Build three-scenario financial models with probability weighting, expected value, and tornado sensitivity charts.
Scenario analysis is a way of making a single-number forecast honest: instead of one projection, you run the model at three settings - best, base, and worst. This tool holds the three scenarios side by side, applies the probability you assign to each, and returns the probability-weighted expected value together with the spread of outcomes. A tornado chart then ranks which assumption - price, volume, cost - moves the result the most. The thing users get wrong first: the expected value rarely equals the base case. It lands wherever the weights pull it, often below base, because downside scenarios get weighted as heavily as upside ones.
A concrete input and expected output from the current implementation.
Input
Best case: 20% probability, profit $130,000. Base case: 55% probability, profit $80,000. Worst case: 25% probability, profit $35,000.
Expected output
Scenario outcomes: Best $130,000 (20%), Base $80,000 (55%), Worst $35,000 (25%). Probability-weighted expected value: 0.20 x 130,000 + 0.55 x 80,000 + 0.25 x 35,000 = $78,750. The weighted result lands $1,250 below the base case of $80,000.
The expected value is the mean of the three outcomes weighted by their probabilities - a weighted average, not the most likely result. The best case swings further from base (+$50,000) than the worst case (-$45,000), but the worst case carries more weight (25% versus 20%), so the weighted mean settles below the base case.
Each assumption or outcome is defined three times - best, base, and worst - and each scenario gets a probability. The expected value is the sum of probability times outcome across the three scenarios, a weighted average of the range you entered. For the weighted average to be meaningful the probabilities must sum to 100%; a best case weighted 20% contributes exactly one fifth of its outcome.
The tornado chart measures each input driver one at a time. A driver is set to its worst value while every other driver stays at base, and the resulting output is recorded; the same driver is then set to its best value. The distance between those two outputs is that driver's range of influence, drawn as a horizontal bar with the largest swing on top. Drivers near the top dominate the result.
They should match your honest judgment, because the expected value is only as good as its inputs. The EV is a summary, not a safety margin. If the worst case would hurt the business, look at the worst-case outcome directly and add a contingency there, rather than quietly inflating the worst-case probability to make the answer look prudent.
Indirectly, yes - compare the spread. One project may swing between break-even and a windfall while the other barely moves, even with identical expected values, because the expected value hides the distribution behind it. The tornado chart adds a second signal: a project whose result rides on one fragile assumption is riskier than one driven by many.
The example and behavioral notes were checked against the browser implementation. Standards and primary references below define the relevant format, formula, or platform behavior.
Calculate gross, operating, and net profit margins from cost and revenue with breakeven analysis.
Convert between annual salary and hourly rate with adjustments for work hours, overtime, and benefits.
Calculate sales tax amounts, tax-inclusive and tax-exclusive prices for any tax rate or jurisdiction.
Calculate return on investment from initial cost and final value with annualized ROI and comparison tables.
Convert between markup and margin percentages, calculate selling price, and compare pricing strategies.
Convert numeric amounts to written words for invoices, checks, and legal documents in multiple currencies.