According to the statement of the European Banking Authority on the prudential framework regarding Default, Forbearance and IFRS 9 in light of COVID-19 measures published on March 25, 2020, saying that the use of COVID-19 aid tools in the form of repayment moratorium, meeting the EBA guidelines on legislative and non-legislative moratoria on loan repayments applied in the light of the COVID-19 crisis published on April 2, 2020 does not automatically classify exposures to default and forbearance, as well as according to the UKNF (Polish Financial Supervision Authority) statement published as a part of the Supervisory Impulse Package for Security and Development that UKNF will apply a flexible approach to the application of EBA guidelines for unsupported and restructured exposures, the Group does not classify the granting of the moratoria due to the COVID-19 crisis as forbearance, with the exception of corporate clients, for whom there is applied an approach based on individual assessment whether classification of such client’s exposure as forborne is required, in accordance with the Group’s internal regulations.
Due to the deterioration of the economic situation in the country resulting from the COVID-19 epidemic, the Group has taken additional actions aimed at including this information in the expected credit losses. Due to the uncertainty caused by dynamic situation changes, the Group’s activities were spread over time and in particular covered:
- review of sectors and individual clients of the corporate portfolio, in particular clients under observation, in order to verify the potential increase in the probability of failure to implement the restructuring plans, which was already carried out in March as the first activity of the Group as part of taking into account the impact of the epidemic on clients' financial situation,
- modification of the weight of macroeconomic scenarios, consisting in assigning a 100% weight to the pessimistic scenario, in the expected credit loss model, applied in the first quarter of 2020,
- updating models of the relationship between the long-term PD parameter and macroeconomic variables, based on historical data and the currently observed economic situation, in the second quarter of 2020,
- updating macroeconomic forecasts, taking into account the impact of COVID-19 and state aid actions, affecting long-term PD, EAD and LGD parameters, as well as the level of exposure allocation to stage 2, in particular by increasing the expected level of allocation for some portfolios due to the expected increase in the loss ratio, in the second quarter of 2020,
- restoration of macroeconomic scenario weights of 60% for the baseline scenario, 20% for the optimistic and 20% for the pessimistic respectively, in the expected credit loss model, while taking into account the current macroeconomic forecasts implemented directly in the risk parameters, in the second quarter of 2020,
- monitoring of macroeconomic forecasts in order to verify the macroeconomic data used in the models in terms of their adequacy to the actual development of the economic situation in Poland. There was no basis for changing the macroeconomic forecasts within the risk parameters, in the third and fourth quarter of 2020.
Due to the uncertainty related to the difficulties in observation of the timeliness of repayment of loans covered by moratoria, Group also decided in the third quarter of 2020 to reclassify, some of the exposures of retail clients covered by this form of support, selected on the basis of behavioral characteristics, to stage 2 despite no evidence of a significant increase of credit risk, which resulted in the recognition of additional cost of credit risk at the end of 2020 in the amount of PLN 53.1 million. The total gross carrying amount of the reclassified portfolio as at December 31, 2020 was PLN 3,227.57 million.
The change had an impact on exposure allocation to the stages. The share of stage 2 in the total exposure of the loan portfolio increased but its coverage with provisions decreased, which is a natural consequence of allocating to stage 2 exposures with a lower probability of default (lower PD).
In addition, at the end of 2020, the Group decided to automatically reclassify exposures subject to the relief in the form of the statutory moratorium to stage 3, or, in justified cases, to stage 2. The final allocation of the exposure to stage 2 was possible after conducting additional analyzes taking into account quantitative and qualitative factors, such as: co-borrower in the contract, credit quality of all customer exposures, the amount of cash flow after the date of the application for a moratorium. The reclassification resulted in the recognition of additional cost of credit risk in the amount of PLN 1.7 million. The total gross carrying amount of the reclassified portfolio as at December 31, 2020 was PLN 9.97 million.
The above-mentioned activities resulted in recognition of additional cost of credit risk in the amount of PLN 330.3 million in the portfolio measured at amortised cost. In addition, these activities had an impact on the valuation of the loan portfolio at fair value through profit or loss, for which the Group recognized an additional cost of PLN 10.3 million.
Due to the fact that changes in risk parameters following the outbreak of the COVID-19 pandemic were implemented over a period of several months in a very dynamically changing macroeconomic environment, the Group decided to present the total value of their impact on December 31, 2020, as presented in the table below.
| PLN (‘000) | Individual customers | Corporate customers | Total |
| Financial asset measured at amortised cost | 134,973 | 195,349 | 330,322 |
| Stage 1 | 3,060 | 4,138 | 7,198 |
| Stage 2 | 114,869 | 51,397 | 166,266 |
| Stage 3 | 17,044 | 139,814 | 156,858 |
| Financial assets measured at fair value through profit or loss | 9,414 | 838 | 10,252 |
The most important write-offs in stage 3 were concerned in corporate clients from the following sectors: accommodation and food service activities, real estate activities and transport and storage.
As at December 31, 2020, the Group applied management corrections (overlays) in the amount of PLN 12.61 million. These adjustments do not result from the methodology used to calculate the allowances for expected credit losses.
The Group will continue to analyze the impact of COVID-19 and state aid programs on the result of the cost of credit risk in the upcoming quarters.
In order to assess expected credit loss (ECL) sensitivity to the future macroeconomic conditions, the Group determined the ECL value separately for each of the scenarios used for the purposes of calculating the expected credit risk losses. The impact of each of the scenarios is presented in the table in the next chapters of the document.
The table below presents forecasts of the main macroeconomic indicators used in the expected credit loss model as of December 31, 2020 and December 31, 2019:
| Scenario as at 31.12.2020 | base | optimistic | pessimistic | ||||
|---|---|---|---|---|---|---|---|
| Probability | 60% | 20% | 20% | ||||
| The first year of the forecast | The average for the next two years | The first year of the forecast | The average for the next two years | The first year of the forecast | The average for the next two years | ||
| GDP | YoY | -4.2 | 4.4 | 0.0 | 3.9 | -6.4 | 0.4 |
| Unemployment rate | end of the year | 7.0% | 5.5% | 3.3% | 2.9% | 9.2% | 11.9% |
| WIBOR3M | end of the year | 0.31 | 0.38 | 0.7 | 0.7 | 0.1 | 0.1 |
| Real estate price index | YoY | 101.0 | 105.5 | 103.0 | 105.9 | 91.9 | 102.8 |
| CHFPLN | end of the year | 4.21 | 4.03 | 4.11 | 3.93 | 4.43 | 4.43 |
| Scenario as at 31.12.2019 | base | optimistic | pessimistic | ||||
|---|---|---|---|---|---|---|---|
| Probability | 60% | 20% | 20% | ||||
| The first year of the forecast | The average for the next two years | The first year of the forecast | The average for the next two years | The first year of the forecast | The average for the next two years | ||
| GDP | YoY | 3.3 | 2.8 | 4.3 | 3.4 | 0.6 | 1.4 |
| Unemployment rate | end of the year | 3.9% | 4.0% | 3.0% | 3.5% | 6.5% | 8.0% |
| WIBOR3M | end of the year | 2.0 | 2.3 | 1.7 | 1.7 | 0.5 | 0.5 |
| Real estate price index | YoY | 102.0 | 101.0 | 103.8 | 103.2 | 100.0 | 102.6 |
| CHFPLN | end of the year | 3.62 | 3.48 | 3.72 | 3.62 | 4.48 | 4.15 |
The value of credit risk cost is the result of all presented macroeconomic scenarios and the weights assigned to them. Impact of individual scenarios on the credit risk costs is as shown in the table below (weight of a given scenario 100%):
| Scenario as at 31.12.2020 |
The change of the value of credit risk costs (PLN (‘000)) 31.12.2020 |
| optimistic | 47,136 |
| base | 10,316 |
| pessimistic | -135,596 |
The above results were estimated taking into account the equal allocation to the stage 2 based on the weighted average of all 3 macroeconomic scenarios, without and assumption of additional potential migrations between stages. The ECL sensitivity analysis was performed on 84% of the assets of the portfolio of loans and advances to customers and off-balance sheet liabilities granted to them.
The reason for changes in the key values in the Group’s risk models were changes in macroeconomic indicators following the outbreak of the COVID-19 pandemic.
Apart from the activities related to the updating of the credit risk models mentioned above, the Bank did not introduce any other dedicated changes into the models used for the purposes of calculating the expected credit risk losses. Due to:
- lack of significant impact of the current economic situation (resulting mainly from the applied support measures) on parameters such as default rate or level of portfolio losses,
- results of consultations with other units of the Bank’s risk division indicating that there is no need to take into account additional effects of the COVID-19 impact on the models,
In the model management process, the Group has carried out cyclical activities such as:
- cyclical recalibration of the short-term PD models reflecting the current level of the portfolio’s default rate,
- cyclical recalibration of the long-term PD models and quantitative staging model,
- recalibration of the long-term LGD model for corporates adjusting the estimated level of losses to the observed in recent years.