Key insights
- The article highlights the significant impact of terminal value assumptions on DCF valuations, often overshadowing the explicit forecast period. It demonstrates how small changes in terminal growth rates can drastically alter valuation outcomes, especially in low-interest-rate environments. The author suggests stress-testing terminal growth assumptions to ensure valuation robustness and avoid overoptimistic forecasts.
After 12 years of investing and building hundreds of DCF models, I still find it striking how little attention terminal value gets relative to how much of the answer it determines. The explicit forecast period gets careful attention. Revenue growth, margins, capex. The terminal value gets a quick sanity check and whatever growth rate feels reasonable. I ran the numbers. In a typical 10-year DCF, terminal value drives 60 to 80% of the total intrinsic value calculation. Take a business generating $5 of free cash flow per share and a 7% discount rate. At 2% terminal growth the fair value is roughly $87. Move the terminal growth to 4% and the same business is worth roughly $123. Same company, same forecast, one assumption changed by 2 percentage points. The answer moves by 42%! The sensitivity gets worse as the discount rate drops. At 8% the same 2 percentage point shift moves the answer by 29%. At 7% it moves by 42%. At 6% by over 60%. In a low interest rate environment, when discount rates compress and investors reach for growth, terminal value assumptions become even more dangerous than usual. The assumption nobody scrutinizes becomes the one doing the most damage. And a 4% terminal growth rate above long-term GDP means you are implicitly forecasting that this single business will eventually become larger than the entire economy. A math error with a spreadsheet around it. The Gordon Growth Model has a structural contradiction built into it. It assumes ROIC equals cost of capital in perpetuity. Which means your wonderful moat business with 30% ROIC in years one through ten has, according to the model, become a completely average company by year eleven. That directly contradicts the thesis you spent ten years building. Most retail DCF templates hide the terminal value in a single cell. That cell determines the answer…If you cannot defend it with conviction, you are not doing valuation. My suggestion: next time you build a model, run it at 2%, 3%, and 4% terminal growth before you trust the number.