How do you carry out scenario and sensitivity analysis?
A plan built on a single number prepares for a future that will not happen. Scenario and sensitivity analysis show how dependent the plan is on each assumption and where the company will struggle if things go wrong.
Every financial plan rests on a set of assumptions about the future, and none of them is certain. Sensitivity analysis measures how the result changes when a single assumption changes. Scenario analysis compares alternative futures in which several assumptions change together within a consistent story.
The aim of both is not to predict the future but to make the plan’s fragile points visible.
Sensitivity analysis
In sensitivity analysis, one assumption is changed while everything else is held constant, and the effect on the result is measured. For example:
| Assumption changed | Example change | Effect on profit | Effect on cash |
|---|---|---|---|
| Sales volume | −10% | Falls by the lost contribution | Partly offset by lower receivables and inventory |
| Sales price | −5% | Falls by the full price difference | Close to the profit effect |
| Raw material cost | +10% | Gross margin narrows | Profit effect plus higher inventory value |
| Collection period | +15 days | None | Cash tied up in higher receivables |
| Interest rate | +5 points | Interest expense rises | Interest payments rise |
Once each row is filled in with numbers from the model, the table shows which assumption the result is most sensitive to. In most companies, price and gross margin assumptions have a bigger effect than volume; the collection period affects cash, not profit. Management’s attention should go to the assumptions with the greatest effect.
Scenario analysis
In real life, assumptions change together, not one at a time. When demand falls, prices come under pressure, customers pay later and inventory moves slowly. Scenario analysis captures that combination.
A common structure has three scenarios:
- Base case: the future management considers most likely.
- Downside case: reasonably worse conditions (lower demand, higher costs and slower collections together).
- Stress case: severe, low-probability conditions that test the company’s resilience.
An upside case can be added, but the real information value for management lies in the downside and stress cases: under which conditions does the company fall below its cash threshold, and under which does it breach the financial covenants in its loan agreements?
Making scenarios meaningful
- Start with a story, then the numbers. Instead of “sales fall 15%”, write a story such as “demand shrinks in our main customer group” and work out how it affects each assumption.
- Always show cash. If scenarios are judged only on profit, the most critical information is lost. The model should produce the income statement, balance sheet and cash flow together; I explain why in Why does a business plan need three statements?
- Define triggers. Which indicator reaching which level signals that the company has entered the downside case?
- Plan actions in advance. If the downside case materialises, which spending is postponed and which source of financing is used?
The last two points turn scenario analysis from a report into a management tool.
Break-even analysis
Another practical and powerful method is to look for the breaking point: how far can sales fall before the company drops below its cash threshold? How far can interest rates rise before the debt service coverage ratio falls below 1? The answers tell management directly how much margin for error it has.
You can see how scenarios run alongside the budget and forecast in the Budget & Forecast Management Model and the Five-Year Business Plan.
Frequently asked questions
How many scenarios should be prepared?
Three are enough for most companies. The more scenarios there are, the more management’s attention is spread; what matters is that they differ meaningfully from each other.
What is the difference between scenario analysis and Monte Carlo simulation?
Scenario analysis compares a few consistent alternative futures. Monte Carlo simulation assigns probability distributions to assumptions and calculates thousands of outcomes. The latter is more detailed but needs reliable information about the distributions; for most management decisions, well-built scenarios are enough.