1. For the purpose of MIR statistics, ‘revolving loans and overdrafts’, ‘convenience credit card credit’ and ‘extended credit card credit’ have the same meaning as described in Part 2 of Annex II to Regulation (EU) 2021/379 (ECB/2021/2) ( 13 ), regardless of their initial period of interest rate fixation. Penalties on overdrafts applied as component of other charges, e.g. in the form of special fees, are not covered by the annualized agreed rate (AAR) as defined in Part 1 of Annex I to Regulation (EU) No 1072/2013 (ECB/2013/34) of the European Central Bank ( 14 ). An AAR or narrowly defined effective rate (NDER) is reported for the categories included in Table 1 of this Annex. The reporting of the AAR/NDER is accompanied by the related new business volumes.

2. In the case of revolving loans and overdrafts, convenience credit card credit and extended credit card credit, the concept of new business volumes is equivalent to outstanding amounts. Indicators 86 and 87 are calculated on the basis of items 12, 23, 32 and 36 in Appendix 2 of Annex I to Regulation (EU) No 1072/2013 (ECB/2013/34), and the outstanding amounts reported for revolving loans and overdrafts, convenience credit card credit and extended credit card credit in accordance with Annex I to Regulation (EU) 2021/379 (ECB/2021/2). Interest rates are calculated as weighted averages of the corresponding items in Appendix 2 to Annex I to Regulation (EU) No 1072/2013 (ECB/2013/34), taking a zero interest rate for convenience credit card credit. Indicators 86 and 87 are intended to provide continuity with indicators 12 and 23 (‘overdrafts’) as previously defined in (now repealed) Regulation (EC) No 63/2002 (ECB/2001/18) ( 15 ), i.e. prior to their amendment by Regulation (EC) No 290/2009 (ECB/2009/7) ( 16 ).

3. For the purpose of MFI interest rate statistics, renegotiated loans to households and non-financial corporations 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. 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. Only an AAR or NDER is reported for the categories included in Table 2 on a best efforts basis in addition to the volumes required under Regulation (EU) No 1072/2013 (ECB/2013/34).

1. NCBs apply the procedure illustrated in the figure below to select the reporting agents for the collection of MFI interest rate statistics in accordance with Regulation (EU) No 1072/2013 (ECB/2013/34). This procedure is defined as follows:

2. Each NCB selects its reporting agents from among the MFIs, except central banks and MMFs, in the reference reporting population that are resident in the same Member State whose currency is the euro (‘euro area Member State’) as the NCB.

3. In order to select the reporting agents, NCBs must either apply a census or follow a sampling approach in line with the requirements established in the following paragraphs.

4. In the case of a census, the NCB asks each resident MFI in the reference reporting population to report MFI interest rate statistics. The variables to be collected by means of the census are the interest rates and the amounts of new business and the interest rates on outstanding amounts.

5. In the case of a sample, only a selection of the MFIs in the reference reporting population will be asked to report. The variables to be estimated by means of the sample are the interest rates and the amounts of new business and the interest rates on outstanding amounts. They are referred to as sampling variables. In order to minimise the risk that the results of a sample survey deviate from the true (unknown) values in the reference reporting population, the sample should be constructed in such a way that it is representative of the reference reporting population. For the purpose of MFI interest rate statistics a sample is considered representative if all the characteristics that are relevant for MFI interest rate statistics and inherent in the reference reporting population are also reflected in the sample. To draw the initial sample, NCBs may use suitable proxies and models to produce the sampling scheme even if the underlying data, which are derived from existing sources, do not perfectly match the definitions in Regulation (EU) No 1072/2013 (ECB/2013/34).

6. In order to ensure that the sample is representative, each NCB that chooses the sampling approach for MFI interest rate statistics should suitably stratify the reference reporting population prior to the selection of any reporting agents. Stratification implies that the reference reporting population N is subdivided into sub-populations or strata N1, N2, N3… NL. These subdivisions into sub-populations or strata must not overlap and together comprise the reference reporting population:

7. NCBs define stratification criteria that allow the subdivision of the reference reporting population into homogeneous strata. Strata are considered homogeneous if the sum of the intra-stratum variances of the sampling variables is substantially lower than the total variance in the entire actual reporting population ( 17 ). The stratification criteria are linked to MFI interest rate statistics, i.e. there is a relationship between the stratification criteria and the interest rates and amounts to be estimated from the sample.

8. Each NCB that chooses the sampling approach must identify at least one stratification criterion to ensure that the sample of MFIs is representative of the relevant euro area Member State and the sampling error is small. NCBs should make efforts to define a hierarchy of stratification criteria. These must take national circumstances into account and be specific for each euro area Member State.

9. The selection of the reporting agents takes place in the form of single-stage sampling after all strata have been defined. Only at this stage are reporting agents drawn from the reference reporting population. No intermediate drawing should be carried out.

10. After defining the national strata in accordance with paragraphs 6 and 7, NCBs that choose the sampling approach draw the sample by selecting the actual reporting agents from each stratum. The total national sample size n is the sum of the sample sizes n1, n2, n3, …, nL for each of the strata:

11. Each NCB chooses the most appropriate allocation of the national sample size n among the strata. Therefore, for each stratum, NCBs define how many reporting agents nh are drawn from the total of MFIs, Nh. The sampling rate nh/Nh for each stratum h permits the estimation of the variance of each stratum. This implies that at least two reporting agents are selected from each stratum.

12. In order to select the actual reporting agents within each stratum, NCBs either include all institutions in the stratum, carry out random sampling or select the largest institutions per stratum. In the case of a random sample, the random drawing of the institutions within each stratum is either carried out with equal probability for all institutions or with probability proportional to the size of the institution. NCBs using random sampling or selection of the largest institutions may choose to include all institutions for some strata.

13. Information about the size of each credit institution and other institutions in the reference reporting population is available at national level from the MFI balance sheet statistics collected in accordance with Regulation (EU) 2021/379 (ECB/2021/2). NCBs must use total euro-denominated deposits and loans vis-à-vis households and non-financial corporations resident in the euro area Member States, which is that part of the balance sheet that is relevant for MFI interest rate statistics, or a close proxy.

14. MFI interest rate statistics must be based on a selection without replacement, i.e. each MFI in the reference reporting population may only be selected once.

15. If an NCB decides on a census of all MFIs in a stratum, the NCB may sample in that stratum at the level of branches. The precondition is that the NCB has a full list of branches that covers all the business of the credit and other institutions in the stratum, and has appropriate data in order to be able to assess the variance of interest rates on new business vis-à-vis households and non-financial corporations across branches. For the selection and maintenance of the branches all of the requirements laid down in this Guideline apply. The selected branches become notional reporting agents subject to all of the reporting requirements laid down in Annex I of Regulation (EU) No 1072/2013 (ECB/2013/34). This procedure is without prejudice to the obligation of each MFI to which branches belong, to be reporting agents.

16. The minimum national sample size is defined differently depending on whether the relevant NCB applies random sampling or selects the largest institutions per stratum.

17. If an NCB applies random sampling in the selection of the actual reporting institutions, the minimum national sample size should be such that the maximum random error for interest rates on new business on average over all instrument categories does not exceed 10 basis points at a confidence level of 90 % ( 18 ).

18. The maximum random error is defined as , with D as the maximum random error, zα/2 as the 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 1-α, where is the variance of the estimator of parameter θ, and is the estimated variance of the estimator of parameter θ.

19. If an NCB selects the largest institution per stratum, the sample quality should be based on a synthetic mean absolute error (MAE) measure. The actual synthetic MAE should not exceed a time-varying MAE threshold assuming a 10 basis points error difference in each stratum and indicator.

20. The synthetic MAES for a given estimator in a particular period should be defined as:

21. The maximum random error and the synthetic MAE are separately calculated for new business and outstanding amounts. For new business, the maximum random error and the synthetic MAE should be calculated on the basis of indicators 1 to 11, 13 to 22, and 24 to 29 as described in Appendix 2 to Annex I to Regulation (EU) No 1072/2013 (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 to Annex I to Regulation (EU) No 1072/2013 (ECB/2013/34).

22. The minimum national sample size refers both to the minimum initial sample and to the minimum sample after maintenance as defined in the next section on maintenance of the sample of the actual reporting population. Due to mergers and leavers, the sample may reduce in size over time until the next maintenance period.

23. NCBs may select more reporting agents than defined as the minimum national sample size, in particular where necessary to increase the representativeness of the national sample in the light of the structure of the national financial system.

24. There needs to be consistency between the number of MFIs in the reference reporting population and the minimum sample size. NCBs may allow MFIs which are resident within a single euro area Member State and individually included in the list of MFIs as established and updated in accordance with the classification principles set out in Section 1 of Part 1 of Annex I to Regulation (EU) 2021/379 (ECB/2021/2) to report MFI interest rate statistics together as a group. The group becomes a notional reporting agent. This means that the group reports MFI interest rate statistics as if it was a single MFI, i.e. it reports one average interest rate per instrument category covering the whole group instead of one rate for each MFI included in the list of MFIs. At the same time, the MFIs within the group are still counted as individual institutions in the reference reporting population and in the sample.

25. NCBs that choose the sampling approach must ensure that the sample remains representative over time.

26. NCBs should therefore check the representativeness of their sample at least once a year. If there are significant changes in the reference reporting population, these must be reflected in the sample after this annual check.

27. At intervals of at most three years, NCBs must carry out a regular review of the sample, taking account of joiners to the reference reporting population, leavers from the reference and actual reporting population, and other changes in the characteristics of the reporting agents, and apply the provisions of Section 5 on minimum national sample size. The regular review of the sample will be based on an assessment of compliance with the provisions on the selection of the actual reporting population as contained in Section 1, based on the monthly data corresponding to the end of each quarter in the year in which the review takes place. However, NCBs may check and refresh their sample more frequently.

28. Over time, the sample will be adjusted to take account of joiners to the reference reporting population in order to remain representative of the reference reporting population. NCBs must therefore draw a sample nb from the population of all joiners Nb. The complementary selection of joining institutions nb among the total number of joiners Nb is referred to as incremental sampling over time.

29. Over time, the sample will be adjusted to take account of leavers from the reference and the actual reporting population. No adjustment is necessary if the number of leavers in the reference reporting population Nd is proportionate to the number of leavers in the sample nd (case 1). If institutions leave the reference reporting population and they are not in the sample, the sample becomes larger 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). In cases 2 and 3, if random sampling is used for the selection of the actual reporting institutions the weights attached to each institution in the sample has to be adjusted by means of an established statistical method derived from the theory of sampling. The weight attached to each reporting agent is the inverse of its selection probability and thereby the expansion factor. In case 2, where the sample is relatively larger for the population, no reporting agent is taken out of the sample. In case 3, if the largest institutions are selected, the sample is adjusted by selecting additional institutions according to their size.

30. Over time, the sample will be adjusted to take account of changes in the characteristics of the reporting agents. These changes can occur because of mergers, divisions, growth of the institutions, etc. Some reporting agents may change stratum. As in cases 2 and 3 for leavers, the sample is adjusted by means of an established statistical method derived from the theory of sampling. New selection probabilities and weights are allocated where NCBs carry out random sampling.

31. In order to achieve consistency between MFI interest rate statistics on outstanding amounts referring to deposits and to loans and on new business referring to deposits and to loans, NCBs that choose the sampling approach must use the same reporting agents to collect these statistics. NCBs may also use the sampling approach for a sub-set of MFI interest rate statistics and a census for the rest. They may not, however, use two or more different samples.

32. In the sample procedure, NCBs do not need to cover each product that exists at national level. However, they cannot exclude a whole instrument category on the grounds that the amounts involved are very small. Therefore, if an instrument category is only offered by one institution, then this institution should be included in the sample. If an instrument category did not exist in a euro area Member State at the time of the initial drawing of the sample, but is introduced by an institution thereafter, this institution should be included in the sample at the time of the next representability check. If a new product is created, the institutions in the sample should cover it in the next report, as all reporting agents are required to report on all their products.

33. NCBs receive weighted average interest rates and related amounts of business from all their actual resident reporting agents and calculate national weighted average interest rates for each instrument category on the basis of grossed-up business volumes per stratum. The data are reported to the ECB.

34. Where random sampling is applied, the estimator of the interest rate at stratum and national levels must be consistent with the sampling procedure, simple random sample or with the probability proportional to size used, implying that grossed up amounts are used for weighting of interest rates.

35. Where the largest institutions are selected, the estimator of the rates should aggregate rates across institutions in the same stratum by weighting by the reported amounts, and aggregates across strata should be done by applying the grossed-up volumes in each stratum.

36. For each of the instrument categories on outstanding amounts, i.e. indicators 1 to 26 in Appendix 1 to Annex I to Regulation (EU) No 1072/2013 (ECB/2013/34), NCBs provide a national weighted average interest rate.

37. For each of the instrument categories on new business, i.e. indicators 1 to 23 and 30 to 85 in Appendix 2 to Annex I to Regulation (EU) No 1072/2013 (ECB/2013/34), NCBs provide a national weighted average interest rate. In addition, for each of indicators 2 to 4, 8 to 11, 13 to 22 , 33 to 35 and 37 to 85 in Appendix 2 to Annex I to Regulation (EU) No 1072/2013 (ECB/2013/34), NCBs provide the amount of new business conducted at national level in each instrument category during the reference month. For the instrument categories referring to renegotiated loans to households and non-financial corporations (indicators 88 to 91 in Appendix 2 to Annex I to Regulation (EU) No 1072/2013 (ECB/2013/34), only information on volumes is required, and information on interest rates is collected on a best efforts basis. These amounts of new business refer to the population total, i.e. to the entire reference reporting population and, similarly to the other new business volumes, this figure is estimated through the grossing-up procedure illustrated in paragraphs 38 to 40.

38. If random sampling or the selection of the largest institutions is used to select the reporting agents, expansion factors are used for grossing up the business volumes. Grossing up is applied at stratum level.

39. If the random sampling approach is used, the expansion factors are defined as the inverse of the selection probabilities πi , i.e. 1/πi . The estimated amount of new business for the population total B is then computed by means of the following generic formula:

40. 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:

41. Expansion factors EFj as defined in paragraph 40 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.

42. 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 decisions 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.

43. NCBs document any (changes in) regulatory arrangements affecting MFI interest rate statistics in the methodological notes that are provided with the national data.

44. NCBs that choose a sampling approach for the selection of the reporting agents provide an estimate of the sampling error for the initial sample. A new estimate is provided after each maintenance of the sample.

1. The treatment of the products defined in the following paragraphs should be used as a reference for products with similar characteristics.

2. A step-up (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. Step-up (step-down) deposits and loans are instruments with fixed interest rates over the whole maturity. The interest rate for the whole maturity of the deposit or loan and the other terms and conditions are agreed in advance at time t0 when the contract is signed. An example of a step-up deposit is a deposit with an agreed maturity of four years, which receives 5 % interest in the first year, 7 % in the second, 9 % in the third and 13 % in the fourth. The AAR on new business, which is covered at time t0 in MFI interest rate statistics, is the geometric average of the factors ‘1 + interest rate’. In line with Annex I to Regulation (EU) No 1072/2013 (ECB/2013/34), NCBs may request reporting agents to implement the NDER for this type of product. The AAR on outstanding amounts that is covered from time t0 to t3 is the rate applied by the reporting agent at the time of calculation of the MFI interest rate, i.e., using the example of a deposit with an agreed maturity of four years, 5 % at time t0, 7 % at time t1, 9 % at time t2 and 13 % at time t3.

3. For the purpose of MFI interest rate statistics, loans taken as part of lines of credit have the same meaning as defined in and are classified as in Part 2 of Annex II to Regulation (EU) 2021/379 (ECB/2021/2). Only outstanding amounts, i.e. amounts withdrawn and not yet repaid in the context of a credit line are covered as new business and reflected in the MFI interest rate statistics. Amounts available through a line of credit that have not been withdrawn or have already been repaid are not considered, whether as new business or as outstanding amounts.

4. 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 and/or the date on which the loan will be drawn and/or the interest rate are not specified, but a range of possibilities may be agreed. Such an umbrella contract is not covered by MFI interest rate statistics. However, as soon as a loan agreed under an umbrella contract is drawn, it is covered under the corresponding item in MFI interest rate statistics, both in new business and outstanding amounts.

5. 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 unknown future attitude towards saving of the household or non-financial corporation. As a convention, such fidelity or growth premiums are not included in the AAR on new business. The AAR on outstanding amounts always covers the rates applied by the reporting agent at the time of calculation of MFI interest rates. Consequently, if such a fidelity or growth premium is granted by the reporting agent, this is reflected in the statistics on outstanding amounts.

6. Loans may be offered to households or non-financial corporations with associated derivative contracts, i.e. an interest rate swap/cap/floor etc. As a convention, such associated derivative contracts are not included in the AAR on new business. The AAR on outstanding amounts always covers the rates applied by the reporting agent at the time the MFI interest rates are calculated. Consequently, if such a derivative contract is exercised and the reporting agent adjusts the interest rate charged to the household or non-financial corporation, this is reflected in the statistics on outstanding amounts.

7. Deposits may be offered comprising two components: a deposit with an agreed maturity to which a fixed interest rate is 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 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. Therefore, only the guaranteed minimum return (usually 0 %) is covered. The AAR on outstanding amounts always covers the interest rate applied by the reporting agent at the time the MFI interest rates are calculated. Until the day of maturity, the rate on the deposit with 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.

8. Deposits with a maturity of over two years as defined in Part 2 of Annex II to Regulation (EU) 2021/379 (ECB/2021/2) may contain pension savings accounts. The main part of the 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 the pension savings accounts may be held in cash and the interest rate determined by the credit or other institution 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 may also be negative. In addition, at the time the deposit is placed, an interest rate is agreed between the household and the credit or other institution which applies only to the deposit part; this does not apply to the part invested in securities. Therefore, 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.

9. Savings plans for housing loans are long-term saving schemes that may provide a low return but, after a certain period of saving, give the household or non-financial corporation the right to a housing loan at a discounted rate. In line with Part 2 of Annex II to Regulation (EU) 2021/379 (ECB/2021/2), these savings plans are classified under deposits with agreed maturity over two years as long as they are used as a deposit. As soon as they are transformed into a loan, they are classified as loans to households for house purchases. 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 this amount on the deposit over time is only covered by 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 or non-financial corporation.

10. In line with Part 2 of Annex II to Regulation (EU) 2021/379 (ECB/2021/2), deposits placed pursuant to the French regulated housing plan plan d’épargne-logement (PEL) are classified as deposits with an agreed maturity of over two years. The government regulates the conditions governing these PELs and fixes the interest rate, which remains unchanged over the entire maturity of the deposit, i.e. each ‘generation’ of PELs has the same interest rate attached to it. PELs are held for at least four years and each year the customer deposits a minimal pre-fixed amount but is allowed to increase the payments at any time during the course of the scheme. Reporting agents report as new business the initial deposit at the opening of a new PEL. The amount of money that is initially placed in the PEL may be very low, which means that the weight attached to the new business rate will also be relatively low. This approach ensures that the new business rate always reflects the conditions governing the current generation of PELs. Changes in the interest rate applied to the new PELs are reflected in the new business rate. The reaction of consumers in terms of a portfolio shift from other long-term deposits to pre-existing PELs is not reflected in the new business rates but only in the rates on outstanding amounts. At the end of the period of four years, the customer may either ask for a loan at a discounted rate or renew the contract. Since this renewal of the PEL is carried out automatically without any active involvement on the part of the customer, and since the terms and conditions of the contract including the interest rate are not renegotiated, in line with Part 2 of Annex I to Regulation (EU) No 1072/2013 (ECB/2013/34) this renewal is not considered as new business. On renewing the contract, the customer is allowed to make additional deposits, provided that the outstanding amount does not exceed a defined ceiling and the contract does not exceed a defined maximum years of maturity. If the ceiling or maximum maturity are reached, the contract is frozen. The household or non-financial corporation retains the borrowing rights and is still granted interest in accordance with the conditions prevailing at the time of opening the PEL as long as the money is left in the bank’s book. The government grants a subsidy in terms of an interest payment on top of the interest rate offered by the credit or other institution. In line with Part 1 of Annex I to Regulation (EU) No 1072/2013 (ECB/2013/34) only that part of the interest payment offered by the credit or other institution is captured in MFI interest rate statistics. The government subsidy, which is paid through but not by the credit or other institution, is ignored.

11. Negative interest rates on deposits should be included in MFI interest rates provided that such rates are not exceptional having regard to market conditions.

1.  Number of offices (local branches) of all credit institutions (CIs) in the Member State. ‘Office’ refers to a geographically identifiable place of business. The offices of institutional units that are not CIs must be excluded, even if they belong to a group of CIs (banking group).

2.  Number of employees of CIs. This indicator refers to the number of employees in all resident CIs during the reference year. NCBs shall follow the concept used when reporting the ‘Employment domestic’ for RIAD pursuant to Guideline (EU) 2018/876 of the European Central Bank (ECB/2018/16). Employees of institutional units that are not CIs must be excluded, even if these institutions belong to a banking group. Where actual data are not available, NCBs shall provide an estimate.

3.  Number of branches of CIs from other Member States. This indicator refers to the number of CI branches resident in the reporting Member State whose ultimate controlling CIs were resident in other Member States as at the end of the reference period. Branches that are ultimately controlled by domestic CIs must be excluded.

4.  Total assets of branches of CIs from other Member States. This indicator refers to the aggregated balance sheet total of the branches covered by the indicator ‘number of branches of CIs from other Member States’ as at the end of the reference period.

5.  Number of subsidiaries of CIs from other Member States. This indicator refers to the number of CI subsidiaries resident in the reporting Member State whose ultimate controlling CIs were resident in other Member States as at the end of the reference period. Subsidiaries that are ultimately controlled by domestic CIs must be excluded.

6.  Total assets of subsidiaries of CIs from other Member States. This indicator refers to the aggregated balance sheet total of the subsidiaries covered by the indicator ‘number of subsidiaries of CIs from other Member States’.

7.  Number of branches of CIs from third countries. This indicator refers to the number of CI branches resident in the reporting Member State whose ultimate controlling CIs were resident in countries that were not Member States as at the end of the reference period. Branches that are ultimately controlled by domestic CIs must be excluded.

8.  Total assets of branches of CIs from third countries. This indicator refers to the aggregated balance sheet total of the branches covered by the indicator ‘number of branches of CIs from third countries’.

9.  Number of subsidiaries of CIs from third countries. This indicator refers to the number of CI subsidiaries resident in the reporting Member State whose ultimate controlling CIs were resident in countries that were not Union Member States as at the end of the reference period. Subsidiaries that are ultimately controlled by domestic CIs must be excluded.

10.  Total assets of subsidiaries of CIs from third countries. This indicator refers to the aggregated balance sheet total of the subsidiaries covered by the indicator ‘number of subsidiaries of CIs from third countries’.

11.  Number of branches of CIs from euro area Member States. This indicator refers to the number of CI branches resident in the reporting Member State whose ultimate controlling CIs were resident in euro area Member States (other than the reporting Member State) at the end of the reference period. Branches that are ultimately controlled by domestic CIs must be excluded.

12.  Total assets of branches of CIs from euro area Member States. This indicator refers to the aggregated balance sheet total of the branches covered by the indicator ‘number of branches of CIs from euro area Member States’.

13.  Number of subsidiaries of CIs from euro area Member States. This indicator refers to the number of subsidiaries resident in the reporting Member State whose ultimate controlling CIs were resident in euro area Member States (other than the reporting Member State) at the end of the reference period.

14.  Total assets of subsidiaries of CIs from euro area Member States. This indicator refers to the aggregated balance sheet total of the subsidiaries covered by the indicator ‘number of subsidiaries of CIs from euro area Member States’.

15.  Share of the five largest CIs in total assets (‘CR5’). This indicator refers to the concentration of the banking business. NCBs must adopt the following unconsolidated aggregated approach to derive it: (a) rank the balance sheet totals of all reporting CIs; (b) calculate (i) the sum of the five largest balance sheet totals; and (ii) the sum of all balance sheet totals; and (c) calculate the proportion of (i) over (ii). Data to be reported to the ECB must be expressed as percentages, e.g. a value of 72,4296% must be reported as 72,4296 and not as 0,7243. Although the composition of the five largest banks may change over time, NCBs must only provide the share of the five largest CIs at a specific point in time (end-December of the reference year). For the purposes of this indicator, NCBs may, where specifically agreed with the ECB, treat specific group-connected CIs as a single CI where this would provide a more meaningful representation of the banking sector concentration.

16.  Herfindahl index (HI) for CIs’ total assets. Similar to the previous indicator, this refers to the concentration of banking business. NCBs must follow as far as possible an aggregated approach. In this case, the calculation of the HI must include the aggregated balance sheet of each CI included in the group, perhaps using the accounting information contained in these institutions’ annual financial statements. Where not all CIs are in the tail report data, data must be grossed up.

17.  Herfindahl index (HI) for total domestic credit. For this indicator, the methodological principles adopted for the calculation of Herfindahl index for CIs’ total assets shall be followed. However, the total assets shall be replaced by total credit, which is defined as loans and debt securities assets vis-à-vis domestic non-MFIs, excluding general government ( 20 ). For the purposes of this indicator, NCBs may, where specifically agreed with the ECB, treat specific group-connected CIs as a single CI where this would provide a more meaningful representation of the banking sector concentration.