How to Build a Multi-Stage DCF Model

A single-stage DCF model assumes one constant growth rate forever. That is a useful simplification for stable, mature companies, but it breaks down for businesses that are growing rapidly today and will eventually slow down. Most real companies do not grow at the same rate for the next 30 years.

A multi-stage DCF solves this by splitting the future into distinct phases, each with its own growth assumptions. This guide walks you through building a two-stage and three-stage DCF model from scratch. If you are new to DCF, start with our DCF Model Explained guide first, then come back here for the advanced version.

Why Single-Stage DCF Falls Short

A single-stage DCF projects free cash flow at one growth rate and discounts everything at a constant WACC. The terminal value uses the same perpetuity growth assumption. This works for utilities, consumer staples, and other businesses with steady, predictable cash flows.

But consider a high-growth technology company growing free cash flow at 25 percent per year. Projecting 25 percent growth into a terminal value formula produces an absurdly high valuation because the math assumes that growth rate continues forever. No company sustains 25 percent growth indefinitely. The economy itself only grows at 2 to 3 percent long-term.

A multi-stage model fixes this by explicitly modeling the transition from high growth to mature growth. It forces you to think about when and how fast growth will decelerate, which is often the most important assumption in the entire valuation.

Two-Stage DCF Model

The two-stage DCF is the most common multi-stage approach. It divides the future into two phases.

Stage 1 is the high-growth phase, typically 5 to 10 years. During this period, you project free cash flow using the company's current or expected near-term growth rate. This rate should reflect the company's competitive position, market opportunity, and recent trajectory.

Stage 2 is the terminal value phase. At the end of Stage 1, you assume growth immediately drops to a long-term sustainable rate (typically 2 to 3 percent) and calculate the terminal value using the Gordon Growth Model.

Where g1 is the high-growth rate, t is each year in Stage 1, n is the final year of Stage 1, and Terminal Value uses the perpetuity growth rate g2.

Three-Stage DCF Model

The three-stage DCF adds a transition phase between high growth and terminal value. This is more realistic because growth rarely drops from 25 percent to 3 percent overnight. It typically decelerates gradually over several years.

Stage 1 is the high-growth phase (years 1 to 5). Growth reflects the company's current trajectory.

Stage 2 is the transition phase (years 6 to 10). Growth decelerates linearly from the Stage 1 rate to the terminal rate. If Stage 1 growth is 20 percent and terminal growth is 3 percent, Stage 2 might step down by roughly 3.4 percentage points per year: 20, 16.6, 13.2, 9.8, 6.4, 3.

Stage 3 is the terminal value, calculated at the stable long-term growth rate.

Step-by-Step: Building a Three-Stage DCF

Let us build a complete three-stage DCF for a hypothetical high-growth technology company.

Step 1: Project Stage 1 Cash Flows

Apply the high-growth rate to each year's free cash flow and discount to present value.

Step 2: Project Stage 2 Transition Cash Flows

The growth rate declines linearly from 20 percent to 3 percent over five years. Calculate the growth rate for each year, project the free cash flow, and discount.

Step 3: Calculate Terminal Value

At the end of year 10, apply the Gordon Growth Model using the terminal growth rate.

Discount the terminal value back to present value.

Step 4: Sum Everything and Calculate Per-Share Value

Add the present values of all three stages.

If the stock trades at $75, it is 19 percent below your intrinsic value estimate, offering a meaningful margin of safety. If it trades at $120, it is 29 percent above, suggesting overvaluation.

Terminal Value: The Most Critical Assumption

In our example, terminal value accounts for 60 percent of total enterprise value. This is typical. Even with a three-stage model, the bulk of the valuation depends on what happens after year 10.

This is why the terminal growth rate deserves extreme scrutiny. Using 4 percent instead of 3 percent increases the terminal value by roughly 17 percent. Using 2 percent decreases it by 12 percent. A one percentage point change in a single assumption shifts the entire valuation by double digits.

Conservative analysts cap the terminal growth rate at the long-term nominal GDP growth rate of the country where the company operates. For the US, this is approximately 2 to 3 percent. For emerging markets, it might be 4 to 5 percent, but using a higher rate requires a correspondingly higher WACC to compensate for the additional risk.

Sensitivity Analysis: Testing Your Assumptions

A DCF model is only as good as its assumptions. Sensitivity analysis tests how the fair value changes when you vary key inputs.

Build a sensitivity table with WACC on one axis (8 to 12 percent in 0.5 percent increments) and Stage 1 growth on the other axis (15 to 25 percent). Fill in the resulting fair value per share for each combination. This creates a range of outcomes rather than a single point estimate.

If the stock only looks undervalued under the most optimistic scenario, the margin of safety is thin. If it looks undervalued even in the bear case, you have a much stronger investment thesis.

Common Mistakes in Multi-Stage DCF

Mistake one: using too high a terminal growth rate. Anything above 3 percent for a developed market company requires strong justification. Above 4 percent almost always inflates the valuation unrealistically.

Mistake two: making Stage 1 too long. A 15-year high-growth phase assumes you can predict the competitive landscape and company execution for over a decade. Five to seven years is a more honest projection horizon for most companies.

Mistake three: not adjusting WACC for changing risk. As a company matures, its risk profile changes. A high-growth startup might warrant a 14 percent WACC today but only 9 percent when it reaches maturity. Some advanced models use different WACCs for each stage.

Mistake four: ignoring reinvestment requirements. High growth requires capital. If a company grows free cash flow at 20 percent, it likely needs significant reinvestment. Make sure your projections account for the capital expenditure needed to fund that growth, not just the revenue side.

Mistake five: treating the model output as a precise number. A three-stage DCF produces a single fair value, but it is an estimate built on estimates. Always think in ranges and use sensitivity analysis to understand the confidence interval around your number.

When to Use Multi-Stage DCF

Use a two-stage model for companies with above-average but moderately predictable growth: large-cap tech, established healthcare, strong consumer brands. These companies are past hyper-growth but still growing faster than GDP.

Use a three-stage model for companies in rapid growth phases: high-growth tech, biotech with new drug launches, emerging market leaders, and companies undergoing transformational change. The transition phase makes the deceleration explicit rather than assuming an abrupt cliff.

For mature, stable companies with predictable cash flows (utilities, consumer staples, REITs), a single-stage DCF or simple P/E comparison is often sufficient. The added complexity of multi-stage modeling does not improve accuracy when growth is already near terminal rates.

Multi-Stage DCF at Fair Price Index

Fair Price Index uses DCF analysis as 50 percent of its blended valuation model. The DCF component incorporates growth-adjusted projections calibrated to each company's stage and sector, combined with relative valuation (30%) and analyst consensus (20%) for a more robust fair price.

You can test your own multi-stage assumptions using the DCF Calculator, which lets you set custom growth rates and projection periods. For a simpler starting point, see the DCF Model Explained guide. Explore fair values for over 37,000 stocks at fairpriceindex.com.