Sensitivity Analysis in DCF Models: WACC, Growth Rate, and Terminal Value Tables

Every DCF model rests on assumptions. You choose a discount rate, a growth rate, a terminal value multiple, and a projection window, and out comes an intrinsic value estimate. The problem is that each of those inputs is uncertain. Change the WACC by one percentage point and the estimated fair value might swing 20%. Change the terminal growth rate by half a point and the result moves again.

Sensitivity analysis is how professional analysts handle this uncertainty honestly. Instead of presenting a single number as 'the answer,' a well-built sensitivity table shows how the estimated fair value changes across a range of plausible input combinations. The result is a band of outcomes rather than a false point estimate, which is far more useful for building conviction in a valuation.

This guide covers how sensitivity analysis works in DCF models, how to build and read a two-way sensitivity table, why terminal value assumptions dominate, and what a complete DCF output should include.

Why Sensitivity Analysis Matters in a DCF Model

A DCF model is not a prediction machine. It is a structured way to translate assumptions about the future into a present-value estimate. The output is only as valid as the inputs, and most inputs require judgment.

Consider a simple three-stage DCF for a mid-cap technology company. You project revenue growth at 12% for five years, then 8% for three years, then assume the business grows in perpetuity at 3%. You use a WACC of 10%. That set of assumptions produces an intrinsic value of, say, $85 per share.

But what if the long-run growth rate is 2.5% instead of 3%? What if WACC is 11% because the market risk premium rises? What if your free cash flow margin assumptions are slightly off in years three through five?

Every one of those shifts moves the output, sometimes dramatically. Without a sensitivity table, you hold a single number that gives no indication of how much confidence it deserves. With a sensitivity table, you can see immediately whether the stock looks attractive across most reasonable input combinations or only in a narrow best-case scenario.

Professional equity research analysts almost always include sensitivity tables in their valuation work for exactly this reason. They are a practical acknowledgment that the future is unknowable and that honest valuation requires showing the range.

One-Way vs. Two-Way Sensitivity Tables

There are two main formats for sensitivity analysis in DCF models.

One-Way Sensitivity Tables

A one-way (or univariate) sensitivity table holds all inputs constant except one. You vary that single input across a range and observe how the estimated fair value responds.

For example, you might vary only the WACC from 8% to 13% in half-point increments while holding everything else fixed. The output is a single column showing how fair value changes as the discount rate moves:

This is a useful diagnostic tool. It tells you how sensitive the model is to changes in that one input, but it only captures one dimension of uncertainty at a time.

Two-Way Sensitivity Tables

A two-way (or bivariate) sensitivity table varies two inputs simultaneously, creating a matrix. The most common pairing in equity valuation is WACC on one axis and the terminal growth rate on the other.

The result is a grid where every cell shows the estimated fair value for a specific combination of those two inputs. You can scan the grid to see which combinations produce values above the current market price and which fall below it, giving you a clearer picture of how much margin of safety exists under different assumptions.

Two-way tables are almost always more informative than one-way tables because the two most uncertain and impactful inputs in any DCF, the discount rate and the long-run growth assumption, interact with each other. Looking at them together is the right approach.

The WACC vs. Terminal Growth Rate Table

The WACC and terminal growth rate table is the most widely used sensitivity matrix in equity DCF analysis. Here is an example using a hypothetical company with a base-case fair value of $85:

Estimated Fair Value: WACC (rows) vs. Terminal Growth Rate (columns)

In this table, the base-case assumptions (WACC = 10.5%, terminal growth = 3.5%) appear at the intersection that shows $85, roughly where the current market price sits.

Higher values (upper-right region) represent scenarios where the model suggests potential undervaluation relative to current price. Lower values (lower-left region) represent scenarios where the model suggests the stock may be overvalued. The spread across the full matrix, from $55 to $156 in this example, illustrates just how wide the uncertainty band is.

How to Read a Sensitivity Table

Reading a sensitivity table well requires understanding its structure.

The center of the table is usually the base case. Your most likely assumptions sit at the intersection of the row and column you consider most plausible. The cells surrounding that intersection represent deviations from the base case.

Look at the direction of movement. In a WACC vs. terminal growth table, moving down any column (higher WACC) always decreases estimated fair value, because higher discount rates reduce the present value of future cash flows. Moving right across any row (higher terminal growth) always increases estimated fair value, because a faster-growing perpetuity is worth more.

Measure the sensitivity gradient. In the table above, moving from 10.5% to 11.0% WACC while holding terminal growth constant at 3.5% drops estimated fair value from $85 to $79, a 7% decline from a half-point WACC change. That tells you WACC has significant leverage over the output.

Count how many cells support a given conclusion. If you are assessing whether a stock may be attractively valued, count the table cells that produce an estimated fair value above the current price. If 18 out of 24 cells show values above the current price across a realistic input range, the conclusion is relatively robust. If only 4 out of 24 cells do, the conclusion depends heavily on assumptions landing precisely right.

The Disproportionate Impact of Terminal Value

The most important thing to understand about DCF sensitivity analysis is that the terminal value dominates.

In a typical 5-year or 10-year DCF model, the terminal value, the estimated present value of all cash flows beyond the explicit forecast period, can represent 60% to 80% of the total estimated fair value. Sometimes more.

This creates a structural reality: the inputs that drive terminal value (WACC and the terminal growth rate) are also the inputs with the widest uncertainty range. Small changes compound dramatically over an infinite horizon.

The mathematical relationship is captured in the Gordon Growth Model formula for terminal value:

Terminal Value = Final Year FCF x (1 + Terminal Growth Rate) / (WACC - Terminal Growth Rate)

The denominator is (WACC minus terminal growth rate). If WACC is 10% and terminal growth is 3%, the denominator is 7%. If you change those to 9.5% and 3.5%, the denominator drops to 6%, a 14% smaller divisor, producing a 14% larger terminal value before any discounting adjustment.

This is why a sensitivity table that varies both WACC and terminal growth simultaneously is so much more informative than varying just one input. The interaction between these two numbers is where most of the valuation uncertainty lives.

Practical implication: if a stock's estimated fair value is only attractive when you use a terminal growth rate near or above nominal GDP growth, examine that assumption carefully. Most businesses do not grow faster than the overall economy in perpetuity.

Tornado Charts: Ranking Input Sensitivity

A tornado chart is a bar chart that shows how much the estimated fair value changes when each model input is varied by a standardized amount, typically plus or minus one standard deviation or a fixed percentage shift. The bars are sorted from longest to shortest, creating the tornado shape.

The purpose is to rank inputs by their impact on the output. In a standard DCF model, a tornado chart almost always puts WACC and terminal growth rate at the top (largest impact), followed by near-term free cash flow growth rate, then margin assumptions, then working capital and capex details at the bottom.

This ranking tells you where to focus your analytical effort. If a two-percentage-point change in revenue growth in year three barely moves the needle, you do not need to spend hours perfecting that forecast. If a half-point change in WACC swings the estimate by 15%, getting your WACC calculation right is worth serious effort.

For self-directed investors, the mental model matters more than producing an actual chart: direct the bulk of your scrutiny toward the inputs at the top of the sensitivity ranking.

Scenario Analysis vs. Sensitivity Analysis

Sensitivity analysis and scenario analysis are related but distinct tools.

Sensitivity analysis varies one or two inputs while holding everything else constant. It answers the question: how does the output change if this specific assumption is wrong? It is a mechanical test of model responsiveness.

Scenario analysis builds multiple complete, internally consistent versions of the model. A bear-case scenario does not just change the WACC in isolation. It changes revenue growth, margin assumptions, capex intensity, and working capital together in a way that reflects a coherent business outcome.

Bear, Base, and Bull Case Structure

The base case uses the most likely set of assumptions: moderate revenue growth, stable margins, and a terminal growth rate of 2% to 3.5% for established businesses.

The bear case reflects a plausible downside: slower growth from competitive pressure, margin compression, a higher WACC from rising rates or increased risk perception, and a lower terminal growth rate.

The bull case reflects a plausible upside: faster revenue growth from new products or market share gains, expanding margins from operating leverage, and a slightly higher terminal growth rate justified by durable competitive advantages.

The three scenarios produce three estimated fair values. You can weight them by probability (for example, 20% bear, 60% base, 20% bull) to produce a probability-weighted estimated fair value. That is a more nuanced output than any single-case estimate.

The key discipline is internal consistency. A bull case that raises revenue growth while simultaneously lowering capex and improving working capital may be unrealistically optimistic. A good scenario tells a coherent business story.

Monte Carlo Simulation

Monte Carlo simulation is the most sophisticated form of DCF sensitivity analysis. Instead of testing a discrete set of input combinations or building three manual scenarios, Monte Carlo generates thousands of random trials, sampling each input from a probability distribution.

You might define WACC as normally distributed with a mean of 10% and a standard deviation of 1.5%. Terminal growth rate might be uniformly distributed between 1.5% and 4.0%. Near-term free cash flow growth might follow a triangular distribution with a most-likely value of 8%, a low of 2%, and a high of 15%.

The simulation runs the DCF thousands of times, drawing a random value for each input every trial. The output is a distribution of estimated fair values rather than a single number, a histogram showing, for example, that 70% of trials produce a value between $70 and $110 with a median of $87.

Monte Carlo has two advantages over manual tables. First, it captures correlations between inputs (WACC tends to be higher when growth is weaker). Second, it produces explicit probability statements about the output range, which is more intellectually honest than a single estimate.

Practical Steps to Build a Sensitivity Table

Building a two-way WACC vs. terminal growth sensitivity table in a spreadsheet is straightforward once your base-case DCF model is complete.

Step one: make sure WACC and terminal growth rate are each a standalone cell reference in your model, not hard-coded inside formulas.

Step two: define the input ranges. For WACC, base case plus or minus 2 percentage points in half-point increments is standard. For terminal growth rate, 1.5% to 4.5% in half-point increments covers most realistic scenarios for established businesses.

Step three: set up the table headers. Place WACC values as row headers and terminal growth rate values as column headers. In the top-left data cell, enter a formula that references your DCF model's final estimated fair value.

Step four: use your spreadsheet's data table function (in Excel: Data, What-If Analysis, Data Table). Set the row input cell to your WACC cell and the column input cell to your terminal growth rate cell. The function populates the entire matrix automatically.

Step five: apply conditional formatting. Color cells where the estimated fair value exceeds the current market price (or current price plus a margin of safety) in green. Color cells below the threshold in red. The distribution of outcomes becomes immediately visible.

The entire process takes 15 to 20 minutes once you understand the data table function.

How Wide Should the Input Ranges Be?

A common mistake is using a range that is too narrow. A WACC range of 9.9% to 10.1% tells you almost nothing useful.

WACC Range

A reasonable range is base case plus or minus 2 to 3 percentage points. For a company with a base-case WACC of 10%, that means testing values from 7% to 13%. Reasonable analysts can disagree on equity risk premium (4% to 7% is a common spread) and beta estimation, which alone creates multi-percentage-point variation.

Terminal Growth Rate Range

A reasonable range for established businesses is 0% to 5%. The lower bound reflects a mature company growing at roughly the rate of inflation. The upper bound is near long-run nominal GDP growth in developed economies.

A terminal growth rate above 5% implies the company will indefinitely grow faster than the overall economy, which is mathematically impossible over infinite time. Most valuation textbooks recommend using a terminal growth rate at or below the expected long-run nominal GDP growth rate of the company's primary market.

Near-Term Growth Rates

For the explicit forecast period, a range of plus or minus 3 to 5 percentage points around the base case is generally appropriate. For highly cyclical businesses or early-stage companies with less predictable cash flows, the range should be wider.

What a Well-Designed DCF Output Should Include

A rigorous DCF analysis is not a single number. A well-designed output communicates the full structure of the estimate alongside its uncertainty.

The base-case estimated fair value should be presented prominently, with the exact inputs that produced it: WACC, terminal growth rate, projection period, and the terminal value method used (Gordon Growth Model, exit multiple, or both).

The two-way sensitivity table should appear directly alongside the base case, showing the full matrix of estimated fair values across the tested input range. The current market price should be visible in context so a reader can immediately assess how many scenarios point to potential undervaluation versus overvaluation.

A brief narrative should explain the most important assumptions: why the terminal growth rate was set where it was, what drives the WACC estimate, and what would have to change for the bear or bull case to materialize.

The output should not claim to predict where the stock price will go. A DCF model produces an estimate of intrinsic value under a set of assumptions. The market price may deviate from that estimate for extended periods. Sensitivity analysis is the tool that makes clear how confident, or uncertain, the analyst is in that estimate.

Equity Rank combines DCF with multiple other valuation methods, including asset-based, earnings-based, and market-comparable approaches, aggregating the results through the SAVE score to reduce dependence on any single model's assumptions. Understanding sensitivity analysis helps you interpret those outputs more critically and use them more effectively in your own research process.