Where stress-testing applies generic shocks, scenario analysis tells stories with numbers. What happens to this portfolio if the central bank surprises with 50 basis points, if the sector leader guides down 20%, if the exchange halts withdrawals for a week?
Build scenarios from your actual exposures rather than from headlines. A book of three regional banks and two homebuilders has one dominant scenario - a rate shock - and the analysis should price it properly across all five positions at once, including the correlation that will show up on that day. Historical episodes make good templates because the co-movements are real rather than imagined.
Its main value is prospective. Deciding in advance what you would do at each stage of a scenario converts a panic into a procedure, and most of the damage in real events comes from decisions made without one.
Related: stress-testing, correlation-breakdown, tail-hedge, worst-case-loss