Reverse stress testing
Definition
A stress testing approach that starts from a defined adverse outcome — such as a breach of a regulatory threshold, a capital shortfall, or business model failure — and works backwards to identify the scenarios or combinations of assumptions that would cause that outcome to materialise.
What This Actually Means
Standard stress testing asks: if rates move by X, what happens to our EVE and NII? Reverse stress testing inverts the question: what combination of rate moves and assumption changes would cause a defined threshold to be breached?
For IRRBB, reverse stress testing is more appropriately framed around breaching a risk appetite level than around breaking the bank entirely — IRRBB in isolation is rarely sufficient to cause institutional failure, but it can readily breach NII or EVE risk appetite limits if the wrong combination of rate and assumption shifts occurs simultaneously. The objective is to find the most plausible scenario that breaches that threshold — not the most severe scenario, but the most credible one. In practice, this is typically a combination of a rate shift and a shift in deposit beta, since these are the two variables with the most direct and correlated impact on NII.
Where It Matters
The EVE SOT limitation: the EVE SOT threshold can be breached with relatively small shifts in deposit behavioural assumptions — a modest change in NMD repricing profile or deposit volume can move EVE sensitivity materially without any change in rates. This makes the SOT a less useful target for reverse stress testing in two respects: it is too easy to breach through assumption changes alone, and it is a gone concern metric that does not reflect how a bank actually manages its balance sheet on a going concern basis. Reverse stress tests are more informative when focused on NII and earnings risk appetite limits, which are the metrics management can act on.
The most plausible failure pathway: the genuinely dangerous scenario for most retail banks is not a single extreme rate shock but a combination of a meaningful rate move and a beta shift — rates move by a moderate amount while competitive dynamics force more pass-through than the model assumed, compressing NII beyond what the structural hedge can offset. This compound scenario is more plausible than a 500bp rate shock and more likely to catch a bank off guard because neither element individually would breach a limit.