DCF sensitivity analysis
A DCF valuation is only as good as its assumptions. Revenue growth, operating margins, capital expenditure, the discount rate, and the terminal value all carry estimation error. Sensitivity analysis answers the question: how much does my enterprise value change if one assumption moves 1%? Build a two-way sensitivity table, test the key drivers, and you'll see which assumptions matter most. More importantly, you'll discover whether your valuation conclusion is robust or depends on a narrow band of assumptions. That clarity prevents overconfidence and keeps you honest.
Quick definition
Sensitivity analysis in a DCF shows how the valuation output (enterprise value or implied share price) changes when individual inputs—perpetuity growth, WACC, terminal year margins, revenue growth—vary. The standard format is a two-way table showing enterprise value for a grid of assumptions. Sensitivity analysis forces you to think probabilistically: if your base case perpetuity growth is 2.5%, what's the valuation if it's 2.0% or 3.0%?
Key takeaways
- Sensitivity analysis is not a substitute for a good base-case forecast; it's a risk measurement tool that shows which assumptions matter most.
- The most common two-way tables pair WACC and perpetuity growth because terminal value is usually the largest valuation component.
- Build sensitivity tables for revenue growth, operating margins, and capex intensity if your valuation is growth-dependent.
- If your valuation conclusion (buy, hold, sell) changes based on a 0.5–1% swing in WACC or a 0.5% change in perpetuity growth, your conviction should be lower.
- Sensitivity analysis is a forcing function: it reveals which bets you're really making.
Why sensitivity analysis prevents overconfidence
A base-case DCF produces a point estimate: enterprise value of $50 billion, implying a share price of $75. That point estimate can seduce you into false precision. You believe you've calculated the value. But you haven't; you've calculated one scenario given a specific set of assumptions. The real valuation is a range, and sensitivity analysis maps that range.
When you learn that a 50 basis points increase in WACC reduces enterprise value by $8 billion (16%), you realize how model-dependent your conclusion is. If the market's implied discount rate is 50bps higher than you thought, you're overvaluing. Sensitivity analysis forces intellectual humility: it shows the assumptions on which your case rests and how fragile those rests might be.
The two-axis sensitivity table (WACC and perpetuity growth)
The most common sensitivity table pairs WACC (discount rate) on one axis and perpetuity growth rate (terminal growth) on the other. Here's why: these two variables usually drive 70–90% of DCF variance because terminal value dominates the valuation.
Build it in a spreadsheet:
- Rows: WACC from 7.0% to 10.0% in 0.5% increments
- Columns: Perpetuity growth from 1.5% to 3.5% in 0.5% increments
- Cells: Enterprise value for each combination
A typical table for a mature software company might look like this:
| WACC | 1.5% Growth | 2.0% Growth | 2.5% Growth | 3.0% Growth | 3.5% Growth |
|---|---|---|---|---|---|
| 7.0% | $72B | $82B | $95B | $112B | $133B |
| 7.5% | $68B | $77B | $88B | $102B | $120B |
| 8.0% | $64B | $72B | $82B | $94B | $110B |
| 8.5% | $61B | $68B | $77B | $87B | $101B |
| 9.0% | $57B | $64B | $72B | $81B | $93B |
| 9.5% | $54B | $61B | $68B | $76B | $86B |
| 10.0% | $51B | $57B | $64B | $71B | $80B |
The table reveals the valuation's sensitivity to the two key long-term drivers. Notice the wide range: from $51B (9.5% WACC, 1.5% growth) to $133B (7% WACC, 3.5% growth). That's a 2.6x spread. Your confidence in your base-case valuation of $82B (8% WACC, 2% growth) depends on how likely that specific combination is.
Building a sensitivity table: step by step
-
Set your base case. Run your full DCF and record enterprise value. Note the WACC, perpetuity growth, terminal margins, and capex assumptions.
-
Choose your ranges. For WACC, a ±1% range is typical. For perpetuity growth, ±1% is standard. For a margin-sensitive company, ±2 percentage points is reasonable.
-
Create a grid. In Excel, use a two-way table formula. Use the base-case DCF as a template, then substitute values for the two axes. Example:
=DCF($B$1:$B$50, WACC_cell, Growth_cell)allows you to paste one formula and vary inputs. -
Color-code the results. Use a color gradient from red (low valuation) to green (high valuation). This makes the visual patterns jump out.
-
Mark your base case. Highlight the cell corresponding to your base-case assumptions so you can see where the valuation sits within the range.
Sensitivity to revenue growth (for high-growth companies)
If you're valuing a fast-growing company, sensitivity to revenue growth matters more than perpetuity growth because near-term growth compounds over the forecast period. Build a two-way table pairing explicit forecast period growth (e.g., years 1–7 CAGR) and terminal growth. Example:
| 7-Year CAGR | 1.5% Terminal | 2.0% Terminal | 2.5% Terminal |
|---|---|---|---|
| 8% Growth | $65B | $72B | $81B |
| 10% Growth | $78B | $87B | $98B |
| 12% Growth | $93B | $104B | $118B |
This table shows that for a high-growth company, a 2–4 percentage point difference in CAGR is often more impactful than terminal growth assumptions.
Sensitivity to operating margins
For a company where margin improvement is your key thesis (e.g., a turnaround or scale story), build sensitivity to terminal-period operating margin:
| Terminal EBITDA Margin | 2% TV Growth | 3% TV Growth |
|---|---|---|
| 18% | $52B | $61B |
| 20% | $62B | $74B |
| 22% | $73B | $87B |
A 2–4 percentage point margin compression (a realistic risk) can cut valuation by 20–30%. If your case depends on margin improvement, make sure it's visible in sensitivity and you have conviction it will happen.
Sensitivity to capex intensity
For capital-intensive businesses (industrials, telecom, utilities), capex as a percentage of revenue drives free cash flow. Build sensitivity:
| Terminal Capex/Revenue | 2% TV Growth | 3% TV Growth |
|---|---|---|
| 3% | $68B | $81B |
| 4% | $61B | $72B |
| 5% | $54B | $63B |
A 1–2 percentage point swing in capex intensity can be material. This is especially important for companies claiming margin expansion; ensure they're not assuming unrealistically low capex reinvestment.
Interpreting a sensitivity table
High sensitivity (wide valuation range). If your valuation ranges from $40B to $120B across reasonable assumption ranges, your base case is model-dependent. Small assumption shifts produce 20–30% swings. Use this as a reason to demand a larger margin of safety before investing.
Low sensitivity (tight valuation range). If valuation ranges from $70B to $85B across the same ranges, your conclusion is more robust. The company looks cheap or expensive regardless of modest assumption changes. This is ideal.
Skewed sensitivity. Notice if one assumption produces much larger swings than another. If varying WACC by ±1% swings valuation by ±30% but varying perpetuity growth by ±1% swings it by ±15%, WACC is the key driver. Build conviction around WACC; be less precious about perpetuity growth precision.
Non-linear sensitivity. At very high growth rates or very low discount rates, small assumption changes produce outsized valuation swings. This is a feature of present value math, not a sign of model error. But it signals that betting on extremely optimistic assumptions is risky.
Real-world example: Microsoft sensitivity
Imagine you're valuing Microsoft. Base case: 8% WACC, 2.5% perpetuity growth, FCF year 10 of $100B, exiting at 18x FCF multiple. Your valuation is $85B enterprise value.
Run sensitivity on WACC and perpetuity growth:
| WACC | 1.5% | 2.0% | 2.5% | 3.0% | 3.5% |
|---|---|---|---|---|---|
| 7.0% | $82B | $94B | $109B | $128B | $152B |
| 7.5% | $77B | $88B | $101B | $118B | $140B |
| 8.0% | $73B | $83B | $95B | $110B | $129B |
| 8.5% | $69B | $78B | $89B | $102B | $119B |
| 9.0% | $65B | $74B | $84B | $95B | $110B |
Your base case ($85B) is in the middle of the range. But notice: if WACC rises to 9% (not unrealistic in a higher-rate environment), valuation drops to $84B. If it rises to 9.5%, valuation is $78B, a 8% decline. That's not huge. But if perpetuity growth is actually 2% instead of 2.5%—a reasonable disagreement—your valuation at 8% WACC is $83B, not $95B. The two-way table shows that either assumption being slightly different materially impacts your valuation.
Common mistakes in sensitivity analysis
Mistake one: using unrealistic ranges. If your perpetuity growth range is 1% to 5%, you're including absurd scenarios. Most developed-market companies have perpetuity growth between 1.5% and 3%. Tighten your ranges to test realistic disagreements, not fantasy scenarios.
Mistake two: ignoring correlation. WACC and perpetuity growth are not independent. If WACC is high, it usually means the market is risk-averse; perpetuity growth should be lower, not higher. A 7% WACC paired with 4% perpetuity growth is inconsistent. Build ranges that make economic sense together.
Mistake three: anchoring to your base case. Sensitivity analysis is most useful when it changes your view, not confirms it. If every scenario in your sensitivity table supports your investment thesis, you've probably rigged the ranges. Include scenarios that would prove you wrong.
Mistake four: not documenting probability. A sensitivity table shows outcomes, not probabilities. One more step: estimate the probability of each assumption. If 8% WACC is most likely, 7.5–8.5% is 70% probable, and 7% or 9% is 30%, your expected value is a probability-weighted average. Most investors skip this, but it produces more realistic valuation ranges.
Mistake five: too many variables at once. Don't build a three-way table (WACC, growth, and margin) unless you really need it. It becomes unreadable. Stick to two-way tables and run multiple tables if needed.
Using sensitivity to guide your diligence
Sensitivity analysis is a diligence roadmap. If sensitivity is high, focus your research on the key drivers. For a margin-sensitive company, dig deep into competitive dynamics and cost structure. For a growth-sensitive company, understand the TAM, market share trajectory, and competitive threats. Use sensitivity to allocate your research time and conviction.
FAQ
Q: Should I use my base-case assumptions as the center of my sensitivity range?
A: Usually yes. Center your range on your best estimate, then extend ±1–1.5 standard deviations (or 50% wider). This ensures your base case is not at the edges of your range, which would suggest overconfidence.
Q: How fine-grained should the sensitivity grid be?
A: For WACC and perpetuity growth, 0.5% increments are standard. For margin or revenue growth, 1–2 percentage point increments work. Fine-grained grids (0.1% increments) are overkill and suggest false precision.
Q: Can sensitivity analysis tell me whether a stock is cheap or expensive?
A: Only in combination with your base case. Sensitivity shows the range of possible valuations. If your base-case intrinsic value is $75 and the stock trades at $60, it looks cheap in most sensitivity scenarios. If it trades at $95, it looks expensive. Sensitivity reveals how much wiggle room your valuation has.
Q: What if my sensitivity table shows the stock is always cheap (or always expensive)?
A: That's a strong signal. If valuation is $50–$100 in your sensitivity range and the stock trades at $55, it's likely cheap across reasonable assumptions. Conversely, if it trades at $120, it's likely overvalued. Tight agreement across scenarios is confidence-building.
Q: Should I publish a sensitivity table or keep it for myself?
A: If you're writing an investment report, include a sensitivity table. It builds credibility by showing you've stress-tested assumptions. If you're managing your own portfolio, building the table for yourself is sufficient; publication is not necessary.
Q: How does sensitivity analysis relate to scenario analysis?
A: Sensitivity analysis changes one or two variables at a time in a spreadsheet grid. Scenario analysis (covered in article 22) builds fully coherent scenarios (bull case, base case, bear case) where multiple assumptions move together consistently. They're complementary tools.
Related concepts
- Terminal value dominance — understanding which assumptions matter most in your DCF.
- Scenario analysis — a narrative-driven extension of sensitivity analysis.
- Margin of safety — using sensitivity to determine how much discount you need before investing.
- WACC calculation — the most important sensitivity input.
- Perpetuity growth rate — the other critical long-term assumption.
Summary
Sensitivity analysis maps the range of valuations that emerge from reasonable assumption changes. Build two-way tables pairing your highest-impact assumptions (usually WACC and perpetuity growth, but sometimes revenue growth, margins, or capex). Use sensitivity to identify which bets you're really making, stress-test your conviction, and avoid overconfidence. A narrow valuation range across sensitivity tests is confidence-building; a wide range is a signal to demand a larger margin of safety. Sensitivity analysis is not a crystal ball, but it's an essential tool for honest self-assessment of model risk.
Next
Scenario analysis around a DCF — Read article 22
Completed articles: 3 of 6 / Article word count: 2,181