Regarding the treatment of the new business in convertible deposits, there is no
capital certainty for the deposits; it is thus considered that these products are more
like financial instruments such as securities than deposits. Therefore, they are not
covered by MFI interest rate statistics on new business. If at maturity the interest is
paid in euro, it should be reported under outstanding amounts in MIR statistics.
8.20 Leasing contracts
Leasing contracts are financial contracts between a lessor (owner) and lessee (user)
that give the lessee rights to use property owned by the lessor for a predefined
period.
Question:
How should leasing, and in particular leasing contracts purchased from a leasing
company, be treated in MIR statistics?
Answer:
Leasing contracts should be reported under loans in MIR statistics and new leasing
contracts should thus be included in new business. Contracts that are purchased
from a non-MFI leasing company by a bank should be reported under new business,
but not as renegotiated loans as these loans are transferred from outside the MFI
sector
121
. The reasoning behind this is that while recording the transferred loans
under renegotiations would be correct for the analysis of the credit market as a
whole, within the MFI sector these loans are “pure new loans”. Nevertheless, if the
analysis of the credit market refers to the whole economy, the exclusion of these
operations from renegotiations can be a significant drawback, should the figures of
these operations be relevant.
8.21 Recording of syndicated loans
Syndicated loans are loans granted by a group of lenders (referred to as a syndicate)
that together provide funds for a single borrower. The loan is usually arranged and
coordinated by one institution (often called the “lead manager”). Each participant,
including the lead manager, records its share of the loan vis-à-vis the borrower, i.e.
not vis-à-vis the lead manager, on its balance sheet. The recording of syndicated
loans in MIR statistics should however be clarified, in particular for the recording in
new business.
121
See also Section 5.5.4.
Manual on MFI interest rate statistics − Specific instruments and national products
110
For example, three banks (A, B and C) grant a loan to an NFC of EUR 1.5 million.
Bank A grants 1.1 million, bank B grants 0.3 million and bank C grants 0.1 million.
Question:
What is the new business volume and interest rate recorded by bank B? In which
category is this loan recorded?
Answer:
The new business volume reported by bank B is EUR 300,000, as this is the amount
that has been granted and which is in the contract for bank B. However, different
practices exist regarding the category in which the new business volume and
corresponding rate are reported: (i) the loan could be reported in the category of new
loans to NFCs of an amount above EUR 1 million, that is to say according to the size
of the whole syndicated loan; or (ii) in the category of loans to NFCs of an amount
above EUR 0.25 million and below EUR 1 million, that is to say according to the
amount actually granted by bank B.
The terms and conditions of the syndicated loan, including the interest rate, are
determined on the basis of the credit risk associated with this loan, which depends
on the borrower and on the full amount of the loan, not on the tranche that a
participating bank grants. It is standard practice that the full amount of the loan has a
single interest rate without any differentiation among participating banks. Therefore,
in this case, the interest rate reported by each of the banks will be the same and is
expected to be, in terms of its level, close to the rates offered for loans above EUR 1
million. For this reason and because of the nature of syndicated loans, it could make
sense for each bank to report the amount granted (for bank B, EUR 300,000) in the
category of the syndicated loan (above EUR 1 million).
In the example given, if banks were to categorise the loan according to the amount
granted, bank C would record this loan in the category up to EUR 0.25 million, which
might distort the national aggregate for that category given how the conditions of this
loan are determined. In addition, if the size of the loan is used as a proxy for the size
of the firm, one would conclude that a loan of EUR 0.1 million was granted to an
SME. Although this Manual recommends categorising the loan according to the size
of the syndicated loan, it is acknowledged that it may be more difficult for reporting
agents to identify the conditions of such loans compared with more standard loans
and therefore, although not optimal, the practice of categorising the syndicated loans
according to the amount granted is accepted for practical reasons.
Manual on MFI interest rate statistics − Aggregation of the data and reporting
obligations
111
9 Aggregation of the data and reporting
obligations
122
9.1 Overview
To derive euro area aggregates for each of the instrument categories in Tables 1 to 9
of the Appendix, three levels of aggregation are necessary: the first at the level of the
reporting agents, the second at that of the NCBs and the third at the ECB level. The
three levels are illustrated in Figure 3.
Figure 3
Aggregation of the data and reporting obligations
The starting point for the compilation of MFI interest rate statistics is the individual
deposit and loan products within the reporting agents. As a first step, each reporting
agent collects data on all relevant products, aggregates them as appropriate and
sends them to the NCB of the Member State in which it is resident. In general and as
indicated in Figure 3,
123
each reporting agent provides for each instrument category
one average interest rate referring to the reporting agent as a whole. In the case of
new business it also provides the volume of the new deposits and loans.
As a second step in aggregation, each NCB compiles, for each instrument category,
a weighted average interest rate for each instrument category referring to the
Member State as a whole and reports it to the ECB. In the case of new business, the
NCB also reports the volume of the new contracts at national level.
The ECB carries out the third and last step of aggregation by compiling, for each
instrument category, a weighted average interest rate referring to the euro area as a
whole. The reporting obligations are discussed in more detail in the following
sections.
122
This chapter refers mainly to Part 5 of Annex I to the Regulation.
123
NCBs may also choose the compilation of implicit rates (see Section 6.1) or require reporting agents to
provide data on the level of individual deposit and loans. In both cases the reporting obligations differ
from those illustrated in Figure 3.
Interest rate
and volume
per product
per reporting
agent
Interest rate
and volume
per instrument
category per
reporting
agent
Instrument
categories per
reporting
agent
Instrument
categories at
euro area
level
Instrument
categories per
country
Deposit and
loan products
Interest rate
and volume
per instrument
category per
country
Manual on MFI interest rate statistics − Aggregation of the data and reporting
obligations
112
9.2 Statistical information at the level of the reporting
agents
124
Each reporting agent needs to submit the data collected for MFI interest rate
statistics to the NCB of the Member State in which it is resident.
As explained in Section 5.2, MFI interest rates on outstanding amounts, i.e.
indicators 1 to 26 in Table 1, may be compiled as a snapshot of end-month
observations or as implicit rates referring to the average over the month. The same
applies to overnight deposits, deposits redeemable at notice extended credit card
credit and revolving loans and overdrafts i.e. indicators 1, 5, 6, 7, 12, 23, 32 and 36
in Table 2 and indicators 86 and 87 in Table 8. It is up to the NCBs to decide which of
the two methods is better suited to the national context. The data that reporting
agents need to provide to the NCB depend on the chosen method:
• In the case of a snapshot of end-month observations, the reporting agents need
to provide, for each instrument category, a weighted average interest rate
referring to the last day of the month. The weighted average interest rate covers
all outstanding deposits that have been placed and not yet withdrawn by
customers or all outstanding loans that have been withdrawn and not yet repaid
by customers in all the periods up to and including the reporting date.
125
• In the case of implicit rates referring to the average of the month, reporting
agents provide for each instrument category the accrued interest payable or
receivable during the month and the stock of deposits and loans on average
during the same month.
As indicated in the corresponding tables, 1 to 9, for each instrument category where
AAR or NDER is required, except for indicators 1, 5, 6, 7, 12, 23, 32 and 36 in Table
2 and indicators 86 and 87 in Table 8, all reporting agents need to provide a
weighted average interest rate. This weighted average interest rate refers to all
interest rates on new business operations in the instrument category during the
entire reference month.
126
Also, as indicated in the corresponding tables, the
reporting agents need to report the volumes of new business and, in the case of the
indicators in Table 6, also the amounts of renegotiated loans. Reporting agents do
not need to provide the amount of new business for loans to households for
consumption and for house purchase compiled as APRC, i.e. indicators 30 and 31 in
Table 5. These amounts can be derived as aggregates from the more detailed data
that are provided for the other new business rates.
Instead of asking reporting agents to provide weighted average interest rates on new
business and the corresponding volume referring to the institution as a whole, NCBs
may also request data at the level of individual deposits and loans.
124
See also paragraphs 66 to 72 of Annex I to the Regulation.
125
Bad loans and loans for debt restructuring at rates below market conditions are excluded. See also
Section 5.1.
126
See also Section 6.2.
Manual on MFI interest rate statistics − Aggregation of the data and reporting
obligations
113
NCBs may allow MFIs which are resident in a single national territory and individually
included in the list of MFIs to report MFI interest rate statistics together as a group.
127
The group then becomes a notional reporting agent and has the same reporting
requirements as the other (individual) MFIs that are reporting agents. So they have
to provide the data on outstanding amounts and new business as defined in this
section referring to the group as a whole. In addition they need to report every year
for each instrument category the number of reporting institutions in the group and the
variance of interest rates across these institutions. The number of reporting
institutions and the variance should refer to the month of October and be transmitted
with the October data.
9.3 National weighted average interest rates and national
total business volumes
128
NCBs receive the data from the reporting agents and aggregate them to MFI interest
rate statistics at the national level. These data are then submitted to the ECB:
• For each instrument category on outstanding amounts, and for each instrument
category on new business, each NCB provides a weighted average interest rate
referring to the Member State as a whole. Interest rates on outstanding
amounts are weighted using outstanding amounts as weights. As regards
indicators 15 to 26 in Table 1 of Appendix 1, outstanding amounts are only
transmitted to the ECB at a monthly frequency. If the NCB does not have the
data available at a monthly frequency, the data shall be carried forward into the
missing periods by repeating them or by applying appropriate statistical
techniques to reflect any trend in the data or seasonal pattern.
• In addition, each NCB provides for each of the new business indicators, except
for overnight deposits, deposits redeemable at notice, extended credit card
credit and revolving loans and overdrafts the amount of new business
conducted at national level in each instrument category during the reference
month. These amounts of new business refer to the population total, i.e. to the
entire reference reporting population, not only to the sample, i.e. the actual
reporting population.
129
Therefore, where a sampling approach is chosen at
national level for selecting the reporting agents, the NCB must apply expansion
factors at national level to derive the population total for the amount of new
business. The computation of selection probabilities and their use as expansion
factors for deriving population totals is explained in generic terms in
Section 12.7.
The NCBs provide the interest rates on outstanding amounts and on new business
to the ECB with a detail of four decimal places, i.e. 12.3456%. The level of detail at
127
Further discussed in Section 12.6.
128
See also paragraphs 33 to 44 in Part 14 of Annex II to the Guideline.
129
The reference and actual reporting population is defined in Chapter 3 and further discussed in
Section 12.1.
Manual on MFI interest rate statistics − Aggregation of the data and reporting
obligations
114
which the NCBs wish to collect the data from the reporting agents is up to them to
define. The published ECB results do not contain more than two decimal places.
Together with the national data, NCBs need to provide methodological notes
130
that
document any important special national practices, comprising regulatory
arrangements, national conventions, institutional arrangements and specific products
affecting MFI interest rate statistics. This information is essential in order to interpret
the level and the development of the deposit and lending rates set by MFIs.
NCBs carrying out a sampling approach for the selection of the reporting agents also
provide an estimate of the sampling error for the initial sample. A new estimate is
required after each revision
131
of the sample. The sampling error depends on the
national sampling method used, i.e. the stratification, the sample size and its
allocation, and the way the reporting agents are selected in each stratum.
9.4 Aggregated results for the euro area
132
The ECB carries out the final level of aggregation of the instrument categories for
each euro area Member State to euro area MFI interest rate statistics. It compiles
weighted average interest rates for each instrument category referring to the euro
area as a whole:
• In the case of outstanding amounts, i.e. indicators 1 to 26 in Table 1, and
overnight deposits, deposits redeemable at notice, extended credit card credit
and revolving loans and overdrafts, i.e. indicators 1, 5, 6, 7, 12, 23, 32 and 36 in
Table 2 and indicators 86 and 87 in Table 8, the weighting information is derived
from the MFI balance sheet statistics. The interest rates on outstanding
amounts compiled as a snapshot of end-month observations have the same
time reference point as MFI balance sheet statistics, whereas implicit rates refer
to the average over the month. Independent of the compilation procedure at
national level, the ECB uses the data on the size of each balance sheet item at
the end of the month in each Member State to calculate from the national
interest rates weighted averages for the euro area.
• In the case of new business other than overnight deposits, deposits redeemable
at notice, credit card debt and revolving loans and overdrafts, the weighting
information is provided by the NCBs, which collect the data from the reporting
agents.
The system of interest rates and weighting information is fully additive, in order to
allow combinations of rates such as synthetic deposit and lending rates.
130
Their content is further discussed in Section 4.3.2.
131
Further discussed in Section 12.8.
132
Further discussed in Section 12.2.
Manual on MFI interest rate statistics − Validation rules
115
10 Validation rules
Some linear constraints can be identified within the MIR dataset. Together with the
NCBs, it has been decided to clearly identify those constraints and to implement
them, wherever possible, as validation rules.
Validation rules can be implemented either at the level of data reception in the form
of automatic rejection or warnings and/or after data have been received. It should be
emphasised that such validation rules are valid for reporting agents reporting the
MIR indicators exactly as specified in the Regulation. For reporting agents reporting
MIR data in whole or in part at a more granular level, those validation rules may not
apply as such. In this case, the NCBs concerned apply different checks tailored to
their national collection systems.
The list of validation rules by rule category is provided below:
• For countries not granting a derogation to reporting agents for the reporting of
the APRC for consumer credit and loans to households for house purchase vis-
à-vis non-profit institutions serving households (NPISHs), in other words, for
those countries that collect the APRC vis-à-vis households including NPISHs,
the APRC should be higher than or equal to the AAR:
No Description # Validation rule Description #
1
APRC on loans in EUR to households for consumption
MIR (NB30) ≥ Loans in EUR to households for consumption MIR (NB13,
NB14, NB15)
2
APRC on loans in EUR to households for house
purchase
MIR (NB31) ≥ Loans in EUR to households for house purchase MIR (NB16,
NB17, NB18,
NB19)
It is acknowledged that in practice there are a few cases where these two rules may
not hold. For instance, if the APRC is not recalculated when a contract is
renegotiated, the AAR could be higher than the APRC. There are also a few cases
where rounding makes the AAR slightly higher than the APRC, since the AAR is
calculated as a weighted average of more granular MIR breakdowns. In such cases
where the validation rules do not hold, reporting agents are required to explain the
inconsistencies.
• For those countries granting a derogation to reporting agents for the reporting of
APRC on loans to non-profit institutions serving households (NPISHs), new
business volumes excluding NPISHs should be lower or equal to new business
volumes including NPISHs:
No Description # Validation rule Description #
3
APRC on loans in EUR to households for consumption
Weight (NB30) ≤ Loans in EUR to households for consumption Weight (NB13,
NB14, NB15)
4
APRC on loans in EUR to households for house
purchase
Weight (NB31) ≤ Loans in EUR to households for house purchase Weight (NB16,
NB17, NB18,
NB19)
Manual on MFI interest rate statistics − Validation rules
116
• Volumes of renegotiated loans should be lower than or equal to volumes of new
business loans:
No Description # Validation rule Description #
5
Renegotiated loans in EUR to households for
consumption
Weight (NB88) ≤ Loans in EUR to households for consumption Weight (NB13,
NB14, NB15)
6
Renegotiated loans in EUR to households for house
purchase
Weight (NB89) ≤ Loans in EUR to households for house purchase Weight (NB16,
NB17, NB18,
NB19)
7
Renegotiated loans in EUR to households for other
purposes
Weight (NB90) ≤ Loans in EUR to households for other purposes Weight (NB20,
NB21, NB22)
8
Renegotiated loans in EUR to non-financial
corporations
Weight (NB91) ≤ Loans in EUR to non-financial corporations Weight (NB37 to
NB54)
• Volumes of new business loans with collateral and/or guarantees should be
lower than or equal to volumes of new business loans:
No Description # Validation rule Description #
9
Loans in EUR to households for consumption with a
floating rate and up to 1 year initial rate fixation,
only collateralised/guaranteed loans
Weight (NB55) ≤ Loans in EUR to households for consumption with a
floating rate and up to 1 year initial rate fixation
Weight (NB13)
10
Loans in EUR to households for consumption with
over 1 and up to 5 years initial rate fixation,
only collateralised/guaranteed loans
Weight (NB56) ≤ Loans in EUR to households for consumption with
over 1 and up to 5 years initial rate fixation
Weight (NB14)
11
Loans in EUR to households for consumption with
over 5 years initial rate fixation,
only collateralised/guaranteed loans
Weight (NB57) ≤ Loans in EUR to households for consumption with
over 5 years initial rate fixation
Weight (NB15)
12
Loans in EUR to households for house purchase with
a floating rate and up to 1 year initial rate fixation,
only collateralised/guaranteed loans
Weight (NB58) ≤ Loans in EUR to households for house purchase with
a floating rate and up to 1 year initial rate fixation
Weight (NB16)
13
Loans in EUR to households for house purchase with
over 1 and up to 5 years initial rate fixation,
only collateralised/guaranteed loans
Weight (NB59) ≤ Loans in EUR to households for house purchase with
over 1 and up to 5 years initial rate fixation
Weight (NB17)
14
Loans in EUR to households for house purchase with
over 5 and up to 10 years initial rate fixation,
only collateralised/guaranteed loans
Weight (NB60) ≤ Loans in EUR to households for house purchase with
over 5 and up to 10 years initial rate fixation
Weight (NB18)
15
Loans in EUR to households for house purchase with
over 10 years initial rate fixation,
only collateralised/guaranteed loans
Weight (NB61) ≤ Loans in EUR to households for house purchase with
over 10 years initial rate fixation
Weight (NB19)
16
Loans in EUR to non-financial corporations up to an
amount of EUR 0.25 mn with a floating rate and up to
3 months initial rate fixation,
only collateralised/guaranteed loans
Weight (NB62) ≤ Loans in EUR to non-financial corporations up to an
amount of EUR 0.25 mn with a floating rate and up to
3 months initial rate fixation
Weight (NB37)
17
Loans in EUR to non-financial corporations up to an
amount of EUR 0.25 mn with over 3 months and up to
1 year initial rate fixation,
only collateralised/guaranteed loans
Weight (NB63) ≤ Loans in EUR to non-financial corporations up to an
amount of EUR 0.25 mn with over 3 months and up to
1 year initial rate fixation
Weight (NB38)
18
Loans in EUR to non-financial corporations up to an
amount of EUR 0.25 mn with over 1 and up to 3 years
initial rate fixation,
only collateralised/guaranteed loans
Weight (NB64) ≤ Loans in EUR to non-financial corporations up to an
amount of EUR 0.25 mn with over 1 and up to 3 years
initial rate fixation
Weight (NB39)
19
Loans in EUR to non-financial corporations up to an
amount of EUR 0.25 mn with over 3 and up to 5 years
initial rate fixation,
only collateralised/guaranteed loans
Weight (NB65) ≤ Loans in EUR to non-financial corporations up to an
amount of EUR 0.25 mn with over 3 and up to 5 years
initial rate fixation
Weight (NB40)
20
Loans in EUR to non-financial corporations up to an
amount of EUR 0.25 mn with over 5 and up to 10
years initial rate fixation,
only collateralised/guaranteed loans
Weight (NB66) ≤ Loans in EUR to non-financial corporations up to an
amount of EUR 0.25 mn with over 5 and up to 10
years initial rate fixation
Weight (NB41)
Manual on MFI interest rate statistics − Validation rules
117
21
Loans in EUR to non-financial corporations up to an
amount of EUR 0.25 mn with over 10 years initial rate
fixation,
only collateralised/guaranteed loans
Weight (NB67) ≤
Loans in EUR to non-financial corporations up to an
amount of EUR 0.25 mn with over 10 years initial rate
fixation
Weight (NB42)
22
Loans in EUR to non-financial corporations over an
amount of EUR 0.25 mn and up to EUR 1 mn with a
floating rate and up to 3 months initial rate fixation,
only collateralised/guaranteed loans
Weight (NB68) ≤ Loans in EUR to non-financial corporations over an
amount of EUR 0.25 mn and up to EUR 1 mn with a
floating rate and up to 3 months initial rate fixation
Weight (NB43)
23
Loans in EUR to non-financial corporations over an
amount of EUR 0.25 mn and up to EUR 1 mn with
over 3 months and up to 1 year initial rate fixation,
only collateralised/guaranteed loans
Weight (NB69) ≤ Loans in EUR to non-financial corporations over an
amount of EUR 0.25 mn and up to EUR 1 mn with
over 3 months and up to 1 year initial rate fixation
Weight (NB44)
24
Loans in EUR to non-financial corporations over an
amount of EUR 0.25 mn and up to EUR 1 mn with
over 1 and up to 3 years initial rate fixation,
only collateralised/guaranteed loans
Weight (NB70) ≤ Loans in EUR to non-financial corporations over an
amount of EUR 0.25 mn and up to EUR 1 mn with
over 1 and up to 3 years initial rate fixation
Weight (NB45)
25
Loans in EUR to non-financial corporations over an
amount of EUR 0.25 mn and up to EUR 1 mn with
over 3 and up to 5 years initial rate fixation,
only collateralised/guaranteed loans
Weight (NB71) ≤ Loans in EUR to non-financial corporations over an
amount of EUR 0.25 mn and up to EUR 1 mn with
over 3 and up to 5 years initial rate fixation
Weight (NB46)
26
Loans in EUR to non-financial corporations over an
amount of EUR 0.25 mn and up to EUR 1 mn with
over 5 and up to 10 years initial rate fixation,
only collateralised/guaranteed loans
Weight (NB72) ≤ Loans in EUR to non-financial corporations over an
amount of EUR 0.25 mn and up to EUR 1 mn with
over 5 and up to 10 years initial rate fixation
Weight (NB47)
27
Loans in EUR to non-financial corporations over an
amount of EUR 0.25 mn and up to EUR 1 mn with
over 10 years initial rate fixation,
only collateralised/guaranteed loans
Weight (NB73) ≤ Loans in EUR to non-financial corporations over an
amount of EUR 0.25 mn and up to EUR 1 mn with
over 10 years initial rate fixation
Weight (NB48)
28
Loans in EUR to non-financial corporations over an
amount of EUR 1 mn with a floating rate and up to 3
months initial rate fixation,
only collateralised/guaranteed loans
Weight (NB74) ≤ Loans in EUR to non-financial corporations over an
amount of EUR 1 mn with a floating rate and up to 3
months initial rate fixation
Weight (NB49)
29
Loans in EUR to non-financial corporations over an
amount of EUR 1 mn with over 3 months and up to 1
year initial rate fixation,
only collateralised/guaranteed loans
Weight (NB75) ≤ Loans in EUR to non-financial corporations over an
amount of EUR 1 mn with over 3 months and up to 1
year initial rate fixation
Weight (NB50)
30
Loans in EUR to non-financial corporations over an
amount of EUR 1 mn with over 1 and up to 3 years
initial rate fixation,
only collateralised/guaranteed loans
Weight (NB76) ≤ Loans in EUR to non-financial corporations over an
amount of EUR 1 mn with over 1 and up to 3 years
initial rate fixation
Weight (NB51)
31
Loans in EUR to non-financial corporations over an
amount of EUR 1 mn with over 3 and up to 5 years
initial rate fixation,
only collateralised/guaranteed loans
Weight (NB77) ≤ Loans in EUR to non-financial corporations over an
amount of EUR 1 mn with over 3 and up to 5 years
initial rate fixation
Weight (NB52)
32
Loans in EUR to non-financial corporations over an
amount of EUR 1 mn with over 5 and up to 10 years
initial rate fixation,
only collateralised/guaranteed loans
Weight (NB78) ≤ Loans in EUR to non-financial corporations over an
amount of EUR 1 mn with over 5 and up to 10 years
initial rate fixation
Weight (NB53)
33
Loans in EUR to non-financial corporations over an
amount of EUR 1 mn with over 10 years initial rate
fixation,
only collateralised/guaranteed loans
Weight (NB79) ≤ Loans in EUR to non-financial corporations over an
amount of EUR 1 mn with over 10 years initial rate
fixation
Weight (NB54)
34
Loans in EUR to non-financial corporations up to an
amount of EUR 0.25 mn with a floating rate and up to
1 year initial rate fixation, with original maturity over 1
year,
only collateralised/guaranteed loans
Weight (NB81) ≤ Loans in EUR to non-financial corporations up to an
amount of EUR 0.25 mn with a floating rate and up to
1 year initial rate fixation, with original maturity over 1
year
Weight (NB80)
35
Loans in EUR to non-financial corporations over an
amount of EUR 0.25 mn and up to EUR 1 mn with a
floating rate and up to 1 year initial rate fixation, with
original maturity over 1 year,
only collateralised/guaranteed loans
Weight (NB83) ≤ Loans in EUR to non-financial corporations over an
amount of EUR 0.25 mn and up to EUR 1 mn with a
floating rate and up to 1 year initial rate fixation, with
original maturity over 1 year
Weight (NB82)
36
Loans in EUR to non-financial corporations over an
amount of EUR 1 mn with a floating rate and up to 1
year initial rate fixation, with original maturity over 1
year,
only collateralised/guaranteed loans
Weight (NB85) ≤ Loans in EUR to non-financial corporations over an
amount of EUR 1 mn with a floating rate and up to 1
year initial rate fixation, with original maturity over 1
year
Weight (NB84)
Manual on MFI interest rate statistics − Validation rules
118
• Volumes of new business loans for other purposes to sole proprietors should be
lower than or equal to volumes of new business loans for other purposes:
No Description # Validation rule Description #
37
Loans in EUR to households for other purposes, of
which: sole proprietors and unincorporated
partnerships with a floating rate and up to 1 year initial
rate fixation
Weight (NB33) ≤ Loans in EUR to households for other purposes with a
floating rate and up to 1 year initial rate fixation
Weight (NB20)
38
Loans in EUR to households for other purposes, of
which: sole proprietors and unincorporated
partnerships with a floating rate and over 1 and up to
5 years initial rate fixation
Weight (NB34) ≤ Loans in EUR to households for other purposes with a
floating rate and over 1 and up to 5 years initial rate
fixation
Weight (NB21)
39
Loans in EUR to households for other purposes, of
which: sole proprietors and unincorporated
partnerships with a floating rate and over 5 years
initial rate fixation
Weight (NB35) ≤ Loans in EUR to households for other purposes with a
floating rate and over 5 years initial rate fixation
Weight (NB22)
• Volumes of new business loans to NFCs with a period of interest rate fixation
below one year and an original maturity over one year should be lower than or
equal to volumes of new business loans to NFCs with a period of interest rate
fixation below one year:
No Description # Validation rule Description #
40
Loans in EUR to non-financial corporations up to an
amount of EUR 0.25 mn with a floating rate and up to
1 year initial rate fixation, with original maturity over 1
year
Weight (NB80) ≤ Loans in EUR to non-financial corporations up to an
amount of EUR 0.25 mn with up to 1 year initial rate
fixation
Weight (NB37,
NB38)
41
Loans in EUR to non-financial corporations up to an
amount of EUR 0.25 mn with a floating rate and up to
1 year initial rate fixation, with original maturity over 1
year,
only collateralised/guaranteed loans
Weight (NB81) ≤ Loans in EUR to non-financial corporations up to an
amount of EUR 0.25 mn with up to 1 year initial rate
fixation,
only collateralised/guaranteed loans
Weight (NB62,
NB63)
42
Loans in EUR to non-financial corporations over an
amount of EUR 0.25 mn and up to EUR 1 mn with a
floating rate and up to 1 year initial rate fixation, with
original maturity over 1 year
Weight (NB82) ≤ Loans in EUR to non-financial corporations over an
amount of EUR 0.25 mn and up to EUR 1 mn with up
to 1 year initial rate fixation
Weight (NB43,
NB44)
43
Loans in EUR to non-financial corporations over an
amount of EUR 0.25 mn and up to EUR 1 mn with a
floating rate and up to 1 year initial rate fixation, with
original maturity over 1 year,
only collateralised/guaranteed loans
Weight (NB83) ≤ Loans in EUR to non-financial corporations over an
amount of EUR 0.25 mn and up to EUR 1 mn with up
to 1 year initial rate fixation,
only collateralised/guaranteed loans
Weight (NB68,
NB69)
44
Loans in EUR to non-financial corporations over an
amount of EUR 1 mn with a floating rate and up to 1
year initial rate fixation, with original maturity over 1
year
Weight (NB84) ≤ Loans in EUR to non-financial corporations over an
amount of EUR 1 mn with up to 1 year initial rate
fixation
Weight (NB49,
NB50)
45
Loans in EUR to non-financial corporations over an
amount of EUR 1 mn with a floating rate and up to 1
year initial rate fixation, with original maturity over 1
year,
only collateralised/guaranteed loans
Weight (NB85) ≤ Loans in EUR to non-financial corporations over an
amount of EUR 1 mn with up to 1 year initial rate
fixation,
only collateralised/guaranteed loans
Weight (NB74,
NB75)
Manual on MFI interest rate statistics − Validation rules
119
• Volumes of new business loans to NFCs over an amount of EUR 1 million
should be higher than EUR 1 million:
No Description
# Validation rule Description
46
Loans in EUR to non-financial corporations over an amount of EUR 1 mn with a floating rate and up to 3
months initial rate fixation
Weight (NB49) > EUR 1 mn
47
Loans in EUR to non-financial corporations over an amount of EUR 1 mn with over 3 months and up to 1 year
initial rate fixation
Weight (NB50) > EUR 1 mn
48
Loans in EUR to non-financial corporations over an amount of EUR 1 mn with over 1 and up to 3 years initial
rate fixation
Weight (NB51) > EUR 1 mn
49
Loans in EUR to non-financial corporations over an amount of EUR 1 mn with over 3 and up to 5 years initial
rate fixation
Weight (NB52) > EUR 1 mn
50
Loans in EUR to non-financial corporations over an amount of EUR 1 mn with over 5 and up to 10 years initial
rate fixation
Weight (NB53) > EUR 1 mn
51
Loans in EUR to non-financial corporations over an amount of EUR 1 mn with over 10 years initial rate fixation Weight (NB54) > EUR 1 mn
52
Loans in EUR to non-financial corporations over an amount of EUR 1 mn with a floating rate and up to 3
months initial rate fixation,
only collateralised/guaranteed loans
Weight (NB74) > EUR 1 mn
53
Loans in EUR to non-financial corporations over an amount of EUR 1 mn with over 3 months and up to 1 year
initial rate fixation,
only collateralised/guaranteed loans
Weight (NB75) > EUR 1 mn
54
Loans in EUR to non-financial corporations over an amount of EUR 1 mn with over 1 and up to 3 years initial
rate fixation,
only collateralised/guaranteed loans
Weight (NB76) > EUR 1 mn
55
Loans in EUR to non-financial corporations over an amount of EUR 1 mn with over 3 and up to 5 years initial
rate fixation,
only collateralised/guaranteed loans
Weight (NB77) > EUR 1 mn
56
Loans in EUR to non-financial corporations over an amount of EUR 1 mn with over 5 and up to 10 years initial
rate fixation,
only collateralised/guaranteed loans
Weight (NB78) > EUR 1 mn
57
Loans in EUR to non-financial corporations over an amount of EUR 1 mn with over 10 years initial rate fixation,
only collateralised/guaranteed loans
Weight (NB79) > EUR 1 mn
58
Loans in EUR to non-financial corporations over an amount of EUR 1 mn with a floating rate and up to 1 year
initial rate fixation, with original maturity over 1 year
Weight (NB84) > EUR 1 mn
59
Loans in EUR to non-financial corporations over an amount of EUR 1 mn with a floating rate and up to 1 year
initial rate fixation, with original maturity over 1 year,
only collateralised/guaranteed loans
Weight (NB85) > EUR 1 mn
Manual on MFI interest rate statistics − Validation rules
120
• Volumes of new business loans to NFCs over an amount of EUR 0.25 million
and up to EUR 1 million should be higher than EUR 0.25 million:
No Description # Validation rule Description
60
Loans in EUR to non-financial corporations over an amount of EUR 0.25 mn and up to EUR 1 mn with a
floating rate and up to 3 months initial rate fixation
Weight (NB43) > EUR 0.25 mn
61
Loans in EUR to non-financial corporations over an amount of EUR 0.25 mn and up to EUR 1 mn with over 3
months and up to 1 year initial rate fixation
Weight (NB44) > EUR 0.25 mn
62
Loans in EUR to non-financial corporations over an amount of EUR 0.25 mn and up to EUR 1 mn with over 1
and up to 3 years initial rate fixation
Weight (NB45) > EUR 0.25 mn
63
Loans in EUR to non-financial corporations over an amount of EUR 0.25 mn and up to EUR 1 mn with over 3
and up to 5 years initial rate fixation
Weight (NB46) > EUR 0.25 mn
64
Loans in EUR to non-financial corporations over an amount of EUR 0.25 mn and up to EUR 1 mn with over 5
and up to 10 years initial rate fixation
Weight (NB47) > EUR 0.25 mn
65
Loans in EUR to non-financial corporations over an amount of EUR 0.25 mn and up to EUR 1 mn with over 10
years initial rate fixation
Weight (NB48) > EUR 0.25 mn
66
Loans in EUR to non-financial corporations over an amount of EUR 0.25 mn and up to EUR 1 mn with a
floating rate and up to 3 months initial rate fixation,
only collateralised/guaranteed loans
Weight (NB68) > EUR 0.25 mn
67
Loans in EUR to non-financial corporations over an amount of EUR 0.25 mn and up to EUR 1 mn with over 3
months and up to 1 year initial rate fixation,
only collateralised/guaranteed loans
Weight (NB69) > EUR 0.25 mn
68
Loans in EUR to non-financial corporations over an amount of EUR 0.25 mn and up to EUR 1 mn with over 1
and up to 3 years initial rate fixation,
only collateralised/guaranteed loans
Weight (NB70) > EUR 0.25 mn
69
Loans in EUR to non-financial corporations over an amount of EUR 0.25 mn and up to EUR 1 mn with over 3
and up to 5 years initial rate fixation,
only collateralised/guaranteed loans
Weight (NB71) > EUR 0.25 mn
70
Loans in EUR to non-financial corporations over an amount of EUR 0.25 mn and up to EUR 1 mn with over 5
and up to 10 years initial rate fixation,
only collateralised/guaranteed loans
Weight (NB72) > EUR 0.25 mn
71
Loans in EUR to non-financial corporations over an amount of EUR 0.25 mn and up to EUR 1 mn with over 10
years initial rate fixation,
only collateralised/guaranteed loans
Weight (NB73) > EUR 0.25 mn
72
Loans in EUR to non-financial corporations over an amount of EUR 0.25 mn and up to EUR 1 mn with a
floating rate and up to 1 year initial rate fixation, with original maturity over 1 year
Weight (NB82) > EUR 0.25 mn
73
Loans in EUR to non-financial corporations over an amount of EUR 0.25 mn and up to EUR 1 mn with a
floating rate and up to 1 year initial rate fixation, with original maturity over 1 year,
only collateralised/guaranteed loans
Weight (NB83) > EUR 0.25 mn
Manual on MFI interest rate statistics − Revision policy
121
11 Revision policy
11.1 Principles of the revision policy
Reporting agents are required to fulfil the minimum standards for transmission,
accuracy, compliance with concepts and revisions as specified in Annex II of
Regulation ECB/2013/34. In particular, the Regulation states that statistical
information must be correct and must comply with the definitions and classifications
contained therein.
Sometimes errors occur when data are reported to the NCB. For instance, data
might be misclassified (e.g. instruments are not reported in the correct category or
maturity bucket) or some information can be omitted by mistake. Furthermore, there
can be errors in the calculation of interest rates, etc. Continuous discussion between
the reporting agents and the NCBs is a key element for a swift resolution of errors in
reporting and correction of data. It is also very important to correct incorrect data not
only for the reference periods following the discovery of the error, but also for the
past periods during which the error occurred. In particular, as MIR statistics are
heavily used in time series analysis, breaks in the series due to errors in data
collection should be avoided as they hamper the analysis and distort the
interpretation of the figures.
Reporting agents are strongly encouraged to correct their data for the past reporting
periods if an error is discovered. This means correcting and resubmitting the data to
the NCB. It is acknowledged that it is sometimes difficult to rebuild the whole history
of the data, and for this reason in such cases NCBs may as a minimum advise the
reporting agent to correct the data for the periods for which the error occurred or
over 13 months of data from the latest reference period or back to the starting date
of the data collection for the given indicator, whichever is the shortest period. For
instance, if data are collected with reference to February 2016 and a mistake is
discovered which occurs in the periods October 2015 and before, as a minimum data
should be corrected back to January 2015, which corresponds to 13 months prior to
the time of the discovery of the error (not back to September 2014, which would be
13 months back from the last period in which the mistake was observed in the data).
In another example where data are collected with reference to February 2016 and a
mistake is discovered which affects the reference periods from December 2015 to
January 2016, data should be corrected for the period from December 2015 to
January 2016.
Paragraph 4 of Annex II to Regulation ECB/2013/34 on the minimum standards for
revisions states that the “revisions policy and procedures set by the ECB and the
relevant NCB must be followed…” (by the reporting agents). Each NCB has
therefore defined specific revision policies and procedures. It is however important to
make sure that, overall, across the Eurosystem but also at the ESCB level, common
minimum standards are followed in terms of revision policy to ensure the high quality
of the MIR data. To this end, NCBs in the ESCB have together defined certain high-
Manual on MFI interest rate statistics − Revision policy
122
level criteria to help in assessing whether revisions of past data should be requested
from reporting agents. These criteria have been designed with the aim to correct
major errors, but they do not prevent NCBs from requesting reporting agents to
transmit even smaller revisions, as is currently the practice for some NCBs. Overall,
it is considered that errors that fulfil these criteria are such that if not corrected over a
longer time period, they could hamper the analysis of the monetary policy
transmission mechanism and the regular analysis by economists of financial stability
or by supervisors working for European banking supervision. These criteria therefore
offer a good balance between the costs faced by reporting agents in reporting
revisions of past data and the benefits in terms of improving the quality of MFI
interest rate statistics.
The underlying principle behind the criteria is that the error of a single reporting
agent should not have a major impact on the national aggregates. If it does, the
quality and therefore the reliability of the corresponding national aggregate could be
seriously compromised. As a consequence, the criteria are based on the impact of
the error of a single reporting agent on national aggregates. Rather than using the
granular breakdowns as they are defined in the Regulation, the impact is assessed
based on national intermediate aggregates. These are typically indicators that
aggregate the data collected from reporting agents, for example for maturities of, for
instance, total new loans for house purchase. The reason for this is that the detailed
breakdowns as defined in the Regulation may not always have the same economic
importance in different countries. For instance, in some countries loans for house
purchase with a floating rate or an initial rate fixation period of up to one year prevail,
whereas in other countries the rates with an initial rate fixation over ten years prevail.
Using a more aggregated level, in other words “all interest rate fixations combined”
offers a good benchmark and ensures a level playing field for all countries.
A simple way to measure the reporting error is to calculate the difference between
the corrected figure and the previously transmitted figure. The assessment of the
need for further revisions in the past is undertaken on the basis of the impact of the
error on the intermediate aggregates. The detailed calculations are explained in the
section below.
11.2 The revision criteria
To calculate the impact of a single reporting agent on national intermediate
aggregates, every reporting agent is treated separately and, for each one, it is
assessed whether the revisions that have been transmitted by this reporting agent
impact intermediate aggregates (including grossing up) by more than 5% of the
figure before correction, all other parameters being equal. This threshold is
considered to provide an appropriate balance between the costs associated with the
revision of data and the benefits that accrue from maintaining accurate historical
data series.
Manual on MFI interest rate statistics − Revision policy
123
(1) For interest rates, revisions on past periods
133
should be transmitted if the
following condition is met:
The absolute value of the change between the interest rate on the national
intermediate aggregate after the correction (period t) and the interest rate on the
national intermediate aggregate before the correction (period t) should be equal to or
larger than 5% of the level of the interest rate on the national intermediate aggregate
before the correction (period t).
|
|
|
0.05
|
[11]
Where:
• r
1
(expressed as a percentage) is the weighted average rate before correction
of the national intermediate aggregate:
=
• r
2
(expressed as a percentage) is the weighted average rate after correction of
the national intermediate aggregate including the revision of reporting agent
“R”:
=
+
+
also written as r
=
where v
=
v
r
is the rate that has been revised by reporting agent R and v
is the corresponding
volume. N is the total number of reporting agents included in the actual reporting
population, while i represents a single reporting agent included in the actual reporting
population i = 1, 2, 3 … N. Correspondingly, r
i
and v
i
are the individual interest rates
and volumes reported by the reporting agents.
(2) For the new business volumes, revisions have to be transmitted on past
periods if the following condition is met:
The absolute value of the change between the new business volume of the national
intermediate aggregate after the correction (period t) and the new business volume
of the national intermediate aggregate before the correction (period t) should be
equal to or larger than 5% of the level of the new business volume of the national
intermediate aggregate before the correction (period t)
|
|
|
0.05
|
[12]
v
1
is the new business volume before correction and v
2
is the new business volume
after correction.
133
That is to say, for the periods for which the error occurred or over 13 months of data from the time of
the revision or back to the starting date of the data collection for the given indicator, whichever is the
shortest period.
Manual on MFI interest rate statistics − Revision policy
124
Where v
=
v
and v
= v
+ (v
v
)
It is important to emphasise that revisions to new business volumes are likely to lead
to revisions to interest rates. Interest rates on outstanding amounts are collected
under the MIR statistics framework, whereas the corresponding outstanding amounts
(i.e. weights used in the calculation of aggregated interest rates) are collected under
the BSI statistics framework, which follows a separate revision policy. NCBs may
nonetheless encourage reporting agents on a best efforts basis to revise outstanding
amounts when they revise the corresponding interest rates on outstanding amounts.
If the revision occurs because of a mistake in reporting that is common to several
reporting agents, in other words in the case of a systematic error, all reporting agents
are strongly encouraged to submit revisions even if their individual impact falls below
the predefined threshold. If the error is present over an extended period but
breaches the threshold only for a few non-consecutive months, the standard practice
is that data should be corrected for all months displaying the error as far back as
possible, but as a minimum for the periods for which the error occurred or over 13
months of data from the time of the revision or back to the starting date of the data
collection for the given indicator, whichever is the shortest period.
The intermediate aggregates for which the criteria will be applied are listed below:
For new business (interest rates and new business volumes):
Description Series key
Loans to corporations of up to EUR 1 mn
MIR.M.XX.B.A2A.A.R.0.2240.EUR.N
Loans to corporations of over EUR 1 mn
MIR.M.XX.B.A2A.A.R.1.2240.EUR.N
Loans to households for consumption (APRC)
MIR.M.XX.B.A2B.A.C.A.2250.EUR.N
Loans to households for house purchase (APRC):
MIR.M.XX.B.A2C.A.C.A.2250.EUR.N
Deposits from corporations
MIR.M.XX.B.L22.A.R.A.2240.EUR.N
Deposits from households
MIR.M.XX.B.L22.A.R.A.2250.EUR.N
For outstanding amounts (interest rates only):
Description Series key
Loans to corporations
MIR.M.XX.B.A20.A.R.A.2240.EUR.O
Loans to households for consumption and other purposes
MIR.M.XX.B.A25.A.R.A.2250.EUR.O
Loans to households for house purchase
MIR.M.XX.B.A22.A.R.A.2250.EUR.O
Deposits from corporations including overnight deposits and
deposits with agreed maturity
MIR.M.XX.B.L22.A.R.A.2240.EUR.O and
MIR.M.XX.B.L21.A.R.A.2240.EUR.N
Deposits from households including overnight deposits,
deposits with agreed maturity and deposits redeemable at notice
MIR.M.XX.B.L22.A.R.A.2250.EUR.O,
MIR.M.XX.B.L21.A.R.A.2250.EUR.N and
MIR.M.XX.B.L23.A.R.A.2250.EUR.N
Furthermore, applying the same criteria to the underlying breakdowns will almost
certainly ensure that the criteria are met at the intermediate level. Therefore,
applying the criteria to all the underlying breakdowns (i.e. at a more granular level)
should also fulfil these criteria at the intermediate level. NCBs, when applying these
criteria, may apply them at the intermediate level or for all the underlying
breakdowns, or a combination of the two, for instance at the intermediate level and
for a few additional underlying breakdowns.
Manual on MFI interest rate statistics − Revision policy
125
For non-euro area EU countries, the revision criteria are applied on a best efforts
basis. It is suggested to apply them to the indicators referring to loans and deposits
denominated in national currency and, if feasible, to the indicators for euro-
denominated instruments, depending on the local importance of the latter.
Manual on MFI interest rate statistics − Selection of the reporting agents
126
12 Selection of the reporting agents
134
12.1 Selection of the actual reporting population
135
MFI interest rate statistics provide for the euro area as a whole and individually for
each Member State detailed information about the interest rates that resident MFIs
apply to euro-denominated deposits and loans vis-à-vis households and non-
financial corporations resident in the euro area. In each Member State, the reporting
agents for MFI interest rate statistics are selected by the respective NCB. Each NCB
has the choice of declaring all or only a subset of resident MFIs as reporting agents.
The starting point for the selection of the reporting agents is the reference reporting
population, which provides the sampling frame. For each NCB, the reference
reporting population comprises all resident MFIs (except central banks and money
market funds) which take euro-denominated deposits from and/or grant euro-
denominated loans to households and/or non-financial corporations resident in the
euro area Member States.
136
In order to select from the reference reporting
population the actual reporting population, i.e. the reporting agents for MFI interest
rate statistics, NCBs follow the procedure defined in Part 12c of the Guideline, which
is outlined as a “decision tree” in Figure 4.
Figure 4
Decision tree
134
This chapter refers mainly to Part 14 of Annex II to the Guideline.
135
See also Article 2 of the Regulation.
136
Defined in Article 1(7) of the Regulation. See also Chapter 3.
Maintenance of sample over time
Reporting
requirements
Census
Sample
Stratification of the reference reporting
population
Random sampling
within each stratum
Selection of the
biggest institutions
within each stratum
Minimum
sample size
Reference reporting population
Actual reporting population
Manual on MFI interest rate statistics − Selection of the reporting agents
127
An NCB first has the choice between a census and a sample. In the case of a
census, the NCB requires each resident MFI in the reference reporting population to
report MFI interest rate statistics. In the case of a sample, the NCB stratifies the
reference reporting population, which means that the entire reference population is
sub-divided into sub-populations including institutions that as a result of their limited
banking activity are of minor significance for MFI balance sheet statistics, i.e. so-
called “tail institutions”. After the stratification, the NCB ideally selects the reporting
agents at random from each stratum. The alternative to drawing at random is to
select the largest institutions within each stratum. In order to keep the sample
representative over time, the NCB needs to refresh the sample with joiners and
adjust it for leavers and other changes in the characteristics of the reporting agents.
To ensure the quality of the output, the Regulation defines a minimum size for the
national sample.
The sampling approach aims to reduce the reporting burden on the whole banking
sector, as only some MFIs have to report, but it might require some small institutions
to report MFI interest rate statistics to ensure a representative range of results. Small
specialised institutions might offer different interest rates than big universal banks.
Following national procedures, the NCB needs to inform the selected reporting
agents about their reporting obligations.
12.2 Census versus sampling approach
137
As stated in Section 12.1, NCBs first have the choice between a census and a
sample. In the case of a census, the NCB asks all MFIs in that Member State to
report MFI interest rate statistics. In other words, the actual reporting population is
identical to the reference reporting population. In the case of a sample, only a
selection of the reference population is asked to report, which implies that the actual
reporting population is smaller than the reference reporting population. In general,
the advantages of a sample survey compared with a census are as follows:
• Costs: A sample is less costly than a census, since data are collected only from
a subset of the reference reporting population.
138
The total costs depend on the
size of the sample, which is determined by the desired degree of precision
139
of
the results.
137
See also paragraphs 2 to 5 of Part 14 of Annex II to the Guideline.
138
Council Regulation (EC) No 2533/98 of 23 November 1998 concerning the collection of statistical
information by the European Central Bank (OJ L 318, 27.11.1998, p. 8), as amended by Council
Regulation (EC) No 951/209 of 9 October 2009 (OJ L 269, 14.10.2009, p. 1) specifies in Article 3 that
“without prejudice to the fulfilment of its statistical reporting requirements, the ECB may fully or partly
exempt specific classes of reporting agents from its statistical reporting requirements”. The ESCB “shall
assess the merits and costs of the collection of the new statistical information in question”.
139
Variance and standard error, also referred to as sampling error, are measures for the precision of the
sampling procedure:
=
Manual on MFI interest rate statistics − Selection of the reporting agents
128
• Timeliness: A sample gives more timely results than a census. Sampling limits
the number of reporting agents and therefore the number of reports that the
NCBs need to collect, check and process.
• Accuracy:
140
Sampling procedures may lead to higher quality data. The smaller
volume of work (fewer reports), make possible more careful guidance and
monitoring of the reporting institutions, and more careful checking of the
responses and data processing. In general, sampling procedures tend to
reduce measurement errors
141
in surveys resulting from coding, editing,
processing, non-response, incorrect answers, etc.
The disadvantages of sampling procedures compared with a census are partly the
costs for setting up and maintaining the sample, but above all the uncertainty
resulting from the fact that only a part of the reference reporting population is being
observed. The results derived from a sample, i.e. from the actual reporting
population, might therefore differ from the true (unknown) values in the reference
reporting population. The errors that may occur because sampling procedures are
used are referred to as sampling errors.139 In contrast to measurement errors,
which may occur both in a sample and in a census, sampling errors only occur in
samples.
As explained in Section 9.4, the ECB performs the final stage in aggregating the
instrument categories for each Member State to euro area results. To do so, the ECB
requires both interest rates and weights, i.e. the deposit and loan amounts per
instrument category and Member State. Whereas for outstanding amounts the
weights can be derived from the MFI balance sheet statistics, for new business the
weights are based on data collected from the actual reporting population. Hence, in
the case of a census and a sample the interest rates on outstanding amounts and on
new business, as well as the amount of new business, need to be collected from the
reporting agents. This implies that, in the case of a sample, not only the interest
rates on outstanding amounts and new business but also the amounts of new
business are sampling variables, i.e. variables measured with certainty after the
selection of the reporting agents. Sampling variables are estimates and therefore
subject to sampling errors.
To minimise the risk that the results of a sample survey deviate from the true
(unknown) value in the reference reporting population, the aim is to compile a
representative sample of MFIs. A sample is considered representative of the
reference reporting population if it comprises all the characteristics of the institutions
that are relevant for MFI interest rate statistics and found in the reference reporting
population. In other words, the sample, i.e. the actual reporting population, should
reflect the relevant characteristics of all MFIs in the reference reporting population.
There are two main ways of creating a sample:
140
The mean square error aims to quantify the accuracy of the estimator:
=
=
+
= +
141
Measurement errors are sometimes referred to as “non-sampling errors”.
Manual on MFI interest rate statistics − Selection of the reporting agents
129
• In the case of non-random or purposive sampling, a number of typical units are
selected from the reference reporting population. A unit is typical if it conforms
closely, in the sampler’s view, to the characteristics of the reference reporting
population. The probability of selecting any particular unit cannot be quantified
and therefore the computation of the precision of the sample is impossible.
• In the case of random sampling, each unit in the population has a chance of
being selected in the sample and this selection probability can be quantified.
The advantages of random sampling over purposive sampling are as follows:
• The reporting agents are selected by chance, i.e. objectively, and not subject to
the sampler’s view.
• The standard error and hence the precision of the sampling procedure can be
quantified.
• The degree of accuracy of the estimates from the sample can be estimated.
• The optimal sample size can be defined, which is further discussed in
Section 12.5.
• Selection probabilities and hence expansion factors can be defined, which is
further discussed in Section 12.1.
12.3 Stratification of the reference reporting population
142
To ensure that the sample of MFIs is representative of the potential reporting
population, the Regulation requires that if a Member State decides on a sample, then
the reference reporting population needs to be stratified before any reporting agents
are selected. Stratification implies that the reference reporting population N is sub-
divided into sub-populations or strata N
1
, N
2
, N
3
, …, N
L
. These are not overlapping
and together comprise the entire reference population:
+
+
+ +
= [13]
The advantage of stratification and hence of stratified random sampling is that
“extreme” samples, i.e. samples which are not representative of the reference
reporting population, can be avoided, resulting in a lower sampling error. For
example, in the case of MFI interest rate statistics, a sample containing only
institutions specialising in housing loans would be an extreme sample that did not
represent all institutions in the reference reporting population. A further advantage of
stratified random sampling is that, in addition to the estimated values referring to the
entire sample, information can also be gathered about each stratum.
To avoid extreme samples, first the heterogeneous reference reporting population is
divided into homogeneous strata. Next, from each stratum a certain number of
142
See also paragraphs 6 to 9 of Part 14 of Annex II to the Guideline.
Manual on MFI interest rate statistics − Selection of the reporting agents
130
reporting agents are drawn, which together constitute the sample. By selecting
reporting agents from each of the homogeneous strata, the sample becomes
heterogeneous and is therefore representative of the heterogeneous reference
reporting population. The heterogeneity refers to the variance of the sampling
variables.
A stratum is homogeneous if the measurements for the sampling variables, i.e. the
interest rates on outstanding amounts and the interest rates and amounts of new
business vary little from one institution to the other. For example, in the case of MFI
interest rate statistics, a stratum containing only institutions specialising in housing
loans could be a homogeneous stratum, as interest rates might be similar. The result
of such homogeneous strata is that the variance of the sampling variables within
each stratum, i.e. the intra-stratum variance, is lower than the variance between
strata, i.e. the extra-stratum variance. The decomposition of the total variance into
intra-stratum and an extra-stratum variance is also known as variance analysis or
Huygens theorem. The univariate equation
143
is as follows:
(
)
=
(
)
+
(
)
(
)
=
1
()
+
1
(
)
[14]
(
)
=
1
1
(
)
+
1
(
)
(
)
=
1
(
)
+
1
(
)
More generally:
(
)
=
(
)
+
(
)
143
The same kind of relationship exists for multiple variables. In vectorial notation it reads as follows:
=
+
Meaning of letters and symbols:
w
i
Weight of institution i
w
h
Weight of the stratum h
y
i
Vector of observations for institution i
g
h
Vector which is the barycentre (centre of gravity) of the stratum h
g Vector which is the barycentre of the whole set of data
•
A kind of distance: Generally, for numerical data, the family of distances of Minkowski is
applied, where the distance between two items, e.g. institutions or barycentres is computed with the
following formula, where i denotes an institution and j a variable:
’
=
(
;
’
)
=
’
The Euclidean distance is given with = 2 and
= 1
Manual on MFI interest rate statistics − Selection of the reporting agents
131
where:
• Var(y) is the total variance of y in the reference reporting population, with y
as the sampling variable
• Var
within
(y) is the variance of y within the strata (intra-stratum variance)
• Var
between
(y) is the variance of y between the strata (extra-stratum variance)
• N is the number of institutions in the reference reporting population
• N
h
is the number of institutions in stratum h, with h = 1, …, L
•
is the average y in stratum h
• is the average y in the reference reporting population
• w
i
is the weight of the individual institution i and
• w
h
is the weight of the stratum h
Stratification requires that suitable stratification criteria be defined, which allows the
sub-division of all MFIs into homogeneous strata. Information on the stratification
criteria must be available for each institution in the reference reporting population.
The stratification criteria must relate to the purpose of the survey, i.e. to the sampling
variables that are to be estimated from the sample. Sampling variables for MFI
interest rate statistics are the interest rates on outstanding amounts and on new
business as well as the amount of new business.
144
Suitable stratification criteria can
be derived, for example, from the MFI balance sheet statistics, national surveys
conducted for supervisory purposes, national retail interest rate statistics, or the list
of MFIs.
If the institutions within each stratum show little variance in terms of the stratification
criteria, and if there is a strong relationship between the stratification criteria and the
sampling variables, then the institutions within each stratum are also likely to show
little variance in interest rates and amounts of new business. If a small number of
MFIs is selected from one stratum, the interest rates and the amounts of new
business collected from these institutions can be assumed to be representative of
that stratum. Combining the data from the different strata should give results that are
representative of the reference reporting population.
Within each Member State at least one stratification criterion is required as a
minimum standard by the Regulation. The aim is to ensure that the sample of MFIs
is representative of this Member State’s reference reporting population and the
standard error small. Ideally each Member State defines a hierarchy of stratification
criteria. These should take into account the national circumstances and hence be
specific to each Member State.
144
See also Section 12.2.
Manual on MFI interest rate statistics − Selection of the reporting agents
132
In many countries the banking business with households and non-financial
corporations is highly concentrated. The organisational structure of the national
banking business may amplify this effect. Some credit institutions might have a legal
organisation where only the “head office” is included in the list of MFIs whereas all
the regional offices are treated as “branches”. In this case, the head office submits
one statistical report covering the whole organisation. In other banking groups, each
regional office might be organised as an independent credit institution and as such
included in the list of MFIs. In this case, one statistical report covers only the
business of the reporting entity. Concentration in the banking business and
organisational differences need to be taken into account when designing the sample.
It might be necessary to work with unequal probabilities for the selection of the
reporting agents.
145
Bigger countries might consider the region in which the MFI is located as a
stratification criterion. Without aiming at the compilation of regional statistics, which
goes beyond the user requirements, regional (ex ante) stratification
146
reduces the
sampling error where regional differences in interest rates or in the type of customer
exist. However, depending on the organisational structure, regional differences might
be apparent at branch level rather than at the level of credit institutions. Regional
differences are captured in interest rates and the new business amounts, if the head
office provides data covering all branches. The alternative is to draw a sample at
branch level.
147
In the latter case, the same minimum standards as for the sampling
of MFIs apply.
For the construction of the strata, the use of quantitative data analysis techniques,
such as principal component or factor analysis, and of regrouping techniques, such
as cluster analysis, is recommended. These statistical methods help to allocate the
institutions according to their statistical proximity, with similar ones allocated to the
same stratum and dissimilar ones to different strata.
Since MFI interest rate statistics are aimed at providing data on the level and
development of interest rates both at euro area and at national level, the country of
residence of the MFIs is chosen as the first “natural” stratification criterion.
Geographical ex ante stratification gives NCBs flexibility to choose within the
framework set by the Regulation the most suitable procedure for selection the
reporting agents. Furthermore, it enables the national samples or census procedures
to be combined into a euro area sample ensuring reliable statistics at euro area level
and national level. In practice, the data are collected by NCBs from reporting agents
at national level and then aggregated to euro area results. The stratification at euro
area level can be illustrated in a schematic way
148
as follows:
145
Further discussed in Section 12.4.
146
Stratification according to region, which is decided by the sample, before any application of quantitative
data analysis and regrouping techniques.
147
Sampling at branch level is allowed under certain conditions and discussed in Section 12.6. It is carried
out without any intermediate drawing, i.e. without sampling (in a first step) credit institutions and (in a
second step) branches of the selected institutions; this is further discussed at the end of this chapter.
148
A brief description of the actual euro area sample for MFI interest rate statistics is given in Section
12.10
Manual on MFI interest rate statistics − Selection of the reporting agents
133
Table 11
The potential stratification at euro area level
Potential stratification criteria
Resulting strata
Member States Region Size of institution Type of business
Member State A
A1
A2
A3
A4
Member State B
B1
B2
B3
B4
B5
B6
Member State C
C1
C2
Member State D
D1
D2
D3
D4
etc.
The random selection of the reporting agents takes place after all strata are defined.
MFIs are drawn from the sampling frame only at this stage (indicated as shaded cells
in Table 11). Since there is no intermediate drawing, for example at the level of
regions or by the size of the institutions, the approach is referred to as single-stage
sampling. If there were an intermediate drawing, the sampling would occur in several
stages (multi-stage sampling) which would require more complex formulae and
increase the variance of the estimator.
12.4 Allocation of the sample across strata
149
After defining the national strata and the national sample size n, the sample is drawn
by selecting the reporting agents from each stratum. In this way, the actual reporting
population is defined. The total sample size n is the sum of the sample sizes n1, n2,
n3, …, nL for each of the strata:
+
+
+ +
= [15]
Two things need to be decided before drawing the reporting agents:
• the allocation of the sample size n among the strata; and
149
See also paragraphs 15 to 20 of Part 14 of Annex II to the Guideline.
Manual on MFI interest rate statistics − Selection of the reporting agents
134
• the method for selecting the reporting agents.
Each NCB may choose the most appropriate allocation of the national sample size n
among the strata. In other words, it is up to the NCBs to define how many reporting
agents nh are drawn from the total MFIs Nh in each stratum, as long as the sampling
rate nh/Nh for each stratum h fulfils the following condition:
0 <
1 [16]
The minimum standard is to select at least one reporting agent from each stratum,
i.e. the sampling rate needs to be above zero (0 < nh/Nh), which implies that it is not
possible to exclude one entire stratum from the actual reporting population. The
number of reporting agents nh may be the same for each stratum (constant
allocation), proportional to the size of the stratum (proportional allocation), or
dependent on the variance of the sampling variables or another closely linked
variable in each stratum (optimal allocation
150
). The sampling rate may also be one
(nh/Nh = 1), which means that all MFIs in a stratum are selected as reporting agents.
The allocation of the national sample size across strata has an influence on the
variance of the estimator. The best results and hence the lowest total sampling error
is in general achieved with optimal allocation.
Regarding the method for selecting the reporting agents within each stratum, the
Regulation gives NCBs the choice between random sampling and the selection of
the largest institutions per stratum,
151
where one method may be used for one part of
the strata and the other method for the rest:
• The statistically ideal case is the random selection of reporting agents in each
stratum. In a random sample, each MFI in the stratum has a known probability
above zero of being selected as a reporting agent. The random drawing of the
institutions in each stratum can then be carried out with equal probability for all
institutions or with probability proportional to the size of the institution. In the
latter case, big institutions are more likely to be drawn. However, the small
institutions also have a probability of selection above zero and could hence in
theory also be selected. Random selection with probability proportional to size
is highly recommended with extremely skewed populations.
• The alternative to random sampling is the selection of the largest institutions in
each stratum. The aim of this procedure is to exclude small MFIs from reporting
so that they do not have to bear disproportionately high costs. The selection of
the largest institutions per stratum is not random sampling, as small institutions
have a probability of selection equal to zero.
The precondition for sampling with probability proportional to size, i.e. for random
sampling and the selection of the largest institutions, is a strong statistical
relationship between the sampling variables and the size of the MFIs in the reference
reporting population. Sampling variables are the interest rates and the amount of
150
Also known as the Neyman optimum.
151
Advantages and drawbacks are discussed in Section 12.2.
Manual on MFI interest rate statistics − Selection of the reporting agents
135
new business. The size of the MFI is approximated by the size of the relevant
balance sheet items for each institution.
12.5 Minimum national sample size
152
The sample of reporting agents for MFI interest rate statistics has to be of a size that
ensures reliable euro area and national statistics simultaneously. Trying to achieve
reliable results for the euro area only could lead to a situation where small countries
in particular have too few reporting agents to compile a national set of statistics,
which would conflict with the aim of providing information about the level and
development of interest rates both at euro area and at national level. So, for the
purpose of MFI interest rate statistics, the minimum national sample size is defined.
The aggregate of the national samples is then big enough to derive reliable data on
the level and development of interest rates in the euro area.
The Regulation gives flexibility to NCBs, while ensuring that the national results are
comparable and both the national and euro area results are of high quality. It defines
that the minimum national sample size in the case of random sampling selection
shall be such that the maximum random error at
=
national level for interest rates on new
business on average over all instrument categories does not exceed 10 basis
points
153
at a confidence level of 90%, where:
•
is a factor computed from the normal distribution or any suitable
distribution according to the structure of the data (e.g. t-distribution) assuming a
confidence level of 1-α
• (
) is the variance of the estimator of parameter θ and
•
(
) is
the estimated variance of the estimator of parameter θ
But the NCBs may opt to select the minimum national sample size through the
selection of the largest institutions within each stratum. In this case, the minimum
national sample size has to comply with a quality measure which is a function of the
estimated mean absolute value of the errors. More precisely, the sample quality
should be based on a synthetic mean absolute error (MAE). The actual synthetic
MAE should not exceed a time-varying threshold assuming a 10 basis point error
difference in each stratum and indicator.
154
The synthetic MAES for a given estimator
in a particular period are defined as:
152
See also paragraphs 16 to 24 of Part 14 of Annex II to the Guideline.
153
The absolute measure of 10 basis points at a confidence level of 90% may be directly translated into a
relative measure in terms of the acceptable maximum variation coefficient of the estimator.
154
More details about this method for selecting the sample can be found in “Quality measures in non-
random sampling: MFI interest rate statistics”, Statistics Paper Series, No 3, European Central Bank,
September 2013 Statistics Paper Series
Manual on MFI interest rate statistics − Selection of the reporting agents
136
=
(
(
(
)
)
[17]
with:
•
(
) as the synthetic MAE
•
,
as the volume in a particular MFI interest rate category
•
as the average interest rate estimated in category c
•
=
(
(
)(
)
as the MAE for a given MIR category on the
basis of estimator
•
as the volume corresponding to the actual non-reporting in a particular
stratum j
•
as the volume corresponding to the actual reporting in a particular stratum j
(the process of grossing up is further described in Section 12.1)
• as the total volume for all strata, i.e. the sum of
and
across all strata
•
= (
+
)
+
as the estimation of the total
error within a stratum j
•
as the weighted average interest rate corresponding to the actual reporting
in a particular stratum j
•
as the value of the estimator
for the take-none sub-stratum of stratum j (in
the event of zero volume coverage in one of the reported strata, the average
of the other stratum should be used to avoid a MAE equal to zero)
•
as the average of the first and third quartiles within the stratum, which are
defined as the interest rate reported for the MFI interest rate statistics category
for which 25% and 75%, respectively, of the reported interest rates are lower
than that number. The first and third quartiles are calculated by previously
weighting the volume in that category by the institutions in the stratum. Hence,
the average between the two MAE estimators – the first (Q1) and the third (Q3)
MAE estimators – needs to be used as an estimation for the parameter
.
155
The maximum random error and the synthetic MAE should be calculated separately
for new business and for outstanding amounts. For new business, the maximum
random error and the synthetic MAE should be calculated on the basis of indicators
2 to 4, 8 to 11, 13 to 22, and 24 to 29 as described in Appendix 2 of Annex I to
Regulation ECB/2013/34. For outstanding amounts, the maximum random error and
the synthetic MAE should be calculated on the basis of indicators 1 to 14 as
described in Appendix 1 of Annex I to Regulation ECB/2013/34.
155
Note that Tables 1 and 2 in “Quality measures in non-random sampling: MFI interest rate statistics”,
Statistics Paper Series, No 3, European Central Bank, September 2013 Statistics Paper Series
highlight the results of the synthetic MAE for the first and the third quartile estimators applied in each
country.
Manual on MFI interest rate statistics − Selection of the reporting agents
137
The minimum national sample size defined in the Regulation refers to the initial
sample and the sample after maintenance. Due to the effect of mergers and leavers,
the sample might shrink over time, but it needs to be refreshed at least in the next
maintenance period.
156
When selecting new reporting agents, NCBs need to
consider that they will require some time to implement the reporting system for MFI
interest rate statistics. Hence it is accepted that between maintenance periods the
sample might drop below the threshold given by the minimum national sample size.
12.6 Special provisions in the case of group reporting
The reference reporting population for MFI interest rate statistics includes all MFIs
except central banks and money market funds that take deposits from and grant
loans to households and non-financial corporations identified in the list of MFIs.
However, as mentioned in Section 9.2, NCBs may allow MFIs which are resident in a
single national territory and individually included in the list of MFIs to report MFI
interest rate statistics together as a group. Such groups could, for example, be the
Rabobanks in the Netherlands or the Caixas de Crédito Agrícola Mútuo in Portugal.
The group becomes a notional reporting agent and has the same reporting
requirements as the other (individual) MFIs that are reporting agents.
The counting of the number of institutions in the reference reporting population must
be consistent with the counting in the minimum sample size. For example, if the 400
or so Rabobanks in the Netherlands are reporting together as a group, they should
either be counted as one entity in both the reference reporting population and the
sample or they should each be counted individually in both the reference reporting
population and the sample.
Groups that are reporting together need to report to the NCB each year, for each
instrument category, the number of reporting institutions and the variance of interest
rates across these institutions in the group. The number of reporting agents and the
variance must refer to the month of October and be transmitted with the October
data.
157
The reporting of variances in the group is intended to compensate for the
loss of information that results from reporting as a group, and is required by NCBs so
they can estimate the total variance of interest rates in the national reference
reporting population. The variance of interest rates across individual reporting agents
is likely to be higher than the variance across groups. The reason is that each group
only reports an average interest rate for the whole group and these rates are
supposed to be more homogeneous than the rates offered by the individual MFIs in
the group. The average of a group could, for example, even out regional differences
if those exist. If group reporting leads to an underestimation of the variance of
interest rates in the national reference reporting population, then the required sample
size would also be underestimated, as ceteris paribus the sample needs to be
smaller (larger) if the variance is lower (higher). Likewise, the sampling error would
156
Further discussed in Section 12.8.
157
October was chosen as a “normal” month, because it is not a month of quarterly production as are
March, June, September and December, and not a “holiday” month.
Manual on MFI interest rate statistics − Selection of the reporting agents
138
be underestimated. So it is necessary to report variances within the group to achieve
comparable results for all Member States based on the same legal definition of a
reporting agent.
12.7 Estimation of total new business volume
158
MFI interest rates for the euro area are compiled as weighted averages of the
interest rates applied in the Member States. For interest rates on outstanding
amounts, the weighting information is derived from the MFI balance sheet statistics.
For interest rates on new business, the new business amount per instrument
category and Member State is needed to compute weighted average new business
interest rates for the euro area. The amount of new business is collected from the
reporting agents together with the interest rates. Assuming that a sample is applied,
the total amount of new business per Member State and instrument category needs
to be estimated from the results in the sample by applying expansion, raising or
inflation factors. The estimation of the population total is also referred to as grossing-
up.
159
This section describes the grossing-up procedure for the amount of new business,
including the computation of selection probabilities and their use as expansion
factors. Generic terms are used because the precise formulae for a Member State
depend on the NCBs’ choice of strata, the allocation of the sample across the strata,
the method for drawing the reporting agents per stratum and also, in the case of the
initial sample, on the auxiliary information available. First, the estimation of the
population total from amounts derived by simple random sampling is explained.
Then, the more complex procedure for sampling with probability proportional to size
is illustrated, which also applies to the selection of the largest institutions within each
stratum. As explained in Section 12.4, both the random selection and the selection of
the biggest institutions in a stratum are treated as random procedures for the
purpose of MFI interest rate statistics. Therefore, the same general method may be
used to estimate the total amount of new business in the reference reporting
population from the results of the sample.
In the case of simple random sampling within a stratum, each MFI has the same
chance of being selected. In each draw, the procedure gives an equal chance of
selection to every institution in the reference reporting population, i.e. to each
institution that has not already been drawn. Hence, each institution has a known
probability of selection. Before the first draw, the probability of selecting institution i
is:
=
[18]
158
Already mentioned in Section 9.3. See also Section 4 of Part 14 of Annex II to the Guideline.
159
No grossing-up is required for simple averages and ratios, because it is assumed that the estimate
from the sample is also the estimate for the population.
Manual on MFI interest rate statistics − Selection of the reporting agents
139
with n as the size of the sample and N as the size of the reference reporting
population. The inverse of this selection probability is then used as the expansion
factor to estimate the total amount of new business in the population from the sample:
=
[19]
The Horvitz-Thompson estimator for the population total derived from a sample is:
=
[20]
with
as the estimated total amount of new business in the reference reporting
population, yi as the amount of new business of institution i and
as the probability
of selecting institution i. In the case of simple random sampling, the estimator from
Equation 26 for the total amount of new business in the population becomes:
=
=
=
[21]
The starting point for sampling with probability proportional to size is also the
calculation of the selection probability
for each institution i. The selection
probabilities are computed separately for each stratum from the fixed sample size nh
per stratum; the variable Ui indicates the size of the institution i, with
being the
size of institution i as a share of all institutions:
=
=
[22]
Variable U signifies auxiliary information, which can differ between Member States. It
is up to NCBs to define the most suitable national variable U, which must be strongly
correlated with the amount of new business, i.e. variable Y, in that Member State. For
example, Ui could be the outstanding amount for each instrument category for
institution i from the MFI balance sheet statistics. An important restriction for the
calculation of the selection probabilities is that the product of the fixed sample size
nh and the size Ui of institution i has to be smaller than the total of all institutions in
the same instrument category:
<
[23]
This restriction guarantees the coherence of the selection probabilities. If the
restriction of Equation 29 is not fulfilled for institution i, this institution is selected
automatically. The probability of its selection is hence set as 1. If Equation 29 is not
fulfilled, the other selection probabilities are recalculated based on the exclusion of i
with the new restriction:
(
1
)
<
[24]
If this restriction is again not fulfilled, institution k is selected automatically and its
selection probability is set to 1. The other probabilities are then calculated as follows:
(
2
)
,
<
,
[25]
etc.
Manual on MFI interest rate statistics − Selection of the reporting agents
140
Table 12 gives an example of the calculation of the selection probabilities in the case
of sampling with probabilities proportional to size.
Table 12
(N=15; n=4)
Institution Size Xi Xi/(sum Xi) Xi/(sum Xi-A) 4*Xi 3*Xi
Selection
probability
A
16,271,700 0.3414 65,086,800
1
B
9,293,500 0.195 0.296 27,880,500
0.8881
=3*0.2960
C
6,627,700 0.1391 0.2111 19,883,100
0.6334
=3*0.2111
D
4,822,600 0.1012 0.1536 14,467,800
0.4609
E
3,420,100 0.0718 0.1089 10,260,300
0.3268
F
2,975,100 0.0624 0.0948 8,925,300
0.2843
G
2,418,600 0.0507 0.077 7,255,800
0.2311
H
669,732 0.0141 0.0213 2,009,196
0.064
I
470,700 0.0099 0.015 1,412,100
0.045
J
260,775 0.0055 0.0083 782,325
0.0249
K
203,760 0.0043 0.0065 611,280
0.0195
L
126,591 0.0027 0.004 379,773
0.0121
M
47,600 0.001 0.0015 142,800
0.0045
N
34,987 0.0007 0.0011 104,961
0.0033
O
20,000 0.0004 0.0006 60,000
0.0019
Sum Xi
47,663,445
Sum Xi-A
31,391,745
As in the case of simple random sampling, the Horvitz-Thompson estimator in
Equation 18 is used to estimate the total amount of new business Y in the
referencereporting population from the result of the sample. Hence, inverses of the
selection probabilities calculated by means of Equations 28 to 31, i.e.
, are used as
expansion factors. Assuming that there is a strong statistical relationship between
the size of the institutions U and the amount of new business Y, the Horvitz-
Thompson formula provides an unbiased estimate of the population total, i.e. the
total amount of new business in the reference reporting population. It is important to
note that
, i.e. the size of institution i as a share of all institutions, in Equation 28,
changes over time. In order to derive an unbiased estimate of the total amount of
new business from the sample,
and hence
need to be recalculated each month.
If the selection of the largest institution approach is used, the expansion factors for
each stratum j are defined as the inverse of the stratum coverage ratio by means of
the following formula:
=
=
=
[26]
Manual on MFI interest rate statistics − Selection of the reporting agents
141
with:
•
as the total volume within stratum j
•
as the volume within each stratum j for the institution i
•
as the number of credit institutions not sampled in the stratum j
•
as the number of credit institutions sampled in the stratum j
Expansion factors EFj as defined in the previous paragraph in respect of new
business are calculated by replacing new business volumes by the related
outstanding amounts. The grossed-up volume of stratum j is then calculated as the
expansion factor for stratum j multiplied by the reported volume for stratum j.
MFI interest rate statistics are based on a selection without replacement, i.e. each
MFI can only be selected once. In random sampling with probability proportional to
size, care has to be taken that the selection probabilities are proportional to the size
of the institutions after each draw. Hence, after the first institution is drawn, the
selection probabilities of the remaining institutions need to be adjusted, the same
after the second institution is drawn, etc. For stratified random sampling, practical
methods have been developed. If the largest institutions within each stratum are
chosen, a selection with replacement can be assumed. In this case, the selection
probabilities only need to be calculated once before the first draw.
Each Member State either carries out a census or has one sample for MFI interest
rate statistics covering all instrument categories.
160
It is hence possible that a
reporting agent has no business in some instrument categories. For example, a
sample comprises 50 reporting agents, but for one instrument category, only 30 of
them carry out business and have outstanding amounts. The outstanding amount,
however, is used as auxiliary variable U indicating the size of the institutions. For
estimating the total amount of new business for this instrument category, the
selection probabilities and hence expansion factors only take into account the 30
reporting agents that carry out business and have outstanding amounts. Whether the
reporting agents also have new business or not in that month is irrelevant in
calculating the selection probabilities and expansion factors.
12.8 Maintenance of the sample
161
NCBs that choose the sampling approach must ensure that the sample remains
representative over time. Maintenance of the sample over time is inherent to panel
surveys that aim to collect two or more measures from the same sample units over
time. The fact that the same units report repeatedly over time ensures the
consistency of the answers, also known as the panel effect.
162
160
See also Section 12.9.
161
See also paragraphs 25 to 30 of Part 14 of Annex II to the Guideline.
162
See also Section 12.9.
Manual on MFI interest rate statistics − Selection of the reporting agents
142
The selection procedure for a panel does not differ from the selection of any other
sample. The same methods are applied, but adapted to the panel situation as
necessary. What is difficult to capture in a panel, but of no importance in a one-off
sample, is the variability of the structure of the reference reporting population
compared with the actual reporting population, i.e. the reporting agents. The sample
must therefore be adjusted for joiners of the reference reporting population, for
leavers from the reference and actual reporting population, as well as for changes in
the characteristics of the reporting agents. The following table describes the
procedures that need to be followed:
Table 13
Maintenance of the sample over time
t1 Leavers Joiners t2
Panel of reporting agents
n1 nd (nb) n2 nc = n1 - nd
Reference reporting population
N1 Nd Nb N2 = Nc + Nb Nc = N1 – Nd
Selection probability
π1i π’2i π2i
At time t
1
of the drawing of the initial sample, n
1
institutions of the total N
1
have been
selected as reporting agents. At time t
2
, the reference reporting population has
changed to N
2
institutions consisting of the common institutions Nc, which are
included in the reference reporting population at t
1
and at t
2
, as well as the joiners of
the reference reporting population N
b
during the period from t
1
to t
2
. The common
institutions N
c
consist of all institutions N
1
in the reference reporting population at t
1
less the leavers N
d
during the period from t
1
to t
2
.
If the initial sample n
1
was drawn as a simple random sample and the sample n
2
not
adjusted for joiners and leavers, then the sample n
2
includes only the common
institutions nc, i.e. the n
1
institutions of the initial sample less the leavers nd. Hence,
although the structure of the reference reporting population changes from t
1
to t
2
, this
is not reflected in the sample. The situation is as follows:
=
=
[27]
The sample n
2
needs to be adjusted for the joiners of the reference reporting
population to remain representative of the reference reporting population over time.
In order to do so, it is necessary to draw a sample nb from the population of common
institutions (N
c
) plus all joiners (N
b
) The sample n
2
then includes the common
institutions nc, i.e. the n1 institutions of the initial sample less the leavers nd, as well
as the joiners nb in the period from t
1
to t
2
. In this way, the changes in the structure of
the reference reporting population resulting from joiners from t
1
to t
2
are also
reflected in the sample. The complementary selection of joining institutions nb
among the total number of joiners Nb is referred to as incremental sampling over
time. The situation is then as follows:
Manual on MFI interest rate statistics − Selection of the reporting agents
143
=
+
=
+
+
+
=
[28]
=
1
+
1
The sample n
2
also needs to be adjusted for the leavers from the reference and the
actual reporting population. No adjustment is necessary if there is proportionality
between the leavers in the reference reporting population Nd and the leavers in the
sample nd (Case 1). If the institutions are leaving the reference reporting population
and these institutions are not in the sample, the sample becomes too large relative to
the size of the reference reporting population (Case 2). If relatively more institutions
leave the sample than the reference reporting population, the sample becomes too
small over time and might cease to be representative (Case 3). These three
situations are illustrated in Figure 5, where the circle signifies the reference reporting
population N, the uncoloured square the sample n and the dark square the leavers in
the period from t
1
to t
2
:
Figure 5
Three cases of the sample compared with reporting population
Case 2: Leavers under-represented in the sample
Case 3: Leavers over-represented in the sample
Whereas in Case 1 the number of leavers from the reference reporting population
can be estimated using the formula
1
, this is not possible in Cases 2
and 3. In Case 2 the number of leavers in the population is underestimated and in
N
n
leavers
N
N
n
leavers
Case 1: Proportionality
N
n
leavers
Manual on MFI interest rate statistics − Selection of the reporting agents
144
Case 3 overestimated. If, in Cases 2 and 3, the inverses of the selection probabilities
were to be used as expansion factors, this would lead to biased results for the
population total. Therefore, in Cases 2 and 3 the weights attached to each reporting
agent in the sample must be adjusted, for example by means of post stratification,
i.e. stratification after the selection of the sample. The weight attached to each
reporting agent is the inverse of its selection probability and hence the expansion
factor for estimating the population total. With post stratification, the sample is
restratified. New selection probabilities and hence weights are allocated.
Even in Case 2, where the sample is relatively too big for the reference reporting
population, none of the common institutions nc is taken out of the sample. This
means that an institution that once implemented the MFI interest rate reporting
scheme will only be relieved of the reporting burden if it leaves the list of MFIs or if
the NCBs in agreement with the ECB decide to reselect all reporting agents.
Finally, the sample needs to be adjusted for changes in the characteristics of the
reporting agents. These changes can occur because of mergers, divisions, growth of
the institution, etc. Some reporting agents might change the stratum. As in Cases 2
and 3 for leavers, the sample needs to be adjusted, for example by means of post
stratification. In this case, the sample is restratified and new selection probabilities
and hence weights are allocated.
According to the Regulation, NCBs that choose the sampling approach must check
the representativity of their sample at least every year. If there are significant
changes in the reference reporting population, these are reflected in the sample after
the annual check. At intervals of at most three years, the sample must be refreshed
to take account of joiners, leavers and other changes in the characteristics of the
reporting agents. NCBs may check and refresh their sample more often. Such
adjustments of the sample over time are statistically necessary to ensure the quality
of panel surveys; in general, they will not lead to breaks in the time series.
The ECB leaves it to the discretion of the NCB when in the year it reviews the
national sample. The ECB also leaves it to the discretion of the NCB how much time
it grants to new reporting agents for implementing the reporting requirements, but
expects that there are not more than 12 months between the identification of a new
reporting agent and the first reporting of data.
12.9 Further sampling issues
163
To achieve consistency between MFI interest rate statistics on (i) outstanding
amounts referring to deposits, (ii) outstanding amounts referring to loans, (iii) new
business referring to deposits and (iv) new business referring to loans, NCBs should
use the same sample of MFIs for collecting these sets of statistics. It is, however,
possible to use a sample for a subset of MFI interest rate statistics and a census for
the rest, for example a sample for new business and a census for outstanding
163
See also paragraphs 31 and 32 of Part 14 of Annex II to the Guideline.
Manual on MFI interest rate statistics − Selection of the reporting agents
145
amounts, or a sample for new lending business and a census for new deposit
business and outstanding amounts. The Regulation does not allow the use of two or
more different samples.
By definition, the interest rates on outstanding amounts cover all deposits placed and
not yet withdrawn and all loans withdrawn and not yet repaid by customers in all the
periods up to and including the reporting date. Moreover, as the MFI interest rates on
new business and on outstanding amounts are weighted averages, not only the
interest rates but also the associated quantities are interlinked. The fact that the
same MFIs report the statistics both on outstanding amounts and new business and
do so repeatedly period after period leads to a panel.
164
This effect ensures the
consistency of answers for the two sets of statistics at one particular time (cross-
sectional analysis) and for each set of statistics over time (time-series analysis). If
two samples are drawn, one for the reporting of new business and the other for that
of outstanding amounts, sampling effects might lead to a different composition of the
two actual reporting populations, which might then lead to discrepancies between the
two sets of statistics.
NCBs need to cover each instrument category that exists in the banking business of
resident MFIs with euro area households and non-financial corporations, but not
each product offered at national level.
165
An instrument category is inapplicable at
national level only if MFIs do not offer any such products to resident non-financial
corporations and households. NCBs must provide data if some business exists,
however limited this business may be. Hence, if an instrument category is only
offered by one institution, then this institution needs to be represented in the sample.
If an instrument category did not exist in a Member State at the time of the initial
drawing of the sample, but afterwards one institution introduces a new product
belonging to this instrument category, then this institution needs to be added to the
sample at the time of the next representativity check. If a new product is created that
belongs to an existing instrument category, the institutions in the sample need to
cover it with the next reporting, as all reporting agents are required to report for each
instrument category all interest rates applied to all the products that come under this
category.
Under certain conditions, sampling is allowed at the level of branches rather than at
the level of MFIs. Sampling at the level of branches can be undertaken for one or
more of the strata in the sample. The first precondition is that the NCB decides on a
census for that stratum, i.e. that all MFIs in the stratum are subject to reporting. The
second precondition is that the NCB has a full list of branches that covers the entire
business of the MFIs in the stratum. The third precondition is that the NCB has
appropriate data to assess the variance of interest rates on new business with
households and non-financial corporations across branches, and based on this
variance the number of branches that should report. The selected branches become
notional reporting agents and hence have the same reporting obligations as MFIs
that are selected as reporting agents. Reporting at branch level does not affect the
164
See also Section 12.8.
165
See also Section 7.2.
Manual on MFI interest rate statistics − Selection of the reporting agents
146
liability as a reporting agent of the MFI to which a branch belongs, i.e. if a branch
reports incorrectly or fails to report, the institution included in the list of MFIs is liable,
rather than the branch.
12.10 Description of the euro area sample
166
In the euro area, around half of the Member States apply a sampling approach for
selecting the reporting agents, whereas the rest (mostly smaller countries) apply a
census or almost a census, since only a set of very small MFIs belonging to the
reference reporting population is excluded. Hence, only a subset of the reference
reporting population has to implement the requirement of the Regulation and
provides data for MFI interest rate statistics.
NCBs draw on a variety of stratification criteria in order to divide the reference
reporting population into homogeneous strata prior to drawing the sample of
reporting institutions that forms the actual reporting population. The variety of
stratification criteria reflects the diversity of the banking business in the euro area.
Some Member States stratify with respect to bank categories; others apply regional
components to group the MFIs in the reference reporting population. Furthermore,
the type of product and customer, the degree of specialisation, the size of the
institution and the number of branches are used to build up homogeneous strata.
Some countries use principal component or factor analysis to determine the relevant
stratification criteria and apply cluster analysis to establish the strata.
After setting up the most convenient stratification, NCBs select the reporting
institutions within each stratum. As stated in Section 10.4, there are three
possibilities: (a) to include all institutions in the stratum; (b) to draw the institutions
following a random sample method with equal probability for all institutions or with
probability proportional to size; or (c) to select the largest institutions per stratum.
Almost all NCBs adopting a sample select as reporting agents the largest institutions
within each stratum. The procedure is often combined with, or leads to, a census in
at least one stratum in such a way that all main MFIs in terms of size are included in
the sample. In several cases, the application of a census in some strata has the
purpose of capturing all institutions specialised in some type of business. Few NCBs
select reporting agents randomly with probability proportional to size.
Some Member States allow group reporting. Therefore, some of their MFIs, which
are individually included in the list of MFIs, report MFI interest rate statistics together
as a group, i.e. as if they were a single MFI. The group becomes a notional reporting
agent.
The selection procedure leads to a euro area sample of nearly 30% of all institutions
included in the reference reporting population. Despite this limited percentage, the
coverage in terms of deposits and loans is broad. With regard to deposits, the euro
166
A very useful complement to this section is “Quality measures in non-random sampling: MFI interest
rate statistics”, Statistics Paper Series, No 3, European Central Bank, September 2013.
Manual on MFI interest rate statistics − Selection of the reporting agents
147
area sample covers about 80% of the stock of euro-denominated deposits received
from households resident in the euro area and the same percentage of the ones
received from non-financial corporations resident in the euro area. Regarding loans
granted, the corresponding percentages are slightly above 85% in both sectors.
Manual on MFI interest rate statistics − Derived indicators
148
13 Derived indicators
13.1 Reason for including this chapter in the Manual
As outlined in Chapter 1, the aim of the Manual is to be a useful tool mainly to assist
the compilers of statistics on MFI interest rates and the reporting agents, since it
contains explanations complementary to the Regulation and the Guideline as well as
detailed examples aiming at clarifying doubtful or complex cases. However, as
experience has demonstrated, the Manual is also very useful to other users and the
public at large to illustrate and make transparent the work on the compilation of
indicators derived from MFI interest rate statistics. Focusing on this use, the Manual
should offer some explanations on the derived indicators that statisticians and other
users regularly use to make their analysis, in particular on the relevance of these
indicators and the way they are calculated.
As the Manual shows, the MFI interest rate statistics contain information that is very
wide in scope. The wide range of instrument series available allows users to analyse
a very broad set of characteristics of retail interest rates on deposits and loans vis-à-
vis households and non-financial corporations. However, at the same time, this
richness of information constitutes a potentially serious obstacle when it is necessary
to have a general overview of interest rates in different EU countries. In particular,
the differences that still exist in the various features of the respective financial
systems across euro area countries call for the building of indicators which allow the
users to compare interest rates in the various countries making up the euro area.
That said, when analysing the changes of MFI interest rates in the euro area as a
whole, it is necessary to go a step further to correctly assess the causes (changes in
interest rates or changes in country weights) of such changes. To facilitate these
sorts of analysis, several derived indicators such as Bennet binary indices for rates
and weights, Bennet chain indices, coefficients of variation, and cost-of-borrowing
indicators are regularly used by the ECB as well as other users. The following
sections explain further the main reasons for the use of these derived indicators and
the way they have been calculated. All these indicators are available in the ECB
Statistical Data Warehouse (SDW) at the euro area level, and the cost-of-borrowing
indicators are also published at the national level.
13.2 Bennet binary indices for rates and weights
As stated in Chapter 9, the final level of aggregation to obtain interest rates for the
euro area for each instrument category is carried out by the ECB. The euro area
interest rates are weighted averages of national MFI interest rates of euro area
countries using the respective business volumes as weights. When this stage is
finished, the analysis of the euro area data begins. The first objective is to observe
and assess the changes in the MFI interest rates of the different instruments in the
euro area as a whole. The euro area interest rates are calculated as a weighted
average of the corresponding interest rates in each of the euro area Member States,
Manual on MFI interest rate statistics − Derived indicators
149
each of them with very different weights in terms of the respective business volumes.
A crucial point for the statisticians and other users is to know which part of the
variations in the euro area interest rates is attributable to a price component (the
pure change in the national interest rates) and which part is explained by the
weights, i.e. changes in the business volumes. Taking, for example, one instrument
category, it may be that the interest rate of this instrument category has not changed
in any Member State for two consecutive periods, but the resulting interest rate for
the euro area displays a change. This can be the case because of changes in
relative business volumes across countries between these two consecutive periods.
When trying to disentangle these two components – the change in the rate of interest
(price component) and in the country weights (quantity component) – the statisticians
of the ECB have to resort to the application of index number theory.
Most of the standard index number theories decompose a value ratio into the
product of a price index and a quantity index. The price index is interpreted as an
aggregate price ratio and the same applies to the other component of the product
(the quantity ratio). However, a change in an interest rate is better understood and
usually communicated in absolute values of the change rather than expressing it as
a percentage. These indices, calculated in terms of differences, were first presented
in the 1920s by T. L. Bennet and J. K. Montgomery who decomposed a value
difference into the sum of a price difference and a quantity difference. In 2005, W. E.
Diewert further developed this theory
167
.
The ways in which a difference in interest rates between two periods can be
decomposed are diverse. Such a difference can be broken down into a “pure”
interest rate change and a “pure” weight change and, in some cases, a mixed or
composite effect of prices and weights. If the total change is only split into two terms
(price and weight effect), this implies that the mixed effect is distributed between
them, either symmetrically or not.
Among these different decomposition techniques, the Marshall-Edgeworth type,
which Diewert calls the “Bennet indicator”, uses the simple average of the previous
and present period weights to calculate the interest rate effect. The weight effect is
calculated in the same way, resulting in a decomposition with only two terms. In this
decomposition, the composite effect is distributed equally between the interest rate
effect and the weight effect. The precise formula is the following:
,
=
()
,
(
)
()
+
()
,
(
)
()
[29]
The ECB uses a Marshall-Edgeworth-type decomposition with an extended weight
effect. Compared with the previous decomposition method, the weight term includes
the difference between the country interest rate and the aggregated (euro area)
interest rate instead of the country interest rate alone, for both subsequent periods.
This expansion does not change the weight effect at the aggregated level, but offers
other more detailed interpretation possibilities. This index is called the Bennet binary
167
More details can be found in Huerga, J. and Steklacova, L., “An application of index numbers theory to
interest rates”, Working Paper Series, No 939, European Central Bank, September 2008.
Manual on MFI interest rate statistics − Derived indicators
150
index, with the term “binary” meaning that the comparison is between two
consecutive periods. The precise calculation method for this index is as follows:
,
=
= ""
,
=
(
)
,
+
(
)
,
[30]
where:
the first addend is the interest rate effect of country k;
the second addend is the weight effect of country k;
K represents all of the euro area countries.
()
,
= ()
,
()
()
[31]
where:
()
,
= ()
()
(
)
=
()
=
()
()
()
,
= ()
()
[32]
where:
()
,
= ()
()
Considering all the decompositions together:
,
=
(
)
,
(
)
(
)
+
(
)
,
(
(()
)(
(
)
)
[33]
where:
I is the MFI interest rate referring to the euro area
i(k) is the national MFI interest rate of country k
w(k)) is the national weight of country k
Therefore, the first part of the sum is the interest rate component of the euro area
month-to-month level change and the second part is the corresponding weight
(business volume) component.
13.3 Bennet chain indices
The Bennet binary index only compares two consecutive periods at a time. However,
when having multiple periods the question arises as to which period should be
compared with which period. Here again, there are multiple alternatives. In pure
Manual on MFI interest rate statistics − Derived indicators
151
conceptual terms, these alternatives range from comparing each period with a direct
index for the whole length of the series, to a comparison of each period with the
consecutive one and “chaining” the results to form a series, covering intermediate
solutions in which the fixed period would change at a certain frequency but not every
period, also implying “chaining” at the time of change in the fixed period.
For MFI interest rates, the price component (the pure interest rate effect) direct index
would be constructed as the interest rate component of the decomposition of the
changes between each month “t” and month “0”. It is noted that between t and 0 a
number of t-1 periods have occurred, with their corresponding rates and weights,
which are not considered in the direct index at time t.
The total change in interest rate
,
is decomposed according to the simplest
Marshall-Edgeworth decomposition:
,
=
()
,
(
)
()
+
()
,
(
)
()
[34]
Therefore, the difference direct index of interest rates on the basis of the interest rate
component of this Marshall-Edgeworth-type decomposition is:
=
()
,
(
)
()
[35]
where DDME indicates the direct difference Marshall-Edgeworth-type (index).
As the index is calculated as the sum of the interest rate effects for all countries, to
better understand it we can focus on one summand, corresponding to the particular
country k.
National contribution of country k:
() = ()
,
(
)
()
[36]
The chain index goes one step further which means that the index is calculated for
each period by comparing the price (the interest rate) level of this period with the
previous period, and then linking the results by multiplying (chaining) each individual
link with the previous one. The procedure is similar to the case of a difference index,
with the particularity that each link is chained to the previous one by adding them.
Chart 11 shows an example of the difference chain index of interest rates on the
basis of the Marshall-Edgeworth-type decomposition:
Manual on MFI interest rate statistics − Derived indicators
152
Chart 11
Euro area, loans to NFCs, up to one million, interest rate fixation up to to one year
(percent per nannum)
13.4 Coefficients of cross-country variation
The differences in the financial system, the economic cycle and the degree of
relevance of the retail interest rate instrument categories across the euro area
Member States are still significant and, as a consequence, the transmission of the
monetary policy from the key operational instruments of the Eurosystem to the
interest rates applied by the different MFIs to households and non-financial
corporations is not uniform. Therefore, for the ECB it is very important to measure
the dispersion of the MFI interest rates of the individual euro area countries with
respect to the euro area interest rate. This measure is applied through the coefficient
of variation. Such a coefficient is calculated for each euro area MFI interest rate
instrument category and is computed as the standard deviation divided by the euro
area interest rate, thus adjusting for the fact that the standard deviation is influenced
by the level of the euro area rate.
The standard deviation is computed as the square root of the weighted variance of
the national MFI interest rates with respect to the euro area interest rate. The
national business volumes serve as country weights.
The concrete formula is as follows:
=
_
__
[37]
The weighted variance is obtained as:
(
)
(
(
)
__
)
[38]
The euro area MFI interest rate is obtained as:
(
)
(
)
[39]
-3
-2.5
-2
-1.5
-1
-0.5
0
0.5
1
1.5
2
interest rate component of euro area month-to-month level change
index on interest rates
Manual on MFI interest rate statistics − Derived indicators
153
where:
()
is the national MFI interest rate level of euro area country k at month t
()
is the national weight of euro area country k at month t, i.e. the volume of
national business in relation to the euro area total
The weighted variance is the squared deviation between the national and euro area
MFI interest rates, calculated according to the national share in the total euro area
business volume for a given instrument category.
By measuring, on a monthly basis, the variation of national interest rates around the
euro area MFI interest rate adjusted for the level of the euro area rate, the coefficient
of variation allows a further comparison of different MFI interest rate indicators.
13.5 Cost-of-borrowing indicators
168
As mentioned in Chapter 2, one of the main uses of MFI interest rate statistics is to
analyse the monetary policy transmission mechanism, since these rates are the final
link from monetary policy actions to the consumption and investment expenditure of
households and non-financial corporations and ultimately prices. Stable, efficient and
integrated financial markets are the basis for a smooth transmission of monetary
policy within a country or across countries belonging to a monetary union like the
euro area.
In the period following the start of Economic and Monetary Union (EMU) at the
beginning of 1999 until the beginning of the financial crisis in September 2008 with
the default of Lehman Brothers, the rates charged by euro area MFIs to households,
mainly for new house purchases, and to non-financial corporations for new loans
recorded a low level of dispersion. Even though a certain degree of heterogeneity in
MFI interest rates still persisted as a consequence of differences in regulatory and
fiscal frameworks across countries, different degrees of competition between banks
or differences in the position in the economic cycle across countries, among other
factors, the level of integration of the financial markets could be qualified as highly
satisfactory
169
.
This situation suddenly changed with the eruption of the financial crisis, which
fragmented the financial markets of the euro area. This complicates the assessment
of the monetary policy transmission mechanism, since in some countries the
expansive monetary policy adopted by the ECB during the crisis was mirrored, more
or less, by the expected correspondence of growth in bank credit to the non-financial
private sectors, while in other countries this variable recorded a much lower
response compared with the results foreseen in periods prior to the crisis. These
168
Further information on the usage of these indicators can be found in the August 2013 ECB Monthly
Bulletin article entitled “Assessing the retail bank interest rate pass-through in the euro area at times of
financial fragmentation” ECB Monthly Bulletin article.
169
For more information on the level of integration of the financial markets in Europe, please refer to the
Financial integration in Europe annual publications Financial integration in Europe.
Manual on MFI interest rate statistics − Derived indicators
154
different effects across countries can be explained as follows: on the one hand, the
pass-through of monetary policy decisions to the real economy experienced before
the crisis cannot explain the levels of heterogeneity in bank lending rates during the
crisis; on the other hand, the relative importance of loan instruments changed with
the crisis.
Regarding the pass-through models (i.e. models where policy interest rates and
market interest rates are considered the most important determinants of retail bank
lending rates), they have failed to correctly measure the rate of adjustment of bank
lending rates to changes in reference rates, because they do not include risk factors
and sovereign debt spreads among the explanatory variables. These two kinds of
variables have had a strong impact on bank lending rates in some countries.
However, users require less volatile instruments than the broad range of MFI interest
rate instrument categories for modelling and forecasting purposes.
With respect to the relative weight of the different loan instruments, the financial
crisis has increased the relative weight of short-term instruments which in a situation
of growing uncertainty cover better credit and interest rate risks than long-term
instruments. Regarding bank loans to non-financial corporations, there has been a
significant increase in overdrafts and other short-term loans in the countries more
affected by the crisis compared with the other countries. Therefore, it is necessary to
use a measure of the borrowing costs of non-financial corporations and households
which is accurate and more comparable across countries.
The reasons mentioned above have led the ECB together with the NCBs to develop
a set of aggregate indicators to measure the cost of borrowing of non-financial
corporations and households.
The composite cost-of-borrowing indicators are based on MFI interest rate statistics.
Thus there are arguments for using monetary policy econometric models with
smoothed rate volatility and which allow the comparison of credit conditions across
euro area countries with different household and NFC borrowing structures.
There are four categories of cost-of-borrowing indicators, which are available in the
ECB SDW at the euro area and the national levels:
1. Cost-of-borrowing indicator for households for house purchase
2. Cost-of-borrowing indicator for non-financial corporations
3. Cost-of-borrowing indicator for short-term loans to households and non-financial
corporations
4. Cost-of-borrowing indicator for long-term loans to households and non-financial
corporations
For most countries, the data are available as of January 2003, although for certain
countries, due to the lack of backdata, the cost-of-borrowing indicators have only
been calculated for a shorter period of time. The indicators are derived following the
methodology explained in the following sections.
Manual on MFI interest rate statistics − Derived indicators
155
13.5.1 Cost-of-borrowing indicator for households for house purchase
The cost of borrowing for households includes only loans for house purchase. Loans
for consumption and other purposes have been excluded as they are very volatile
and less relevant for macroeconomic projections. “Short-term” refers to loans with an
initial period of interest rate fixation of up to one year regardless of maturity.
Accordingly, “long-term” refers to loans with an initial period of interest rate fixation
over one year.
This indicator is calculated as a weighted average of MFI interest rates on short-term
and long-term loans to households for house purchase, where the new business
volumes used are smoothed with a moving average of the previous 24 months’
observations. The precise formula is as follows:
At time t:
=
(
)
(
)
(
)
(
)
(
)
(
)
[40]
where:
is the cost of borrowing for households for house purchase
() are interest rates on new business of loans to households for house
purchase with a floating rate or an initial rate fixation up to 1 year (MFI interest rate
indicator 16)
() are volumes of new business of loans to households for house purchase
with a floating rate or an initial rate fixation up to 1 year (MFI interest rate indicator
16)
() are interest rates on new business of loans to households for house
purchase with an initial rate fixation over 1 year (MFI interest rate indicators 17, 18
and 19)
() are volumes of new business of loans to households for house purchase
with an initial rate fixation over 1 year (MFI interest rate indicators 17, 18 and 19)
13.5.2 Cost-of-borrowing indicator for non-financial corporations
The aggregated cost-of-borrowing indicator for non-financial corporations is
calculated in a similar way to the one for households above, i.e. as a weighted
average of rates on short-term and long-term loans to non-financial corporations.
As regards short-term loans to non-financial corporations, the MFI interest rate data
on new business do not include overdrafts, revolving loans, or convenience and
extended credit. For companies in some euro area countries, these instruments
(mainly overdrafts) are however a significant source of short-term finance. Thus,
since interest rates on overdrafts are, on average, higher than other short-term bank
lending rates, their exclusion tends to lower average short-term rates in these
Manual on MFI interest rate statistics − Derived indicators
156
countries. To improve comparability across countries, for cost-of-borrowing
purposes, revolving loans and overdrafts and extended credit card credit are
incorporated into the calculation of short-term lending rates as described below.
At time t:
=
(
)
(
)
(
)
(
)
[41]
where:
is the cost of borrowing for non-financial corporations
(
)
are interest rates on new business of long-term loans (i.e. loans
with interest rate fixation over one year) to non-financial corporations (MFI interest
rate indicators 39 to 42; 45 to 48; and 51 to 54)
(
)
are volumes of new business of long-term loans (i.e. loans with
interest rate fixation over one year) to non-financial corporations (MFI interest rate
indicators 39 to 42; 45 to 48; and 51 to 54)
is the estimated interest rate on new business of short-term loans to non-
financial corporations adjusted to take into account overdrafts
is the estimated volume of new business of short-term loans to non-financial
corporations slightly inflated to take into account overdrafts, as these are an
important source of short-term funding for non-financial corporations. The way to do
this is to increase the volume of new business in short-term loans by the share of
overdrafts in the total amounts outstanding of short-term loans. When calculating the
total amounts outstanding of short-term loans, long-term loans with a residual
maturity over one year and interest reset below one year are considered short-term
and therefore their amounts are added to the outstanding amounts of short-term
loans as taken from BSI statistics.
The estimated volume is calculated as follows:
=
(
)
× 1 +
(
)
(
)
,
(
)
[42]
=
where:
(
)
are volumes of new business of short-term loans to non-financial
corporations (MFI interest rate indicators 37, 38, 43, 44, 49 and 50)
is the share of overdrafts in the total amounts outstanding of short-term
loans
() is the volume of overdrafts, revolving loans, and convenience and
extended credit (outstanding amounts)
Manual on MFI interest rate statistics − Derived indicators
157
(
)
are volumes of total short-term loans to non-financial corporations
(i.e. loans with contractual maturity up to one year – taken from the BSI statistics)
,
(
)
is the estimated volume of long-term loans with original maturity
over 1 year, residual maturity over 1 year and with interest rate reset within a year
,
(
)
=
,
(
)
(
)
×
(
)
[43]
with:
(
)
as volumes of total long-term loans (loans with original maturity
over 1 year) to non-financial corporations (taken from the BSI statistics)
,
(
)
as the real volume of long-term loans with original maturity over 1
year, residual maturity over 1 year and interest rate reset within a year, taken from
the BSI statistics
The estimated rate
is calculated as follows:
=
(
)
+
(
1
)
(
)
[44]
where:
(
)
are interest rates on overdrafts, revolving loans, convenience and
extended credit to non-financial corporations (MFI interest rate indicators 23 and 36)
(
)
are interest rates on new business of short-term loans to non-
financial corporations (MFI interest rate indicators 37, 38, 43, 44, 49 and 50)
13.5.3 Cost-of-borrowing indicator for short-term loans to households and
non-financial corporations
,
=
(
)
×
(
)
(
)
+
× 1
(
)
(
)
[45]
where all the components of the formula are already defined in the previous sections
13.5.1 and 13.5.2.
13.5.4 Cost-of-borrowing indicator for long-term loans to households and
non-financial corporations
,
=
(
)
(
)
(
)
(
)
(
)
(
)
[46]
where all the components of the formula are already defined in the previous sections
13.5.1 and 13.5.2.
Manual on MFI interest rate statistics − Appendix
158
Appendix
Table 1
Indicators for rates on outstanding amounts
170
Sector Type of instrument Original maturity Residual maturity Interest rate reset
Outstanding amount
indicator number:
Reporting
obligation
Deposits in
EUR
From households With agreed maturity
Up to 2 years 1 AAR
Over 2 years 2 AAR
From non-financial
corporations
With agreed maturity
Up to 2 years 3 AAR
Over 2 years 4 AAR
Repos
5 AAR
Loans in EUR
To households
For house purchases
Up to 1 year 6 AAR
Over 1 and up to 5
years
7 AAR
Over 5 years 8 AAR
For consumption and
other purposes
Up to 1 year 9 AAR
Over 1 and up to 5
years
10 AAR
Over 5 years 11 AAR
Total
Over 1 year
15 AAR
Up to 1 year 16 AAR
Over 1 year In the next 12 months 17 AAR
Over 2 year
18 AAR
Up to 2 years 19 AAR
Over 2 years In the next 24 months 20 AAR
To non-financial corporations
Up to 1 year 12 AAR
Over 1 and up to 5
years
13 AAR
Over 5 years 14 AAR
Over 1 year
21 AAR
Up to 1 year 22 AAR
Over 1 year In the next 12 months 23 AAR
Over 2 years
24 AAR
Up to 2 years 25 AAR
Over 2 years In the next 24 months 26 AAR
170
In the following table “up to” means “up to and including” and “households” include NPISHs.
Manual on MFI interest rate statistics − Appendix
159
Table 2
Indicators on new business
171
Sector Type of instrument
Original maturity, period of notice,
initial rate fixation
New business
indicator number:
Reporting
obligation
Deposits in EUR
From households
Overnight (1) AAR
With agreed maturity
Up to 1 year maturity 2 AAR, amount
Over 1 and up to 2 years maturity 3 AAR, amount
Over 2 years maturity 4 AAR, amount
Redeemable at notice
Up to 3 months notice (5) AAR
Over 3 months notice (6) AAR
From
non-financial
corporations
Overnight (7) AAR
With agreed maturity Up to 1 year maturity 8 AAR, amount
Over 1 and up to 2 years maturity 9 AAR, amount
Over 2 years maturity 10 AAR, amount
Repos
11 AAR, amount
Loans in EUR
To households
Revolving loans and overdrafts (12) AAR
Extended credit card credit (32) AAR
For consumption
Floating rate and up to 1 year initial rate fixation 13 AAR, amount
Over 1 and up to 5 years initial rate fixation 14 AAR, amount
Over 5 years initial rate fixation 15 AAR, amount
For house purchases
Floating rate and up to 1 year initial rate fixation 16 AAR, amount
Over 1 and up to 5 years initial rate fixation 17 AAR, amount
Over 5 and up to 10 years initial rate fixation 18 AAR, amount
Over 10 years initial rate fixation 19 AAR, amount
For other purposes
Floating rate and up to 1 year initial rate fixation 20 AAR, amount
Over 1 and up to 5 years initial rate fixation 21 AAR, amount
Over 5 years initial rate fixation 22 AAR, amount
For other purposes, of
which: Sole proprietors
Floating rate and up to 1 year initial rate fixation 33 AAR, amount
Over 1 and up to 5 years initial rate fixation 34 AAR, amount
Over 5 years initial rate fixation 35 AAR, amount
To non-financial
corporations
Revolving loans and overdrafts (23) AAR
Extended credit card credit (36) AAR
Loans up to an amount of
EUR 0.25 mn
Floating rate and up to 3months period of initial rate fixation 37 AAR, amount
Over 3 months and up to 1 year period of initial rate fixation 38 AAR, amount
Over 1 and up to 3 years initial rate fixation 39 AAR, amount
Over 3 and up to 5 years initial rate fixation 40 AAR, amount
Over 5 and up to 10 years initial rate fixation 41 AAR, amount
Over 10 years initial rate fixation 42 AAR, amount
171
In this table “up to” means “up to and including” and “households” include NPISHs. Furthermore,
“variable rate” is meant as a synonym for “floating rate”, the latter being the expression used in the
Regulation. For indicators 5 and 6, households and non-financial corporations are merged and
allocated to the household sector, since it owns about 98% of the outstanding amount of deposits
redeemable at notice in all participating Member States combined.
Manual on MFI interest rate statistics − Appendix
160
Loans over an amount of
EUR 0.25 mn and up to
EUR 1 mn
Floating rate and up to 3months period of initial rate fixation 43 AAR, amount
Over 3 months and up to 1 year period of initial rate fixation 44 AAR, amount
Over 1 and up to 3 years initial rate fixation 45 AAR, amount
Over 3 and up to 5 years initial rate fixation 46 AAR, amount
Over 5 and up to 10 years initial rate fixation 47 AAR, amount
Over 10 years initial rate fixation 48 AAR, amount
Loans over an amount of
EUR 1 mn
Floating rate and up to 3months period of initial rate fixation 49 AAR, amount
Over 3 months and up to 1 year period of initial rate fixation 50 AAR, amount
Over 1 and up to 3 years initial rate fixation 51 AAR, amount
Over 3 and up to 5 years initial rate fixation 52 AAR, amount
Over 5 and up to 10 years initial rate fixation 53 AAR, amount
Over 10 years initial rate fixation 54 AAR, amount
Table 3
Indicators on new business loans with collateral and/or guarantees
Sector Type of instrument
Original maturity, period of notice,
initial rate fixation
New business
indicator number: Reporting obligation
Loans in EUR
To households
For consumption
Floating rate and up to 1 year initial rate fixation 55 AAR, amount
Over 1 and up to 5 years initial rate fixation 56 AAR, amount
Over 5 years initial rate fixation 57 AAR, amount
For house purchases
Floating rate and up to 1 year initial rate fixation 58 AAR, amount
Over 1 and up to 5 years initial rate fixation 59 AAR, amount
Over 5 and up to 10 years initial rate fixation 60 AAR, amount
Over 10 years initial rate fixation 61 AAR, amount
To non-financial
corporations
Loans up to an amount of
EUR 0.25 mn
Floating rate and up to 3months period of initial rate fixation 62 AAR, amount
Over 3 months and up to 1 year period of initial rate fixation 63 AAR, amount
Over 1 and up to 3 years initial rate fixation 64 AAR, amount
Over 3 and up to 5 years initial rate fixation 65 AAR, amount
Over 5 and up to 10 years initial rate fixation 66 AAR, amount
Over 10 years initial rate fixation 67 AAR, amount
Loans over an amount of
EUR 0.25 mn and up to
EUR 1 mn
Floating rate and up to 3months period of initial rate fixation 68 AAR, amount
Over 3 months and up to 1 year period of initial rate fixation 69 AAR, amount
Over 1 and up to 3 years initial rate fixation 70 AAR, amount
Over 3 and up to 5 years initial rate fixation 71 AAR, amount
Over 5 and up to 10 years initial rate fixation 72 AAR, amount
Over 10 years initial rate fixation 73 AAR, amount
Loans over an amount of
EUR 1 mn
Floating rate and up to 3months period of initial rate fixation 74 AAR, amount
Over 3 months and up to 1 year period of initial rate fixation 75 AAR, amount
Over 1 and up to 3 years initial rate fixation 76 AAR, amount
Over 3 and up to 5 years initial rate fixation 77 AAR, amount
Over 5 and up to 10 years initial rate fixation 78 AAR, amount
Over 10 years initial rate fixation 79 AAR, amount
Manual on MFI interest rate statistics − Appendix
161
Table 4
Indicators on new business loans to non-financial corporations with period of initial rate fixation below 1 year and
original maturity over 1 year
Sector Type of instrument
All loans/
collateralised/guaranteed loans by original maturity
New business
indicator number: Reporting obligation
Loans in EUR
To non-financial
corporations
Loans up to an amount of
EUR 0.25 mn
Floating rate and up to 1 year period of initial rate fixation,
with original maturity over 1 year
80 AAR, amount
Floating rate and up to 1 year period of initial rate fixation,
with original maturity over 1 year, only
collateralised/guaranteed loans
81 AAR, amount
Loans over an amount of
EUR 0.25 mn and up to
EUR 1 mn
Floating rate and up to 1 year period of initial rate fixation,
with original maturity over 1 year
82 AAR, amount
Floating rate and up to 1 year period of initial rate fixation,
with original maturity over 1 year, only
collateralised/guaranteed loans
83 AAR, amount
Loans over an amount of
EUR 1 mn
Floating rate and up to 1 year period of initial rate fixation,
with original maturity over 1 year
84 AAR, amount
Floating rate and up to 1 year period of initial rate fixation,
with original maturity over 1 year, only
collateralised/guaranteed loans
85 AAR, amount
Table 5
Indicators 30 and 31 referring to the APRC
Sector Type of instrument New business indicator number: Reporting obligation
Loans in EUR
To households
172
For consumption 30 APRC
For house purchases 31 APRC
Table 6
Indicators on new business renegotiated loans
Sector Type of instrument Original maturity, period of notice, initial rate fixation
New business
indicator number: Reporting obligation
Loans in EUR
To households
For consumption Total 88 Amount
For house purchase Total 89 Amount
For other purposes Total 90 Amount
To non-financial corporations Total 91 Amount
172
In general including NPISHs, but NCBs may grant derogations in this respect.
Manual on MFI interest rate statistics − Appendix
162
Table 7
New loans to non-financial corporations (indicators required by the Guideline)
Sector Type of instrument Initial period of interest rate fixation
New business
indicator number Reporting obligation
Loans in EUR
To non-financial
corporations
Loans up to an amount of
EUR 1 million
Floating rate and up to 1 year period of initial rate fixation 24 AAR/NDER, amount
Over 1 and up to 5 years period of initial rate fixation 25 AAR/NDER, amount
Over 5 years period of initial rate fixation 26 AAR/NDER, amount
Loans over an amount of
EUR 1 million
Floating rate and up to 1 year period of initial rate fixation 27 AAR/NDER, amount
Over 1 and up to 5 years period of initial rate fixation 28 AAR/NDER, amount
Over 5 years period of initial rate fixation 29 AAR/NDER, amount
Table 8
Revolving loans and overdrafts and convenience and extended credit card credit
(indicators required by the Guideline)
Sector Type of instrument
New business
indicator number Reporting obligation
Loans in EUR
To households Revolving loans and overdrafts, convenience and extended credit card credit 86 AAR/NDER, amount
To non-financial
corporations
Revolving loans and overdrafts, convenience and extended credit card credit 87
AAR/NDER, amount
Table 9
Interest rates on renegotiated loans to households and non-financial corporations
(indicators required by the Guideline)
Sector Type of instrument
Original maturity, period of notice, initial period of
interest rate fixation
New business
indicator number Reporting obligation
Renegotiated
loans in EUR
To households
For consumption total 88 AAR/NDER
For house purchase total 89 AAR/NDER
For other purposes total 90 AAR/NDER
To non-financial corporations total 91 AAR/NDER
Manual on MFI interest rate statistics − Index of terms
163
Index of terms
A
agio .................................................................................................................................. 24
annual percentage rate of charge (APRC) ........................................................... 15, 31, 67
national legislation ....................................................................................................... 36
annualised agreed rate .............................................................................................. 15, 29
annualising interest rates ................................................................................................. 14
B
bad loans ................................................................................................................... 39, 72
bank overdrafts .............................................................................................. 22, 34, 63, 68
Business coverage ......... See MFI interest rates on new business or outstanding amounts
C
capital certainty, lack of ..................................................................................... 98, 99, 102
ceiling .............................................................................................................................. 59
census ............................................................................................................................. 72
charges
in MFI interest rate statistics .................................................................................. 15, 27
in the APRC ................................................................................................................. 17
indicator of ................................................................................................................... 33
consumer credit ......................................................................................................... 33, 72
Consumer Credit Directive ................................................................................... 30, 32, 36
converting the deposit into shares ................................................................................... 98
cooling off period ............................................................................................................. 61
credit cards ...................................................................................................................... 77
credit institutions ................................................................................................................ 9
D
deposit comprising two components ................................................................................ 99
deposits redeemable at notice ................................................................. 41, 45, 46, 76, 77
deposits with agreed maturity .......................................................... 48, 52, 76, 85, 94, 106
derivative contract...................................................................................................... 59, 99
disagio ....................................................................................................................... 15, 23
E
effective interest rate ....................................................................................................... 14
euro area ........................................................................................................................... 9
euro area sample ........................................................................................................... 132
expansion factors........................................................................................... 113, 129, 138
external index .................................................................................................................. 59
F
favourable rates ............................................................................................................... 26
Manual on MFI interest rate statistics − Index of terms
164
fees .................................................................................................................. See charges
fixed interest rates ............................................................................................... 35, 53, 86
floor .................................................................................................................................. 59
frequency
of interest payments ............................................................................................... 14, 20
of interest rate statistics ................................................................................................. 9
of repayments of principal ............................................................................................ 21
frequency of ........................................................ See frequency:of repayments of principal
G
grossing-up ....................................................................................... See expansion factors
group reporting ...................................................................................................... 137, 146
H
Horvitz-Thompson estimator .......................................................................................... 139
households .................................................................................................................. 9, 73
Huygens theorem .......................................................................................................... 130
I
implicit rates ............................................................................................................... 44, 68
incremental sampling ..................................................................................................... 142
indefinite loans ................................................................................................................. 22
initial period of fixation ......................................................................................... 35, 53, 86
institutional arrangements .............................................................................................. 114
interest rates ..................................................................................... See MFI interest rates
L
loan for debt restructuring ................................................................................................ 39
loan in tranches ......................................................................................................... 55, 86
loan offer .......................................................................................................................... 61
loans for debt restructuring ........................................................................................ 39, 72
loans to households for house purchases ............................................... 34, 35, 38, 82, 87
Loans to non-financial corporations ................................................................................. 81
M
Matured deposit ............................................................................................................... 50
maximum random error ................................................................................................. 136
measurement error ........................................................................................................ 128
methodological notes ..................................................................................................... 114
MFI interest rate statistics
coverage of .................................................................................................................... 8
scope of ......................................................................................................................... 9
uses of ........................................................................................................................... 9
MFI interest rates
linked to share price ..................................................................................................... 99
on new business ...................................................................................... 43, 46, 70, 109
on outstanding amounts ..................................................................................... 9, 71, 83
on the amount granted but not yet withdrawn .............................................................. 57
Manual on MFI interest rate statistics − Index of terms
165
monetary financial institutions (MFIs) ................................................................................ 7
list of ............................................................................................................................ 12
Moratorium on a loan ....................................................................................................... 62
mortgage loans ............................................ See loans to households for house purchases
N
narrowly defined effective rate (NDER) ............................................................... 14, 15, 17
national conventions ........................................................................................................ 29
new business ............................................................... See interest rates on new business
new product ............................................................................................................. 72, 145
non-financial corporation sector ....................................................................................... 73
non-financial corporations ............................................................................................ 9, 13
non-negotiable debt security ............................................................................................ 24
non-profit institutions serving households .................................................................. 37, 75
notional reporting agent ................................................................................................. 113
O
one-off deposits ............................................................................................................... 23
original maturity ............................................................................................... See maturity
other loans to households ................................................................................................ 78
outstanding amounts ................................. See MFI interest rates on outstanding amounts
overnight deposits............................................................................ 29, 41, 43, 68, 76, 106
P
panel surveys ................................................................................................................ 141
participating Member States .............................................................................................. 7
period of notice ................................................................................................................ 85
population total ................................................................................. See expansion factors
preliminary offers ............................................................................................................. 61
prime rates ....................................................................................................................... 14
prolongations of existing contracts .................................................................................. 47
purposive sampling ........................................................................................................ 129
R
random sampling ........................................................................................... 129, 134, 138
regular savings .......................................................................................................... 51, 95
Regulation ECB/2001/18 ................................................................................................... 7
regulatory arrangements .................................................................................................. 26
renegotiation of contract ............................................................................................ 35, 47
repayments
exceptional repayments ............................................................................................... 22
stop of .......................................................................................................................... 62
reporting agents ................................................................................................................. 8
reporting population
actual ................................................................................................... 13, 123, 126, 128
representative ............................................................................................................ 127
stratification ................................................................................................................ 127
repos .......................................................................................................................... 75, 77
Manual on MFI interest rate statistics − Index of terms
166
resident .............................................................................................................................. 9
retail interest rates ........................................................................................................... 39
reverse convertible .......................................................................................................... 98
S
sample maintenance ...................................................................................................... 137
sample size ................................................................................................ 8, 114, 129, 133
sampling error ................................................................................................................ 128
sampling variables ......................................................................................................... 128
sampling with probability proportional to size ................................................................ 134
savings bond ................................................................................................................... 96
Savings plan ........................................................................................ See regular savings
securitisation .................................................................................................................. 103
selection of the largest institutions ......................................................................... 134, 138
selection probabilities ....................................................................... See expansion factors
selection without replacement ....................................................................................... 141
single-stage sampling .................................................................................................... 133
snapshot .............................................................................................................. 43, 44, 68
Sole proprietorships ......................................................................................................... 75
Special national practices ................................................................................................ 29
split of loans ..................................................................................................................... 98
standard year ................................................................................................................... 16
step-up and step-down contracts ..................................................................................... 89
stratification criteria ........................................................................................................ 131
subsidies .................................................................................................................... 26, 36
subsidised loans .............................................................................................................. 29
T
tail institutions ................................................................................................................ 127
taxes ................................................................................................................................ 26
Top-up loans .................................................................................................................... 54
total costs of the credit to the consumer .......................................................................... 15
V
variable interest rates ................................................................................................ 25, 57
variance ................................................................................................. 130, 133, 137, 152
W
weighted average interest rates ..................................................................................... 113
© European Central Bank 2017
Address Sonnemannstrasse 20, 60314 Frankfurt am Main, Germany
Postal address 60640 Frankfurt am Main, Germany
Telephone +49 69 1344 0
Website www.ecb.int
Fax +49 69 1344 6000
All rights reserved. Any reproduction, publication and reprint in the form of a different publication, whether printed or produced
electronically, in whole or in part, is permitted only with the explicit written authorisation of the ECB or the authors.
ISBN 978-92-899-2641-6 (pdf)
DOI 10.2866/84250 (pdf)
EU catalogue No QB-07-16-023-EN-N (pdf)
