Reverse Stress Testing
Instead of picking a scenario and seeing how bad the losses are, reverse stress testing starts from a defined failure and works backward to find what scenario would cause it.
Prerequisites: Supervisory Stress Tests: CCAR and DFAST, Stressed VaR
A conventional stress test picks a scenario — a recession, a rate shock, a market crash — and asks "how much would we lose?" That's useful, but it only ever tests the scenarios someone thought to write down. Reverse stress testing flips the question: start from a defined bad outcome, such as the firm running out of capital or becoming unable to fund itself, and work backward to find what combination of events would actually cause it.
Reverse stress testing starts from the failure and searches backward for the scenario, instead of starting from a scenario and calculating the failure — which is the only way to find the dangerous scenarios nobody thought to test directly.
Working backward instead of forward
The process usually begins by defining the failure point precisely — for a bank, that might be CET1 capital falling below the regulatory minimum, or liquid assets running out under a funding stress. Analysts then search across combinations of shocks (market moves, credit losses, funding cost spikes, operational failures) to identify which combinations are severe enough to breach that point, and which of those combinations are actually plausible given the firm's specific business mix, rather than generic. The output isn't a single number — it's a narrative: "a 25% property price fall combined with a 150 basis point funding spread widening would exhaust our capital buffer," for instance.
Worked example
A mid-size bank defines its failure point as liquid assets falling short of net outflows over a 30-day stress window. Working backward, the analysis finds that a scenario combining a two-notch credit rating downgrade, a 20% deposit outflow from its largest corporate clients, and a freeze in short-term wholesale funding would exhaust liquid assets in 11 days. None of those three shocks alone would breach the threshold — a rating downgrade alone costs the bank a manageable amount of funding — but the combination, which regulators had not specified in the standard CCAR scenario, is identified as the bank's actual vulnerability.
What this means in practice
Reverse stress testing is most valuable precisely where forward scenario testing is weakest: it surfaces idiosyncratic vulnerabilities specific to a firm's business mix — concentration in one client, one funding source, one region — that a generic, industry-wide scenario would never target. Regulators increasingly require it as a complement to standard stress tests, not a replacement, because the two answer different questions.
Reverse stress testing can surface a scenario and then have the firm dismiss it as "too implausible to matter." The value of the exercise depends entirely on taking the identified scenario seriously even when it doesn't resemble any scenario regulators or history have handed the firm before.
Further reading
- Bank of England, 'Guidance on Reverse Stress Testing'