An isolated change is useful as a diagnostic
In a sensitivity analysis, hold cash flow constant and change the rate. The difference shows the mathematical effect of the rate, but not every economic consequence of a change in conditions. If expensive capital also reduces demand and requires investment to be revised, the actual scenario will affect several lines. A one-parameter table should not be called a complete stress test.
Define the model output: operating-asset value, owners’ equity or the price per share. The relative change in equity can be much larger because of debt. Without a label, a reader may mistake stable EV for stable share value. If the number of shares changes, recalculate the denominator rather than leaving the old ownership fraction.
A small matrix without false probabilities
Consider a mature fictional business whose next-year FCFF is 12 million roubles. In this simplified perpetuity model, at a 10% rate and 2% growth, value is 150 million. At the same rate and 4% growth, it is 200 million. At a 12% rate, the corresponding values are 120 million and 150 million.
The table shows high dependence on the gap between rate and growth. It does not say that the four combinations are equally likely or that a future price must fall between 120 and 200 million. Fixed FCFF with changing growth is also a diagnostic simplification: a full scenario must align reinvestment and future returns.
A scenario starts with events
Write a chain: a competitor appears and lowers price; the company changes retention spending; capacity utilization falls; investment is partly delayed. Then translate the chain into volume, margin, working capital and timing. A bad scenario is not simply a 10% reduction in every row: some costs may rise, while investment may not fall quickly enough.
A conditional base scenario may require 15 million of reinvestment, while a stress scenario requires 10 million. Reducing investment does not necessarily soften the problem permanently: worn assets will require replacement in later periods. Research on value creation links growth with resources and returns; a scenario must preserve that causality.[97](#source-cv-value-creation)
How to find genuinely important unknowns
Compare the size of the result change with the reliability of the inputs. A material parameter that is barely disclosed deserves research priority. A precise rate cannot repair an uncertain path to profitability. Conversely, a well-supported operating base does not make the choice of long-term regime irrelevant.
Check boundary conditions separately: negative cash flow, a need for new financing, growth no higher than the rate and an inability to make payments by the selected date. An invalid region of the formula must not be placed in a polished heat map as an ordinary value. Error or inapplicability should be stated in words rather than disguised as a normal negative value for a continuing business.
What to show the reader
Present the original model, the parameter that changes, the conditions of its range and the assumptions that remain fixed. For scenarios, add a short description of events and links between them. A probability-weighted average is appropriate only with a separate probability rationale; subjective weights should be labelled as such.
The review may conclude that the data are insufficient for a narrow valuation. That is useful if it shows which disclosure could change the conclusion. Sensitivity is a map of mathematical dependence; a scenario is a map of economic conditions. Neither is a quote forecast, a proven loss limit or a guarantee that an investment will be profitable.
Reduce the risk of invisible fitting
Save the original parameter bounds before viewing the final valuation. If a range appears only after seeing a desired price, the researcher may unconsciously exclude uncomfortable combinations. Changing bounds is acceptable when new information arrives, but the reason should remain in the log so another reader can see the decision history.
For each scenario, check physical feasibility: is there enough production capacity, staff, material and launch time? A spreadsheet can easily allow sales that the company cannot yet deliver. Then check financial feasibility: is capital available before receipts arrive? If the scenario requires an additional issuance, reflect its effect on owners’ results.
These checks do not provide probabilities, but they separate an internally coherent story from a convenient set of numbers. That is what makes scenario work more substantive than merely stretching a sensitivity table’s boundaries.
Related foundations
[Discounted cash flow: a model with open assumptions](/guides/discounted-cash-flow/); [Portfolio risk: losses, timing and joint asset movements](/guides/portfolio-risk/).