In an interconnected financial system, shocks seldom respect national borders. A liquidity squeeze in one market can ricochet through funding channels, asset prices, and confidence across continents. To anticipate and withstand such contagion, regulators mandate stress tests – structured “what-if” exercises that quantify how banks would fare under severe but plausible scenarios.
While the goal is shared globally – that of resilience – the frameworks differ by jurisdiction. This essay compares three influential regimes: the United States’ Comprehensive Capital Analysis and Review (CCAR), the European Union’s EU-Wide Stress Test (EWST), and the globally adopted, regulator-led Internal Capital Adequacy Assessment Process (ICAAP) under Basel’s Pillar 2.
Foundations of Financial Stress Testing Models
Different methodological approaches have evolved to address the varying objectives of stress testing. A top-down approach is typically regulator-led, where common macroeconomic scenarios are applied centrally across institutions to provide a sector-wide perspective.
In contrast, a bottom-up approach relies on institution-specific models and assumptions, allowing each firm to reflect its unique portfolio characteristics and risk exposures. Complementing these, sensitivity analysis isolates the effects of individual shocks – such as an interest rate hike or a sharp decline in asset prices – to test the volatility of specific risk factors.
Meanwhile, scenario analysis integrates multiple shocks into coherent, multi-variable simulations, offering a more holistic view of how complex interactions might impact both individual firms and the broader financial system.
Comparative Regulatory Practices in Shock Simulation
CCAR (U.S.): Capital Planning Under Fire
CCAR, administered by the Federal Reserve, evaluates large U.S. bank holding companies’ ability to maintain minimum capital ratios while executing their planned capital actions (dividends, share repurchases) under supervisory scenarios. It blends quantitative loss projections with a qualitative review of governance, data, and model risk controls.
- Scenario architecture
The Fed publishes a supervisory severely adverse scenario (and others), specifying trajectories for GDP, unemployment, inflation, interest rates, equity and real estate prices, and certain market shocks. The horizon typically spans nine quarters, allowing a full credit cycle to materialize in the projections. The Fed also imposes market-shock components for trading and counterparty exposures on banks with material trading activity.
- Methodology
Banks project pre-provision net revenue (PPNR), loan losses by portfolio, provisions, risk-weighted assets (RWAs), and capital ratios across the horizon. The Fed independently runs its own models on common data templates, often producing more conservative loss estimates. A distinctive feature is the Stress Capital Buffer (SCB), under which the largest drop in CET1 ratio projected in the Fed’s stress scenario (plus dividends) is translated directly into the bank’s capital buffer requirement. This makes stress testing directly binding for capital distributions.
- Governance and transparency
CCAR places heavy emphasis on model risk management (documentation, back-testing, challenger models), internal controls, and board oversight of the capital plan. Public disclosure is granular, including revenue and loss components and post-stress capital ratios, which can influence market discipline.
- Strengths and trade-offs
CCAR’s rigor and binding SCB align incentives, but reliance on supervisory models can create model opacity for firms and procyclicality if scenarios toughen quickly following expansions. As far model opacity goes, supervisory models (used by the Fed rather than the banks’ own) are black boxes to firms. Banks do not fully know which risk drivers or assumptions push their capital requirements higher, which limits their ability to anticipate or contest results. On procyclicality risks, because stress scenarios are designed to reflect a severe turn in the cycle, they often become harsher after periods of expansion (e.g., sharper unemployment shocks or steeper GDP contractions). This can compel banks to raise capital or restrict lending just as conditions begin to weaken, thereby amplifying the downturn.
The reliance on a nine-quarter horizon and fixed scenario raises concerns about under-representation of slow-moving structural risks. Climate change, for instance, is unlikely to manifest its systemic implications within such a restricted timeframe.
EWST (EU): Consistency and System-Level Transparency
The EU-Wide Stress Test, coordinated by the European Banking Authority (EBA) with scenarios developed by the European Systemic Risk Board (ESRB) and execution overseen by the ECB/SSM (European Central Bank/Single Supervisory Mechanism) for significant institutions, targets comparability across banks and countries. It assesses capital adequacy under a common adverse scenario without being linked to a capital buffer like the SCB (though regulators can use results to set Pillar 2 Guidance).
Note: In contrast to the United States, where the Federal Reserve both designs and administers the CCAR stress tests, the European framework is bifurcated: the EBA develops the common methodology and coordinates EU-wide exercises, while the ECB, through the Single Supervisory Mechanism, applies these standards in practice by directly supervising significant euro area banks and incorporating stress test outcomes into its Supervisory Review and Evaluation Process (SREP).
- Scenario architecture
The EWST adverse scenario is typically a three-year macro-financial path, reflecting EU-specific risks: euro area demand shocks, energy price spikes, sovereign spread widening, and property market corrections. Scenario variables span macroeconomy, markets, and real estate, with country-specific calibrations to capture heterogeneity across member states.
- Methodology
Losses, revenues, risk-weighted assets, and capital are projected by banks within detailed templates designed to standardize outcomes. To enhance comparability, the EBA stipulates methodological guidelines, hurdle rates, and, in certain cases, caps or floors on elasticities. While this promotes consistency, it also raises questions about whether structural differences across banks are fully captured.The EBA does not just give banks a scenario and say, “go run your models”. It provides detailed instructions on how banks should project losses, income, risk-weighted assets, etc. This ensures all banks are following the same playbook, rather than using their own (possibly optimistic) methods.Hurdle rates are the minimum capital thresholds banks must stay above during the stress test. For example, a bank may need to show it can keep its CET1 ratio above, say, 8% in the adverse scenario. Falling below the hurdle rate signals weakness and could trigger supervisory pressure or restrictions.
Elasticities here refer to sensitivities in models, like how credit losses respond to GDP decline, or how interest income reacts to rate changes. To prevent banks from using overly favorable assumptions, the EBA sometimes sets floors (minimum responsiveness) or caps (maximum responsiveness). For example, if GDP falls 5%, banks must assume at least a certain percentage increase in default rates.
For market-risk banks, a granular market shock is included. The ECB may apply top-down models to challenge bank results, but the framework is primarily constrained bottom-up. The constrained part refers to the fact that this bottom-up modeling is not free-form — it’s subject to methodological guidelines, hurdle rates, caps/floors.
Banks with significant trading books (market-risk banks) face not just credit and macro shocks but also detailed shocks to financial markets. These include shocks to interest rates, equity prices, credit spreads, FX rates, and commodities. Granular means it’s not one blanket shock — the scenario specifies many different shocks across asset classes and regions.
While banks submit their own bottom-up projections (using their internal models and data), the ECB runs top-down benchmark models. These are supervisory models that estimate what should happen given the scenario. If a bank’s projections look too optimistic (e.g., low trading losses compared to peers), the ECB can challenge or adjust the results.
- Governance and disclosure
Public transparency is an EWST hallmark – the EBA publishes bank-by-bank outcomes and rich data annexes, enabling peer benchmarking by investors and analysts. While not tied to an automatic buffer, regulators can translate findings into Pillar 2 Guidance (P2G) and qualitative remediation.
- Strengths and trade-offs
EWST delivers clear value in terms of cross-system comparability and disclosure, but the burden of its template-driven framework remains considerable. Its outcomes do not directly translate into capital buffer requirements, which can limit bindingness; however, the use of P2G and SREP linkages by supervisors effectively maintains its regulatory bite.
ICAAP (Basel Pillar 2): Firm-Specific Resilience by Design
The Internal Capital Adequacy Assessment Process is a Basel Pillar 2 requirement implemented by national regulators (e.g., PRA in the UK, MAS in Singapore, RBI in India). ICAAP’s core idea is bank-specific assessment: each firm identifies material risks (credit, market, IRRBB (Interest Rate Risk in the Banking Book), liquidity, conduct, model risk, operational, climate), quantifies them under baseline and stress, and demonstrates capital (and often liquidity) adequacy aligned with its risk appetite and business model.
- Scenario architecture
Unlike CCAR/EWST, ICAAP scenarios are bespoke (custom-designed stress scenarios tailored to a specific bank, market, or risk profile) and multi-lens: macro downturns, idiosyncratic events (counterparty default, cyberattack), reverse stress tests (what breaks the firm), and emerging risks (climate transition/physical risks). Horizons can be longer (often three to five years), matching strategic and funding cycles, with satellite models for impairments, margin compression, and RWA dynamics.Note: Regulators and practitioners use the term “bespoke” rather than “custom” to emphasize that such scenarios are carefully tailored to a bank’s unique risk profile and formally embedded in supervisory practice, rather than merely user-defined adjustments.
- Methodology
ICAAP integrates risk identification, quantification (economic capital and regulatory capital views), and capital planning that includes credible management actions (balance-sheet shrinkage, hedging, dividend cuts). Regulators review the ICAAP through the Supervisory Review and Evaluation Process (SREP), setting Pillar 2 Requirements (P2R) and P2G. Many jurisdictions also require an ILAAP for liquidity, ensuring alignment between capital and funding resilience.
- Governance and disclosure
ICAAP is fundamentally a governance document – it tests whether the board understands its risk profile, validates models, and embeds stress results into strategy. Public disclosure is lighter than EWST; the emphasis is on supervisory dialogue and remedial action plans.
- Strengths and trade-offs
ICAAP’s flexibility captures firm-specific risks and strategic choices, but variability in methods and data quality can hinder cross-bank comparability. The framework’s success is therefore contingent on the rigor of supervisory engagement and the strength of internal risk governance.
Where the Regimes Are Converging and Diverging
Convergence is evident in stronger model risk management, richer PD/LGD/EAD dynamics through the cycle, and explicit capital planning linkages. Divergence persists in bindingness (SCB vs. guidance), horizon length, and flexibility (template rigor vs. bespoke realism). The optimal mix may be hybrid: a common, transparent system-wide test for comparability (EWST-style), a binding capital buffer to sharpen incentives (CCAR-style), and a bank-specific ICAAP to capture idiosyncratic and strategic risks beyond templates.
Final Remarks
Stress testing has matured from a crisis-era diagnostic into a central pillar of prudential policy and bank management. CCAR enforces discipline through a binding stress buffer; EWST delivers comparability and market transparency across a heterogenous banking union; ICAAP ensures that firm-specific risks and strategies are front and center.
In a world where shocks transcend borders swiftly, robust stress testing must combine these strengths – supervisory consistency, binding capital consequences, and bespoke, forward-looking analysis. Designed well, stress testing does more than estimate potential losses – it pre-positions resilience, shifting the exercise from a hypothetical “what if” to an institutional state of readiness.
Additional Information
For more information on SAS Stress Testing visit the software information page here. For more information on curated learnings paths on SAS Solutions and SAS Viya, visit the SAS Training page. You can also browse the catalog of SAS courses here.
Find more articles from SAS Global Enablement and Learning here.
