2018 Registration document and annual fi nancial report - BNP PARIBAS166

4 CONSOLIDATED FINANCIAL STATEMENTS FOR THE YEAR ENDED 31 DECEMBER 2018

4

Notes to the fi nancial statements

General model The Group identifi es three stages that correspond each to a specifi c status with regards to the evolution of counterparty credit risk since the initial recognition of the asset.

■ 12-month expected credit losses ( stage 1 ): If at the reporting date, the credit risk of the fi nancial instrument has not increased signifi cantly since its initial recognition, this instrument is impaired at an amount equal to 12-month expected credit losses (resulting from the risk of default within the next 12 months);

■ lifetime expected credit losses for non-impaired assets ( stage 2 ): The loss allowance is measured at an amount equal to the lifetime expected credit losses if the credit risk of the fi nancial instrument has increased signifi cantly since initial recognition, but the fi nancial asset is not considered credit-impaired or doubtful;

■ lifetime expected credit losses for credit-impaired or doubtful fi nancial assets ( stage 3 ): the loss allowance is also measured for an amount equal to the lifetime expected credit losses.

This general model is applied to all instruments within the scope of IFRS 9 impairment, except for purchased or originated credit-impaired fi nancial assets and instruments for which a simplifi ed model is used (see below).

The IFRS 9 expected credit loss approach is symmetrical, i.e. if lifetime expected credit losses have been recognised in a previous reporting period, and if it is assessed in the current reporting period that there is no longer any signifi cant increase in credit risk since initial recognition, the loss allowance reverts to a 12-months expected credit loss.

As regards interest income, under stage 1 and 2, it is calculated on the gross carrying amount. Under stage 3 , interest income is calculated on the amortised cost (i.e. the gross carrying amount adjusted for the loss allowance).

Definition of default The defi nition of default is aligned with the Basel regulatory default defi nition, with a rebuttable presumption that the default occurs no later than 90 days past-due.

The defi nition of default is used consistently for assessing the increase in credit risk and measuring expected credit losses.

Doubtful credit-impaired financial assets

Definition

A fi nancial asset is considered doubtful and classifi ed in stage 3 when one or more events that have a detrimental impact on the estimated future cash fl ows of that fi nancial asset have occurred.

At an individual level, objective evidence that a fi nancial asset is credit- impaired includes observable data regarding the following events: the existence of accounts that are more than 90 days past due; knowledge or indications that the borrower meets signifi cant fi nancial diffi culties, such that a risk can be considered to have arisen regardless of whether the borrower has missed any payments; concessions with respect to the credit terms granted to the borrower that the lender would not have considered had the borrower not been meeting fi nancial diffi culties (see section Restructuring of fi nancial assets for fi nancial diffi culties ).

Specific cases of purchased or originated credit-impaired assets

In some cases, financial assets are credit-impaired at their initial recognition.

For these assets, there is no loss allowance accounted for at initial recognition. The effective interest rate is calculated taking into account the lifetime expected credit losses in the initial estimated cash fl ows. Any change in lifetime expected credit losses since initial recognition, positive or negative, is recognis ed as a loss allowance adjustment in profi t or loss.

Simplified model The simplifi ed approach consists in accounting for a loss allowance corresponding to lifetime expected credit losses since initial recognition, and at each reporting date.

The Group applies this model to trade receivables with a maturity shorter than 12 months.

Significant increase in credit risk The signifi cant increase in credit risk may be assessed on an individual basis or on a collective basis (by grouping fi nancial instruments according to common credit risk characteristics) taking into account all reasonable and supportable information and comparing the risk of default of the fi nancial instrument at the reporting date with the risk of default of the fi nancial instrument at the date of initial recognition.

Assessment of deterioration is based on the comparison of the probabilities of default or the ratings on the date of initial recognition with those existing at the reporting date.

There is also, according to the standard, a rebuttable presumption that the credit risk of an instrument has signifi cantly increased since initial recognition when the contractual payments are more than 30 days past due.

In the consumer credit specialised business, a signifi cant increase in credit risk is also considered when a past due event has occurred within the last 12 months, even if regularized since.

The principles applied to assess the signifi cant increase in credit risk are detailed in note 3.h Cost of risk.

Measurement of expected credit losses Expected credit losses are defi ned as an estimate of credit losses (i.e. the present value of all cash shortfalls) weighted by the probability of occurrence of these losses over the expected life of fi nancial instruments . They are measured on an individual basis, for all exposures.

In practice, for exposures classifi ed in stage 1 and stage 2, expected credit losses are measured as the product of the probability of default (PD), Loss Given Default (LGD) and exposure at default (EAD), discounted at the effective interest rate of the exposure (EIR). They result from the risk of default within the next 12 months (stage 1), or from the risk of default over the maturity of the facility (stage 2). In the consumer credit specialised business, because of the specifi city of credit exposures, the methodology used is based on the probability of transition to term forfeiture, and on discounted loss rates after term forfeiture. The measurement of these parameters is performed on a statistical basis for homogeneous populations.