# Credit Stress Testing Framework Model credit portfolio stress tests to see tail risk and expected loss without treating stress as an exercise separate from risk management. Covers probability of default, loss given default, exposure at default, and scenario analysis in one framework. - Canonical: https://finamodel.com/templates/stress-testing-model - Excel download: https://finamodel.com/templates/stress-testing.xlsx - Category: Credit - Model type: Sector planning - Difficulty: Advanced - Audiences: Credit & risk, CFOs & FP&A, Risk officers, Credit analysts, Portfolio managers, Regulators - Tags: default-probability, loss-given-default, var, concentration ## Overview Model credit portfolio stress tests to see tail risk and expected loss without treating stress as a separate exercise from risk management. The model applies macroeconomic shock scenarios (Base / Adverse / Severely Adverse) with explicit GDP, rate, unemployment, and house price parameters. It then transmits those shocks to loan-level probability of default (PD) via stress multipliers, calculates expected credit losses (ECL) by segment (corporate, retail, mortgage), and outputs portfolio loss reserve (PCL). The model generates three-scenario outputs: NII (net interest income under rate shocks), PCL (provisions under credit stress), PPOP (pre-provision operating profit), and CET1 capital ratio (regulatory capital). Key mechanics: NII is sensitive to rate repricing speed (assets reprice faster than liabilities in rate rises, creating NIM compression); PCL multipliers are scenario-linked (corporate ECL at 4.5x base rate under severe stress, mortgages at 3.0x due to collateral); RWA increases under stress (risk weights move 10-20% higher); and capital (CET1) falls as both losses reduce retained earnings and RWA growth consumes available capital. The model includes covenant tracking (DSCR 1.2x+, leverage limits) and dividend suspension in stressed scenarios (regulatory requirement). Sensitivity tables show CET1 ratio across a range of ECL multipliers, giving regulators confidence in capital adequacy. Use cases: bank capital planning, regulatory DFAST/EBA submissions, stress testing for corporate credit committees, and evaluating concentration risk in large loan books. Works with scenario publishing (Fed, ECB, PRA parameters). ## What's included - Counterparty ratings and default probability curves - Exposure at default (EAD) and facility-level risk - Loss given default (LGD) and recovery assumptions - Expected loss (EL) by counterparty and portfolio - Portfolio loss distribution and Value at Risk (VaR) - Scenario analysis under economic downturns ## Credit Stress Testing Framework for Bank Capital Planning This credit stress testing framework models a multi-segment loan portfolio under Base, Adverse, and Severely Adverse scenarios to assess capital adequacy, earnings, and losses. It projects net interest income, provisions, pre-provision profit, net income, and CET1 ratio over five years, helping risk and finance teams evaluate tail risk and regulatory thresholds. ### Scenario Design and Macroeconomic Drivers The framework uses three regulatory-style scenarios: Base, Adverse, and Severely Adverse. Each scenario specifies distinct macroeconomic shocks, including GDP paths, unemployment spikes, and interest rate movements. - For example, the Severely Adverse scenario assumes a GDP contraction of −4.0% in Year 1 and a 200 basis point rate cut, while the Base scenario projects modest growth and a 75 basis point rate rise. These parameters feed into all scenario-sensitive drivers, such as loan growth rates, repricing betas, and credit loss multipliers. - Loan growth, for instance, halves in the Adverse scenario and turns negative in the Severely Adverse scenario, reflecting reduced credit demand and faster runoff. The scenario toggle on the Assumptions sheet allows users to switch between scenarios, automatically updating the model outputs. ### How Credit Losses Are Calculated Credit losses are modelled through expected credit loss (ECL) rates per segment—Corporate, Retail, and Mortgage—each with a base ECL rate and a stress multiplier. Base ECL rates are 0.50% for Corporate, 1.50% for Retail, and 0.20% for Mortgage. - Under stress, these rates are multiplied: the Severely Adverse multipliers are 4.5× for Corporate, 3.5× for Retail, and 3.0× for Mortgage. The provision for credit losses (PCL) is then calculated as the sum of average segment balances multiplied by their stressed ECL rates. - This PCL charge flows into the income statement and also builds the loan loss reserve on the balance sheet, which is reduced by write-offs. Write-offs follow a lag factor, typically 0.7× ECL in Year 1 and 1.0× thereafter, reflecting the time it takes for delinquencies to crystallise. ### Capital and Earnings Transmission The model integrates earnings and capital impacts. Net interest income (NII) is driven by average loan and funding balances, with yields and costs adjusted for rate shocks and repricing betas. - Asset repricing betas are higher than liability betas, creating asymmetric NIM compression in stress. Non-interest income is scenario-sensitive, with fee growth turning negative in adverse scenarios. - Operating expenses are largely fixed but include a modest restructuring adjustment in severe stress. Pre-provision operating profit (PPOP) is total revenue minus opex and depreciation, and net income is PPOP minus PCL, wholesale interest expense, and tax. Net income feeds into CET1 capital, which also adjusts for dividends and AT1 coupons. Dividends are suspended in stressed scenarios to preserve capital. ### Capital Adequacy and Practical Use The framework outputs the CET1 ratio, calculated as CET1 capital divided by risk-weighted assets (RWA). RWA incorporates risk weights per asset class and a stress overlay that increases credit RWA by 10% in Adverse and 20% in Severely Adverse scenarios. - The model checks compliance with regulatory thresholds, such as a 10.5% target CET1 ratio and a 7.0% minimum distributable amount trigger. A sensitivity table shows how the Year 3 CET1 ratio varies with uniform ECL multipliers, helping users assess loss severity tolerance. - Validation checks ensure balance sheet integrity, NIM reasonableness, and provision adequacy. This makes the framework suitable for capital planning and risk management teams conducting DFAST or EBA-style assessments. ## Probability of default curves PD is modeled as a function of credit rating, industry, and macroeconomic conditions to capture rating migration and credit cycle risk across the portfolio. ## Concentration and correlation Counterparty correlation is modeled explicitly to show portfolio concentration risk and the diversification benefit of spreading exposure across industries and geographies. ## Recession and tail scenario analysis Recession, financial crisis, and tail scenarios show portfolio loss under stress and identify the key risk drivers most likely to threaten capital adequacy. ## Probability of default curves PD is modeled as a function of credit rating, industry, and macroeconomic conditions to capture rating migration and credit cycle risk across the portfolio. ## Concentration and correlation Counterparty correlation is modeled explicitly to show portfolio concentration risk and the diversification benefit of spreading exposure across industries and geographies. ## Recession and tail scenario analysis Recession, financial crisis, and tail scenarios show portfolio loss under stress and identify the key risk drivers most likely to threaten capital adequacy. ## Features - **Probability of default curves:** Model PD as a function of credit rating, industry, and macroeconomic conditions to capture rating migration and cycle risk. - **Correlation and diversification:** Model counterparty correlation to show portfolio concentration risk and diversification benefit. - **Scenario and sensitivity analysis:** Run recession, financial crisis, and tail scenarios to show portfolio loss under stress and identify key risk drivers. ## Use cases - **Regulatory capital requirements:** Model expected loss and stress loss to support regulatory capital adequacy (Dodd-Frank, CCAR, ICAAP) reporting. - **Portfolio risk management:** Identify concentration risk and guide credit issuance to stay within risk limits and diversification targets. - **Provision and reserve setting:** Model expected loss to support provision calculations and CECL reserve methodology. ## Frequently asked questions ### What is probability of default (PD)? PD is the likelihood a counterparty will default over a one-year period, typically estimated from historical default rates by credit rating and economic cycle. ### What is loss given default (LGD)? LGD is the percentage of exposure lost when a counterparty defaults, accounting for collateral recovery and seniority. Secured loans have lower LGD; unsecured exposure has higher LGD. ### How do you model counterparty correlation? Asset-value correlation models such as Vasicek or Merton, or historical default correlations by industry and geography, capture concentration risk that simple expected loss calculations miss. ### What regulations require credit stress testing? Dodd-Frank CCAR, the Basel ICAAP framework, and CECL reserve methodology all require institutions to model expected and stressed credit losses. This model supports each use case. ### Who uses credit stress testing models? Risk officers, credit analysts, portfolio managers, and regulators use them for capital adequacy reporting, provision setting, and credit limit management. ## Related templates - [Credit Portfolio CDO Model](https://finamodel.com/templates/credit-portfolio-cdo-model) - [Loan Portfolio CDR Model](https://finamodel.com/templates/loan-portfolio-cdr-model) - [Mortgage Portfolio Model](https://finamodel.com/templates/mortgage-portfolio-model)