Société Générale / Risk report 2021 - Pillar 3 2020

7 COUNTERPARTY CREDIT RISK COUNTERPARTY CREDIT RISK MEASURES PILLAR 3 - 2021 | SOCIETE GENERALE GROUP | 142 COUNTERPARTY CREDIT RISKMEASURES 7.3 REPLACEMENT RISK 7.3.1 The measure of replacement risk is based on an internal model that determines the Group’s exposure profiles. PRINCIPLES OF THE MODEL The future fair value of market transactions with each counterparty is estimated from Monte Carlo models based on a historical analysis of market risk factors. The principle of the model is to represent the possible future financial market conditions by simulating the evolutions of the main risk factors to which the institution’s portfolio is sensitive. For these simulations, the model uses different diffusion models to account for the characteristics inherent in the risk factors considered and uses a 10-year history for calibration. The transactions with the various counterparties are then revalued according to these different scenarios at the different future dates until the maturity of the transactions, taking into account the terms and conditions defined in the contractual legal framework agreed, notably in terms of netting and collateralisation. The distribution of the counterparty exposures thus obtained allows the calculation of regulatory capital for counterparty credit risk and the economic monitoring of positions. The Risk Department responsible for Model Risk Management at Group level, assesses the theoretical robustness (review of the design and development quality), the compliance of the implementation and the suitability of the use of the model. This independent review process ends with (i) a report that describes the scope of the review, the tests carried out, the results of the review, the conclusions or recommendations and (ii) review and approval Committees. This process results in reporting to the appropriate bodies. REGULATORY INDICATOR With respect to the calculation of capital requirements for counterparty credit risk, the ECB, following the Targeted Review of Internal Models, has renewed the approval for using the internal model described above to determine the Effective Expected Positive Exposure (EEPE) indicator. For products not covered by the internal model as well as for entities in the Societe Generale Group that have not been authorised by the supervisor to use the internal model, the Group uses the market-price valuation method for derivatives (1) and the general financial security-based method for securities financing transactions (SFT). The effects of compensation agreements and collateralisation are taken into account either by their simulation in the internal model, or by applying the rules as defined in the market-price valuation method or the financial security-based method, and by subtracting the value of the collateral. These exposures are then weighted by rates resulting from the credit quality of the counterparty to compute the Risk-Weighted Assets (RWA). These rates can be determined by the standardised approach or the advanced approach (AIRB). The RWA breakdown for each approach is available in the “Analysis of Counterparty Credit Risk Exposure by Approach” table in Section 7.4 “Quantitative Information”. ECONOMIC INDICATOR For the economic monitoring of positions, Societe Generale relies mainly on a maximum exposure indicator determined from the Monte Carlo simulation, called Credit Value-at-Risk (CVaR) internally or PFE (Potential Future Exposure). This is the maximum amount of loss that could occur after eliminating 1% of the most adverse occurrences. This indicator is calculated at different future dates, which are then aggregated into segments, each of them being framed by limits. The Group has also developed a set of stress test scenarios to determine the exposure that would result from changes in the fair value of transactions with all its counterparties in the event of an extreme shock affecting the market parameters. CREDIT VALUATION 7.3.2 ADJUSTMENT MAIN PRINCIPLES The CVA (Credit Valuation Adjustment) refers to adjustment to Marked-to-market of derivatives and repos portfolio to take into account counterparty credit risk of each counterparty facing the Group in the valuation. This adjustment is equivalent to the counterparty credit risk hedging cost in the Credit Default Swap (CDS) market. For a specific counterparty, the CVA is determined on the basis of: the positive expected exposure to the counterparty, which is the p average of the positive hypothetical future exposure values for a transaction or a group of transactions. It is mainly determined using risk neutral Monte Carlo simulations of risk factors that may affect the valuation of the derivatives transactions. The transactions are revalued through time according to the different scenarios, taking into account the terms and conditions defined in the contractual legal framework agreed, notably in terms of netting and collateralisation ( i.e that transactions with appropriate credit mitigants will generate lower expected exposure compared to transactions without credit mitigants); the probability of default of the counterparty, which is linked to the p level of CDS spreads; the amount of losses in the event of default (LGD – Loss Given p Default taking into account the recovery rate). The Group computes this adjustment for all counterparties which are not subject to a daily margin call or for which collateral only partially covers the exposure. In this method, the EAD (Exposure At Default) relating to the Bank's counterparty credit risk is by aggregating the positive market values of all transactions (replacement (1) cost) supplemented by an add-on factor.