A stress test applies deliberately adverse market, liquidity, correlation, or operational assumptions to a position or portfolio to examine how losses and controls behave outside normal conditions.
How it works
Scenarios can include price gaps, wider spreads, failed stop execution, correlation spikes, liquidity deterioration, or temporary platform unavailability. Each scenario should state its assumptions rather than pretending to be a precise forecast.
The stressed outcome is compared with normal planned risk and account capacity. If an adverse but plausible scenario creates intolerable loss, the pre-trade response may be smaller size, different instruments, diversification changes, or contingency procedures.
Why it matters
Normal stop distance and position size often assume continuous prices, usable liquidity, functioning infrastructure, and diversification that may weaken during stress.
Stress testing does not predict the next crisis. It makes hidden dependencies visible before capital is exposed.
A simple market example
A trade planned to lose 200 is re-evaluated under a price gap, three-times-wider spread, delayed stop fill, and correlated loss in another position. The purpose is to see whether the account still tolerates the combined failure modes.
Common mistakes
Treating a stress scenario as a forecast with a claimed probability it was never designed to estimate.
Using only historical average volatility and ignoring liquidity or operational failure modes.
Frequently asked questions
How is a stress test different from a stop loss?
A stop is an execution/risk rule for a trade; a stress test asks what happens if normal assumptions around that rule fail.
Should shocks be combined?
Combination scenarios can be useful when risks may occur together, but assumptions and double-counting should be explicit.
What should happen after a failed stress test?
Reduce exposure, change the product or controls, add contingencies, or reject the trade before execution.
Educational content only. Definitions describe common market usage and may vary by jurisdiction, instrument, or institution.