Scenario Analysis vs Sensitivity Analysis

The distinction, precisely

A scenario is a coherent alternative story — a connected set of assumption changes that would plausibly happen together. An upside case does not just raise revenue growth in isolation; it tells a story (faster category growth, successful premiumisation) that would plausibly lift growth and margin at once.

A sensitivity holds everything else fixed and asks: if I move just this one input, how much does the answer move?

Scenarios tell you what plausible alternative futures look like. Sensitivities tell you which single assumption the result is most exposed to. Neither substitutes for the other.

Three scenarios, defined before they are built

• —Base: revenue growth and EBITDA margin continue on the historical trend.
• —Upside: growth +2.5pp a year, EBITDA margin +1.0pp — the story being faster category growth and successful premiumisation.
• —Downside: growth −3.0pp a year, margin −1.5pp — input-cost inflation outpacing pricing power in a weak discretionary environment.

Note that each case names a story first and a number second. A scenario with no story behind it is just an arbitrarily adjusted base case.

Building the upside case properly

Base-case growth of about 11.0% plus 2.5pp gives 13.5%. Applied to prior-year revenue of ₹682.88 crore, upside revenue is ₹775.07 crore, against ₹757.93 crore in the base case.

Base-case EBITDA margin of about 17.80% plus 1.0pp gives 18.80%. Applied to that higher revenue, upside EBITDA is ₹145.72 crore, against ₹134.87 crore base.

Two assumption changes, applied together and consistently. That consistency is what makes it a scenario rather than an inflated number.

Why a downside case compounds

Should downside EBITDA land below both other cases, or somewhere between them?

Below both — and by more than people expect. The downside applies a lower growth rate (shrinking the revenue base EBITDA is calculated on) and a lower margin (shrinking the share of that smaller revenue that becomes EBITDA). Two unfavourable changes compound: downside revenue around ₹737.51 crore and downside EBITDA around ₹120.21 crore.

A downside case that lands only marginally below base usually means the assumptions were not actually moved together.

A sensitivity table does a different job

Take the base case and vary only revenue growth — say 9%, 11%, 13% — holding margin fixed. Each produces a different EBITDA, and only because of the one input deliberately changed.

That is what makes a sensitivity table worth building: it shows at a glance which single assumption the bottom line is most exposed to, isolated from every other moving part. A two-dimensional version — WACC against terminal growth — is the standard presentation once a valuation exists to test.

One switch, not three spreadsheets

The wrong way to build scenarios is three separate hardcoded files. Fix an error in one and you have to remember to fix it in the other two — and eventually you won’t.

The right way is one model with a scenario switch: a single input cell selecting which assumption set feeds every downstream formula. Change the switch and the whole model recalculates consistently, with no risk of three versions silently drifting apart.

This is a model-control habit, not a presentation choice. It is far easier to build in from the start than to retrofit onto a model that already has three copies in circulation.

Common mistakes

• —Calling a one-input change a “scenario” when it is a sensitivity.
• —Building an upside case by raising revenue alone, with no corresponding margin story.
• —Maintaining base, upside and downside as three separate hardcoded workbooks.
• —Presenting the base case alone, as though a single forecast were a certainty.
• —Hardcoding a scenario’s numbers into formulas instead of driving them from input cells.