Sensitivity Analysis explained: this guide covers what it means, who it applies to, the step-by-step process, documents required, fees, due dates and penalties in India — so you can stay compliant with confidence and avoid costly mistakes.
Reports present one number because reporting requires one number. That should never be read as suggesting value can be determined with absolute precision.
The problem with "Rs. 847 crore"
A conclusion stated as a single figure may convey a degree of certainty that does not reflect the underlying valuation assumptions. In reality, valuation outcomes are often highly dependent on a limited number of key assumptions, and reasonable — not extreme — changes in those assumptions may produce materially different results.
Presenting a single conclusion without acknowledging the assumptions it rests on is the precision illusion: an impression of exactness that does not reflect economic reality.
Where uncertainty comes from
Uncertainty arises whenever future outcomes are unknown, information is incomplete or professional judgement is required. In insolvency assignments it is amplified by:
- uncertain business continuity;
- distressed market conditions;
- incomplete or evolving information;
- uncertain asset recoverability; and
- compressed timelines for completing the valuation.
The existence of uncertainty does not undermine a valuation conclusion. Material uncertainties should, however, be identified, evaluated and appropriately communicated within the report.
The three things sensitivity analysis does
- It demonstrates that the conclusion has been stress-tested and is not dependent on a single set of assumptions.
- It identifies and quantifies the impact of key value drivers, enabling users to distinguish between assumptions that are critical and those that are relatively robust.
- It anticipates one of the most common questions raised during reviews, inspections and disputes: "What happens to the valuation if a key assumption changes?"
Sensitivity analysis is normally justified as a way of showing how much the answer could move. The more useful function is the opposite: it identifies the assumptions that do not matter.
A report that shows a conclusion barely responds to a contested input has removed that input from the argument. A challenge to an assumption that moves value by a fraction of a percent is answerable from within the report itself.
Without the analysis, every assumption in the report looks equally consequential — which means every assumption is equally available to challenge, and the valuer must defend all of them rather than the two or three that actually drive the conclusion.
What is sensitive, by approach
| Approach | Typically sensitive to |
|---|---|
| Income (DCF) | Revenue growth, EBITDA margins, discount rates, terminal growth assumptions |
| Liquidation value | Asset realisation values, liquidation haircuts, sale horizon assumptions |
Sensitivity analysis enables these impacts to be quantified and disclosed transparently rather than left for a reader to infer.
Why challenges land on assumptions, not arithmetic
Valuation challenges often focus on assumptions rather than calculations. Nobody disputes that a discounted cash flow model computes what it computes. What is disputed is the 12% growth rate and the 16.2% WACC that were fed into it — which is exactly what a sensitivity table addresses.
As a defensibility tool
By quantifying the impact of changes in key assumptions, sensitivity analysis enables stakeholders to identify which assumptions are truly critical to value and to anticipate how the conclusion may change if those assumptions prove inaccurate.
That is a materially stronger position than a single figure defended after the fact. It also connects directly to assumption disclosure: assumptions that are both material and uncertain require higher justification, disclosure and, where appropriate, sensitivity analysis.
Common mistakes
- Presenting a single figure where the conclusion turns on two or three assumptions.
- Treating sensitivity analysis as an appendix rather than part of the conclusion.
- Testing inputs that do not move value and calling the conclusion robust.
- Leaving the reader to infer how much the answer could change.