Anchor every case to a common model
Scenarios are useful only when they change the same drivers. Start with units, price, retention, gross margin, operating investment, and capital needs. The base case should be the most defensible path, not a midpoint chosen for symmetry. Bull and bear cases then express coherent alternatives rather than optimistic and pessimistic adjectives.
Write the mechanism, not just the number
For each changed assumption, explain the event that causes it and the evidence you would observe. Higher margins may require mix shift or utilization; lower growth may follow customer saturation or a weaker channel. Avoid combining peak growth, peak margin, and low investment unless the business mechanics can support all three at once.
- Keep the forecast horizon and accounting definitions consistent.
- Expose revenue, margin, cash, and dilution assumptions together.
- Assign monitoring signals rather than false probabilities when evidence is thin.
Use scenarios to manage uncertainty
The output is a map of what matters. Record the next event capable of changing the case, the early warning signal, and the assumption most sensitive to new evidence. Updating a scenario should change the model and the written rationale at the same time.
Build scenarios whose numbers do not contradict each other
For a company with $100 million of revenue, a base case might combine 12% customer growth with a 3% price increase to produce roughly 15% growth. A bull case should not merely type in 25%; it should explain how a new channel, lower churn, and better mix can coexist. A bear case should trace weaker acquisition or heavier discounting into revenue and gross margin instead of invoking a vague recession.
Changing revenue while freezing operating expense, working capital, and share count creates an internally inconsistent model. Faster growth may require sales hiring or inventory before revenue arrives, and external financing may dilute per-share value. Every case should use the same formulas and accounting definitions so the difference truly comes from the assumptions.
Attach probabilities only when evidence supports them. More often, define which next-quarter observation in net adds, retention, pricing, or gross margin would move the company out of the base range. The model then becomes a way to process evidence rather than decorate a price target.
- Use the same forecast horizon and valuation method in every case.
- Connect revenue, margin, reinvestment, cash, and dilution assumptions.
- Give each assumption evidence, a monitoring signal, and an invalidation condition.
- After earnings, update the failed assumption before updating the target price.
Worked example: from sales assumptions to operating profit
Consider a fictional business with annual revenue of $100 million. Forecast the next full year, not the next quarter. All amounts in the table are USD millions; these are teaching assumptions, not company guidance or a recommendation. Hold the revenue recognition policy constant. Customer volume changes first, then average price changes on that volume. Operating costs below exclude cost of sales, which is already captured in gross margin.
Revenue = 100 × (1 + volume growth) × (1 + price change). The base case is 100 × 1.12 × 1.03 = 115.36, not 115: growth compounds to 15.36%. Gross profit = revenue × gross margin; operating profit = gross profit − operating costs. Thus the base case produces 115.36 × 50% − 40 = 17.68. Every column uses exactly the same formulas.
The bear case links customer losses and discounting to lower gross margin while operating costs remain sticky. The bull case assumes better demand and product mix, but also higher staffing and selling costs. At base-case revenue, one percentage point of gross margin changes operating profit by 1.1536 million if other assumptions stay fixed. This sensitivity isolates one input; a scenario changes a coherent group of inputs. Neither output is free cash flow or equity value.
| Driver / result | Bear | Base | Bull |
|---|---|---|---|
| Customer volume growth | −5% | +12% | +20% |
| Average price change | −2% | +3% | +5% |
| Revenue | 93.10 | 115.36 | 126.00 |
| Gross margin | 45% | 50% | 52% |
| Gross profit | 41.895 | 57.68 | 65.52 |
| Operating costs | 39.00 | 40.00 | 43.00 |
| Operating profit | 2.895 | 17.68 | 22.52 |
Make the example usable after the next earnings release
Before applying it to a real company, collect the revenue policy and segment disclosures, management discussion, cash-flow statement and risk factors from its filings. Put the filing date, page or section, units and your own assumption beside each model input. After earnings, compare volume, price or mix, margin and costs with the same-period assumptions. If the company does not disclose volume and price separately, mark the split as an estimate rather than manufacturing precision. Explain the deviation before changing the forecast.
This deliberately stops at operating profit: interest, tax, capital spending, working capital, debt and dilution are not modeled. A profitable scenario can still require financing. Add those schedules before estimating cash flow or per-share value, and test a downside worse than the bear case. Do not interpret three scenarios as a confidence interval, assign 25/50/25 probabilities without evidence, or treat their midpoint as a price target. The SEC links below explain source documents; they do not endorse this fictional model or its assumptions.
