General
1. In order to assess compliance of an institution with the requirements on quantification of risk parameters, for the purposes of Article 144(1)(a) of Regulation (EU) No 575/2013, competent authorities shall verify the institution’s:
(a)
compliance with the overall requirements for estimation laid down in Article 179 of Regulation (EU) No 575/2013, in accordance with Articles 42, 43 and 44;
(b)
compliance with the requirements specific to PD estimation laid down in Article 180 of Regulation (EU) No 575/2013, in accordance with Articles 45 and 46;
(c)
compliance with the requirements specific to own-LGD estimates laid down in Article 181 of Regulation (EU) No 575/2013, in accordance with Articles 47 to 52;
(d)
compliance with the requirements specific to own-conversion factor estimates laid down in Article 182 of Regulation (EU) No 575/2013, in accordance with Articles 53 to 56;
(e)
compliance with the requirements for assessing the effect of guarantees and credit derivatives laid down in Article 183 of Regulation (EU) No 575/2013, in accordance with Article 57;
(f)
compliance with the requirements for purchased receivables laid down in Article 184 of Regulation (EU) No 575/2013, in accordance with Article 58.
2. For the purposes of the verification under paragraph 1, competent authorities shall apply all of the following methods:
(a)
review the institution’s relevant internal policies;
(b)
review the institution's technical documentation of relevant estimation methodology and process;
(c)
review and challenge the relevant estimation of risk parameter manuals, methodologies and processes;
(d)
review the relevant minutes of the institution’s internal bodies, including the management body, model committee, or other committees;
(e)
review the reports on the performance of risk parameters and the recommendations made by the credit risk control unit, the validation function, the internal audit function or any other control function of the institution;
(f)
assess progress reports on the efforts made by the institution to correct shortcomings and mitigate risks detected during relevant audits, validations and monitoring;
(g)
obtain written statements from or interview the relevant staff and the senior management of the institution.
3. For the purposes of the verification under paragraph 1, competent authorities may also apply any of the following additional methods:
(a)
request the provision of additional documentation or analysis substantiating the institution’s methodological choices and the results obtained;
(b)
conduct their own estimations of risk parameters or replicate those of the institution, using the relevant data supplied by the institution;
(c)
request and analyse the data used in the process of estimation;
(d)
review the functional documentation of the IT systems which are relevant to the scope of the assessment;
(e)
perform their own tests on the data of the institution or request the institution to perform tests proposed by the competent authorities;
(f)
review other relevant documents of the institution.
SECTION 2 — Methodology for assessing overall requirements for quantification of risk parameters
Data requirements
1. When assessing compliance with the overall requirements for estimation laid down in Article 179 of Regulation (EU) No 575/2013, the data used for the quantification of risk parameters and the quality of that data, competent authorities shall verify:
(a)
the completeness of the quantitative and qualitative data and other information in relation to the methods used for the quantification of risk parameters to ensure that all relevant historical experience and empirical evidence are used;
(b)
the availability of quantitative data providing a breakdown of the loss experience by the factors which drive the respective risk parameters as referred to in Article 179(1)(b) of Regulation (EU) No 575/2013;
(c)
the representativeness of the data used to estimate the risk parameters for certain types of exposures;
(d)
the adequacy of the number of exposures in the sample and the length of the historical observation period referred to in Articles 45, 47 and 53, used for the quantification to ensure that the estimates of the institution are accurate and robust;
(e)
the justification for and the documentation of all data cleansing, including any exclusions of observations from the estimation and a confirmation that these exclusions do not bias the risk quantification; for PD estimates, in particular, the justification and the documentation of the impact of the data cleansing on the long-run average default rate;
(f)
the consistency between the data sets used for the risk parameters estimation, in particular with regard to the default definition, treatment of defaults, including multiple defaults as referred to in Articles 46(1)(b) and 49, and the sample composition.
2. For the purposes of the verification under point (c) of paragraph 1, competent authorities shall assess the representativeness of the data used to estimate the risk parameters for certain types of exposures by assessing:
(a)
the structure of exposures covered by each rating model and the different risk characteristics of the obligors or facilities, and whether the current portfolio is, to the degree required, comparable to the portfolios constituting the reference data set;
(b)
the comparability of the current underwriting and recovery standards with the ones applied at the time of the reference data set;
(c)
the consistency of the default definition in the observation period:
(i)
where the default definition has been changed in the observation period, the description of the adjustments performed in order to achieve the required level of consistency with current default definition;
(ii)
where default definitions vary across the jurisdictions in which the institution operates, the adequacy of measures and conservatism adopted by the institution;
(d)
where external data and data pooled across institutions are used in the quantification of risk parameters, the relevance and appropriateness of these data for the institution’s exposures, products and risk profile and the definition of default;
(e)
where the external or pooled data are not consistent with the institution’s internal default definition, the description of adjustments to the external or pooled data performed by the institution in order to achieve the required level of consistency with the internal default definition.
3. When assessing the quality of the data pooled across institutions that is used for quantification of risk parameters, competent authorities shall apply the assessment methodology laid down in paragraphs 1 and 2 in addition to verifying the compliance with the requirements laid down in Article 179(2) of Regulation (EU) No 575/2013.
Review of estimates
When assessing the review of risk parameter estimates by the institution as referred to in Article 179(1)(c) of Regulation (EU) No 575/2013, competent authorities shall verify that:
(a)
the process and the annual plan for the review of estimates provide for a timely review of all estimates;
(b)
criteria for the identification of situations which trigger a more frequent review have been identified;
(c)
the methodologies and data used for the estimation of risk parameters reflect changes in the underwriting process and in the composition of the portfolios;
(d)
the methodologies and data used for the LGD estimation reflect changes in the recovery process, the types of recoveries and the duration of the recovery process;
(e)
the methodologies and data used for the conversion factor estimation reflect changes in the monitoring process of undrawn amounts;
(f)
the data set used for the estimation of risk parameters includes the relevant data from the latest observation period, and are updated at least on an annual basis;
(g)
the technical advances and other relevant information are reflected in the risk parameters estimates.
Margin of conservatism
1. Competent authorities shall assess whether an appropriate margin of conservatism is included in the values of risk parameters used in the calculation of capital requirements as referred to in point (f) of Article 179(1) of Regulation (EU) No 575/2013, in the following situations:
(a)
the methods and data do not provide sufficient certainty of the risk parameter estimates, including where there are high estimation errors;
(b)
relevant deficiencies in the methods, information and data have been identified by the credit risk control unit, validation function or internal audit function or any other function of the institution;
(c)
relevant changes to the standards of underwriting or recovery policies or changes in the institution’s risk appetite.
2. Competent authorities shall assess whether the institutions do not use the margin of conservativism as a substitute to any corrective action applied by the institution under Article 146 of Regulation (EU) No 575/2013.
SECTION 3 — Methodology for assessing requirements specific for PD estimation
Length of the historical observation period
When assessing the length of the historical observation period referred to in point (h) of Article 180(1) and point (e) of Article 180(2) of Regulation (EU) No 575/2013, taking into account conditions laid down in Commission Delegated Regulation (EU) 2017/72 with regard to regulatory technical standards specifying conditions for data waiver permissions ( 6 ) , and the calculation of one year default rates based on internal default experience as referred to in point (e) of Article 180(1), competent authorities shall verify:
(a)
that the length of the historical observation period covers at least the minimum length in accordance with the requirements laid down in point (h) of paragraph 1 and point (e) of paragraph 2 of Article 180 of Regulation (EU) No 575/2013 and, where applicable, Delegated Regulation (EU) 2017/72;
(b)
where the available historical observation period is longer than the minimum period required in point (h) of Article 180(1) or in point (e) of Article 180(2) of Regulation (EU) No 575/2013 for a data source, and the data obtained from it are relevant, that the information for that longer period is used in order to estimate the long-run average of one-year default rates;
(c)
for retail exposures where the institution does not give equal importance to all historical data used, that this is justified by better prediction of default rates and that a zero or very small weight applied to a specific period is either duly justified or leads to more conservative estimates;
(d)
that there is consistency between underwriting standards and the rating systems in place and that comparable underwriting standards were used at the time of generating the internal default data or that changes in underwriting standards and rating systems have been addressed by applying the margin of conservatism as referred to in point (c) of Article 44(1);
(e)
for exposures to corporates, institutions, central governments and central banks, that the definition of obligors that are highly leveraged and obligors whose assets are predominantly traded assets as referred to in point (a) of Article 180(1) of Regulation (EU) No 575/2013 as well as the identification of periods of stressed volatilities for those obligors as referred to in that provision are adequate.
Method of PD estimation
1. When assessing the method of PD estimation, as referred to in Article 180 of Regulation (EU) No 575/2013, competent authorities shall verify that the one-year default rate for each grade or pool is calculated in a manner consistent with the characteristics of the one-year default rate defined in point 78 of Article 4(1) of Regulation (EU) No 575/2013, and they shall verify that:
(a)
the denominator of the one-year default rate includes the obligors or exposures which, at the beginning of a one year period, are not in default and are assigned to that rating grade or pool;
(b)
the numerator of the one-year default rate includes those of the obligors or exposures referred to in point (a) that have defaulted within that one year period; multiple defaults for the same obligor or exposure, which have been observed during the one year period relating to the default rate, are considered to be a single default as referred to in Article 49(b) having occurred on the date of the first of those multiple defaults.
2. Competent authorities shall verify that the method of PD estimation by obligor grade or pool is based on the long-run average of one-year default rates.
For that purpose they shall verify that the period used by the institution to estimate the long-run average of one-year default rates is representative of the likely range of variability of default rates for that type of exposures.
3. Where observed data used for PD estimation are not representative of the likely range of variability of default rates for a type of exposures, competent authorities shall verify that both of the following conditions are met:
(a)
the institution uses an appropriate alternative method for estimating the average of one-year default rates over a period that is representative of the likely range of variability of default rates for that type of exposures;
(b)
an appropriate margin of conservatism is applied where, after applying an appropriate method as referred to in point (a), the estimation of the averages of default rates is found to be unreliable or to have other limitations.
4. For the purposes of the verification under paragraph 1, competent authorities shall verify that all of the following is appropriate for the type of exposures:
(a)
the functional and structural form of the estimation method;
(b)
assumptions on which the estimation method is based;
(c)
the cyclicality of the estimation method;
(d)
the length of the historical observation period used in accordance with Article 45;
(e)
the margin of conservatism applied in accordance with Article 44;
(f)
the human judgement;
(g)
where applicable, the choice of risk drivers.
5. For exposures to corporates, institutions, central governments and central banks, where the obligors are highly leveraged or the assets of the obligor are predominantly traded assets as referred to in point (a) of Article 180(1) of Regulation (EU) No 575/2013, competent authorities shall verify that the PD reflects the performance of the underlying assets in the periods of stressed volatility as referred to in that provision.
6. For exposures to corporates, institutions, central governments and central banks, where the institution makes use of a rating scale of an ECAI, competent authorities shall verify the institution’s analysis of compliance with the requirements laid down in point (f) of Article 180(1) of Regulation (EU) No 575/2013, and check that that analysis addresses the issue of whether the types of exposures rated by the ECAI are representative of the institution’s types of exposures and the time horizon for the credit assessment by the ECAI.
7. For retail exposures, where the institution derives the estimates of PD or LGD from an estimate of total losses and an appropriate estimate of PD or LGD as referred to in point (d) of Article 180(2) of Regulation (EU) No 575/2013, competent authorities shall verify the institution’s analysis of compliance with all relevant criteria on PD and LGD estimation laid down in Articles 178 to 184 of Regulation (EU) No 575/2013.
8. For retail exposures, competent authorities shall verify that the institution regularly analyses and takes into account the expected changes of PD over the life of credit exposures (‘seasoning effects’) as referred to in point (f) of Article 180(2) of Regulation (EU) No 575/2013.
9. In the assessment of statistical models for PD estimation, competent authorities shall, in addition to the methods laid down in paragraphs 1 to 8, apply the methodology for assessing specific requirements for statistical models or other mechanical methods laid down in Articles 37 to 40.
SECTION 4 — Methodology for assessing requirements specific to own-LGD estimates
Length of the historical observation period
When assessing the length of the period used for LGD estimation for the purpose of point (j) of paragraph 1 and subparagraph 2 of paragraph 2 of Article 181 of Regulation (EU) No 575/2013 and Delegated Regulation (EU) 2017/72, (‘historical observation period’), competent authorities shall verify that:
(a)
the length of the historical observation period covers at least the minimum length in accordance with the requirements laid down in paragraph 1(j) and the second subparagraph of paragraph 2 of Article 181 of Regulation (EU) No 575/2013 and, where applicable, Delegated Regulation (EU) 2017/72;
(b)
where the available historical observation period is longer than the minimum period according to point (j) of paragraph 1 of Article 181 and subparagraph 2 of paragraph 2 of Article 181 of Regulation (EU) No 575/2013 for a data source, and the data obtained from it are relevant for the LGD estimation, that the information for that longer period is used;
(c)
for retail exposures, where the institution does not give equal importance to all historical data used, that this is justified by better prediction of loss rates and that a zero or very small weight applied to a specific period is either duly justified or leads to more conservative estimates.
Method of LGD estimation
When assessing the method of own-LGD estimation, as referred to in Article 181 of Regulation (EU) No 575/2013, competent authorities shall verify that:
(a)
the institution assesses LGD by homogenous facility grade or pool;
(b)
the average realized LGD by facility grade or pool is calculated using the number of default weighted average;
(c)
all observed defaults within the data sources are used, in particular that the incomplete recovery processes are taken into account in a conservative manner for the purposes of LGD estimation, and that the choice of workout period and methodologies for estimating additional costs and recoveries after and, where necessary, during that period, are relevant;
(d)
the LGD estimates of secured exposures are not solely based on the estimated market value of the collateral and that they take into account the realised revenues from past liquidations and the potential inability of an institution to gain control of the collateral and liquidate it;
(e)
the LGD estimates of secured exposures take into account the potential decreases in collateral value from the point of time of LGD estimation to the eventual recovery;
(f)
the degree of dependence between the risk of the obligor and that of the collateral as well as the cost of liquidating the collateral are taken into account conservatively;
(g)
any unpaid late fees that have been capitalised in the institution’s income statement before the default are added to the institution’s measure of exposure and loss;
(h)
the possibility of future drawings after the default is taken into account appropriately;
(i)
all of the following aspects are appropriate for the type of exposures to which they are applied:
(i)
the functional and structural form of the estimation method;
(ii)
the assumptions regarding the estimation method;
(iii)
the estimation method for a downturn effect;
(iv)
the length of data series used;
(v)
the margin of conservatism;
(vi)
the use of the human judgement;
(vii)
where applicable, the choice of risk drivers.
Treatment of multiple defaults
For the treatment of obligors that default and recover several times in a limited period of time as defined by the institution (‘multiple defaults’), competent authorities shall assess the adequacy of the methods used by the institution and shall verify that:
(a)
explicit conditions are defined before a facility is considered to have returned to a non-default status;
(b)
multiple defaults identified within a period of time specified by the institution are considered to be a single default for the purpose of LGD estimation, using the default date of the first observed default as the relevant default date and considering the recovery process from that date until the end of the recovery process after the last observed default in this period;
(c)
the length of period within which multiple defaults are recognised as a single default is determined taking into account the institution’s internal policies and analysis of the default experience;
(d)
defaults used for the purpose of PD and conversion factors estimation are treated consistently with defaults used for the purpose of LGD estimation.
Use of LGD estimates appropriate for economic downturn
When assessing whether the requirement to use LGD estimates that are appropriate for an economic downturn as laid down in point (b) of Article 181(1) of Regulation (EU) No 575/2013 is fulfilled, competent authorities shall verify that:
(a)
the institution uses LGD estimates that are appropriate for an economic downturn, where those are more conservative than the long-run average;
(b)
the institution provides both long-run averages and LGD estimates appropriate for an economic downturn for justification of its choices;
(c)
the institution applies a rigorous and well documented process for identifying an economic downturn and assessing its effects on recovery rates and for producing LGD estimates appropriate for an economic downturn;
(d)
the institution incorporates in the LGD estimates any adverse dependencies that have been identified between on the one hand selected economic indicators and on the other hand the recovery rates.
LGD, ELBE and UL estimation for exposures in-default
1. When assessing the requirements for LGD estimates for the exposures in default, and for the best estimate of expected losses (‘EL BE
’) as referred to in Article 181(1)(h) of Regulation (EU) No 575/2013, competent authorities shall verify that the institution uses one of the following approaches and shall assess the approach used by the institution:
(a)
direct estimation of the LGD for defaulted exposures (‘LGD in-default’) and direct estimation of ELBE;
(b)
direct estimation of ELBE and estimation of the LGD in-default as the sum of ELBE and an add-on capturing the unexpected loss related to exposures in default that might occur during the recovery period.
2. When assessing the approach of the institution in accordance with paragraph 1, competent authorities shall verify that:
(a)
the LGD in-default estimation methods, either as a direct estimation or as an add-on to ELBE, take into account possible additional unexpected losses during the recovery period, and in particular consider possible adverse changes in economic conditions during the expected length of the recovery process;
(b)
the LGD in-default, either as a direct estimation or as an add-on to ELBE, and the ELBE estimation methods take into account the information on the time in- default and recoveries realised so far;
(c)
where the institution uses a direct estimation of the LGD in-default, the estimation methods are consistent with the requirements of Articles 47, 48 and 49;
(d)
the LGD in-default estimate is higher than the ELBE, or, where the LGD in- default is equal to the ELBE, that for individual exposures such cases are limited and duly justified by the institution;
(e)
the ELBE estimation methods take into account all currently available and relevant information and in particular consider current economic circumstances;
(f)
where the specific credit risk adjustments exceed the ELBE estimates the differences between the two are analysed and duly justified;
(g)
the LGD in-default, either as a direct estimation or as an add-on to EL BE , and the EL BE estimation methods are clearly documented.
Requirements on collateral management, legal certainty and risk management
When assessing whether the institution has established internal requirements for collateral management, legal certainty and risk management which are generally consistent with those set out in Chapter 4, Section 3 of Regulation (EU) No 575/2013, as referred to in Article 181(1)(f) of that Regulation, competent authorities shall verify that at least the policies and procedures of the institution relating to the internal requirements for collateral valuation and legal certainty are fully consistent with the requirements of Section 3 of Chapter 4 of Title II in Part Three of Regulation (EU) No 575/2013.
SECTION 5 — Methodology for assessing requirements specific to own- conversion factor estimates
Length of the historical observation period
When assessing the length of the period used for the estimation of conversion factors referred to in paragraph 2 and paragraph 3 of Article 182 of Regulation (EU) No 575/2013 and Delegated Regulation (EU) 2017/72 (‘historical observation period’), competent authorities shall verify that:
(a)
the length of the historical observation period covers at least the minimum length required by paragraph 2 and paragraph 3 of Article 182 of Regulation (EU) No 575/2013 and, where applicable, Delegated Regulation (EU) 2017/72;
(b)
where the available observation period is longer than the minimum period required by paragraph 2 and paragraph 3 of Article 182 of Regulation (EU) No 575/2013 for a data source, and the data obtained from it are relevant for the estimation of conversion factors, that the information for that longer period is used;
(c)
for retail exposures, where the institution does not give equal importance to all historical data used, that this is justified by better prediction of drawings on commitments and that, if a zero weight or a very small weight is applied to a specific period, this is either duly justified or leads to more conservative estimates.
Method of conversion factors estimation
When assessing the method of estimating conversion factors as referred to in Article 182 of the Regulation (EU) No 575/2013, competent authorities shall verify that:
(a)
the institution assesses estimates of conversion factors by facility grade or pool;
(b)
the average realised conversion factors by facility grade or pool are calculated using the number of default weighted average;
(c)
all observed defaults within the data sources are used for conversion factors estimation;
(d)
the possibility of additional drawings is taken into account in a conservative manner, except for retail exposures when they are included in the LGD estimates;
(e)
the institution’s policies and strategies regarding account monitoring, including limit monitoring, and payment processing are reflected in the conversion factors estimation;
(f)
all of the following are adequate to the type of exposures to which they are applied:
(i)
the functional and structural form of the estimation method;
(ii)
assumptions on which the estimation method is based;
(iii)
where applicable the method of estimation of the downturn effect;
(iv)
the length of the historical observation period in accordance with Article 53;
(v)
the margin of conservatism applied in accordance with Article 44;
(vi)
the human judgement;
(vii)
where applicable, the choice of risk drivers.
Use of conversion factor estimates appropriate for economic downturn
When assessing whether the requirement to use conversion factor estimates that are appropriate for an economic downturn as laid down in point (b) of Article 182(1) of Regulation (EU) No 575/2013 is fulfilled, competent authorities shall verify that:
(a)
the institution uses conversion factor estimates that are appropriate for an economic downturn, where those are more conservative than the long-run average;
(b)
the institution provides both the long-run averages and the conversion factor estimates appropriate for an economic downturn for justification of its choices;
(c)
the institution applies a rigorous and well documented process for identifying an economic downturn and assessing its effects on the drawing of credit limits and for producing conversion factor estimates appropriate for an economic downturn;
(d)
the institution incorporates in the conversion factor estimates any adverse dependencies that have been identified between on the one hand the selected economic indicators and on the other hand the drawing of credit limits.
Requirements on policies and strategies for account monitoring and payment processing
In order to assess compliance with the requirements regarding the estimation of the conversion factors as referred to in point (d) and (e) of Article 182(1) of Regulation (EU) No 575/2013, competent authorities shall verify that the institution has policies and strategies in place in respect of account monitoring and payment processing, and has adequate systems and procedures to monitor facility amounts on a daily basis.
SECTION 6 — Methodology for assessing the effect of guarantees and credit derivatives
Eligibility of guarantors and guarantees
When assessing compliance with the requirements for assessing the effect of guarantees and credit derivatives on risk parameters as referred to in Article 183 of Regulation (EU) No 575/2013, competent authorities shall verify that:
(a)
the institution has clearly specified criteria for identifying situations where PD estimates or LGD estimates are to be adjusted in order to incorporate mitigating effects of guarantees, and that those criteria are used consistently over time;
(b)
where the PD of the protection provider is to be used for the purpose of adjusting the risk-weighted exposure amounts in accordance with Article 153(3) of Regulation (EU) No 575/2013, the mitigating effects of guarantees are not included in the estimates of LGD or PD of the obligor;
(c)
the institution has clearly specified criteria for recognising guarantors and guarantees for the calculation of risk-weighted exposure amounts, in particular through own estimates of LGD or PD;
(d)
the institution documents the criteria for adjusting own estimates of LGD or PD to reflect the effects of guarantees;
(e)
in its own estimates of LGD or PD the institution recognises only the guarantees that meet the following criteria:
(i)
where the guarantor is internally rated by the institution with a rating system that has already been approved by the competent authorities for the purpose of the IRB Approach, the guarantee meets the requirements laid down in Article 183(1)(c) of Regulation (EU) No 575/2013;
(ii)
where the institution has received permission to use the Standardised Approach pursuant to Articles 148 and 150 of Regulation (EU) No 575/2013 for exposures to entities such as the guarantor both of the following are met:
—
the guarantor is assigned to an exposure class in accordance with Article 147 of Regulation (EU) No 575/2013 as an institution, a central government, a central bank or a corporate entity that has been given a credit assessment by an ECAI,
—
the guarantee meets the requirements set out in Articles 213 to 216 of Regulation (EU) No 575/2013.
(f)
the institution meets the requirements of points (a) and (e) also for the single-name credit derivatives.
SECTION 7 — Methodology for assessing the requirements for purchased receivables
Risk parameter estimates for purchased corporate receivables
1. When assessing the adequacy of PD and LGD estimates for purchased corporate receivables, where the institution derives PD or LGD for purchased corporate receivables from an estimate of EL in accordance with Article 160(2) and point (e) and (f) of Article 161(1) and an appropriate estimate of PD or LGD, competent authorities shall verify that:
(a)
EL is estimated from the long-run average of one-year total loss rates or by another appropriate approach;
(b)
the process for estimating the total loss is consistent with the concept of LGD as set out in Article 181(1)(a) of Regulation (EU) No 575/2013;
(c)
that the institution is able to decompose its EL estimates into PDs and LGDs in a reliable way;
(d)
in the case of purchased corporate receivables where Article 153(6) of Regulation (EU) No 575/2013 is applied, sufficient external and internal data are used.
2. When assessing the adequacy of PD and LGD estimates for purchased corporate receivables in cases other than those referred to in paragraph 1, competent authorities shall:
(a)
assess those estimates in accordance with Articles 42 to 52;
(b)
verify that the requirements of Article 184 of Regulation (EU) No 575/2013 are met.
Source: EUR-Lex (Publications Office of the EU), © European Union, reuse permitted under Commission Decision 2011/833/EU.
