Credit risk models quantify and manage the likelihood and impact of borrower default across individual exposures and portfolios. They inform underwriting, pricing, provisioning, capital allocation, portfolio steering, and performance measurement, making them a core pillar of modern bank risk governance and profitability.
Credit risk models in India support underwriting, provisioning, capital computation, portfolio steering, and pricing under RBI’s Basel III framework and Ind AS 109 expected credit loss regime for applicable entities, aligning risk decisions with solvency and profitability objectives across banks and NBFCs.
Uses of credit models
- Underwriting and limits: Internal models translate borrower risk into grades and PD–LGD–EAD measures for sanctioning, exposure caps, and covenanting in line with Board-approved policies and RBI capital rules.
- Provisioning and impairment: Ind AS 109 requires 12‑month or lifetime ECL based on significant increase in credit risk (SICR), interacting with IRACP prudential norms where the higher provision governs in practice for many lenders.
- Capital and ICAAP: Models feed Basel III Pillar‑1 RWAs under the Standardised or IRB approaches and underpin ICAAP/economic capital for Pillar‑2 risk assessment in India.
Types of models
- PD models: Obligor scorecards (financial, behavioral, bureau), survival/hazard models, and where relevant, structural distance‑to‑default overlays for market‑linked calibrations.
- LGD models: Facility‑level recoveries based on collateral type, seniority, legal process efficiency, costs, and downturn adjustments for capital.
- EAD models: Credit conversion factors for undrawn limits and amortization/curtailment dynamics for term facilities as per Basel components.
- Portfolio models: Factor/ASRF frameworks for correlated default and capital; used within ICAAP and concentration risk management.
Measuring credit risk
- Exposure level: Expected loss computed as EL = PD × LGD × EAD anchors ECL, pricing floors, and risk appetite triggers in Indian practice.
- Portfolio level: Loss distributions, concentration metrics, and stress tests drive capital buffers and sectoral limits under Basel III and RBI expectations.
Probability of Default (PD)
PD is the likelihood of default over a defined horizon (often 12‑month for capital and stage‑1 ECL; lifetime for stage‑2/3), calibrated to observed defaults and mapped to internal grades, with point‑in‑time and through‑the‑cycle variants used for ECL and capital respectively.
Methods for estimating PD
- Statistical scorecards using financial ratios, cash flows, and bureau signals for retail/MSME/corporate pools in Indian banks.
- Survival/hazard models embedding macro covariates for scenario‑conditional PDs in stress tests and ECL staging.
- Structural and hybrid overlays to align internal grades with Basel components for capital use if migrating beyond Standardised Approach.
Exposure at Default (EAD)
EAD reflects the expected outstanding at default, combining drawn amounts and estimated future drawings via credit conversion factors for revolving products and amortization for term loans, consistent with Basel risk components used in India.
Loss Given Default (LGD)
LGD measures the percentage loss conditional on default after recoveries, costs, and discounting; Indian calibrations differentiate collateral classes and apply downturn LGDs for capital while ECL uses unbiased, probability‑weighted recoveries.
Portfolio assessment of credit risk
- Concentration and correlation: Indian banks apply name/sector/geography limits and assess correlated clusters within ICAAP alongside capital buffers that exceed Basel minima.
- Stress testing: Scenario‑based shifts to PD/LGD/EAD and migration matrices inform capital planning and provisioning overlays under RBI expectations.
RAROC: definition and construction
RAROC = Risk‑adjusted net income divided by economic capital, aligning pricing and origination with solvency targets; Indian banks embed RAROC in deal screening and portfolio allocation as part of ICAAP and business performance frameworks.
Economic capital in RAROC
Economic capital represents capital for unexpected loss at a target confidence and horizon from portfolio loss distributions; embedding it in RAROC ensures consistency with RBI’s higher‑than‑Basel minimum capital stance and internal risk appetite.
Uses of RAROC
- Deal acceptance and price floors: Transactions clear only if RAROC exceeds hurdle rates set by Board‑approved frameworks and ICAAP constraints.
- Capital allocation and incentives: Business lines receive capital guided by marginal/sector RAROC subject to diversification and concentration limits.
Risk‑based pricing
Indian pricing frameworks add components for expected loss, cost of funds, liquidity, operating costs, and a capital charge reflecting economic/regulatory capital to reach a minimum all‑in spread, ensuring compliance and sustainable returns.
Methods of risk‑based pricing
- Cost‑plus models using PD–LGD–EAD for EL and a capital charge based on economic capital or regulatory RWAs under the Standardised/IRB options recognized by RBI.
- Market‑implied and option‑adjusted approaches referencing bond/CDS spreads with adjustments for liquidity, prepayment, and utilization common in large Indian lenders.
Credit derivatives in India
Credit derivatives such as single‑name/index CDS, credit‑linked notes, and securitisation tranches are used for hedging concentrations and optimizing capital, with usage aligned to RBI’s Basel III capital treatment and domestic market liquidity constraints.
Implementation notes for India
- Regulatory perimeter: RBI prescribes Basel III capital and recognizes Standardised and IRB approaches, with migration subject to supervisory approval and preparedness.
- Accounting overlay: Ind AS 109 ECL staging (SICR, default at 90‑DPD or unlikely‑to‑pay) complements prudential IRACP; RBI’s ECL discussion paper outlines convergence pathways for banks not yet under Ind AS.
- Capital calibration: India applies a 9% minimum CRAR vs Basel 8% and aligns multipliers/1250% deductions, implying structurally higher capital expectations that affect pricing and RAROC hurdles.
Risk Management Articles related to Model ‘B’ of CAIIB –Elective paper:
