We have finally arrived at the "heart" of the DCF: Unlevered Free Cash Flow (UFCF). This is the actual cash a business generates that is "free and clear" of all internal or external obligations . It is the money that could be used to pay dividends to shareholders, buy back stock, or pay down debt .
UFCF vs. Levered Free Cash Flow (LFCF)
It is vital to understand the difference between "Unlevered" and "Levered" cash flows, as they serve different purposes in valuation .
- Unlevered Free Cash Flow (UFCF): Cash flow before interest payments. It represents the value of the company's operations (Enterprise Value) .
- Levered Free Cash Flow (LFCF): Cash flow after interest and debt payments. It represents the money left specifically for shareholders (Equity Value) .
Most professional DCF models use UFCF because it allows you to value the entire business regardless of its capital structure . If you use UFCF, you discount it using the Weighted Average Cost of Capital (WACC). If you use LFCF, you discount it using the Cost of Equity .
Building the 5-to-10 Year Forecast Table
In a real-world model, you will lay these numbers out in a grid. The first 5 to 10 years are your "Discrete Forecast Period," where you make specific assumptions for every year .
| Line Item | Year 1 | Year 2 | Year 3 | ... | Year 10 |
|---|---|---|---|---|---|
| Revenue | $112M | $125M | $140M | ... | $250M |
| (-) Operating Expenses | ($90M) | ($98M) | ($108M) | ... | ($180M) |
| EBIT | $22M | $27M | $32M | ... | $70M |
| (-) Taxes | ($4M) | ($5M) | ($6M) | ... | ($14M) |
| (+) Depreciation | $2M | $2M | $3M | ... | $5M |
| (-) CapEx | ($5M) | ($6M) | ($6M) | ... | ($10M) |
| (-) Change in Working Cap | ($1M) | ($1M) | ($1M) | ... | ($2M) |
| UFCF | $14M | $17M | $22M | ... | $49M |
The "Terminal Value" Concept
You might wonder: "What happens after Year 10? Does the company just disappear?" Of course not. Because we can't realistically forecast Year 11, 12, or 50 with any accuracy, we use a "Terminal Value" to represent all the cash flows from Year 11 into infinity . This is usually based on a "perpetual growth rate," often around 2% to 4% (matching the long-term growth of the economy) .
Sensitivity Analysis: The "What If" Tool
Because a DCF is based on estimates, it is never "perfect" . A professional model includes a Sensitivity Analysis (or "What-If" table). This shows how the company's value changes if your assumptions are slightly off .
- What if revenue growth is 8% instead of 10%?
- What if the tax rate increases?
- What if the cost of equipment (CapEx) spikes?
By looking at a range of values, you get a much better sense of the investment's risk .
Common Pitfalls in Forecasting
- The "Hockey Stick" Forecast: This is when an analyst predicts that a company that has been struggling for years will suddenly have massive, skyrocketing growth starting next year. Be realistic; business turnarounds are hard .
- Ignoring the Cycle: Many industries (like oil or housing) are cyclical. If you forecast 10 years of "peak" performance without accounting for a potential downturn, you will overvalue the stock .
- Forgetting Reinvestment: You cannot forecast 20% revenue growth while also forecasting $0 in CapEx. To grow, a company almost always needs to spend on new assets .
Frequently Asked Questions: The Final Model
Q: Is a 10-year forecast better than a 5-year forecast?
A: Not necessarily. The further out you go, the more "total shots in the dark" your estimates become
. 5 years is standard for stable companies; 10 years is often used for high-growth companies that need more time to reach a "steady state"
.
Q: How do I know if my UFCF number is "good"?
A: Compare it to the company's history. If your projected UFCF is 5x higher than anything the company has ever achieved, you need a very strong reason (like a revolutionary new patent) to justify it
.
Q: Can I use this for private companies?
A: Yes, but it’s harder. Private companies don't file 10-Ks, so you have to rely on their internal books, which might mix personal and business expenses
. You often have to "clean" their data before you can start your forecast
.

Comments