2019 Universal registration document and annual financial report - BNP PARIBAS 203

4Consolidated finanCial statements for the year ended 31 deCemBer 2019

4

Notes to the financial statements

Risk classes

Balance Sheet valuation

(in millions of euros) Main product types composing the Level 3 stock within the risk class

Valuation technique used for the product types considered

Main unobservable inputs for the product types considered

Range of unobservable input across Level 3

population considered Weighted

averageAsset Liability

Repurchase agreements

134 333 Long-term repo and reverse- repo agreements

Proxy techniques, based amongst other on the funding basis of a benchmark bond pool, that is actively traded and representative of the repo underlying

Long-term repo spread on private bonds (High Yield, High Grade) and on ABSs

0 pb to 77 pb 47 pb(b)

Interest rate derivatives

1,362 1,352

Hybrid Forex/Interest rates derivatives

Hybrid Forex interest rate option pricing model

Correlation between FX rate and interest rates. Main currency pairs are EUR/JPY, USD/JPY, AUD/JPY

10% to 71% 42%(b)

Hybrid inflation rates/Interest rates derivatives

Hybrid inflation interest rate option pricing model

Correlation between interest rates and inflation rates mainly in Europe.

5% to 45% 39%

Floors and caps on inflation rate or on the cumulative inflation (such as redemption floors), predominantly on European and French inflation

Inflation pricing model

Volatility of cumulative inflation

0.7% to 9% (c)

Volatility of the year on year inflation rate

0.3% to 1.8%

Forward Volatility products such as volatility swaps, mainly in euro

Interest rates option pricing model

Forward volatility of interest rates

0.3% to 0.6% (c)

Balance-guaranteed fixed rate, basis or cross currency swaps, predominantly on European collateral pools

Prepayment modelling Discounted cash flows

Constant prepayment rates

1.5% to 18% 4%(b)

Credit Derivatives

259 371

Collateralised Debt Obligations and index tranches for inactive index series

Base correlation projection technique and recovery modelling

Base correlation curve for bespoke portfolios

29% to 79% (c)

Inter-regions default cross correlation

80% to 90% 90%(a)

Recovery rate variance for single name underlyings

0 to 25% (c)

N-to-default baskets Credit default model Default correlation 50% to 85% 60.8%(b)

Single name Credit Default Swaps (other than CDS on ABSs and loans indices)

Stripping, extrapolation and interpolation

Credit default spreads beyond observation limit (10 years)

55 pb to 234 pb(1) 217 pb(a)

Illiquid credit default spread curves (across main tenors)

0 pb to 2.856 bp(2) 110 pb(a)

Equity Derivatives

1,374 6,629 Simple and complex derivatives on multi-underlying baskets on stocks

Various volatility option models

Unobservable equity volatility

0% to 94%(3) 26%(d)

Unobservable equity correlation

15% to 96% 68%(a)

(1) The upper part of the range relates to a significant balance sheet position on an issuer belonging to the telecommunication sector. The remaining positions relate mainly to sovereign and financial issuers.

(2) The upper bound of the range relates to a financial sector issuer that represents an insignificant portion of the balance sheet (CDSs with illiquid underlying instruments).

(3) The upper part of the range relates to one equity instrument representing a non-material portion of the balance sheet on options with equity underlying instruments. Including these inputs, the upper bound of the range would be around 118%.

(a) Weighting is not based on risks, but on an alternative methodology in relation with the Level 3 instruments (present value or notional). (b) Weights based on relevant risk axis at portfolio level. (c) No weighting, since no explicit sensitivity is attributed to these inputs. (d) Simple averaging.