BPCE - 2020 Universal Registration Document

FINANCIAL REPORT

IFRS CONSOLIDATED FINANCIAL STATEMENTS OF BPCE SA GROUP AS AT DECEMBER 31, 2020

Accordingly, Groupe BPCE revised its forward-lookingeconomic forecasts to take account of the Covid-19 situation and the economic support measures taken in response. The Group applied three main scenarios to calculate its IFRS 9 provisions with projections to 2023: the core scenario was updated based on the scenarios • determined by the Group’s economists in September 2020; a pessimistic scenario, corresponding to a more severe • deterioration in the macro-economic variables defined in relation to the core scenario;

an optimistic scenario, corresponding to an improvement in • the macro-economic variables defined in relation to the core scenario; Following the historic economic shock from Covid-19 in 2020,the core scenario sees a sharp recovery in GDP as from 2021, returning gradually to a more normal long-term economic growth trend over the years that follow. Economic activity should get back to pre-crisis (2019) levels by 2023. We show below the 4-year projections for key economic variables used, based on the scenario of the Group's economists for each range limit:

Optimistic

Core

Pessimistic

GDP Unemp.

10Y yld.

GDP Unemp.

10Y yld.

GDP Unemp.

10Y yld.

2020 2021 2022 2023

(5.8%)

7.4% 0.30% (9.6%)

8.5% (0.11%)

(12.3%)

11.5% (0.60%)

10.0% 8.7% 0.70% 7.2% 10.0% 0.01% 4.0% 12.5% (0.40%) 4.3% 7.9% 0.82% 2.6% 9.3% 0.13% 0.9% 11.7% (0.28%) 2.8% 7.6% 0.94% 1.6% 9.0% 0.25% 0.4% 11.4% (0.16%)

In addition, BPCE SA group has extended and adapted this approach by adjusting for a number of factors specific to certain scopes or significant markets: Each scenario is therefore weighted based on how close it is to the consensus forecast for the main economic variables in each scope or significant market of the Group. For Retail Banking, projections are calculated using the main economic variables such as GDP, unemployment and interest rates on 10-year French sovereign debt. For Corporate & Investment Banking, the economic variables applied relate to the international situation and make greater use of financial and market data. For Retail Banking, economic scenarios have been adapted to take account of uncertainties surrounding economic forecasts and economic support measures (SGL, partial furlough, tax breaks). The result of these adaptations is to: mitigate the suddennessand intensity of the crisis with a 60% • moderating factor applied to GDP impacts under the scenarios. For instance, in the central scenario, the GDP figure projected is a weighted average of the scenario's starting value (-9.6%, weighted 40%) and long-term growth in France (+1.4%, weighted 60%). This adjustment is consistent with the ECB's communications on accounting for the Covid-19 under IFRS 9 and the EBA's guidance on moratoria: and spread the effects of the crisis over a longer period, pushing • back the scenario by nine months, which means it takes an additional nine months for the worsening of GDP and other variables to feed through to probabilities of default. These adjustments reflect the positive impacts of the various measures taken by the government to prop up the economy: most notably defaults are reduced and delayed. In summary, the main changes made in 2020 to the calculation of expected credit loss were as following: updating of economic scenarios in September to better reflect • changes in forecasts (pre-crisis, scenarios were only updated annually, in June);

adaptation of the method for weighting economic scenarios to • reflect the high levels of uncertainty in current circumstances, which translate as widely disparate upper and lower limits for each scenario; adaptation of economic variables to take into account • economic uncertaintiesand the impact of government support measures; for Corporate & Investment Banking, the sector-based • adjustment of the probabilities of default replaces the use of the change in the rating of the sector as a criterion for monitoring the deterioration of the risk. This more granular approach better captures the influence of the sector on credit risk by making counterparty ratings more important as a discriminatory factor. It also helps mitigate the pro-cyclical effects of the previous methodology, which systematically downgraded to Stage 2 all counterparty contracts in a sector whose rating had fallen below a given threshold; in consumer finance, the introduction of a model that • translates the new economic assumptions into a more granular segmentation of portfolios. Based on the scenarios mentioned above and after taking into account the methodological adjustments and support measures, the calculation of expected credit losses led the BPCE SA group to recognize a cost of credit risk of €234 million in 2020, an increase of €243 million euros compared to 2019. ECL SENSITIVITY ANALYSIS The sensitivity of Retail Banking's expected credit loss to uncertainty about the level of the moderating factor and a 3-month delay to the economic scenario has been estimated: a variation of the moderating factor of +/- 10% around the • scenario value of 60% has an impact of approximately +/- €4 million;

5

401

UNIVERSAL REGISTRATION DOCUMENT 2020 | GROUPE BPCE