Banking Risk

Integrated Stress Testing: Connecting Credit, Market, Liquidity and Operational Risk

By Jonas (Yonas) Mohamed Osman Abdelghafour · August 2026

Executive introduction

The most serious weakness in many stress-testing frameworks is not the absence of sophisticated individual models. It is the absence of interaction between them. A bank can maintain strong credit, market, liquidity, IRRBB and operational models and still underestimate a crisis if each function uses inconsistent scenarios and independent assumptions.

Integrated stress testing begins with one coherent adverse narrative and translates it through macro-financial variables, portfolio risk factors, financial statements, liquidity, capital and management actions. Recent ECB supervisory work has reinforced the importance of consistent scenario transmission and the solvency-liquidity nexus.

Key takeaways

One scenario, several transmission engines

A scenario might combine geopolitical disruption, energy inflation, weaker growth, persistent rates, spread widening and higher cyber activity. The same paths for GDP, inflation, rates, spreads, FX and commodities should feed the relevant risk models rather than being reinterpreted independently by each function.

This creates traceability from narrative to risk factor and from risk factor to P&L, liquidity and capital. Inconsistencies become visible rather than being hidden inside separate submissions.

Credit, market and IRRBB transmission

Credit satellite models can translate macro variables into PD, LGD, EAD, rating migration and expected credit loss. Market models apply rates, spread, equity, FX, commodity and volatility shocks. IRRBB models translate the same rate scenario through deposit behaviour, repricing, optionality and hedges.

The results should reconcile with the balance-sheet and earnings engine. Stress testing is not complete when individual risk numbers are calculated; it is complete when they explain the evolution of the institution's financial position.

Liquidity and operational risk

Market volatility can generate margin calls and haircut changes before accounting losses appear. Capital deterioration can increase wholesale funding costs or affect depositor confidence. These effects should feed the liquidity engine rather than being assessed after solvency is finalised.

Operational scenarios can add cyberattack, payment disruption, third-party failure, fraud or legal cost where these are causally connected to the stress. The purpose is not to combine every risk in every scenario, but to capture interactions that can materially amplify loss.

Management actions and feedback

Capital raising, asset sales, deposit repricing, lending reductions and cost actions all have timing and market-capacity limits. If many institutions attempt the same action, execution may become more expensive or impossible. Stress frameworks should therefore use constrained rather than automatic management actions.

Second-round effects can be represented through amplification scenarios when full system modelling is impractical. This is preferable to assuming away feedback entirely.

Technical framework

A useful architecture has eight layers: scenario engine, risk-factor translation, portfolio models, balance-sheet engine, P&L engine, liquidity engine, capital engine and management-action layer. Reverse stress adds a search process for scenarios where capital or liquidity crosses a critical threshold. Governance should define ownership and reconciliation across every layer.

Practical example

In an energy-shock scenario, higher inflation and persistent rates increase corporate defaults and deposit competition. Credit losses reduce earnings and capital, spread widening creates valuation losses, margin calls consume liquidity and weaker capital raises funding costs. An integrated model captures this chain; siloed models can miss the feedback.

What risk leaders should do now

  1. Use one centrally governed scenario library with consistent macro-financial variables.
  2. Reconcile credit, market, IRRBB, liquidity and operational outputs through the financial statements.
  3. Integrate solvency and liquidity feedback.
  4. Constrain management actions for timing, capacity and execution risk.
  5. Use reverse stress testing to identify structural failure points.
  6. Connect results directly to ICAAP, ILAAP, recovery planning, risk appetite and strategic planning.

Frequently asked questions

Does integrated stress testing require one giant model?

No. It requires coherent interfaces and assumptions across specialised models, not a single monolithic engine.

Why are management actions often overstated?

Because models can assume actions are immediate and unlimited even when markets, operations or regulation constrain execution.

What is the main value of reverse stress testing?

It identifies combinations of risks that can breach a critical outcome even when those combinations were not selected in the standard scenario set.

Conclusion

Integrated stress testing is valuable when it prevents organisational boundaries from becoming modelling boundaries. Credit, capital, liquidity, market conditions and operations interact during real stress and should interact in the scenario architecture.

The objective is not perfect crisis prediction. It is a sufficiently coherent representation of transmission and feedback to improve capital, liquidity and management decisions before stress occurs.

About the author

Jonas (Yonas) Mohamed Osman Abdelghafour writes about financial risk management, quantitative modelling, actuarial science, banking risk, insurance risk, capital modelling, model validation, climate risk and geopolitical risk. His work focuses on translating complex quantitative and regulatory risk issues into practical frameworks for financial institutions. Author profile.