4.2 Data requirements

In order to comply with the requirement of Article 76 of the RTS on IRB assessment methodology that institutions should have sound policies, processes and methods for assessing and improving the quality of data used for the purpose of credit risk measurement and management processes, institutions should ensure that those policies apply to all data used in model development and calibration, as well as to the data used in the application of the risk parameters.

In order for the data used in the model development and in the application of risk parameters as inputs into the model to meet the requirements of accuracy, completeness and appropriateness specified in Article 174(b) of Regulation (EU) No 575/2013, it should be sufficiently precise to avoid material distortions of the outcome of the assignment of exposures to obligors or facility grades or pools, and it should not contain any biases which make the data unfit for purpose. 4.2.2 Governance for data representativeness

In order to comply with the requirement of the representativeness of data used in the PD and LGD models specified in Articles 174(c), 179(1)(d) and 179(2)(b) of Regulation (EU) No 575/2013 as well as in Articles 40 and 45 of the RTS on IRB assessment methodology, institutions should have sound policies, processes and methods for assessing the representativeness of data used for the purpose of estimation of risk parameters. Institutions should specify in their internal policies the statistical tests and metrics to be used for the purpose of assessing the representativeness of data used for risk differentiation and, separately, for data underlying the risk quantification. Institutions should also specify methods for qualitative assessment of data for the cases, defined in their policies, where the application of statistical tests is not possible.

Institutions should use the same standards and methods for the assessment of representativeness of data stemming from different sources, including internal, external and pooled data or a combination of these, unless different methods are justified by the specificity of the data source or availability of information.

Where external or pooled data are used institutions should obtain sufficient information from the data providers to assess the representativeness of such external or pooled data to the institutions’ own portfolios and processes.

4.2.3 Representativeness of data for model development

Institutions should analyse the representativeness of data in the case of statistical models and other mechanical methods used to assign exposures to grades or pools, as well as in the case of statistical default prediction models generating default probability estimates for individual obligors or facilities. Institutions should select an appropriate dataset for the purpose of model development to ensure that the performance of the model on the application portfolio, in particular its discriminatory power, is not significantly hindered by insufficient representativeness of data.

For the purposes of ensuring that the data used in developing the model for assigning obligors or exposures to grades or pools is representative of the application portfolio covered by the relevant model, as required in Article 174(c) of Regulation (EU) No 575/2013 and Article 40(2) of the RTS on IRB assessment methodology institutions should analyse the representativeness of the data at the stage of model development in terms of all of the following:

• (a) the scope of application;
• (b) the definition of default;
• (c) the distribution of the relevant risk characteristics;
• (d) lending standards and recovery policies.

For the purpose of paragraph 21(a) institutions should analyse the segmentation of exposures and consider whether there were any changes to the scope of application of the considered model over the period covered by the data used in developing the model for assigning obligors or exposures to grades or pools. Where such changes were observed institutions should analyse the risk drivers relevant for the change of the scope of application of the model by comparing their distribution in the RDS before and after the change as well as with the distribution of those risk drivers in the application portfolio. For this purpose institutions should apply statistical methodologies such as cluster analysis or similar techniques to demonstrate representativeness. In the case of pooled models the analysis should be performed with regard to the part of the scope of the model that is used by an institution.

For the purpose of paragraph 21(b) institutions should ensure that the definition of default underlying the data used for model development is consistent over time and, in particular, that it is consistent with all of the following:

• (a) that adjustments have been made to achieve consistency with the current default definition where the default definition has been changed during the observation period;
• (b) that adequate measures have been adopted by the institution, where the model covers exposures in several jurisdictions having or having had different default definitions;
• (c) that the definition of default in each data source has been analysed separately;
• (d) that the definition of default used for the purposes of model development does not have a negative impact on the structure and performance of the rating model, in terms of risk differentiation and predictive power, where this definition is different from the definition of default used by the institution in accordance with Article 178 of Regulation (EU) No 575/2013.

For the purpose of paragraph 21(c) institutions should analyse the distribution and range of values of key risk characteristics of the data used in developing the model for risk differentiation in comparison with the application portfolio. With regard to LGD models, institutions should perform such analysis separately for non-defaulted and defaulted exposures.

Institutions should analyse the representativeness of the data in terms of the structure of the portfolio by relevant risk characteristics based on statistical tests specified in their policies to ensure that the range of values observed on these risk characteristics in the application portfolio is adequately reflected in the development sample. Where the application of statistical tests is not possible, institutions should carry out at least a qualitative analysis on the basis of the descriptive statistics of the structure of the portfolio, taking into account the possible seasoning effects referred to in Article 180(2)(f) of Regulation (EU) No 575/2013. When considering the results of this analysis, institutions should take into account the sensitivity of the risk characteristics to economic conditions. Material differences in the key risk characteristics between the data sample and the application portfolio should be addressed, for example by using another data sample or a subset of observations or by adequately reflecting these risk characteristics as risk drivers in the model.

For the purpose of paragraph 21(d) institutions should analyse whether, over the relevant historical observation period, there were significant changes in their lending standards or recovery policies or in the relevant legal environment, including changes in insolvency law, legal foreclosure procedures and any legal regulations related to realisation of collaterals, which may influence the level of risk or the distribution or ranges of the risk characteristics in the portfolio covered by the considered model. Where institutions observe such changes they should compare the data included in the RDS before and after the change of the policy. Institutions should ensure comparability of the current underwriting or recovery standards with those applied to the observations included in the RDS and used for model development.

Within the PD model the representativeness of data used in developing the model for risk differentiation does not require that the proportion of defaulted and non-defaulted exposures in this dataset be equal to the proportion of defaulted and non-defaulted exposures in the institution’s application portfolio. However, institutions should have a sufficient number of defaulted and non-defaulted observations in the development dataset and they should document the difference.

4.2.4 Representativeness of data for calibration of risk parameters

In order for institutions to ensure that the data used in risk quantification is representative of the application portfolio covered by the relevant model in accordance with Sub-section 2 of Section 6 of Chapter 3 in Part Three of Regulation (EU) No 575/201 3 and Article 45(2) of the RTS on IRB assessment methodology, institutions should analyse the comparability of the data used for the purpose of calculating long-run average default rates or long-run average LGDs as referred to in Article 179(1)(d) of Regulation (EU) No 575/2013 and, where relevant, the representativeness of the pool in accordance with Article 179(2)(b) of that Regulation, in terms of all of the following:

• (a) the scope of application;
• (b) the definition of default;
• (c) the distribution of the relevant risk characteristics;
• (d) the current and foreseeable economic or market conditions;
• (e) lending standards and recovery policies.

For the purpose of paragraph 28(a) institutions should perform an analysis as specified in paragraph 22.

For the purpose of paragraph 28(b) and in order to ensure that the definition of default underlying the data used for risk quantification from each data source is consistent with the requirements of Article 178 of Regulation (EU) No 575/2013, institutions should compare the definition of default applied by the institution currently with the definitions used for the observations included in the dataset used for risk quantification. Where the definition of default has changed during the historical observation period institutions should assess the representativeness of historical data included in the RDS and used for risk quantification in the same way as specified for external data in Chapter 6 of the EBA Guidelines on the application of the definition of default under Article 178 of Regulation (EU) No 575/2013. Where the definition of default has changed during the historical observation period more than once, institutions should perform the analysis of each of the past definitions of default separately.

For the purpose of paragraph 28(c) institutions should perform an appropriate analysis to ensure that at the level of the calibration segment the ranges of values of the key risk characteristics in the application portfolio are comparable to those in the portfolio constituting the reference data set for risk quantification to the degree required to ensure that the risk quantification is not biased.

For the purpose of paragraph 28(d) institutions should perform the analysis of the market and economic conditions underlying the data in the following manner:

• (a) with regard to the PD estimation, in accordance with section 5.3.4;
• (b) with regard to the LGD estimation, in accordance with section 6.3.2 and taking into account the consideration of economic downturn as required by Article 181(1)(b) of Regulation (EU) No 575/2013.

For the purpose of paragraph 28(e) institutions should analyse whether there were significant changes in the lending standards or recovery policies over the relevant historical observation period that may influence the level of risk or the distribution or ranges of the characteristics of relevant risk drivers in the portfolio covered by the considered model. Where institutions observe such changes they should analyse the potential bias in the estimates of risk parameters resulting from these changes in the following manner:

• (c) with regard to the PD estimation, in terms of the level of default rates and the likely range of variability of default rates;
• (d) with regard to the LGD estimation, in terms of loss rates, average duration of the recovery processes, frequencies of use of certain recovery scenarios and the loss severity distributions.

Where the representativeness of data assessed in accordance with paragraphs 28 to 33 is insufficient and leads to a bias or increased uncertainty of risk quantification, institutions should introduce an appropriate adjustment to correct the bias and they should apply a margin of conservatism in accordance with section 4.4.