Bank stress-testing mathematics asks a forward-looking question: if the economic and financial environment becomes severely adverse, how do losses, earnings, capital, liquidity and balance-sheet behaviour evolve through time? It turns macroeconomic scenarios into quantitative paths for probability of default, loss given default, market prices, funding costs, net interest income, provisions, risk-weighted assets, capital ratios and liquidity. A good stress test is not one shock multiplied by one ratio. It is an internally consistent scenario propagated through a network of models and balance-sheet identities.
For readers searching for bank stress testing mathematics, scenario analysis, reverse stress testing, capital stress test, bank stress scenarios, macro stress testing, credit loss stress, PPNR stress, CET1 stress, liquidity stress test, sensitivity analysis, scenario expansion, bank resilience, stress testing models or stress testing framework, the central structure is simple: define a coherent adverse scenario, translate macro variables into risk factors, project bank earnings and losses, update balance sheets and regulatory denominators, then test whether capital and liquidity remain above relevant constraints.
The Basel Committee’s current consolidated guidance, effective in 2026, states that stress testing is a critical element of bank risk management and supervision, should have clear objectives and governance, should capture material risks, and should apply sufficiently severe and internally consistent adverse scenarios. The guidance also emphasises that stress test results should inform business decisions, capital and liquidity assessment, contingency arrangements and corrective actions. This page develops the mathematics behind that framework. It is educational, not regulatory advice.
50-Second Router
- Baseline: central forecast used as a reference path, not a promise.
- Adverse scenario: severe but plausible internally consistent path for macro/market variables.
- Sensitivity test: isolates one factor or parameter; useful for diagnostics but not a full scenario.
- Scenario expansion: translate high-level macro variables into detailed rates, spreads, prices and sector variables needed by models.
- Credit loss: stress PD, LGD, EAD, migration and concentration through the scenario.
- PPNR/NII: project pre-provision earnings, deposit beta, funding cost, fees and operating expenses.
- Market loss: revalue trading/investment positions under shocked rates, FX, equity, credit spreads and volatilities.
- Capital path: beginning CET1 + earnings − provisions − market/operational losses − distributions ± adjustments, divided by stressed RWA.
- Liquidity path: stressed outflows, HQLA use, collateral/margin calls, funding access and survival horizon.
- Reverse stress: begin with failure condition and solve for scenarios that could cause it.
- Management actions: include only actions that are credible, feasible, timely and consistent with strategy/recovery plans.
- Governance: scenario design, models, data, challenge, validation and use matter as much as the arithmetic.
The Central Proposition: Stress Testing Is a Dynamic Balance-Sheet Projection
A static capital ratio says where a bank stands today. A stress test asks where it could stand after a path of adverse events. The difference matters because losses arrive over time, borrowers migrate through risk grades, deposits can leave, asset values move, provisions change, interest rates reprice loans and funding differently, management can react, and RWA can rise even while assets shrink.
The correct mathematical object is therefore a sequence: state at t=0 → scenario shocks → model outputs at t=1 → updated balance sheet → next-period shocks → new state. Every period’s result becomes the starting point for the next period.
Adrian’s first rule is to reject a stress test that cannot explain how the end-state was reached. A credible stress path needs a bridge from starting balance sheet to every intermediate state.
1. Baseline scenario
Baseline scenario is reference macro-financial path used for comparison. It provides the no-stress trajectory against which adverse effects are measured. Stress testing becomes credible only when the scenario, model translation and balance-sheet effect are visible separately.
Mathematics. Baseline variables x_t feed the same models as adverse variables. Keep time indexes explicit. A stress path is not one number; each period’s losses, earnings, balances, capital and liquidity feed the next period.
Failure mode. A baseline is not a prediction certainty and can itself be wrong. Jo’s diagnostic is to ask whether the result comes from scenario severity, model sensitivity, starting exposure, management action or regulatory denominator.
Connection. This feeds directly into scenario comparison. Ryan would rerun at least one harsher and one alternative-shape scenario to test whether the vulnerability is structural or scenario-specific.
2. Adverse scenario
Adverse scenario is severe but plausible coherent path designed to expose vulnerabilities. It is the main stress-testing input. Stress testing becomes credible only when the scenario, model translation and balance-sheet effect are visible separately.
Mathematics. Scenario={GDP_t, unemployment_t, rates_t, spreads_t, property_t, FX_t,…}. Keep time indexes explicit. A stress path is not one number; each period’s losses, earnings, balances, capital and liquidity feed the next period.
Failure mode. Shocking variables independently can create impossible combinations. Jo’s diagnostic is to ask whether the result comes from scenario severity, model sensitivity, starting exposure, management action or regulatory denominator.
Connection. This feeds directly into capital/liquidity stress. Ryan would rerun at least one harsher and one alternative-shape scenario to test whether the vulnerability is structural or scenario-specific.
3. Severe but plausible
Severe but plausible is design principle balancing material adversity with economic coherence. It avoids trivial stress and implausible fantasy. Stress testing becomes credible only when the scenario, model translation and balance-sheet effect are visible separately.
Mathematics. Severity can be benchmarked against history and forward-looking vulnerabilities. Keep time indexes explicit. A stress path is not one number; each period’s losses, earnings, balances, capital and liquidity feed the next period.
Failure mode. Historical maximum alone is not sufficient for new risks. Jo’s diagnostic is to ask whether the result comes from scenario severity, model sensitivity, starting exposure, management action or regulatory denominator.
Connection. This feeds directly into scenario design. Ryan would rerun at least one harsher and one alternative-shape scenario to test whether the vulnerability is structural or scenario-specific.
4. Scenario narrative
Scenario narrative is economic story explaining why variables move together. It is the glue that makes a scenario internally consistent. Stress testing becomes credible only when the scenario, model translation and balance-sheet effect are visible separately.
Mathematics. Narrative maps cause→macro path→market/credit/funding effects. Keep time indexes explicit. A stress path is not one number; each period’s losses, earnings, balances, capital and liquidity feed the next period.
Failure mode. Numbers without narrative invite contradictory shocks. Jo’s diagnostic is to ask whether the result comes from scenario severity, model sensitivity, starting exposure, management action or regulatory denominator.
Connection. This feeds directly into scenario governance. Ryan would rerun at least one harsher and one alternative-shape scenario to test whether the vulnerability is structural or scenario-specific.
5. Scenario horizon
Scenario horizon is period over which stress is projected. It determines which risks and management responses can emerge. Stress testing becomes credible only when the scenario, model translation and balance-sheet effect are visible separately.
Mathematics. Capital tests may span multiple years; liquidity stresses can be days/weeks. Keep time indexes explicit. A stress path is not one number; each period’s losses, earnings, balances, capital and liquidity feed the next period.
Failure mode. Using annual averages can hide short liquidity cliffs. Jo’s diagnostic is to ask whether the result comes from scenario severity, model sensitivity, starting exposure, management action or regulatory denominator.
Connection. This feeds directly into stress architecture. Ryan would rerun at least one harsher and one alternative-shape scenario to test whether the vulnerability is structural or scenario-specific.
6. Time step
Time step is frequency of model projection. It controls granularity and path dependence. Stress testing becomes credible only when the scenario, model translation and balance-sheet effect are visible separately.
Mathematics. Quarterly/monthly/daily steps depending risk. Keep time indexes explicit. A stress path is not one number; each period’s losses, earnings, balances, capital and liquidity feed the next period.
Failure mode. Too-coarse steps can miss peak losses or margin calls. Jo’s diagnostic is to ask whether the result comes from scenario severity, model sensitivity, starting exposure, management action or regulatory denominator.
Connection. This feeds directly into simulation. Ryan would rerun at least one harsher and one alternative-shape scenario to test whether the vulnerability is structural or scenario-specific.
7. Sensitivity test
Sensitivity test is one-factor shock while other assumptions held fixed. It diagnoses local exposure. Stress testing becomes credible only when the scenario, model translation and balance-sheet effect are visible separately.
Mathematics. ΔOutcome≈Sensitivity×Shock. Keep time indexes explicit. A stress path is not one number; each period’s losses, earnings, balances, capital and liquidity feed the next period.
Failure mode. It does not capture correlated macro feedback. Jo’s diagnostic is to ask whether the result comes from scenario severity, model sensitivity, starting exposure, management action or regulatory denominator.
Connection. This feeds directly into risk diagnostics. Ryan would rerun at least one harsher and one alternative-shape scenario to test whether the vulnerability is structural or scenario-specific.
8. Scenario test
Scenario test is joint path of multiple risk factors. It captures interaction and second-order effects. Stress testing becomes credible only when the scenario, model translation and balance-sheet effect are visible separately.
Mathematics. Outcome=Model(x_1,…,x_n through time). Keep time indexes explicit. A stress path is not one number; each period’s losses, earnings, balances, capital and liquidity feed the next period.
Failure mode. Adding independent sensitivities may miss nonlinearities. Jo’s diagnostic is to ask whether the result comes from scenario severity, model sensitivity, starting exposure, management action or regulatory denominator.
Connection. This feeds directly into stress testing. Ryan would rerun at least one harsher and one alternative-shape scenario to test whether the vulnerability is structural or scenario-specific.
9. Reverse stress test
Reverse stress test is analysis starting from a failure condition and searching for scenarios that cause it. It reveals hidden fragility and challenge assumptions. Stress testing becomes credible only when the scenario, model translation and balance-sheet effect are visible separately.
Mathematics. Find x such that CapitalRatio(x) Failure mode. Reverse stress identifies pathways, not necessarily their probability. Jo’s diagnostic is to ask whether the result comes from scenario severity, model sensitivity, starting exposure, management action or regulatory denominator. Connection. This feeds directly into resilience. Ryan would rerun at least one harsher and one alternative-shape scenario to test whether the vulnerability is structural or scenario-specific. Threshold event is predefined adverse outcome such as capital minimum breach or liquidity exhaustion. It anchors reverse stress. Stress testing becomes credible only when the scenario, model translation and balance-sheet effect are visible separately. Mathematics. Solve for scenario boundary. Keep time indexes explicit. A stress path is not one number; each period’s losses, earnings, balances, capital and liquidity feed the next period. Failure mode. Choosing a threshold too remote from actual decision use reduces relevance. Jo’s diagnostic is to ask whether the result comes from scenario severity, model sensitivity, starting exposure, management action or regulatory denominator. Connection. This feeds directly into reverse stress. Ryan would rerun at least one harsher and one alternative-shape scenario to test whether the vulnerability is structural or scenario-specific. Scenario expansion is mapping coarse macro variables into granular model inputs. It connects scenario to portfolios. Stress testing becomes credible only when the scenario, model translation and balance-sheet effect are visible separately. Mathematics. GDP/unemployment→sector PDs; yield curve→tenor rates; property index→collateral values. Keep time indexes explicit. A stress path is not one number; each period’s losses, earnings, balances, capital and liquidity feed the next period. Failure mode. Poor expansion can create inconsistent model inputs. Jo’s diagnostic is to ask whether the result comes from scenario severity, model sensitivity, starting exposure, management action or regulatory denominator. Connection. This feeds directly into macro models. Ryan would rerun at least one harsher and one alternative-shape scenario to test whether the vulnerability is structural or scenario-specific. Satellite model is model linking scenario variables to portfolio risk parameters. It translates macro stress into PD/LGD, revenue, NII etc. Stress testing becomes credible only when the scenario, model translation and balance-sheet effect are visible separately. Mathematics. PD_t=f(unemployment_t,GDP_t,borrower state). Keep time indexes explicit. A stress path is not one number; each period’s losses, earnings, balances, capital and liquidity feed the next period. Failure mode. Spurious historical correlations can fail in new regimes. Jo’s diagnostic is to ask whether the result comes from scenario severity, model sensitivity, starting exposure, management action or regulatory denominator. Connection. This feeds directly into model risk. Ryan would rerun at least one harsher and one alternative-shape scenario to test whether the vulnerability is structural or scenario-specific. Credit loss model is projection of defaults, migrations, recoveries and exposure. It often drives bank stress losses. Stress testing becomes credible only when the scenario, model translation and balance-sheet effect are visible separately. Mathematics. Loss_t=ΣPD×LGD×EAD or scenario simulation. Keep time indexes explicit. A stress path is not one number; each period’s losses, earnings, balances, capital and liquidity feed the next period. Failure mode. Holding LGD and EAD constant while PD rises can understate downturn loss. Jo’s diagnostic is to ask whether the result comes from scenario severity, model sensitivity, starting exposure, management action or regulatory denominator. Connection. This feeds directly into credit stress. Ryan would rerun at least one harsher and one alternative-shape scenario to test whether the vulnerability is structural or scenario-specific. Point-in-time PD is default probability responsive to current scenario conditions. It is useful for stress sensitivity. Stress testing becomes credible only when the scenario, model translation and balance-sheet effect are visible separately. Mathematics. PD_t changes with borrower/macro state. Keep time indexes explicit. A stress path is not one number; each period’s losses, earnings, balances, capital and liquidity feed the next period. Failure mode. Through-the-cycle ratings need mapping to stressed default probabilities. Jo’s diagnostic is to ask whether the result comes from scenario severity, model sensitivity, starting exposure, management action or regulatory denominator. Connection. This feeds directly into credit models. Ryan would rerun at least one harsher and one alternative-shape scenario to test whether the vulnerability is structural or scenario-specific. Stressed LGD is loss severity under adverse collateral/recovery conditions. It captures downturn recovery deterioration. Stress testing becomes credible only when the scenario, model translation and balance-sheet effect are visible separately. Mathematics. LGD_stress=(EAD−PV stressed recoveries)/EAD. Keep time indexes explicit. A stress path is not one number; each period’s losses, earnings, balances, capital and liquidity feed the next period. Failure mode. Using average benign recovery during property crash is weak. Jo’s diagnostic is to ask whether the result comes from scenario severity, model sensitivity, starting exposure, management action or regulatory denominator. Connection. This feeds directly into credit stress. Ryan would rerun at least one harsher and one alternative-shape scenario to test whether the vulnerability is structural or scenario-specific. Stressed EAD is default exposure after borrower drawdown/amortisation under stress. It captures line utilisation and balance evolution. Stress testing becomes credible only when the scenario, model translation and balance-sheet effect are visible separately. Mathematics. EAD_t depends on utilisation path. Keep time indexes explicit. A stress path is not one number; each period’s losses, earnings, balances, capital and liquidity feed the next period. Failure mode. Revolving borrowers may draw more before default. Jo’s diagnostic is to ask whether the result comes from scenario severity, model sensitivity, starting exposure, management action or regulatory denominator. Connection. This feeds directly into credit stress. Ryan would rerun at least one harsher and one alternative-shape scenario to test whether the vulnerability is structural or scenario-specific. Migration model is transition among ratings/delinquency states under stress. It changes future PD/RWA before default. Stress testing becomes credible only when the scenario, model translation and balance-sheet effect are visible separately. Mathematics. Transition matrix P_t conditional on scenario. Keep time indexes explicit. A stress path is not one number; each period’s losses, earnings, balances, capital and liquidity feed the next period. Failure mode. Using one unchanged transition matrix ignores scenario. Jo’s diagnostic is to ask whether the result comes from scenario severity, model sensitivity, starting exposure, management action or regulatory denominator. Connection. This feeds directly into credit migration. Ryan would rerun at least one harsher and one alternative-shape scenario to test whether the vulnerability is structural or scenario-specific. Vintage model is loss projection by origination cohort. It captures seasoning and underwriting changes. Stress testing becomes credible only when the scenario, model translation and balance-sheet effect are visible separately. Mathematics. Loss_{vintage,t}. Keep time indexes explicit. A stress path is not one number; each period’s losses, earnings, balances, capital and liquidity feed the next period. Failure mode. Mixing old/new underwriting can hide weak recent cohorts. Jo’s diagnostic is to ask whether the result comes from scenario severity, model sensitivity, starting exposure, management action or regulatory denominator. Connection. This feeds directly into retail stress. Ryan would rerun at least one harsher and one alternative-shape scenario to test whether the vulnerability is structural or scenario-specific. Portfolio concentration is clustering of exposure by borrower, sector or geography. It increases tail stress sensitivity. Stress testing becomes credible only when the scenario, model translation and balance-sheet effect are visible separately. Mathematics. HHI or factor exposures quantify concentration. Keep time indexes explicit. A stress path is not one number; each period’s losses, earnings, balances, capital and liquidity feed the next period. Failure mode. Average PD misses common-factor loss clustering. Jo’s diagnostic is to ask whether the result comes from scenario severity, model sensitivity, starting exposure, management action or regulatory denominator. Connection. This feeds directly into credit stress. Ryan would rerun at least one harsher and one alternative-shape scenario to test whether the vulnerability is structural or scenario-specific. Collateral stress is decline in collateral value and liquidation quality. It increases LGD and LTV. Stress testing becomes credible only when the scenario, model translation and balance-sheet effect are visible separately. Mathematics. StressedLTV=EAD/StressedCollateralValue. Keep time indexes explicit. A stress path is not one number; each period’s losses, earnings, balances, capital and liquidity feed the next period. Failure mode. Property price stress can affect default and recovery simultaneously. Jo’s diagnostic is to ask whether the result comes from scenario severity, model sensitivity, starting exposure, management action or regulatory denominator. Connection. This feeds directly into secured lending. Ryan would rerun at least one harsher and one alternative-shape scenario to test whether the vulnerability is structural or scenario-specific. PPNR is pre-provision net revenue from interest and non-interest income minus operating expense. It is core capacity to absorb credit losses before capital. Stress testing becomes credible only when the scenario, model translation and balance-sheet effect are visible separately. Mathematics. PPNR=NII+NonInterestIncome−OperatingExpense. Keep time indexes explicit. A stress path is not one number; each period’s losses, earnings, balances, capital and liquidity feed the next period. Failure mode. Static PPNR can be overly optimistic in deep recession. Jo’s diagnostic is to ask whether the result comes from scenario severity, model sensitivity, starting exposure, management action or regulatory denominator. Connection. This feeds directly into capital stress. Ryan would rerun at least one harsher and one alternative-shape scenario to test whether the vulnerability is structural or scenario-specific. Net interest income is interest income minus interest expense. It responds to rates, deposit beta, loan repricing and funding mix. Stress testing becomes credible only when the scenario, model translation and balance-sheet effect are visible separately. Mathematics. NII_t=ΣAssetBalances×Yields−ΣFunding×Costs. Keep time indexes explicit. A stress path is not one number; each period’s losses, earnings, balances, capital and liquidity feed the next period. Failure mode. Parallel rate shock alone misses curve/basis/behavioural effects. Jo’s diagnostic is to ask whether the result comes from scenario severity, model sensitivity, starting exposure, management action or regulatory denominator. Connection. This feeds directly into earnings stress. Ryan would rerun at least one harsher and one alternative-shape scenario to test whether the vulnerability is structural or scenario-specific. Deposit beta stress is change in pass-through from market rates to deposit rates. It affects funding cost and retention. Stress testing becomes credible only when the scenario, model translation and balance-sheet effect are visible separately. Mathematics. DepositRate_t=base+β_t×benchmark move. Keep time indexes explicit. A stress path is not one number; each period’s losses, earnings, balances, capital and liquidity feed the next period. Failure mode. Stress can raise beta through competition even when historical beta was low. Jo’s diagnostic is to ask whether the result comes from scenario severity, model sensitivity, starting exposure, management action or regulatory denominator. Connection. This feeds directly into NII. Ryan would rerun at least one harsher and one alternative-shape scenario to test whether the vulnerability is structural or scenario-specific. Deposit runoff stress is withdrawal of customer funding. It affects liquidity and replacement funding cost. Stress testing becomes credible only when the scenario, model translation and balance-sheet effect are visible separately. Mathematics. Outflow=Balance×runoff rate. Keep time indexes explicit. A stress path is not one number; each period’s losses, earnings, balances, capital and liquidity feed the next period. Failure mode. NII and liquidity should be stressed jointly. Jo’s diagnostic is to ask whether the result comes from scenario severity, model sensitivity, starting exposure, management action or regulatory denominator. Connection. This feeds directly into funding stress. Ryan would rerun at least one harsher and one alternative-shape scenario to test whether the vulnerability is structural or scenario-specific. Wholesale funding spread is market spread the bank pays to refinance. It can widen during bank/system stress. Stress testing becomes credible only when the scenario, model translation and balance-sheet effect are visible separately. Mathematics. FundingRate=Benchmark+Spread_stress. Keep time indexes explicit. A stress path is not one number; each period’s losses, earnings, balances, capital and liquidity feed the next period. Failure mode. Assuming normal market access throughout stress is optimistic. Jo’s diagnostic is to ask whether the result comes from scenario severity, model sensitivity, starting exposure, management action or regulatory denominator. Connection. This feeds directly into liquidity/NII. Ryan would rerun at least one harsher and one alternative-shape scenario to test whether the vulnerability is structural or scenario-specific. Fee income stress is decline in transaction, wealth, card or underwriting fees. It reduces PPNR. Stress testing becomes credible only when the scenario, model translation and balance-sheet effect are visible separately. Mathematics. Fee_t linked to volumes/assets/markets. Keep time indexes explicit. A stress path is not one number; each period’s losses, earnings, balances, capital and liquidity feed the next period. Failure mode. Focusing only on credit losses misses revenue stress. Jo’s diagnostic is to ask whether the result comes from scenario severity, model sensitivity, starting exposure, management action or regulatory denominator. Connection. This feeds directly into earnings. Ryan would rerun at least one harsher and one alternative-shape scenario to test whether the vulnerability is structural or scenario-specific. Operating expense stress is cost changes during stress. It can rise due to collections, legal, cyber or restructuring. Stress testing becomes credible only when the scenario, model translation and balance-sheet effect are visible separately. Mathematics. Opex_t=base+stress adjustments. Keep time indexes explicit. A stress path is not one number; each period’s losses, earnings, balances, capital and liquidity feed the next period. Failure mode. Assuming immediate cost cuts can be unrealistic. Jo’s diagnostic is to ask whether the result comes from scenario severity, model sensitivity, starting exposure, management action or regulatory denominator. Connection. This feeds directly into PPNR. Ryan would rerun at least one harsher and one alternative-shape scenario to test whether the vulnerability is structural or scenario-specific. Market-risk shock is movement in traded rates, spreads, FX, equities, commodities and vol. It creates mark-to-market loss and margin calls. Stress testing becomes credible only when the scenario, model translation and balance-sheet effect are visible separately. Mathematics. ΔV from full repricing or sensitivities. Keep time indexes explicit. A stress path is not one number; each period’s losses, earnings, balances, capital and liquidity feed the next period. Failure mode. One-day VaR shocks may be too mild for capital stress. Jo’s diagnostic is to ask whether the result comes from scenario severity, model sensitivity, starting exposure, management action or regulatory denominator. Connection. This feeds directly into market stress. Ryan would rerun at least one harsher and one alternative-shape scenario to test whether the vulnerability is structural or scenario-specific. Credit-spread shock is widening of market credit spreads. It reduces value of bonds/loans before default. Stress testing becomes credible only when the scenario, model translation and balance-sheet effect are visible separately. Mathematics. Approx ΔP≈−SpreadDuration×ΔSpread×P. Keep time indexes explicit. A stress path is not one number; each period’s losses, earnings, balances, capital and liquidity feed the next period. Failure mode. Default and spread losses can double count if models overlap. Jo’s diagnostic is to ask whether the result comes from scenario severity, model sensitivity, starting exposure, management action or regulatory denominator. Connection. This feeds directly into market/credit. Ryan would rerun at least one harsher and one alternative-shape scenario to test whether the vulnerability is structural or scenario-specific. Yield-curve shock is parallel/nonparallel interest-rate movement. It affects securities, NII and EVE. Stress testing becomes credible only when the scenario, model translation and balance-sheet effect are visible separately. Mathematics. Reprice discount/projected cash flows. Keep time indexes explicit. A stress path is not one number; each period’s losses, earnings, balances, capital and liquidity feed the next period. Failure mode. Only one parallel shock misses slope/curvature risk. Jo’s diagnostic is to ask whether the result comes from scenario severity, model sensitivity, starting exposure, management action or regulatory denominator. Connection. This feeds directly into IRRBB. Ryan would rerun at least one harsher and one alternative-shape scenario to test whether the vulnerability is structural or scenario-specific. FX shock is currency movement affecting trading, translation and borrowers. It can have direct and indirect credit effects. Stress testing becomes credible only when the scenario, model translation and balance-sheet effect are visible separately. Mathematics. Convert foreign exposures under stressed FX. Keep time indexes explicit. A stress path is not one number; each period’s losses, earnings, balances, capital and liquidity feed the next period. Failure mode. Hedged positions can still face basis/margin liquidity. Jo’s diagnostic is to ask whether the result comes from scenario severity, model sensitivity, starting exposure, management action or regulatory denominator. Connection. This feeds directly into market stress. Ryan would rerun at least one harsher and one alternative-shape scenario to test whether the vulnerability is structural or scenario-specific. Equity shock is decline in equity prices. It affects trading, fee income, collateral and pensions. Stress testing becomes credible only when the scenario, model translation and balance-sheet effect are visible separately. Mathematics. Full revaluation of equity-sensitive positions. Keep time indexes explicit. A stress path is not one number; each period’s losses, earnings, balances, capital and liquidity feed the next period. Failure mode. Second-order effects can matter for leveraged clients. Jo’s diagnostic is to ask whether the result comes from scenario severity, model sensitivity, starting exposure, management action or regulatory denominator. Connection. This feeds directly into market stress. Ryan would rerun at least one harsher and one alternative-shape scenario to test whether the vulnerability is structural or scenario-specific. Volatility shock is increase in implied/realised volatility. It changes options and margin/collateral needs. Stress testing becomes credible only when the scenario, model translation and balance-sheet effect are visible separately. Mathematics. Use vega/gamma/full repricing. Keep time indexes explicit. A stress path is not one number; each period’s losses, earnings, balances, capital and liquidity feed the next period. Failure mode. Price-only shocks can miss derivative stress. Jo’s diagnostic is to ask whether the result comes from scenario severity, model sensitivity, starting exposure, management action or regulatory denominator. Connection. This feeds directly into market stress. Ryan would rerun at least one harsher and one alternative-shape scenario to test whether the vulnerability is structural or scenario-specific. Liquidity shock is cash outflow/funding closure/haircut increase. It tests survival rather than solvency only. Stress testing becomes credible only when the scenario, model translation and balance-sheet effect are visible separately. Mathematics. Liquidity path=Opening buffer+inflows−outflows+actions. Keep time indexes explicit. A stress path is not one number; each period’s losses, earnings, balances, capital and liquidity feed the next period. Failure mode. Capital and liquidity shocks can reinforce each other. Jo’s diagnostic is to ask whether the result comes from scenario severity, model sensitivity, starting exposure, management action or regulatory denominator. Connection. This feeds directly into liquidity stress. Ryan would rerun at least one harsher and one alternative-shape scenario to test whether the vulnerability is structural or scenario-specific. Margin-call shock is collateral outflow caused by derivative/secured funding moves. It can create immediate cash need. Stress testing becomes credible only when the scenario, model translation and balance-sheet effect are visible separately. Mathematics. Margin=stressed exposure−posted collateral. Keep time indexes explicit. A stress path is not one number; each period’s losses, earnings, balances, capital and liquidity feed the next period. Failure mode. Economic hedge can be profitable long term but liquidity-negative today. Jo’s diagnostic is to ask whether the result comes from scenario severity, model sensitivity, starting exposure, management action or regulatory denominator. Connection. This feeds directly into liquidity. Ryan would rerun at least one harsher and one alternative-shape scenario to test whether the vulnerability is structural or scenario-specific. Operational-risk shock is loss from systems, fraud, legal, conduct or process failure. It can be scenario-specific and correlated with stress. Stress testing becomes credible only when the scenario, model translation and balance-sheet effect are visible separately. Mathematics. Operational loss distribution/scenario overlay. Keep time indexes explicit. A stress path is not one number; each period’s losses, earnings, balances, capital and liquidity feed the next period. Failure mode. Historical average operational loss misses severe event. Jo’s diagnostic is to ask whether the result comes from scenario severity, model sensitivity, starting exposure, management action or regulatory denominator. Connection. This feeds directly into operational stress. Ryan would rerun at least one harsher and one alternative-shape scenario to test whether the vulnerability is structural or scenario-specific. Climate/transition shock is scenario affecting borrowers/assets through physical or transition channels. It can operate over longer horizons and sector concentrations. Stress testing becomes credible only when the scenario, model translation and balance-sheet effect are visible separately. Mathematics. Map hazards/policies to cash flow and PD/LGD. Keep time indexes explicit. A stress path is not one number; each period’s losses, earnings, balances, capital and liquidity feed the next period. Failure mode. Precision can be lower than conventional macro stress; document uncertainty. Jo’s diagnostic is to ask whether the result comes from scenario severity, model sensitivity, starting exposure, management action or regulatory denominator. Connection. This feeds directly into emerging risks. Ryan would rerun at least one harsher and one alternative-shape scenario to test whether the vulnerability is structural or scenario-specific. Cyber shock is operational disruption plus fraud/recovery cost/liquidity effects. It can affect payments and confidence. Stress testing becomes credible only when the scenario, model translation and balance-sheet effect are visible separately. Mathematics. Scenario includes downtime, losses, customer runoff. Keep time indexes explicit. A stress path is not one number; each period’s losses, earnings, balances, capital and liquidity feed the next period. Failure mode. Pure capital loss estimate can miss operational liquidity. Jo’s diagnostic is to ask whether the result comes from scenario severity, model sensitivity, starting exposure, management action or regulatory denominator. Connection. This feeds directly into resilience. Ryan would rerun at least one harsher and one alternative-shape scenario to test whether the vulnerability is structural or scenario-specific. Second-round effect is feedback from bank/market responses back into prices and economy. It can amplify initial shock. Stress testing becomes credible only when the scenario, model translation and balance-sheet effect are visible separately. Mathematics. Fire sales→lower prices→more losses; credit contraction→weaker economy. Keep time indexes explicit. A stress path is not one number; each period’s losses, earnings, balances, capital and liquidity feed the next period. Failure mode. Single-bank static tests often omit systemic feedback. Jo’s diagnostic is to ask whether the result comes from scenario severity, model sensitivity, starting exposure, management action or regulatory denominator. Connection. This feeds directly into macroprudential stress. Ryan would rerun at least one harsher and one alternative-shape scenario to test whether the vulnerability is structural or scenario-specific. Fire-sale feedback is forced sales depress market prices. It converts liquidity stress into capital loss. Stress testing becomes credible only when the scenario, model translation and balance-sheet effect are visible separately. Mathematics. PriceImpact=f(volume sold, market depth). Keep time indexes explicit. A stress path is not one number; each period’s losses, earnings, balances, capital and liquidity feed the next period. Failure mode. Assuming sale at normal price under stress is inconsistent. Jo’s diagnostic is to ask whether the result comes from scenario severity, model sensitivity, starting exposure, management action or regulatory denominator. Connection. This feeds directly into systemic stress. Ryan would rerun at least one harsher and one alternative-shape scenario to test whether the vulnerability is structural or scenario-specific. Credit contraction feedback is banks reduce lending to preserve capital. It can worsen macro conditions and future credit losses. Stress testing becomes credible only when the scenario, model translation and balance-sheet effect are visible separately. Mathematics. Lending↓→GDP/investment↓→PD↑. Keep time indexes explicit. A stress path is not one number; each period’s losses, earnings, balances, capital and liquidity feed the next period. Failure mode. Difficult to model but important systemically. Jo’s diagnostic is to ask whether the result comes from scenario severity, model sensitivity, starting exposure, management action or regulatory denominator. Connection. This feeds directly into macro stress. Ryan would rerun at least one harsher and one alternative-shape scenario to test whether the vulnerability is structural or scenario-specific. Management action is decision taken in response to stress such as cutting dividends, issuing capital or reducing RWA. It can restore resilience. Stress testing becomes credible only when the scenario, model translation and balance-sheet effect are visible separately. Mathematics. Capital_t includes action timing/effects. Keep time indexes explicit. A stress path is not one number; each period’s losses, earnings, balances, capital and liquidity feed the next period. Failure mode. Assuming perfect immediate execution overstates resilience. Jo’s diagnostic is to ask whether the result comes from scenario severity, model sensitivity, starting exposure, management action or regulatory denominator. Connection. This feeds directly into stress planning. Ryan would rerun at least one harsher and one alternative-shape scenario to test whether the vulnerability is structural or scenario-specific. Dividend cut is capital-preserving action. It retains earnings/CET1. Stress testing becomes credible only when the scenario, model translation and balance-sheet effect are visible separately. Mathematics. Capital improves by avoided distribution. Keep time indexes explicit. A stress path is not one number; each period’s losses, earnings, balances, capital and liquidity feed the next period. Failure mode. Board/legal/market constraints and timing matter. Jo’s diagnostic is to ask whether the result comes from scenario severity, model sensitivity, starting exposure, management action or regulatory denominator. Connection. This feeds directly into capital actions. Ryan would rerun at least one harsher and one alternative-shape scenario to test whether the vulnerability is structural or scenario-specific. Share issuance is external capital raise. It can restore ratios. Stress testing becomes credible only when the scenario, model translation and balance-sheet effect are visible separately. Mathematics. CET1 increases by net proceeds. Keep time indexes explicit. A stress path is not one number; each period’s losses, earnings, balances, capital and liquidity feed the next period. Failure mode. Market access may be worst in severe stress. Jo’s diagnostic is to ask whether the result comes from scenario severity, model sensitivity, starting exposure, management action or regulatory denominator. Connection. This feeds directly into recovery. Ryan would rerun at least one harsher and one alternative-shape scenario to test whether the vulnerability is structural or scenario-specific. Asset sale is balance-sheet reduction/liquidity action. It can release RWA/cash but crystallise loss. Stress testing becomes credible only when the scenario, model translation and balance-sheet effect are visible separately. Mathematics. Capital effect=price−carrying value plus RWA reduction. Keep time indexes explicit. A stress path is not one number; each period’s losses, earnings, balances, capital and liquidity feed the next period. Failure mode. Selling illiquid assets in stress can worsen losses. Jo’s diagnostic is to ask whether the result comes from scenario severity, model sensitivity, starting exposure, management action or regulatory denominator. Connection. This feeds directly into management action. Ryan would rerun at least one harsher and one alternative-shape scenario to test whether the vulnerability is structural or scenario-specific. RWA reduction is shrinking/rebalancing risk exposures. It can improve ratios. Stress testing becomes credible only when the scenario, model translation and balance-sheet effect are visible separately. Mathematics. CapitalRatio=CET1/RWA. Keep time indexes explicit. A stress path is not one number; each period’s losses, earnings, balances, capital and liquidity feed the next period. Failure mode. Aggressive deleveraging can harm earnings/economy. Jo’s diagnostic is to ask whether the result comes from scenario severity, model sensitivity, starting exposure, management action or regulatory denominator. Connection. This feeds directly into capital action. Ryan would rerun at least one harsher and one alternative-shape scenario to test whether the vulnerability is structural or scenario-specific. Deposit repricing is raise rates to retain funding. It can reduce runoff but compress NII. Stress testing becomes credible only when the scenario, model translation and balance-sheet effect are visible separately. Mathematics. Higher deposit cost vs replacement funding benefit. Keep time indexes explicit. A stress path is not one number; each period’s losses, earnings, balances, capital and liquidity feed the next period. Failure mode. Not all customers require same repricing. Jo’s diagnostic is to ask whether the result comes from scenario severity, model sensitivity, starting exposure, management action or regulatory denominator. Connection. This feeds directly into liquidity action. Ryan would rerun at least one harsher and one alternative-shape scenario to test whether the vulnerability is structural or scenario-specific. Recovery-plan action is predefined measure to restore capital/liquidity during severe stress. It should align with executable plans. Stress testing becomes credible only when the scenario, model translation and balance-sheet effect are visible separately. Mathematics. Examples: capital raise, sale, funding action. Keep time indexes explicit. A stress path is not one number; each period’s losses, earnings, balances, capital and liquidity feed the next period. Failure mode. Counting an action not present in recovery plan weakens credibility. Jo’s diagnostic is to ask whether the result comes from scenario severity, model sensitivity, starting exposure, management action or regulatory denominator. Connection. This feeds directly into governance. Ryan would rerun at least one harsher and one alternative-shape scenario to test whether the vulnerability is structural or scenario-specific. Capital path is projected regulatory capital through scenario. It combines earnings, losses and distributions. Stress testing becomes credible only when the scenario, model translation and balance-sheet effect are visible separately. Mathematics. CET1_t=CET1_{t-1}+NI_t−Distributions±Adjustments. Keep time indexes explicit. A stress path is not one number; each period’s losses, earnings, balances, capital and liquidity feed the next period. Failure mode. Static beginning CET1 minus total loss misses timing and profit regeneration. Jo’s diagnostic is to ask whether the result comes from scenario severity, model sensitivity, starting exposure, management action or regulatory denominator. Connection. This feeds directly into capital stress. Ryan would rerun at least one harsher and one alternative-shape scenario to test whether the vulnerability is structural or scenario-specific. RWA path is projected regulatory denominator through scenario. It changes with credit migration, balance sheet and market/operational risk. Stress testing becomes credible only when the scenario, model translation and balance-sheet effect are visible separately. Mathematics. RWA_t=f(exposure,risk parameters,approach). Keep time indexes explicit. A stress path is not one number; each period’s losses, earnings, balances, capital and liquidity feed the next period. Failure mode. Holding RWA fixed can overstate stressed capital ratios. Jo’s diagnostic is to ask whether the result comes from scenario severity, model sensitivity, starting exposure, management action or regulatory denominator. Connection. This feeds directly into capital stress. Ryan would rerun at least one harsher and one alternative-shape scenario to test whether the vulnerability is structural or scenario-specific. CET1 stress ratio is stressed CET1 divided by stressed RWA. It is a key solvency endpoint/path measure. Stress testing becomes credible only when the scenario, model translation and balance-sheet effect are visible separately. Mathematics. CET1Ratio_t=CET1_t/RWA_t. Keep time indexes explicit. A stress path is not one number; each period’s losses, earnings, balances, capital and liquidity feed the next period. Failure mode. End-horizon ratio alone can hide interim breach. Jo’s diagnostic is to ask whether the result comes from scenario severity, model sensitivity, starting exposure, management action or regulatory denominator. Connection. This feeds directly into capital. Ryan would rerun at least one harsher and one alternative-shape scenario to test whether the vulnerability is structural or scenario-specific. Leverage ratio stress is Tier1 versus stressed leverage exposure. It constrains low-RWA expansion and losses. Stress testing becomes credible only when the scenario, model translation and balance-sheet effect are visible separately. Mathematics. LR_t=Tier1_t/Exposure_t. Keep time indexes explicit. A stress path is not one number; each period’s losses, earnings, balances, capital and liquidity feed the next period. Failure mode. Risk-based ratio and leverage ratio can bind differently. Jo’s diagnostic is to ask whether the result comes from scenario severity, model sensitivity, starting exposure, management action or regulatory denominator. Connection. This feeds directly into capital. Ryan would rerun at least one harsher and one alternative-shape scenario to test whether the vulnerability is structural or scenario-specific. Liquidity survival horizon is time until available cash/liquidity becomes negative. It is an internal liquidity stress metric. Stress testing becomes credible only when the scenario, model translation and balance-sheet effect are visible separately. Mathematics. First t with cumulative stressed liquidity<0. Keep time indexes explicit. A stress path is not one number; each period’s losses, earnings, balances, capital and liquidity feed the next period. Failure mode. Monthly model can miss overnight failure. Jo’s diagnostic is to ask whether the result comes from scenario severity, model sensitivity, starting exposure, management action or regulatory denominator. Connection. This feeds directly into liquidity. Ryan would rerun at least one harsher and one alternative-shape scenario to test whether the vulnerability is structural or scenario-specific. LCR stress path is evolution of HQLA/net outflows during stress. It shows buffer usage. Stress testing becomes credible only when the scenario, model translation and balance-sheet effect are visible separately. Mathematics. LCR_t=HQLA_t/NetOutflows_t. Keep time indexes explicit. A stress path is not one number; each period’s losses, earnings, balances, capital and liquidity feed the next period. Failure mode. Basel intends buffer use under stress; path interpretation matters. Jo’s diagnostic is to ask whether the result comes from scenario severity, model sensitivity, starting exposure, management action or regulatory denominator. Connection. This feeds directly into liquidity. Ryan would rerun at least one harsher and one alternative-shape scenario to test whether the vulnerability is structural or scenario-specific. NSFR stress is structural stable funding under prolonged stress. It captures longer funding change. Stress testing becomes credible only when the scenario, model translation and balance-sheet effect are visible separately. Mathematics. NSFR=ASF/RSF. Keep time indexes explicit. A stress path is not one number; each period’s losses, earnings, balances, capital and liquidity feed the next period. Failure mode. A short stress may not materially change structural ratio immediately. Jo’s diagnostic is to ask whether the result comes from scenario severity, model sensitivity, starting exposure, management action or regulatory denominator. Connection. This feeds directly into liquidity. Ryan would rerun at least one harsher and one alternative-shape scenario to test whether the vulnerability is structural or scenario-specific. Model overlay is judgemental adjustment where model misses known stress mechanism. It can improve realism if governed. Stress testing becomes credible only when the scenario, model translation and balance-sheet effect are visible separately. Mathematics. Output_final=Model+Overlay. Keep time indexes explicit. A stress path is not one number; each period’s losses, earnings, balances, capital and liquidity feed the next period. Failure mode. Permanent unexplained overlay can hide weak model. Jo’s diagnostic is to ask whether the result comes from scenario severity, model sensitivity, starting exposure, management action or regulatory denominator. Connection. This feeds directly into model governance. Ryan would rerun at least one harsher and one alternative-shape scenario to test whether the vulnerability is structural or scenario-specific. Expert judgement is structured qualitative input. It is necessary where data are sparse or new risk emerges. Stress testing becomes credible only when the scenario, model translation and balance-sheet effect are visible separately. Mathematics. Document rationale, challenger and approval. Keep time indexes explicit. A stress path is not one number; each period’s losses, earnings, balances, capital and liquidity feed the next period. Failure mode. Unrecorded optimism/pessimism creates bias. Jo’s diagnostic is to ask whether the result comes from scenario severity, model sensitivity, starting exposure, management action or regulatory denominator. Connection. This feeds directly into governance. Ryan would rerun at least one harsher and one alternative-shape scenario to test whether the vulnerability is structural or scenario-specific. Model validation is independent assessment of conceptual soundness, implementation and performance. It constrains model risk. Stress testing becomes credible only when the scenario, model translation and balance-sheet effect are visible separately. Mathematics. Backtest/challenge sensitivities and outcomes. Keep time indexes explicit. A stress path is not one number; each period’s losses, earnings, balances, capital and liquidity feed the next period. Failure mode. Validation cannot prove unprecedented stress model correct. Jo’s diagnostic is to ask whether the result comes from scenario severity, model sensitivity, starting exposure, management action or regulatory denominator. Connection. This feeds directly into governance. Ryan would rerun at least one harsher and one alternative-shape scenario to test whether the vulnerability is structural or scenario-specific. Benchmark model is simpler alternative used to challenge primary model. It detects extreme or implausible outputs. Stress testing becomes credible only when the scenario, model translation and balance-sheet effect are visible separately. Mathematics. Compare primary vs benchmark. Keep time indexes explicit. A stress path is not one number; each period’s losses, earnings, balances, capital and liquidity feed the next period. Failure mode. Agreement between two similar models is weak evidence if same assumptions. Jo’s diagnostic is to ask whether the result comes from scenario severity, model sensitivity, starting exposure, management action or regulatory denominator. Connection. This feeds directly into validation. Ryan would rerun at least one harsher and one alternative-shape scenario to test whether the vulnerability is structural or scenario-specific. Scenario challenge is independent review of severity and coherence. It prevents convenient scenarios. Stress testing becomes credible only when the scenario, model translation and balance-sheet effect are visible separately. Mathematics. Compare history, vulnerabilities and reverse-stress thresholds. Keep time indexes explicit. A stress path is not one number; each period’s losses, earnings, balances, capital and liquidity feed the next period. Failure mode. Political/management pressure can soften scenarios. Jo’s diagnostic is to ask whether the result comes from scenario severity, model sensitivity, starting exposure, management action or regulatory denominator. Connection. This feeds directly into governance. Ryan would rerun at least one harsher and one alternative-shape scenario to test whether the vulnerability is structural or scenario-specific. Data lineage is traceability from source data to stress outputs. It makes results reproducible. Stress testing becomes credible only when the scenario, model translation and balance-sheet effect are visible separately. Mathematics. Store source/version/transformation. Keep time indexes explicit. A stress path is not one number; each period’s losses, earnings, balances, capital and liquidity feed the next period. Failure mode. A sophisticated model on inconsistent exposure data is unreliable. Jo’s diagnostic is to ask whether the result comes from scenario severity, model sensitivity, starting exposure, management action or regulatory denominator. Connection. This feeds directly into control. Ryan would rerun at least one harsher and one alternative-shape scenario to test whether the vulnerability is structural or scenario-specific. Model inventory is catalogue of all models in stress chain. It exposes dependencies and gaps. Stress testing becomes credible only when the scenario, model translation and balance-sheet effect are visible separately. Mathematics. Map scenario→satellite→portfolio→capital outputs. Keep time indexes explicit. A stress path is not one number; each period’s losses, earnings, balances, capital and liquidity feed the next period. Failure mode. Hidden spreadsheets create control risk. Jo’s diagnostic is to ask whether the result comes from scenario severity, model sensitivity, starting exposure, management action or regulatory denominator. Connection. This feeds directly into governance. Ryan would rerun at least one harsher and one alternative-shape scenario to test whether the vulnerability is structural or scenario-specific. Aggregation is combination of stress results across portfolios/risks. It needs consistent timing and diversification assumptions. Stress testing becomes credible only when the scenario, model translation and balance-sheet effect are visible separately. Mathematics. TotalLoss≠simple sum if interactions/offsets modelled. Keep time indexes explicit. A stress path is not one number; each period’s losses, earnings, balances, capital and liquidity feed the next period. Failure mode. Over-generous diversification can understate stress. Jo’s diagnostic is to ask whether the result comes from scenario severity, model sensitivity, starting exposure, management action or regulatory denominator. Connection. This feeds directly into enterprise stress. Ryan would rerun at least one harsher and one alternative-shape scenario to test whether the vulnerability is structural or scenario-specific. Diversification benefit is reduction in aggregate risk from imperfect correlation. It can disappear in stress. Stress testing becomes credible only when the scenario, model translation and balance-sheet effect are visible separately. Mathematics. Correlation_stress often higher. Keep time indexes explicit. A stress path is not one number; each period’s losses, earnings, balances, capital and liquidity feed the next period. Failure mode. Using calm-period correlation is optimistic. Jo’s diagnostic is to ask whether the result comes from scenario severity, model sensitivity, starting exposure, management action or regulatory denominator. Connection. This feeds directly into aggregation. Ryan would rerun at least one harsher and one alternative-shape scenario to test whether the vulnerability is structural or scenario-specific. Concentration add-on is extra loss/risk from clustered exposures. It captures granularity weaknesses. Stress testing becomes credible only when the scenario, model translation and balance-sheet effect are visible separately. Mathematics. Stress large names/sectors explicitly. Keep time indexes explicit. A stress path is not one number; each period’s losses, earnings, balances, capital and liquidity feed the next period. Failure mode. Portfolio average models can miss single-name loss. Jo’s diagnostic is to ask whether the result comes from scenario severity, model sensitivity, starting exposure, management action or regulatory denominator. Connection. This feeds directly into credit stress. Ryan would rerun at least one harsher and one alternative-shape scenario to test whether the vulnerability is structural or scenario-specific. Scenario probability is probability weight assigned to scenario if probabilistic framework used. Many supervisory stresses are not probability forecasts. Stress testing becomes credible only when the scenario, model translation and balance-sheet effect are visible separately. Mathematics. ExpectedStress=Σw_sOutcome_s in some frameworks. Keep time indexes explicit. A stress path is not one number; each period’s losses, earnings, balances, capital and liquidity feed the next period. Failure mode. Do not call severe scenario ‘1-in-100’ without statistical basis. Jo’s diagnostic is to ask whether the result comes from scenario severity, model sensitivity, starting exposure, management action or regulatory denominator. Connection. This feeds directly into scenario design. Ryan would rerun at least one harsher and one alternative-shape scenario to test whether the vulnerability is structural or scenario-specific. Narrative consistency is qualitative coherence across variables. It is essential because models inherit scenario relationships. Stress testing becomes credible only when the scenario, model translation and balance-sheet effect are visible separately. Mathematics. High unemployment, strong consumer spending and rising property prices may conflict unless story explains. Keep time indexes explicit. A stress path is not one number; each period’s losses, earnings, balances, capital and liquidity feed the next period. Failure mode. Numerical correlation does not guarantee economic coherence. Jo’s diagnostic is to ask whether the result comes from scenario severity, model sensitivity, starting exposure, management action or regulatory denominator. Connection. This feeds directly into scenario design. Ryan would rerun at least one harsher and one alternative-shape scenario to test whether the vulnerability is structural or scenario-specific. Path dependency is outcome depends on order/timing of shocks. It matters for defaults, hedges, funding and management actions. Stress testing becomes credible only when the scenario, model translation and balance-sheet effect are visible separately. Mathematics. State_t feeds State_{t+1}. Keep time indexes explicit. A stress path is not one number; each period’s losses, earnings, balances, capital and liquidity feed the next period. Failure mode. Applying final shock instantly can misstate cumulative effect. Jo’s diagnostic is to ask whether the result comes from scenario severity, model sensitivity, starting exposure, management action or regulatory denominator. Connection. This feeds directly into dynamic stress. Ryan would rerun at least one harsher and one alternative-shape scenario to test whether the vulnerability is structural or scenario-specific. Nonlinearity is response is not proportional to shock size. It arises from options, defaults, thresholds and balance-sheet feedback. Stress testing becomes credible only when the scenario, model translation and balance-sheet effect are visible separately. Mathematics. Outcome(2x)≠2Outcome(x). Keep time indexes explicit. A stress path is not one number; each period’s losses, earnings, balances, capital and liquidity feed the next period. Failure mode. Linear sensitivities can fail in severe stress. Jo’s diagnostic is to ask whether the result comes from scenario severity, model sensitivity, starting exposure, management action or regulatory denominator. Connection. This feeds directly into modeling. Ryan would rerun at least one harsher and one alternative-shape scenario to test whether the vulnerability is structural or scenario-specific. Threshold effect is behaviour/model changes after a boundary is crossed. It creates discontinuity. Stress testing becomes credible only when the scenario, model translation and balance-sheet effect are visible separately. Mathematics. Examples: margin calls, covenant breach, downgrade. Keep time indexes explicit. A stress path is not one number; each period’s losses, earnings, balances, capital and liquidity feed the next period. Failure mode. Smooth models can miss cliff risk. Jo’s diagnostic is to ask whether the result comes from scenario severity, model sensitivity, starting exposure, management action or regulatory denominator. Connection. This feeds directly into stress. Ryan would rerun at least one harsher and one alternative-shape scenario to test whether the vulnerability is structural or scenario-specific. Covenant breach is stress causes contractual threshold failure. It can accelerate debt/funding loss. Stress testing becomes credible only when the scenario, model translation and balance-sheet effect are visible separately. Mathematics. Debt terms change after ratio breach. Keep time indexes explicit. A stress path is not one number; each period’s losses, earnings, balances, capital and liquidity feed the next period. Failure mode. Ignoring legal triggers understates liquidity. Jo’s diagnostic is to ask whether the result comes from scenario severity, model sensitivity, starting exposure, management action or regulatory denominator. Connection. This feeds directly into corporate/bank stress. Ryan would rerun at least one harsher and one alternative-shape scenario to test whether the vulnerability is structural or scenario-specific. Rating downgrade is credit-quality deterioration of bank/borrower. It can increase funding/collateral cost. Stress testing becomes credible only when the scenario, model translation and balance-sheet effect are visible separately. Mathematics. Spread/haircut changes after downgrade. Keep time indexes explicit. A stress path is not one number; each period’s losses, earnings, balances, capital and liquidity feed the next period. Failure mode. Feedback can be nonlinear. Jo’s diagnostic is to ask whether the result comes from scenario severity, model sensitivity, starting exposure, management action or regulatory denominator. Connection. This feeds directly into funding stress. Ryan would rerun at least one harsher and one alternative-shape scenario to test whether the vulnerability is structural or scenario-specific. Wrong-way risk is exposure increases when counterparty credit worsens. It is critical in CCR stress. Stress testing becomes credible only when the scenario, model translation and balance-sheet effect are visible separately. Mathematics. Joint shock exposure and PD. Keep time indexes explicit. A stress path is not one number; each period’s losses, earnings, balances, capital and liquidity feed the next period. Failure mode. Independent stress of exposure and credit misses interaction. Jo’s diagnostic is to ask whether the result comes from scenario severity, model sensitivity, starting exposure, management action or regulatory denominator. Connection. This feeds directly into counterparty stress. Ryan would rerun at least one harsher and one alternative-shape scenario to test whether the vulnerability is structural or scenario-specific. Counterparty concentration is large derivative/financing exposure to few counterparties. It creates idiosyncratic stress. Stress testing becomes credible only when the scenario, model translation and balance-sheet effect are visible separately. Mathematics. Stress top names plus market factors. Keep time indexes explicit. A stress path is not one number; each period’s losses, earnings, balances, capital and liquidity feed the next period. Failure mode. Portfolio averages hide default waterfall. Jo’s diagnostic is to ask whether the result comes from scenario severity, model sensitivity, starting exposure, management action or regulatory denominator. Connection. This feeds directly into CCR. Ryan would rerun at least one harsher and one alternative-shape scenario to test whether the vulnerability is structural or scenario-specific. Recovery rate stress is lower recoveries during downturn. It increases credit loss. Stress testing becomes credible only when the scenario, model translation and balance-sheet effect are visible separately. Mathematics. LGD=1−discounted recovery ratio. Keep time indexes explicit. A stress path is not one number; each period’s losses, earnings, balances, capital and liquidity feed the next period. Failure mode. Recovery and PD often negatively related in stress. Jo’s diagnostic is to ask whether the result comes from scenario severity, model sensitivity, starting exposure, management action or regulatory denominator. Connection. This feeds directly into credit. Ryan would rerun at least one harsher and one alternative-shape scenario to test whether the vulnerability is structural or scenario-specific. Forecast error is difference between modelled and realised outcomes. It is unavoidable and should inform uncertainty. Stress testing becomes credible only when the scenario, model translation and balance-sheet effect are visible separately. Mathematics. Error=Actual−Forecast. Keep time indexes explicit. A stress path is not one number; each period’s losses, earnings, balances, capital and liquidity feed the next period. Failure mode. Backtesting normal periods does not validate crisis tails fully. Jo’s diagnostic is to ask whether the result comes from scenario severity, model sensitivity, starting exposure, management action or regulatory denominator. Connection. This feeds directly into model risk. Ryan would rerun at least one harsher and one alternative-shape scenario to test whether the vulnerability is structural or scenario-specific. Uncertainty band is range around stress output from model/parameter uncertainty. It communicates false precision risk. Stress testing becomes credible only when the scenario, model translation and balance-sheet effect are visible separately. Mathematics. Output interval from alternative assumptions/models. Keep time indexes explicit. A stress path is not one number; each period’s losses, earnings, balances, capital and liquidity feed the next period. Failure mode. Single-point CET1 forecast can look more certain than it is. Jo’s diagnostic is to ask whether the result comes from scenario severity, model sensitivity, starting exposure, management action or regulatory denominator. Connection. This feeds directly into reporting. Ryan would rerun at least one harsher and one alternative-shape scenario to test whether the vulnerability is structural or scenario-specific. Capital depletion rate is speed at which CET1 falls during stress. It matters for management response time. Stress testing becomes credible only when the scenario, model translation and balance-sheet effect are visible separately. Mathematics. ΔCET1 per quarter. Keep time indexes explicit. A stress path is not one number; each period’s losses, earnings, balances, capital and liquidity feed the next period. Failure mode. Same terminal loss with faster path can breach earlier. Jo’s diagnostic is to ask whether the result comes from scenario severity, model sensitivity, starting exposure, management action or regulatory denominator. Connection. This feeds directly into capital. Ryan would rerun at least one harsher and one alternative-shape scenario to test whether the vulnerability is structural or scenario-specific. Liquidity burn rate is speed of HQLA/cash consumption. It determines survival horizon. Stress testing becomes credible only when the scenario, model translation and balance-sheet effect are visible separately. Mathematics. BurnRate=NetStressOutflow per day/week. Keep time indexes explicit. A stress path is not one number; each period’s losses, earnings, balances, capital and liquidity feed the next period. Failure mode. Average monthly burn hides peak days. Jo’s diagnostic is to ask whether the result comes from scenario severity, model sensitivity, starting exposure, management action or regulatory denominator. Connection. This feeds directly into liquidity. Ryan would rerun at least one harsher and one alternative-shape scenario to test whether the vulnerability is structural or scenario-specific. Stress loss attribution is decomposition of total loss by risk/source. It supports management action. Stress testing becomes credible only when the scenario, model translation and balance-sheet effect are visible separately. Mathematics. Total loss=credit+market+operational+earnings shortfall+other. Keep time indexes explicit. A stress path is not one number; each period’s losses, earnings, balances, capital and liquidity feed the next period. Failure mode. Double counting across categories must be avoided. Jo’s diagnostic is to ask whether the result comes from scenario severity, model sensitivity, starting exposure, management action or regulatory denominator. Connection. This feeds directly into reporting. Ryan would rerun at least one harsher and one alternative-shape scenario to test whether the vulnerability is structural or scenario-specific. Risk appetite stress metric is board limit based on stressed outcome. It connects testing to decisions. Stress testing becomes credible only when the scenario, model translation and balance-sheet effect are visible separately. Mathematics. e.g., stressed CET1 floor or loss limit. Keep time indexes explicit. A stress path is not one number; each period’s losses, earnings, balances, capital and liquidity feed the next period. Failure mode. A metric not tied to action has weak governance value. Jo’s diagnostic is to ask whether the result comes from scenario severity, model sensitivity, starting exposure, management action or regulatory denominator. Connection. This feeds directly into risk management. Ryan would rerun at least one harsher and one alternative-shape scenario to test whether the vulnerability is structural or scenario-specific. Contingency plan is predefined response to stress. It converts analysis into action. Stress testing becomes credible only when the scenario, model translation and balance-sheet effect are visible separately. Mathematics. Trigger→Action→Owner→Timing. Keep time indexes explicit. A stress path is not one number; each period’s losses, earnings, balances, capital and liquidity feed the next period. Failure mode. Plans must be operationally tested. Jo’s diagnostic is to ask whether the result comes from scenario severity, model sensitivity, starting exposure, management action or regulatory denominator. Connection. This feeds directly into resilience. Ryan would rerun at least one harsher and one alternative-shape scenario to test whether the vulnerability is structural or scenario-specific. Stress testing governance is roles, approvals, challenge and reporting around entire process. Basel makes governance explicit. Stress testing becomes credible only when the scenario, model translation and balance-sheet effect are visible separately. Mathematics. Board/senior management oversee key assumptions and use. Keep time indexes explicit. A stress path is not one number; each period’s losses, earnings, balances, capital and liquidity feed the next period. Failure mode. Treating stress test as model-team exercise only is weak. Jo’s diagnostic is to ask whether the result comes from scenario severity, model sensitivity, starting exposure, management action or regulatory denominator. Connection. This feeds directly into governance. Ryan would rerun at least one harsher and one alternative-shape scenario to test whether the vulnerability is structural or scenario-specific. A loan portfolio has EAD S$10bn, base PD 1%, LGD 30%. Base expected loss≈S$30m. In recession, stressed PD rises to 4%, LGD to 45%, and EAD rises 5% from line drawdowns. Stressed expected loss≈10.5bn×0.04×0.45=S$189m. Loss rises by more than six times because PD, LGD and EAD worsen together. Stress testing should not assume only one component moves when the scenario affects all three. This is a simple expected-loss lens; tail and migration effects can make capital stress larger. Starting CET1 S$12bn, RWA S$100bn, ratio 12%. Year-1 pre-provision earnings S$2bn, provisions S$3bn, other losses S$0.5bn, no distributions. CET1 before adjustments falls by S$1.5bn to S$10.5bn. If stressed RWA rises to S$110bn because downgrades increase risk weights, CET1 ratio becomes 9.55%, not 10.5%. Denominator stress matters. Bank loses S$20bn deposits costing 1.5%, replaces S$15bn with wholesale funding at 5%, and funds remaining S$5bn from HQLA runoff. Annualised funding-cost increase on replacement is about S$525m before other effects. NIM falls, liquidity buffer shrinks, and asset sales may create market losses. One funding shock therefore touches earnings, liquidity and capital. Integrated stress testing should capture these channels together. Bank management sets a reverse-stress threshold CET1 ratio 7%. Starting CET1 S$14bn, RWA S$120bn. Holding RWA fixed initially, maximum capital loss before hitting 7% is 14−0.07×120=S$5.6bn. If stress also pushes RWA to S$140bn, allowable loss falls to 14−9.8=S$4.2bn. Reverse stress then asks what combination of credit, market, operational and earnings shocks could produce that loss. The exercise turns ‘what could break us?’ into a solvable boundary problem. A planned asset sale releases S$10bn RWA but can only close at end of quarter 3. Capital ratio breaches internal floor in quarter 2. Counting the sale from quarter 1 creates false resilience. Stress actions must be time-stamped and operationally feasible. Delayed actions cannot repair an earlier breach retroactively. Path matters as much as end state. Rate-sensitive assets S$50bn reprice +150bp over a year; rate-sensitive liabilities S$45bn reprice +220bp due to high deposit beta. Approx gross annual interest-income increase S$750m; interest-expense increase S$990m; NII falls S$240m before balance changes. A rising-rate scenario does not automatically improve bank earnings. Relative repricing speeds and funding mix determine outcome. This is a stress-test version of gap/beta mathematics. Liquidity stress forces sale of S$5bn bonds at 8% below carrying value, crystallising S$400m loss. The sale raises cash but reduces CET1 through earnings/equity and can lower future NII. If the same bonds also counted as HQLA, the bank trades future liquidity buffer for present cash. Management actions have side effects that must be modelled. A supervisory adverse scenario may specify unemployment 10%, property prices −25%, severe spread widening and recession. The exercise can be useful without assigning a precise 2.3% probability. Stress severity and statistical likelihood are different dimensions. False probability precision can be more misleading than a clearly labelled deterministic severe scenario. Use scenario weights only where the framework genuinely requires/justifies them. A corporate bond portfolio is stressed for spread widening and also for default losses. If spread shock already reflects default expectations and the model also writes down bonds for expected defaults over the same horizon without separating paths, losses can be double counted. A clean architecture defines pre-default mark-to-market, default event losses and recovery consistently. Aggregation requires semantic as well as arithmetic consistency. Opening liquid resources S$12bn. Daily stressed net outflows: 1.5bn,1.8bn,2.2bn,2.5bn,3.0bn over five days. Cumulative after day 4=4.0bn remaining; after day 5=1.0bn remaining. Add a day-6 margin call S$1.5bn and survival horizon ends before completion unless funding action occurs. A 30-day aggregate outflow total would not reveal the exact day of failure. Liquidity stress needs granular timing. A stress test that produces a colourful report but does not affect limits, pricing, capital plans, liquidity buffers, funding strategy, concentrations or contingency plans is incomplete. Basel’s current guidance explicitly states stress testing should inform business decisions and risk management, not remain a compliance exercise. This use test is powerful. If management cannot state what action would change when a stress metric deteriorates, the metric may not be connected to governance. Mira’s final question in every stress review is: what would we do differently because we know this? It lacks joint scenario interaction. The repair is to rebuild the scenario-to-balance-sheet chain and identify the exact missing channel. Interim breaches can occur earlier. The repair is to rebuild the scenario-to-balance-sheet chain and identify the exact missing channel. Migration and balance changes can raise denominator. The repair is to rebuild the scenario-to-balance-sheet chain and identify the exact missing channel. Execution delay matters. The repair is to rebuild the scenario-to-balance-sheet chain and identify the exact missing channel. New risks may exceed history. The repair is to rebuild the scenario-to-balance-sheet chain and identify the exact missing channel. Narrative and macro relationships should cohere. The repair is to rebuild the scenario-to-balance-sheet chain and identify the exact missing channel. Downturn components can worsen together. The repair is to rebuild the scenario-to-balance-sheet chain and identify the exact missing channel. Replacement funding cost compresses NII. The repair is to rebuild the scenario-to-balance-sheet chain and identify the exact missing channel. Liquidity actions can reduce capital. The repair is to rebuild the scenario-to-balance-sheet chain and identify the exact missing channel. Correlations can rise in stress. The repair is to rebuild the scenario-to-balance-sheet chain and identify the exact missing channel. Severe tails often have highest model error. The repair is to rebuild the scenario-to-balance-sheet chain and identify the exact missing channel. Testing without action is weak governance. The repair is to rebuild the scenario-to-balance-sheet chain and identify the exact missing channel. Situation. Property values fall, unemployment rises and mortgage defaults increase. The task is to connect a coherent narrative to bank-specific financial outcomes. Method. Stress mortgage PD/LGD, property collateral, NII and capital together. Adrian maps scenario variables, Jo checks satellite models, Aisha projects balance-sheet paths, and Ryan tests management actions and reverse thresholds. Boundary. Property shock should affect both default likelihood and recovery. Mira then asks what the model cannot observe directly and how uncertainty should be communicated. Situation. GDP falls and credit spreads widen. The task is to connect a coherent narrative to bank-specific financial outcomes. Method. Stress corporate migration, default, fee income and bond market values. Adrian maps scenario variables, Jo checks satellite models, Aisha projects balance-sheet paths, and Ryan tests management actions and reverse thresholds. Boundary. Credit and market channels can overlap; avoid double counting. Mira then asks what the model cannot observe directly and how uncertainty should be communicated. Situation. Rumour triggers deposit outflow and wholesale spread widening. The task is to connect a coherent narrative to bank-specific financial outcomes. Method. Model runoff, deposit repricing, HQLA usage and replacement funding. Adrian maps scenario variables, Jo checks satellite models, Aisha projects balance-sheet paths, and Ryan tests management actions and reverse thresholds. Boundary. Liquidity and profitability interact immediately. Mira then asks what the model cannot observe directly and how uncertainty should be communicated. Situation. Yield curve rises nonparallel while deposit betas jump. The task is to connect a coherent narrative to bank-specific financial outcomes. Method. Project NII and EVE plus securities values. Adrian maps scenario variables, Jo checks satellite models, Aisha projects balance-sheet paths, and Ryan tests management actions and reverse thresholds. Boundary. Earnings and economic value can move differently. Mira then asks what the model cannot observe directly and how uncertainty should be communicated. Situation. Payments fail for hours and fraud losses occur. The task is to connect a coherent narrative to bank-specific financial outcomes. Method. Model operational loss, customer runoff and intraday liquidity. Adrian maps scenario variables, Jo checks satellite models, Aisha projects balance-sheet paths, and Ryan tests management actions and reverse thresholds. Boundary. Operational resilience can become liquidity stress. Mira then asks what the model cannot observe directly and how uncertainty should be communicated. Situation. Large derivatives counterparty weakens as market exposure rises. The task is to connect a coherent narrative to bank-specific financial outcomes. Method. Stress exposure and creditworthiness jointly. Adrian maps scenario variables, Jo checks satellite models, Aisha projects balance-sheet paths, and Ryan tests management actions and reverse thresholds. Boundary. Wrong-way risk invalidates independent shocks. Mira then asks what the model cannot observe directly and how uncertainty should be communicated. Situation. Vacancy rises, rents fall, cap rates widen. The task is to connect a coherent narrative to bank-specific financial outcomes. Method. Stress DSCR, collateral value, PD/LGD and bank concentration. Adrian maps scenario variables, Jo checks satellite models, Aisha projects balance-sheet paths, and Ryan tests management actions and reverse thresholds. Boundary. Sector concentration matters beyond average portfolio loss. Mira then asks what the model cannot observe directly and how uncertainty should be communicated. Situation. Currency depreciates and local rates rise. The task is to connect a coherent narrative to bank-specific financial outcomes. Method. Stress FX positions, borrower debt service and funding costs. Adrian maps scenario variables, Jo checks satellite models, Aisha projects balance-sheet paths, and Ryan tests management actions and reverse thresholds. Boundary. Indirect borrower FX mismatch can dominate direct bank position. Mira then asks what the model cannot observe directly and how uncertainty should be communicated. Situation. Stress causes CET1 decline; management proposes issuance. The task is to connect a coherent narrative to bank-specific financial outcomes. Method. Time action and test market access/price dilution. Adrian maps scenario variables, Jo checks satellite models, Aisha projects balance-sheet paths, and Ryan tests management actions and reverse thresholds. Boundary. Assuming full raise instantly overstates resilience. Mira then asks what the model cannot observe directly and how uncertainty should be communicated. Situation. Liquidity stress triggers bond sales. The task is to connect a coherent narrative to bank-specific financial outcomes. Method. Model cash raised, realised loss and reduced future income. Adrian maps scenario variables, Jo checks satellite models, Aisha projects balance-sheet paths, and Ryan tests management actions and reverse thresholds. Boundary. Management actions can worsen capital. Mira then asks what the model cannot observe directly and how uncertainty should be communicated. Situation. Board wants scenarios that breach internal CET1 floor. The task is to connect a coherent narrative to bank-specific financial outcomes. Method. Solve combinations of loss, RWA growth and funding shock that cross threshold. Adrian maps scenario variables, Jo checks satellite models, Aisha projects balance-sheet paths, and Ryan tests management actions and reverse thresholds. Boundary. Reverse stress exposes hidden fragility without pretending probability. Mira then asks what the model cannot observe directly and how uncertainty should be communicated. Situation. Many banks deleverage simultaneously. The task is to connect a coherent narrative to bank-specific financial outcomes. Method. Add price impact and credit contraction feedback. Adrian maps scenario variables, Jo checks satellite models, Aisha projects balance-sheet paths, and Ryan tests management actions and reverse thresholds. Boundary. Second-round effects can amplify first-round losses. Mira then asks what the model cannot observe directly and how uncertainty should be communicated. Stress testing is not about predicting the next crisis. It is about discovering whether the bank can survive adverse worlds it may not have experienced before. The mathematics connects scenario paths to credit loss, earnings, capital, liquidity and management action. The governance connects those outputs to real decisions. Both are necessary. A strong framework therefore treats stress testing as a dynamic balance-sheet simulation, not a static haircut exercise. It asks how risks interact, how quickly buffers deplete, when thresholds are crossed and whether planned actions arrive in time. That leads directly to the next owner: market-risk mathematics, where VaR, Expected Shortfall, liquidity horizons, backtesting and stress scenarios quantify losses from market prices. Start with balance sheet, CET1 and RWA. Project PPNR, provisions, market losses, taxes, distributions and RWA each quarter. Identify first threshold breach. Complete the lab with a bridge from scenario input to model output to financial statement to regulatory ratio. Ben should reconcile arithmetic, Clara should document scenario source/version, and Ethan should identify which assumption is least observable in stress. Then alter the path while keeping the same final macro endpoint. If losses change, explain the path dependency. This shows why stress testing needs time series rather than endpoint shocks alone. Specify GDP, unemployment, property prices, rates and credit spreads for eight quarters. Explain causal narrative and check variables for contradictions. Complete the lab with a bridge from scenario input to model output to financial statement to regulatory ratio. Ben should reconcile arithmetic, Clara should document scenario source/version, and Ethan should identify which assumption is least observable in stress. Then alter the path while keeping the same final macro endpoint. If losses change, explain the path dependency. This shows why stress testing needs time series rather than endpoint shocks alone. Choose CET1 floor and solve combinations of credit loss, RWA inflation and earnings decline required to breach it. Compare with historical and hypothetical scenarios. Complete the lab with a bridge from scenario input to model output to financial statement to regulatory ratio. Ben should reconcile arithmetic, Clara should document scenario source/version, and Ethan should identify which assumption is least observable in stress. Then alter the path while keeping the same final macro endpoint. If losses change, explain the path dependency. This shows why stress testing needs time series rather than endpoint shocks alone. Model deposit run, asset sales, market haircuts and realised losses. Track both cash survival and CET1 through days/weeks. Complete the lab with a bridge from scenario input to model output to financial statement to regulatory ratio. Ben should reconcile arithmetic, Clara should document scenario source/version, and Ethan should identify which assumption is least observable in stress. Then alter the path while keeping the same final macro endpoint. If losses change, explain the path dependency. This shows why stress testing needs time series rather than endpoint shocks alone. List proposed actions, timing, dependencies and side effects. Remove each action in turn and measure resilience difference. Complete the lab with a bridge from scenario input to model output to financial statement to regulatory ratio. Ben should reconcile arithmetic, Clara should document scenario source/version, and Ethan should identify which assumption is least observable in stress. Then alter the path while keeping the same final macro endpoint. If losses change, explain the path dependency. This shows why stress testing needs time series rather than endpoint shocks alone. Start with balance sheet, CET1 and RWA. Project PPNR, provisions, market losses, taxes, distributions and RWA each quarter. Identify first threshold breach. Complete the lab with a bridge from scenario input to model output to financial statement to regulatory ratio. Ben should reconcile arithmetic, Clara should document scenario source/version, and Ethan should identify which assumption is least observable in stress. Then alter the path while keeping the same final macro endpoint. If losses change, explain the path dependency. This shows why stress testing needs time series rather than endpoint shocks alone. Specify GDP, unemployment, property prices, rates and credit spreads for eight quarters. Explain causal narrative and check variables for contradictions. Complete the lab with a bridge from scenario input to model output to financial statement to regulatory ratio. Ben should reconcile arithmetic, Clara should document scenario source/version, and Ethan should identify which assumption is least observable in stress. Then alter the path while keeping the same final macro endpoint. If losses change, explain the path dependency. This shows why stress testing needs time series rather than endpoint shocks alone. Choose CET1 floor and solve combinations of credit loss, RWA inflation and earnings decline required to breach it. Compare with historical and hypothetical scenarios. Complete the lab with a bridge from scenario input to model output to financial statement to regulatory ratio. Ben should reconcile arithmetic, Clara should document scenario source/version, and Ethan should identify which assumption is least observable in stress. Then alter the path while keeping the same final macro endpoint. If losses change, explain the path dependency. This shows why stress testing needs time series rather than endpoint shocks alone. Model deposit run, asset sales, market haircuts and realised losses. Track both cash survival and CET1 through days/weeks. Complete the lab with a bridge from scenario input to model output to financial statement to regulatory ratio. Ben should reconcile arithmetic, Clara should document scenario source/version, and Ethan should identify which assumption is least observable in stress. Then alter the path while keeping the same final macro endpoint. If losses change, explain the path dependency. This shows why stress testing needs time series rather than endpoint shocks alone. List proposed actions, timing, dependencies and side effects. Remove each action in turn and measure resilience difference. Complete the lab with a bridge from scenario input to model output to financial statement to regulatory ratio. Ben should reconcile arithmetic, Clara should document scenario source/version, and Ethan should identify which assumption is least observable in stress. Then alter the path while keeping the same final macro endpoint. If losses change, explain the path dependency. This shows why stress testing needs time series rather than endpoint shocks alone. Start with balance sheet, CET1 and RWA. Project PPNR, provisions, market losses, taxes, distributions and RWA each quarter. Identify first threshold breach. Complete the lab with a bridge from scenario input to model output to financial statement to regulatory ratio. Ben should reconcile arithmetic, Clara should document scenario source/version, and Ethan should identify which assumption is least observable in stress. Then alter the path while keeping the same final macro endpoint. If losses change, explain the path dependency. This shows why stress testing needs time series rather than endpoint shocks alone. Specify GDP, unemployment, property prices, rates and credit spreads for eight quarters. Explain causal narrative and check variables for contradictions. Complete the lab with a bridge from scenario input to model output to financial statement to regulatory ratio. Ben should reconcile arithmetic, Clara should document scenario source/version, and Ethan should identify which assumption is least observable in stress. Then alter the path while keeping the same final macro endpoint. If losses change, explain the path dependency. This shows why stress testing needs time series rather than endpoint shocks alone. Choose CET1 floor and solve combinations of credit loss, RWA inflation and earnings decline required to breach it. Compare with historical and hypothetical scenarios. Complete the lab with a bridge from scenario input to model output to financial statement to regulatory ratio. Ben should reconcile arithmetic, Clara should document scenario source/version, and Ethan should identify which assumption is least observable in stress. Then alter the path while keeping the same final macro endpoint. If losses change, explain the path dependency. This shows why stress testing needs time series rather than endpoint shocks alone. Model deposit run, asset sales, market haircuts and realised losses. Track both cash survival and CET1 through days/weeks. Complete the lab with a bridge from scenario input to model output to financial statement to regulatory ratio. Ben should reconcile arithmetic, Clara should document scenario source/version, and Ethan should identify which assumption is least observable in stress. Then alter the path while keeping the same final macro endpoint. If losses change, explain the path dependency. This shows why stress testing needs time series rather than endpoint shocks alone. List proposed actions, timing, dependencies and side effects. Remove each action in turn and measure resilience difference. Complete the lab with a bridge from scenario input to model output to financial statement to regulatory ratio. Ben should reconcile arithmetic, Clara should document scenario source/version, and Ethan should identify which assumption is least observable in stress. Then alter the path while keeping the same final macro endpoint. If losses change, explain the path dependency. This shows why stress testing needs time series rather than endpoint shocks alone. Start with balance sheet, CET1 and RWA. Project PPNR, provisions, market losses, taxes, distributions and RWA each quarter. Identify first threshold breach. Complete the lab with a bridge from scenario input to model output to financial statement to regulatory ratio. Ben should reconcile arithmetic, Clara should document scenario source/version, and Ethan should identify which assumption is least observable in stress. Then alter the path while keeping the same final macro endpoint. If losses change, explain the path dependency. This shows why stress testing needs time series rather than endpoint shocks alone. Specify GDP, unemployment, property prices, rates and credit spreads for eight quarters. Explain causal narrative and check variables for contradictions. Complete the lab with a bridge from scenario input to model output to financial statement to regulatory ratio. Ben should reconcile arithmetic, Clara should document scenario source/version, and Ethan should identify which assumption is least observable in stress. Then alter the path while keeping the same final macro endpoint. If losses change, explain the path dependency. This shows why stress testing needs time series rather than endpoint shocks alone. Choose CET1 floor and solve combinations of credit loss, RWA inflation and earnings decline required to breach it. Compare with historical and hypothetical scenarios. Complete the lab with a bridge from scenario input to model output to financial statement to regulatory ratio. Ben should reconcile arithmetic, Clara should document scenario source/version, and Ethan should identify which assumption is least observable in stress. Then alter the path while keeping the same final macro endpoint. If losses change, explain the path dependency. This shows why stress testing needs time series rather than endpoint shocks alone. Model deposit run, asset sales, market haircuts and realised losses. Track both cash survival and CET1 through days/weeks. Complete the lab with a bridge from scenario input to model output to financial statement to regulatory ratio. Ben should reconcile arithmetic, Clara should document scenario source/version, and Ethan should identify which assumption is least observable in stress. Then alter the path while keeping the same final macro endpoint. If losses change, explain the path dependency. This shows why stress testing needs time series rather than endpoint shocks alone. List proposed actions, timing, dependencies and side effects. Remove each action in turn and measure resilience difference. Complete the lab with a bridge from scenario input to model output to financial statement to regulatory ratio. Ben should reconcile arithmetic, Clara should document scenario source/version, and Ethan should identify which assumption is least observable in stress. Then alter the path while keeping the same final macro endpoint. If losses change, explain the path dependency. This shows why stress testing needs time series rather than endpoint shocks alone. Start with balance sheet, CET1 and RWA. Project PPNR, provisions, market losses, taxes, distributions and RWA each quarter. Identify first threshold breach. Complete the lab with a bridge from scenario input to model output to financial statement to regulatory ratio. Ben should reconcile arithmetic, Clara should document scenario source/version, and Ethan should identify which assumption is least observable in stress. Then alter the path while keeping the same final macro endpoint. If losses change, explain the path dependency. This shows why stress testing needs time series rather than endpoint shocks alone. Specify GDP, unemployment, property prices, rates and credit spreads for eight quarters. Explain causal narrative and check variables for contradictions. Complete the lab with a bridge from scenario input to model output to financial statement to regulatory ratio. Ben should reconcile arithmetic, Clara should document scenario source/version, and Ethan should identify which assumption is least observable in stress. Then alter the path while keeping the same final macro endpoint. If losses change, explain the path dependency. This shows why stress testing needs time series rather than endpoint shocks alone. Choose CET1 floor and solve combinations of credit loss, RWA inflation and earnings decline required to breach it. Compare with historical and hypothetical scenarios. Complete the lab with a bridge from scenario input to model output to financial statement to regulatory ratio. Ben should reconcile arithmetic, Clara should document scenario source/version, and Ethan should identify which assumption is least observable in stress. Then alter the path while keeping the same final macro endpoint. If losses change, explain the path dependency. This shows why stress testing needs time series rather than endpoint shocks alone. Model deposit run, asset sales, market haircuts and realised losses. Track both cash survival and CET1 through days/weeks. Complete the lab with a bridge from scenario input to model output to financial statement to regulatory ratio. Ben should reconcile arithmetic, Clara should document scenario source/version, and Ethan should identify which assumption is least observable in stress. Then alter the path while keeping the same final macro endpoint. If losses change, explain the path dependency. This shows why stress testing needs time series rather than endpoint shocks alone. List proposed actions, timing, dependencies and side effects. Remove each action in turn and measure resilience difference. Complete the lab with a bridge from scenario input to model output to financial statement to regulatory ratio. Ben should reconcile arithmetic, Clara should document scenario source/version, and Ethan should identify which assumption is least observable in stress. Then alter the path while keeping the same final macro endpoint. If losses change, explain the path dependency. This shows why stress testing needs time series rather than endpoint shocks alone. Start with balance sheet, CET1 and RWA. Project PPNR, provisions, market losses, taxes, distributions and RWA each quarter. Identify first threshold breach. Complete the lab with a bridge from scenario input to model output to financial statement to regulatory ratio. Ben should reconcile arithmetic, Clara should document scenario source/version, and Ethan should identify which assumption is least observable in stress. Then alter the path while keeping the same final macro endpoint. If losses change, explain the path dependency. This shows why stress testing needs time series rather than endpoint shocks alone. Specify GDP, unemployment, property prices, rates and credit spreads for eight quarters. Explain causal narrative and check variables for contradictions. Complete the lab with a bridge from scenario input to model output to financial statement to regulatory ratio. Ben should reconcile arithmetic, Clara should document scenario source/version, and Ethan should identify which assumption is least observable in stress. Then alter the path while keeping the same final macro endpoint. If losses change, explain the path dependency. This shows why stress testing needs time series rather than endpoint shocks alone. Choose CET1 floor and solve combinations of credit loss, RWA inflation and earnings decline required to breach it. Compare with historical and hypothetical scenarios. Complete the lab with a bridge from scenario input to model output to financial statement to regulatory ratio. Ben should reconcile arithmetic, Clara should document scenario source/version, and Ethan should identify which assumption is least observable in stress. Then alter the path while keeping the same final macro endpoint. If losses change, explain the path dependency. This shows why stress testing needs time series rather than endpoint shocks alone. Model deposit run, asset sales, market haircuts and realised losses. Track both cash survival and CET1 through days/weeks. Complete the lab with a bridge from scenario input to model output to financial statement to regulatory ratio. Ben should reconcile arithmetic, Clara should document scenario source/version, and Ethan should identify which assumption is least observable in stress. Then alter the path while keeping the same final macro endpoint. If losses change, explain the path dependency. This shows why stress testing needs time series rather than endpoint shocks alone. List proposed actions, timing, dependencies and side effects. Remove each action in turn and measure resilience difference. Complete the lab with a bridge from scenario input to model output to financial statement to regulatory ratio. Ben should reconcile arithmetic, Clara should document scenario source/version, and Ethan should identify which assumption is least observable in stress. Then alter the path while keeping the same final macro endpoint. If losses change, explain the path dependency. This shows why stress testing needs time series rather than endpoint shocks alone. Start with balance sheet, CET1 and RWA. Project PPNR, provisions, market losses, taxes, distributions and RWA each quarter. Identify first threshold breach. Complete the lab with a bridge from scenario input to model output to financial statement to regulatory ratio. Ben should reconcile arithmetic, Clara should document scenario source/version, and Ethan should identify which assumption is least observable in stress. Then alter the path while keeping the same final macro endpoint. If losses change, explain the path dependency. This shows why stress testing needs time series rather than endpoint shocks alone. Specify GDP, unemployment, property prices, rates and credit spreads for eight quarters. Explain causal narrative and check variables for contradictions. Complete the lab with a bridge from scenario input to model output to financial statement to regulatory ratio. Ben should reconcile arithmetic, Clara should document scenario source/version, and Ethan should identify which assumption is least observable in stress. Then alter the path while keeping the same final macro endpoint. If losses change, explain the path dependency. This shows why stress testing needs time series rather than endpoint shocks alone. Choose CET1 floor and solve combinations of credit loss, RWA inflation and earnings decline required to breach it. Compare with historical and hypothetical scenarios. Complete the lab with a bridge from scenario input to model output to financial statement to regulatory ratio. Ben should reconcile arithmetic, Clara should document scenario source/version, and Ethan should identify which assumption is least observable in stress. Then alter the path while keeping the same final macro endpoint. If losses change, explain the path dependency. This shows why stress testing needs time series rather than endpoint shocks alone. Model deposit run, asset sales, market haircuts and realised losses. Track both cash survival and CET1 through days/weeks. Complete the lab with a bridge from scenario input to model output to financial statement to regulatory ratio. Ben should reconcile arithmetic, Clara should document scenario source/version, and Ethan should identify which assumption is least observable in stress. Then alter the path while keeping the same final macro endpoint. If losses change, explain the path dependency. This shows why stress testing needs time series rather than endpoint shocks alone. List proposed actions, timing, dependencies and side effects. Remove each action in turn and measure resilience difference. Complete the lab with a bridge from scenario input to model output to financial statement to regulatory ratio. Ben should reconcile arithmetic, Clara should document scenario source/version, and Ethan should identify which assumption is least observable in stress. Then alter the path while keeping the same final macro endpoint. If losses change, explain the path dependency. This shows why stress testing needs time series rather than endpoint shocks alone. Start with balance sheet, CET1 and RWA. Project PPNR, provisions, market losses, taxes, distributions and RWA each quarter. Identify first threshold breach. Complete the lab with a bridge from scenario input to model output to financial statement to regulatory ratio. Ben should reconcile arithmetic, Clara should document scenario source/version, and Ethan should identify which assumption is least observable in stress. Then alter the path while keeping the same final macro endpoint. If losses change, explain the path dependency. This shows why stress testing needs time series rather than endpoint shocks alone. Specify GDP, unemployment, property prices, rates and credit spreads for eight quarters. Explain causal narrative and check variables for contradictions. Complete the lab with a bridge from scenario input to model output to financial statement to regulatory ratio. Ben should reconcile arithmetic, Clara should document scenario source/version, and Ethan should identify which assumption is least observable in stress. Then alter the path while keeping the same final macro endpoint. If losses change, explain the path dependency. This shows why stress testing needs time series rather than endpoint shocks alone. Choose CET1 floor and solve combinations of credit loss, RWA inflation and earnings decline required to breach it. Compare with historical and hypothetical scenarios. Complete the lab with a bridge from scenario input to model output to financial statement to regulatory ratio. Ben should reconcile arithmetic, Clara should document scenario source/version, and Ethan should identify which assumption is least observable in stress. Then alter the path while keeping the same final macro endpoint. If losses change, explain the path dependency. This shows why stress testing needs time series rather than endpoint shocks alone. Model deposit run, asset sales, market haircuts and realised losses. Track both cash survival and CET1 through days/weeks. Complete the lab with a bridge from scenario input to model output to financial statement to regulatory ratio. Ben should reconcile arithmetic, Clara should document scenario source/version, and Ethan should identify which assumption is least observable in stress. Then alter the path while keeping the same final macro endpoint. If losses change, explain the path dependency. This shows why stress testing needs time series rather than endpoint shocks alone. List proposed actions, timing, dependencies and side effects. Remove each action in turn and measure resilience difference. Complete the lab with a bridge from scenario input to model output to financial statement to regulatory ratio. Ben should reconcile arithmetic, Clara should document scenario source/version, and Ethan should identify which assumption is least observable in stress. Then alter the path while keeping the same final macro endpoint. If losses change, explain the path dependency. This shows why stress testing needs time series rather than endpoint shocks alone. Start with balance sheet, CET1 and RWA. Project PPNR, provisions, market losses, taxes, distributions and RWA each quarter. Identify first threshold breach. Complete the lab with a bridge from scenario input to model output to financial statement to regulatory ratio. Ben should reconcile arithmetic, Clara should document scenario source/version, and Ethan should identify which assumption is least observable in stress. Then alter the path while keeping the same final macro endpoint. If losses change, explain the path dependency. This shows why stress testing needs time series rather than endpoint shocks alone. Specify GDP, unemployment, property prices, rates and credit spreads for eight quarters. Explain causal narrative and check variables for contradictions. Complete the lab with a bridge from scenario input to model output to financial statement to regulatory ratio. Ben should reconcile arithmetic, Clara should document scenario source/version, and Ethan should identify which assumption is least observable in stress. Then alter the path while keeping the same final macro endpoint. If losses change, explain the path dependency. This shows why stress testing needs time series rather than endpoint shocks alone. Choose CET1 floor and solve combinations of credit loss, RWA inflation and earnings decline required to breach it. Compare with historical and hypothetical scenarios. Complete the lab with a bridge from scenario input to model output to financial statement to regulatory ratio. Ben should reconcile arithmetic, Clara should document scenario source/version, and Ethan should identify which assumption is least observable in stress. Then alter the path while keeping the same final macro endpoint. If losses change, explain the path dependency. This shows why stress testing needs time series rather than endpoint shocks alone. Model deposit run, asset sales, market haircuts and realised losses. Track both cash survival and CET1 through days/weeks. Complete the lab with a bridge from scenario input to model output to financial statement to regulatory ratio. Ben should reconcile arithmetic, Clara should document scenario source/version, and Ethan should identify which assumption is least observable in stress. Then alter the path while keeping the same final macro endpoint. If losses change, explain the path dependency. This shows why stress testing needs time series rather than endpoint shocks alone. List proposed actions, timing, dependencies and side effects. Remove each action in turn and measure resilience difference. Complete the lab with a bridge from scenario input to model output to financial statement to regulatory ratio. Ben should reconcile arithmetic, Clara should document scenario source/version, and Ethan should identify which assumption is least observable in stress. Then alter the path while keeping the same final macro endpoint. If losses change, explain the path dependency. This shows why stress testing needs time series rather than endpoint shocks alone. Start with balance sheet, CET1 and RWA. Project PPNR, provisions, market losses, taxes, distributions and RWA each quarter. Identify first threshold breach. Complete the lab with a bridge from scenario input to model output to financial statement to regulatory ratio. Ben should reconcile arithmetic, Clara should document scenario source/version, and Ethan should identify which assumption is least observable in stress. Then alter the path while keeping the same final macro endpoint. If losses change, explain the path dependency. This shows why stress testing needs time series rather than endpoint shocks alone. Specify GDP, unemployment, property prices, rates and credit spreads for eight quarters. Explain causal narrative and check variables for contradictions. Complete the lab with a bridge from scenario input to model output to financial statement to regulatory ratio. Ben should reconcile arithmetic, Clara should document scenario source/version, and Ethan should identify which assumption is least observable in stress. Then alter the path while keeping the same final macro endpoint. If losses change, explain the path dependency. This shows why stress testing needs time series rather than endpoint shocks alone. Choose CET1 floor and solve combinations of credit loss, RWA inflation and earnings decline required to breach it. Compare with historical and hypothetical scenarios. Complete the lab with a bridge from scenario input to model output to financial statement to regulatory ratio. Ben should reconcile arithmetic, Clara should document scenario source/version, and Ethan should identify which assumption is least observable in stress. Then alter the path while keeping the same final macro endpoint. If losses change, explain the path dependency. This shows why stress testing needs time series rather than endpoint shocks alone. Model deposit run, asset sales, market haircuts and realised losses. Track both cash survival and CET1 through days/weeks. Complete the lab with a bridge from scenario input to model output to financial statement to regulatory ratio. Ben should reconcile arithmetic, Clara should document scenario source/version, and Ethan should identify which assumption is least observable in stress. Then alter the path while keeping the same final macro endpoint. If losses change, explain the path dependency. This shows why stress testing needs time series rather than endpoint shocks alone. List proposed actions, timing, dependencies and side effects. Remove each action in turn and measure resilience difference. Complete the lab with a bridge from scenario input to model output to financial statement to regulatory ratio. Ben should reconcile arithmetic, Clara should document scenario source/version, and Ethan should identify which assumption is least observable in stress. Then alter the path while keeping the same final macro endpoint. If losses change, explain the path dependency. This shows why stress testing needs time series rather than endpoint shocks alone. Start with balance sheet, CET1 and RWA. Project PPNR, provisions, market losses, taxes, distributions and RWA each quarter. Identify first threshold breach. Complete the lab with a bridge from scenario input to model output to financial statement to regulatory ratio. Ben should reconcile arithmetic, Clara should document scenario source/version, and Ethan should identify which assumption is least observable in stress. Then alter the path while keeping the same final macro endpoint. If losses change, explain the path dependency. This shows why stress testing needs time series rather than endpoint shocks alone. Specify GDP, unemployment, property prices, rates and credit spreads for eight quarters. Explain causal narrative and check variables for contradictions. Complete the lab with a bridge from scenario input to model output to financial statement to regulatory ratio. Ben should reconcile arithmetic, Clara should document scenario source/version, and Ethan should identify which assumption is least observable in stress. Then alter the path while keeping the same final macro endpoint. If losses change, explain the path dependency. This shows why stress testing needs time series rather than endpoint shocks alone. Choose CET1 floor and solve combinations of credit loss, RWA inflation and earnings decline required to breach it. Compare with historical and hypothetical scenarios. Complete the lab with a bridge from scenario input to model output to financial statement to regulatory ratio. Ben should reconcile arithmetic, Clara should document scenario source/version, and Ethan should identify which assumption is least observable in stress. Then alter the path while keeping the same final macro endpoint. If losses change, explain the path dependency. This shows why stress testing needs time series rather than endpoint shocks alone. Model deposit run, asset sales, market haircuts and realised losses. Track both cash survival and CET1 through days/weeks. Complete the lab with a bridge from scenario input to model output to financial statement to regulatory ratio. Ben should reconcile arithmetic, Clara should document scenario source/version, and Ethan should identify which assumption is least observable in stress. Then alter the path while keeping the same final macro endpoint. If losses change, explain the path dependency. This shows why stress testing needs time series rather than endpoint shocks alone. List proposed actions, timing, dependencies and side effects. Remove each action in turn and measure resilience difference. Complete the lab with a bridge from scenario input to model output to financial statement to regulatory ratio. Ben should reconcile arithmetic, Clara should document scenario source/version, and Ethan should identify which assumption is least observable in stress. Then alter the path while keeping the same final macro endpoint. If losses change, explain the path dependency. This shows why stress testing needs time series rather than endpoint shocks alone. Start with balance sheet, CET1 and RWA. Project PPNR, provisions, market losses, taxes, distributions and RWA each quarter. Identify first threshold breach. Complete the lab with a bridge from scenario input to model output to financial statement to regulatory ratio. Ben should reconcile arithmetic, Clara should document scenario source/version, and Ethan should identify which assumption is least observable in stress. Then alter the path while keeping the same final macro endpoint. If losses change, explain the path dependency. This shows why stress testing needs time series rather than endpoint shocks alone. Specify GDP, unemployment, property prices, rates and credit spreads for eight quarters. Explain causal narrative and check variables for contradictions. Complete the lab with a bridge from scenario input to model output to financial statement to regulatory ratio. Ben should reconcile arithmetic, Clara should document scenario source/version, and Ethan should identify which assumption is least observable in stress. Then alter the path while keeping the same final macro endpoint. If losses change, explain the path dependency. This shows why stress testing needs time series rather than endpoint shocks alone. Choose CET1 floor and solve combinations of credit loss, RWA inflation and earnings decline required to breach it. Compare with historical and hypothetical scenarios. Complete the lab with a bridge from scenario input to model output to financial statement to regulatory ratio. Ben should reconcile arithmetic, Clara should document scenario source/version, and Ethan should identify which assumption is least observable in stress. Then alter the path while keeping the same final macro endpoint. If losses change, explain the path dependency. This shows why stress testing needs time series rather than endpoint shocks alone. Model deposit run, asset sales, market haircuts and realised losses. Track both cash survival and CET1 through days/weeks. Complete the lab with a bridge from scenario input to model output to financial statement to regulatory ratio. Ben should reconcile arithmetic, Clara should document scenario source/version, and Ethan should identify which assumption is least observable in stress. Then alter the path while keeping the same final macro endpoint. If losses change, explain the path dependency. This shows why stress testing needs time series rather than endpoint shocks alone. List proposed actions, timing, dependencies and side effects. Remove each action in turn and measure resilience difference. Complete the lab with a bridge from scenario input to model output to financial statement to regulatory ratio. Ben should reconcile arithmetic, Clara should document scenario source/version, and Ethan should identify which assumption is least observable in stress. Then alter the path while keeping the same final macro endpoint. If losses change, explain the path dependency. This shows why stress testing needs time series rather than endpoint shocks alone. Start with balance sheet, CET1 and RWA. Project PPNR, provisions, market losses, taxes, distributions and RWA each quarter. Identify first threshold breach. Complete the lab with a bridge from scenario input to model output to financial statement to regulatory ratio. Ben should reconcile arithmetic, Clara should document scenario source/version, and Ethan should identify which assumption is least observable in stress. Then alter the path while keeping the same final macro endpoint. If losses change, explain the path dependency. This shows why stress testing needs time series rather than endpoint shocks alone. Specify GDP, unemployment, property prices, rates and credit spreads for eight quarters. Explain causal narrative and check variables for contradictions. Complete the lab with a bridge from scenario input to model output to financial statement to regulatory ratio. Ben should reconcile arithmetic, Clara should document scenario source/version, and Ethan should identify which assumption is least observable in stress. Then alter the path while keeping the same final macro endpoint. If losses change, explain the path dependency. This shows why stress testing needs time series rather than endpoint shocks alone. Choose CET1 floor and solve combinations of credit loss, RWA inflation and earnings decline required to breach it. Compare with historical and hypothetical scenarios. Complete the lab with a bridge from scenario input to model output to financial statement to regulatory ratio. Ben should reconcile arithmetic, Clara should document scenario source/version, and Ethan should identify which assumption is least observable in stress. Then alter the path while keeping the same final macro endpoint. If losses change, explain the path dependency. This shows why stress testing needs time series rather than endpoint shocks alone. Model deposit run, asset sales, market haircuts and realised losses. Track both cash survival and CET1 through days/weeks. Complete the lab with a bridge from scenario input to model output to financial statement to regulatory ratio. Ben should reconcile arithmetic, Clara should document scenario source/version, and Ethan should identify which assumption is least observable in stress. Then alter the path while keeping the same final macro endpoint. If losses change, explain the path dependency. This shows why stress testing needs time series rather than endpoint shocks alone. List proposed actions, timing, dependencies and side effects. Remove each action in turn and measure resilience difference. Complete the lab with a bridge from scenario input to model output to financial statement to regulatory ratio. Ben should reconcile arithmetic, Clara should document scenario source/version, and Ethan should identify which assumption is least observable in stress. Then alter the path while keeping the same final macro endpoint. If losses change, explain the path dependency. This shows why stress testing needs time series rather than endpoint shocks alone. Start with balance sheet, CET1 and RWA. Project PPNR, provisions, market losses, taxes, distributions and RWA each quarter. Identify first threshold breach. Complete the lab with a bridge from scenario input to model output to financial statement to regulatory ratio. Ben should reconcile arithmetic, Clara should document scenario source/version, and Ethan should identify which assumption is least observable in stress. Then alter the path while keeping the same final macro endpoint. If losses change, explain the path dependency. This shows why stress testing needs time series rather than endpoint shocks alone. Specify GDP, unemployment, property prices, rates and credit spreads for eight quarters. Explain causal narrative and check variables for contradictions. Complete the lab with a bridge from scenario input to model output to financial statement to regulatory ratio. Ben should reconcile arithmetic, Clara should document scenario source/version, and Ethan should identify which assumption is least observable in stress. Then alter the path while keeping the same final macro endpoint. If losses change, explain the path dependency. This shows why stress testing needs time series rather than endpoint shocks alone. Choose CET1 floor and solve combinations of credit loss, RWA inflation and earnings decline required to breach it. Compare with historical and hypothetical scenarios. Complete the lab with a bridge from scenario input to model output to financial statement to regulatory ratio. Ben should reconcile arithmetic, Clara should document scenario source/version, and Ethan should identify which assumption is least observable in stress. Then alter the path while keeping the same final macro endpoint. If losses change, explain the path dependency. This shows why stress testing needs time series rather than endpoint shocks alone. Model deposit run, asset sales, market haircuts and realised losses. Track both cash survival and CET1 through days/weeks. Complete the lab with a bridge from scenario input to model output to financial statement to regulatory ratio. Ben should reconcile arithmetic, Clara should document scenario source/version, and Ethan should identify which assumption is least observable in stress. Then alter the path while keeping the same final macro endpoint. If losses change, explain the path dependency. This shows why stress testing needs time series rather than endpoint shocks alone. List proposed actions, timing, dependencies and side effects. Remove each action in turn and measure resilience difference. Complete the lab with a bridge from scenario input to model output to financial statement to regulatory ratio. Ben should reconcile arithmetic, Clara should document scenario source/version, and Ethan should identify which assumption is least observable in stress. Then alter the path while keeping the same final macro endpoint. If losses change, explain the path dependency. This shows why stress testing needs time series rather than endpoint shocks alone. Start with balance sheet, CET1 and RWA. Project PPNR, provisions, market losses, taxes, distributions and RWA each quarter. Identify first threshold breach. Complete the lab with a bridge from scenario input to model output to financial statement to regulatory ratio. Ben should reconcile arithmetic, Clara should document scenario source/version, and Ethan should identify which assumption is least observable in stress. Then alter the path while keeping the same final macro endpoint. If losses change, explain the path dependency. This shows why stress testing needs time series rather than endpoint shocks alone. Specify GDP, unemployment, property prices, rates and credit spreads for eight quarters. Explain causal narrative and check variables for contradictions. Complete the lab with a bridge from scenario input to model output to financial statement to regulatory ratio. Ben should reconcile arithmetic, Clara should document scenario source/version, and Ethan should identify which assumption is least observable in stress. Then alter the path while keeping the same final macro endpoint. If losses change, explain the path dependency. This shows why stress testing needs time series rather than endpoint shocks alone. Choose CET1 floor and solve combinations of credit loss, RWA inflation and earnings decline required to breach it. Compare with historical and hypothetical scenarios. Complete the lab with a bridge from scenario input to model output to financial statement to regulatory ratio. Ben should reconcile arithmetic, Clara should document scenario source/version, and Ethan should identify which assumption is least observable in stress. Then alter the path while keeping the same final macro endpoint. If losses change, explain the path dependency. This shows why stress testing needs time series rather than endpoint shocks alone. Model deposit run, asset sales, market haircuts and realised losses. Track both cash survival and CET1 through days/weeks. Complete the lab with a bridge from scenario input to model output to financial statement to regulatory ratio. Ben should reconcile arithmetic, Clara should document scenario source/version, and Ethan should identify which assumption is least observable in stress. Then alter the path while keeping the same final macro endpoint. If losses change, explain the path dependency. This shows why stress testing needs time series rather than endpoint shocks alone. List proposed actions, timing, dependencies and side effects. Remove each action in turn and measure resilience difference. Complete the lab with a bridge from scenario input to model output to financial statement to regulatory ratio. Ben should reconcile arithmetic, Clara should document scenario source/version, and Ethan should identify which assumption is least observable in stress. Then alter the path while keeping the same final macro endpoint. If losses change, explain the path dependency. This shows why stress testing needs time series rather than endpoint shocks alone. Start with balance sheet, CET1 and RWA. Project PPNR, provisions, market losses, taxes, distributions and RWA each quarter. Identify first threshold breach. Complete the lab with a bridge from scenario input to model output to financial statement to regulatory ratio. Ben should reconcile arithmetic, Clara should document scenario source/version, and Ethan should identify which assumption is least observable in stress. Then alter the path while keeping the same final macro endpoint. If losses change, explain the path dependency. This shows why stress testing needs time series rather than endpoint shocks alone. Specify GDP, unemployment, property prices, rates and credit spreads for eight quarters. Explain causal narrative and check variables for contradictions. Complete the lab with a bridge from scenario input to model output to financial statement to regulatory ratio. Ben should reconcile arithmetic, Clara should document scenario source/version, and Ethan should identify which assumption is least observable in stress. Then alter the path while keeping the same final macro endpoint. If losses change, explain the path dependency. This shows why stress testing needs time series rather than endpoint shocks alone. Choose CET1 floor and solve combinations of credit loss, RWA inflation and earnings decline required to breach it. Compare with historical and hypothetical scenarios. Complete the lab with a bridge from scenario input to model output to financial statement to regulatory ratio. Ben should reconcile arithmetic, Clara should document scenario source/version, and Ethan should identify which assumption is least observable in stress. Then alter the path while keeping the same final macro endpoint. If losses change, explain the path dependency. This shows why stress testing needs time series rather than endpoint shocks alone.10. Threshold event
11. Scenario expansion
12. Satellite model
13. Credit loss model
14. Point-in-time PD
15. Stressed LGD
16. Stressed EAD
17. Migration model
18. Vintage model
19. Portfolio concentration
20. Collateral stress
21. PPNR
22. Net interest income
23. Deposit beta stress
24. Deposit runoff stress
25. Wholesale funding spread
26. Fee income stress
27. Operating expense stress
28. Market-risk shock
29. Credit-spread shock
30. Yield-curve shock
31. FX shock
32. Equity shock
33. Volatility shock
34. Liquidity shock
35. Margin-call shock
36. Operational-risk shock
37. Climate/transition shock
38. Cyber shock
39. Second-round effect
40. Fire-sale feedback
41. Credit contraction feedback
42. Management action
43. Dividend cut
44. Share issuance
45. Asset sale
46. RWA reduction
47. Deposit repricing
48. Recovery-plan action
49. Capital path
50. RWA path
51. CET1 stress ratio
52. Leverage ratio stress
53. Liquidity survival horizon
54. LCR stress path
55. NSFR stress
56. Model overlay
57. Expert judgement
58. Model validation
59. Benchmark model
60. Scenario challenge
61. Data lineage
62. Model inventory
63. Aggregation
64. Diversification benefit
65. Concentration add-on
66. Scenario probability
67. Narrative consistency
68. Path dependency
69. Nonlinearity
70. Threshold effect
71. Covenant breach
72. Rating downgrade
73. Wrong-way risk
74. Counterparty concentration
75. Recovery rate stress
76. Forecast error
77. Uncertainty band
78. Capital depletion rate
79. Liquidity burn rate
80. Stress loss attribution
81. Risk appetite stress metric
82. Contingency plan
83. Stress testing governance
Worked Example 1: Credit Loss Under Recession
Worked Example 2: CET1 Path
Worked Example 3: Deposit Funding Shock
Worked Example 4: Reverse Stress Threshold
Worked Example 5: Management Action Timing
Worked Example 6: NII Rate Shock
Worked Example 7: Fire Sale
Worked Example 8: Scenario Probability Is Optional, Not Automatic
Worked Example 9: Credit-Market Double Count
Worked Example 10: Liquidity Survival
Stress Testing Should Change Decisions
A Professional Bank Stress-Testing Workflow
Common Failure Modes
1. One-factor sensitivity called stress test
2. Final-year ratio only
3. RWA held constant
4. Management actions assumed immediate
5. Historical worst used as universal stress
6. Scenario variables inconsistent
7. Credit PD stressed without LGD/EAD
8. Funding shock separated from earnings
9. Fire-sale cash counted without loss
10. Diversification assumed from calm periods
11. Model output reported without uncertainty
12. Stress results not linked to decisions
Formula Map
Measure Simplified formula Meaning Stressed expected loss ΣPD_s×LGD_s×EAD_s Scenario credit-loss expectation. PPNR NII+Noninterest income−Operating expense Pre-provision earning capacity. CET1 path CET1_{t-1}+NI_t−Distributions±Adjustments Capital numerator evolution. Stressed CET1 ratio CET1_t/RWA_t Solvency after scenario. Liquidity path Opening liquidity+ΣInflows−ΣOutflows+Actions Cash survival through stress. Reverse stress loss capacity Starting capital−Threshold×Stressed RWA Approx loss to threshold. Authoritative Reference Map
Connected Banking And Finance Mathematics Route
Applied Case Study 1: Severe housing downturn
Applied Case Study 2: Corporate recession
Applied Case Study 3: Bank-specific run
Applied Case Study 4: Rate shock
Applied Case Study 5: Cyber outage
Applied Case Study 6: Counterparty default
Applied Case Study 7: Commercial real estate stress
Applied Case Study 8: Sovereign/FX shock
Applied Case Study 9: Capital raise action
Applied Case Study 10: Fire-sale response
Applied Case Study 11: Reverse stress
Applied Case Study 12: System-wide stress
Final Principle
Deep Practice Lab 1: Build a three-year capital stress
Deep Practice Lab 2: Create a coherent macro scenario
Deep Practice Lab 3: Reverse-stress capital
Deep Practice Lab 4: Run liquidity-capital interaction
Deep Practice Lab 5: Challenge management actions
Deep Practice Lab 6: Build a three-year capital stress
Deep Practice Lab 7: Create a coherent macro scenario
Deep Practice Lab 8: Reverse-stress capital
Deep Practice Lab 9: Run liquidity-capital interaction
Deep Practice Lab 10: Challenge management actions
Deep Practice Lab 11: Build a three-year capital stress
Deep Practice Lab 12: Create a coherent macro scenario
Deep Practice Lab 13: Reverse-stress capital
Deep Practice Lab 14: Run liquidity-capital interaction
Deep Practice Lab 15: Challenge management actions
Deep Practice Lab 16: Build a three-year capital stress
Deep Practice Lab 17: Create a coherent macro scenario
Deep Practice Lab 18: Reverse-stress capital
Deep Practice Lab 19: Run liquidity-capital interaction
Deep Practice Lab 20: Challenge management actions
Deep Practice Lab 21: Build a three-year capital stress
Deep Practice Lab 22: Create a coherent macro scenario
Deep Practice Lab 23: Reverse-stress capital
Deep Practice Lab 24: Run liquidity-capital interaction
Deep Practice Lab 25: Challenge management actions
Deep Practice Lab 26: Build a three-year capital stress
Deep Practice Lab 27: Create a coherent macro scenario
Deep Practice Lab 28: Reverse-stress capital
Deep Practice Lab 29: Run liquidity-capital interaction
Deep Practice Lab 30: Challenge management actions
Deep Practice Lab 31: Build a three-year capital stress
Deep Practice Lab 32: Create a coherent macro scenario
Deep Practice Lab 33: Reverse-stress capital
Deep Practice Lab 34: Run liquidity-capital interaction
Deep Practice Lab 35: Challenge management actions
Deep Practice Lab 36: Build a three-year capital stress
Deep Practice Lab 37: Create a coherent macro scenario
Deep Practice Lab 38: Reverse-stress capital
Deep Practice Lab 39: Run liquidity-capital interaction
Deep Practice Lab 40: Challenge management actions
Deep Practice Lab 41: Build a three-year capital stress
Deep Practice Lab 42: Create a coherent macro scenario
Deep Practice Lab 43: Reverse-stress capital
Deep Practice Lab 44: Run liquidity-capital interaction
Deep Practice Lab 45: Challenge management actions
Deep Practice Lab 46: Build a three-year capital stress
Deep Practice Lab 47: Create a coherent macro scenario
Deep Practice Lab 48: Reverse-stress capital
Deep Practice Lab 49: Run liquidity-capital interaction
Deep Practice Lab 50: Challenge management actions
Deep Practice Lab 51: Build a three-year capital stress
Deep Practice Lab 52: Create a coherent macro scenario
Deep Practice Lab 53: Reverse-stress capital
Deep Practice Lab 54: Run liquidity-capital interaction
Deep Practice Lab 55: Challenge management actions
Deep Practice Lab 56: Build a three-year capital stress
Deep Practice Lab 57: Create a coherent macro scenario
Deep Practice Lab 58: Reverse-stress capital
Deep Practice Lab 59: Run liquidity-capital interaction
Deep Practice Lab 60: Challenge management actions
Deep Practice Lab 61: Build a three-year capital stress
Deep Practice Lab 62: Create a coherent macro scenario
Deep Practice Lab 63: Reverse-stress capital
Deep Practice Lab 64: Run liquidity-capital interaction
Deep Practice Lab 65: Challenge management actions
Deep Practice Lab 66: Build a three-year capital stress
Deep Practice Lab 67: Create a coherent macro scenario
Deep Practice Lab 68: Reverse-stress capital
Deep Practice Lab 69: Run liquidity-capital interaction
Deep Practice Lab 70: Challenge management actions
Deep Practice Lab 71: Build a three-year capital stress
Deep Practice Lab 72: Create a coherent macro scenario
Deep Practice Lab 73: Reverse-stress capital
Deep Practice Lab 74: Run liquidity-capital interaction
Deep Practice Lab 75: Challenge management actions
Deep Practice Lab 76: Build a three-year capital stress
Deep Practice Lab 77: Create a coherent macro scenario
Deep Practice Lab 78: Reverse-stress capital