Sensitivity analysis

What it means

Sensitivity analysis examines how much the result of a plan, budget or valuation changes when one assumption changes and the rest stay the same. It answers what happens to profit if the price drops five percent, or to a project’s value if interest rises by a point.

It is most often used for investment decisions. Inputs such as price, sales volume, material costs, the interest rate or the discount rate are varied to see how NPV, IRR or a DCF valuation responds.

The result is usually shown in a table or a tornado chart ranking assumptions by how strongly they move the outcome. Unlike scenario analysis, which changes several assumptions at once, it isolates the effect of each one.

How it works in practice

For a forecast or loan application we show clients which three or four assumptions decide whether the plan works. Often it is not revenue but margin, customer payment terms or the price of energy.

It also helps to calculate a break-point for individual inputs, meaning how far an assumption can worsen before the project stops being profitable. This extends the classic break-even to several variables, and banks like to see it in a feasibility study.

For companies trading in foreign currency we include exchange rate sensitivity: what exchange difference and margin effect a move of a few percent would cause. The result decides whether hedging the currency risk is worthwhile.

What to watch out for

Sensitivity analysis assumes only one thing changes, but in reality assumptions interact: a lower price raises volume, higher inflation pushes up both wages and prices. The results should therefore be read as a guide and combined with scenarios.

The range tested is a risk too. If only a one-percent move is tested, the analysis is reassuring even though real swings are often ten times larger. We therefore derive the range from the company’s and the market’s history, not from what looks comfortable.