A single DCF output is not a valuation — it is an artifact of specific assumptions. McKinsey insists that every valuation must include sensitivity analysis, scenario analysis, and ideally Monte Carlo simulation to understand how the value conclusion responds to key assumption changes. The goal is not to produce a number but to understand the shape of value: what drives it, what kills it, and where the critical uncertainty lies.
| WACC \ g | 1.5% | 2% | 2.5% | 3% | 3.5% |
|---|---|---|---|---|---|
| 9.0% | $1.84B | $1.97B | $2.13B | $2.33B | $2.59B |
| 9.5% | $1.63B | $1.74B | $1.86B | $2.01B | $2.21B |
| 10.0% | $1.46B | $1.55B | $1.7B | $1.76B | $1.91B |
| 10.5% | $1.32B | $1.39B | $1.47B | $1.56B | $1.68B |
| 11.0% | $1.19B | $1.26B | $1.33B | $1.4B | $1.49B |
Sensitivity analysis asks: 'If I change one assumption while holding all others constant, how does enterprise value change?' It isolates the marginal impact of each driver. One-way sensitivity changes a single variable across a range; two-way sensitivity varies two variables simultaneously, producing a matrix of EV outcomes. The critical inputs for a DCF sensitivity: (1) terminal growth rate; (2) WACC; (3) the NOPAT margin in year 5 (or terminal year); (4) revenue growth in the explicit period. These four variables capture virtually all material valuation uncertainty.
| WACC \ Growth | 1.5% | 2.0% | 2.5% | 3.0% | 3.5% | 4.0% |
|---|---|---|---|---|---|---|
| 8.0% | $2,420 | $2,640 | $2,910 | $3,250 | $3,720 | $4,410 |
| 8.5% | $2,100 | $2,270 | $2,480 | $2,740 | $3,090 | $3,600 |
| 9.0% | $1,840 | $1,970 | $2,130 | $2,330 | $2,590 | $2,960 |
| 9.5% | $1,630 | $1,740 | $1,860 | $2,010 | $2,210 | $2,490 |
| 10.0% | $1,460 | $1,550 | $1,645 | $1,760 | $1,905 | $2,110 |
| 10.5% | $1,315 | $1,390 | $1,470 | $1,560 | $1,675 | $1,830 |
| 11.0% | $1,190 | $1,255 | $1,325 | $1,400 | $1,490 | $1,615 |
The table above shows EV varying from $1,190M to $4,410M — a 3.7× range from the most bearish to the most bullish WACC/growth assumptions. This is not model noise; it is genuine uncertainty about long-run assumptions. The analyst's job is to identify which WACC and growth rates are defensible given the company's competitive position and industry dynamics — not to simply report the range. A common mistake: presenting the full table and implying all outcomes are equally probable. The table should be accompanied by a clear view of where the base case sits and why the analyst believes that specific WACC and growth rate are most likely.
| Terminal NOPAT Margin | 8% | 9% | 10% | 11% | 12% | 13% | 14% | 15% |
|---|---|---|---|---|---|---|---|---|
| Enterprise Value ($M) | $1,050 | $1,185 | $1,320 | $1,510 | $1,695 | $1,880 | $2,065 | $2,250 |
| % Change from Base | −38% | −30% | −22% | −11% | — | +11% | +22% | +33% |
Sensitivity analysis changes one variable at a time, which can produce internally inconsistent combinations — e.g., simultaneously assuming high revenue growth (bullish) and a high WACC (bearish market). Scenario analysis avoids this by defining complete, coherent narratives where all assumptions are consistent with each other. A bear case is not just low growth — it includes lower margins (scale doesn't materialize), higher capex intensity (competitive investment required), and potentially higher WACC (if the bear case includes sector stress).
| Assumption | Bear Case | Base Case | Bull Case |
|---|---|---|---|
| Revenue CAGR (5-year) | 3% | 8% | 14% |
| Terminal NOPAT Margin | 8% | 12% | 16% |
| Capex / Revenue (mature) | 9% | 7% | 5% |
| NWC / Revenue | 18% | 14% | 11% |
| Terminal Growth Rate | 1.5% | 2.5% | 3.5% |
| WACC | 11.5% | 10.0% | 9.0% |
| Implied ROIC (terminal) | 9% | 14% | 22% |
| Enterprise Value | $780M | $1,695M | $3,850M |
| Equity Value per Share (73M diluted) | $4.80 | $17.31 | $47.80 |
| Narrative | Competition intensifies, pricing erodes, growth stalls | Base assumptions hold, moderate expansion, stable margins | Market share gains, scale benefits materialize, platform business emerges |
Some analysts assign probabilities to scenarios (e.g., bear 25%, base 50%, bull 25%) and compute a probability-weighted intrinsic value: 0.25×$4.80 + 0.50×$17.31 + 0.25×$47.80 = $21.81/share. This is a useful framework for communicating the expected value, but the probabilities themselves are subjective — the precision of the arithmetic should not obscure the uncertainty in the scenario definitions. McKinsey cautions against treating the probability-weighted value as a single true estimate; it is a decision-support tool, not a precise prediction.
The tornado chart is a structured visualization of one-way sensitivity analysis, ranking assumptions from most to least impactful on enterprise value. By showing the EV range from each assumption's low and high bound side by side, it reveals which variables drive the most valuation uncertainty. The widest bars (most impactful assumptions) deserve the most analytical attention; the narrowest bars can be treated as given and require no further investigation.
| Assumption | Low Bound | EV at Low | EV at High | High Bound | Total EV Range |
|---|---|---|---|---|---|
| Terminal Growth Rate | 1.0% | $1,155M | $2,465M | 4.0% | $1,310M |
| WACC | 8.0% | $2,465M | $1,155M | 12.0% | $1,310M |
| Terminal NOPAT Margin | 7% | $910M | $2,250M | 16% | $1,340M |
| Revenue CAGR (Yrs 1-5) | 2% | $1,340M | $2,050M | 15% | $710M |
| Capex / Revenue | 5% | $1,820M | $1,560M | 10% | $260M |
| Tax Rate | 20% | $1,765M | $1,620M | 30% | $145M |
| NWC / Revenue | 10% | $1,750M | $1,640M | 18% | $110M |
Instead of point estimates at the low and high bounds, Monte Carlo simulation treats each assumption as a probability distribution (e.g., terminal growth = Normal(2.5%, 0.8%); WACC = Uniform(8.5%, 11.5%)). Running 10,000 simulations produces a distribution of enterprise values, from which analysts compute: expected EV, percentile ranges (10th–90th), probability of EV exceeding or falling below key thresholds (e.g., P(EV > $2B) = 34%). In practice, Monte Carlo is most valuable when assumptions are genuinely uncertain and correlated — for example, in resources and mining, oil price uncertainty correlates with both revenue and WACC. For most corporate DCFs, the three-scenario framework captures the essential uncertainty at lower analytical cost.
The reverse DCF inverts the valuation question: instead of asking 'what is this company worth given my assumptions?', it asks 'what assumptions must be true for the current market price to be justified?' This is one of the most powerful tools in the applied valuation toolkit because it tells the analyst exactly what the market is pricing in — and whether those implied assumptions are realistic.
For growth stocks trading at high multiples, the reverse DCF is far more informative than asking 'is the P/E too high?' The reverse DCF tells you exactly what growth rate and margin profile are required — and you can assess whether those are achievable. Damodaran's approach: compute the implied revenue (or FCFF) growth rate at which the DCF equals the market price; then apply a probability to that growth rate being achieved; if (probability × EV at implied growth) + (1−probability) × EV at realistic growth < market price, the stock is overvalued. This is how systematic investors quantify the risk priced into a stock, not just whether it is 'cheap' or 'expensive' on a multiple.
Key Takeaways
A two-way sensitivity table shows EV ranging from $800M to $3,200M when WACC varies from 8%–12% and terminal growth varies from 1%–4%. Your base case is WACC=10%, g=2.5%, giving EV=$1,695M. The company is trading at EV=$2,100M. What can you conclude?