Scenario Analysis vs Sensitivity Analysis

Sensitivity analysis answers 'how much does the output move if this one input changes?' — for example, how IRR responds to the exit multiple, holding growth and leverage fixed. It isolates the model's sensitivity to a single driver. Scenario analysis answers 'what happens under a realistic combination of conditions?' — a recession bear case might pair lower revenue growth with margin compression, a higher discount rate, and a lower exit multiple all at once, because in the real world those move together. Sensitivity is ceteris paribus (all else equal); scenario deliberately violates ceteris paribus to tell a consistent story.

Sensitivity analysis is typically built with an Excel Data Table (Data > What-If Analysis > Data Table), which produces a grid showing the output across a range of one input (one-variable) or two inputs (two-variable, the classic IRR-by-entry-and-exit-multiple grid in an LBO). Scenario analysis is usually built with a CHOOSE or INDEX function tied to a scenario-selector cell: you toggle a single cell to '1' (base), '2' (bull), or '3' (bear), and CHOOSE pulls the matching set of assumptions into the live model. Excel also has a built-in Scenario Manager, but bankers almost always use a CHOOSE/INDEX switch instead because it's transparent and auditable.

Use sensitivity analysis to identify which assumptions matter most — if IRR barely moves when you flex growth but swings 800bps on the exit multiple, you know where the value (and risk) really lives, and where to focus diligence. Use scenario analysis to communicate downside and upside to an investment committee in a digestible, story-driven way: 'in our bear case, with a recession, the deal still returns a 12% IRR.' In practice, models contain both: a scenario switch for the headline cases, plus sensitivity tables on the one or two outputs everyone cares about.