DCF Sensitivity Analysis: Building Valuation Ranges

The same model, three answers

Your DCF says the company is worth $50/share. Your friend's DCF says $45/share. Your boss says $60/share. You're all using the same model. So who's right? Nobody. That's why smart analysts use sensitivity analysis.

Why point estimates are dangerous

A single DCF value assumes every assumption in your model is exactly right. They are not. WACC could reasonably be 8% or 10%; terminal perpetuity growth could be 2% or 3% — and those bands are not nitpicking, they are standard disagreement in a live process.

Sensitivity analysis asks: "If I am wrong on WACC, how wrong is my valuation?" (The same logic applies to growth, margins, and other drivers.) That teaches you which assumptions matter most and builds defensibility in your pitch — you can show the range implied by reasonable disagreement instead of defending one fragile number.

• It surfaces which inputs actually move the needle, so you prepare for the questions that will come.
• It replaces a single printout with a valuation band you can reconcile to comps and a football field.
• It mirrors how seniors push back: they rarely attack the spreadsheet layout; they attack the assumptions.

You are not abandoning rigor. You are being explicit about uncertainty — which is what separates a student output from a professional work product.

The two-way sensitivity table

The workhorse format is a two-way data table: hold your base case forecast fixed, then vary two inputs across rows and columns. For a standard DCF, WACC and terminal growth are the usual pair — both feed directly into discounting and the Gordon-growth terminal value, so small changes compound.

Below, rows are WACC (8% through 12%). Columns are terminal value perpetuity growth (2.0% through 3.5%). Each cell shows an illustrative implied enterprise value for the same operating model (round numbers for teaching; your model will produce different levels).

How to read it: Move down a column (holding g fixed): higher WACC → lower present value → lower EV in every cell. Move right along a row (holding WACC fixed): higher long-run growth → higher terminal cash flows → higher EV. The spread from the southwest to the northeast corner is your valuation range under reasonable disagreement on those two inputs.

Example: In the grid above, illustrative EV runs from about $10.4B (12% WACC, 2.0% g) to about $17.6B (8% WACC, 3.5% g). Your job in the room is not to defend the midpoint — it is to explain which corner is more plausible given rates, company quality, and competitive moat, and what would have to be true for the bear case to win.

From table to narrative

Turn the grid into language a PM or client retains: "Fair value is not a point; it is a band." Anchor on a base case (often near the center of what you believe is reasonable), then show downside and upside driven by WACC and growth — the two inputs audiences question first.

Pair this with other methods so no one method carries the whole story. A football field chart stacks DCF ranges next to trading comps and precedents; sensitivity explains why your DCF bar has width. Same idea: valuation is a distribution of outcomes, not a laser dot.

One-way sensitivities (only WACC, or only margin) are great for slides; two-way tables show interaction — when both inputs move against you at once, the hit is worse than either line item suggests alone.

Practical tips

• Keep the operating forecast fixed when building the table, or you will mix operating errors with discounting errors.
• Use ranges peers will accept (e.g., WACC ±1–2%, g capped modestly above long-run GDP). Extreme corners are useful for math, not for credibility.
• Reconcile to per-share if the audience thinks in stock price: divide equity value by diluted shares after the same bridge every time.
• Document your base case so someone can reproduce the middle cell without guessing which version of the model you ran.

If you have not already, walk through the core DCF mechanics in DCF Model: Discounted Cash Flow — sensitivity is the layer you add once the single-case model is sound.

Try it on Briefed

Open the DCF for any public ticker, set your base assumptions, then move WACC and terminal growth and watch how implied value responds — that live feedback is the same intuition the two-way table formalizes for your deck.