Société Générale / Risk Report - Pillar III

6 CREDIT AND COUNTERPARTY CREDIT RISK RISK MEASUREMENT AND INTERNAL RATINGS

Credit risk measurements of retail clients The Group has implemented the following system for the retail portfolio made up of individual customers, SCIs (real estate investment companies - Sociétés civiles immobilières ) and professional customers: RATING SYSTEM AND ASSOCIATED PROBABILITY OF DEFAULT The modelling of the probability of default of retail client counterparties is carried out specifically by each of the Group’s business lines recording its assets using the AIRB method. The models incorporate data on the payment behaviour of counterparties. They are segmented by type of customer and distinguish between retail customers, professional customers, very small businesses and real estate investment companies (Sociétés civiles immobilières) . The counterparties of each segment are classified automatically, using statistical models, into homogeneous risk pools, each of which is assigned a probability of default. These estimates are adjusted by a safety margin to estimate as best as possible a complete default cycle, using a through-the-cycle (TTC) approach.

LGD MODELS The models for estimating the Loss Given Default (LGD) of retail customers are specifically applied by business line portfolio and by product, according to the existence or not of collateral. Consistent with operational recovery processes, estimate methods are generally based on a two-step modelling process that initially estimates the proportion of defaulted loans in loan termination, followed by the loss incurred in case of loan termination. The expected losses are estimated using internal long-term historical recovery data for exposures that have defaulted. These estimates are adjusted by safety margins in order to reflect the possible impact of a downturn. CCF MODELS For its off-balance sheet exposures, Societe Generale applies its estimates for revolving loans and overdrafts on current accounts held by retail and professional customers.

TABLE 27: RETAIL CLIENTS - MAIN CHARACTERISTICS OF MODELS ANDMETHODS USED

Parameter modelled

Portfolio/Category of Basel assets

Methodology Number of years of default/loss

Number of models

RETAIL CLIENTS

8 models according to entity, type of guarantee (security, mortgage), type of counterparty: individuals or professionals/VSB, real-estate investment company (SCI). 15 models according to entity and to the nature and object of the loan: personal loan, consumer loan, car loan, etc. 5 models according to entity and nature of the loan: overdraft on current account, revolving credit or consumer loan. 10 models according to entity, nature of the loan (medium- and long-term investment credits, short-term credit, car loans), and type of counterparty (individual or real-estate investment company (SCI)). 8 models according to entity, type of guarantee (security, mortgage), and type of counterparty: individuals or professionals/VSB, real-estate investment company (SCI). 17 models according to entity and to the nature and object of the loan: personal loan, consumer loan, car loan, etc. 7 models according to entity and nature of the loan: overdraft on current account, revolving credit or consumer loan. 12 models according to entity, nature of the loan (medium- and long-term investment credits, short-term credit, car loans), and type of counterparty (individual or real- estate investment company (SCI)). 12 calibrations by entity for revolving products and personal overdrafts.

Residential real estate

Statistical model (regression), behavioural score. Defaults observed over a period of more than 5 years.

Other loans to individual customers

Statistical model (regression), behavioural score. Defaults observed over a period of more than 5 years.

Probability of Default (PD)

Renewable exposures

Statistical model (regression), behavioural score. Defaults observed over a period of more than 5 years.

Professionals and very small businesses

Statistical model (regression or segmentation), behavioural score. Defaults observed over a period of more than 5 years.

Statistical model of expected recoverable flows based on the current flows. Model adjusted by expert opinions if necessary. Losses and recoverable flows observed over a period of more than 10 years. Statistical model of expected recoverable flows based on the current flows. Model adjusted by expert opinions if necessary. Losses and recoverable flows observed over a period of more than 10 years. Statistical model of expected recoverable flows based on the current flows. Model adjusted by expert opinions if necessary. Losses and recoverable flows observed over a period of more than 10 years. Statistical model of expected recoverable flows based on the current flows. Model adjusted by expert opinions if necessary. Losses and recoverable flows observed over a period of more than 10 years.

Residential real estate

Other loans to individual customers

Loss Given Default (LGD)

Renewable exposures

Professionals and very small businesses

Credit Conversion Factor (CCF)

Renewable exposures

Models calibrated by segment over a period of observation of defaults of more than 5 years.

80

PILLAR 3 - 2020 | SOCIETE GENERALE GROUP |