Building a DCF
Read firstFree cash flowCost of capital. Valuation without accounting is memorization.
Every assumption below is arguable and I will say which ones I am least sure of. That is the difference between a model and a spreadsheet: a model comes with an account of where it could be wrong.
The assumptions, one at a time
Revenue growth: 3%, 3%, 2.5%, 2.5%, 2%. Target Corporation did $104,780M last year, down about 1.7% from the year before. Forecasting recovery to modest growth is a judgment. A mature US retailer growing with the economy is the base case, and it fades to 2% because a $100B revenue base cannot compound above nominal GDP for long.
: 5.2%. Actual was 4.9%. A small recovery assumes the mix and markdown pressure of the last year eases. This is the assumption I am least confident in, and it is also the one with the largest linear effect on the answer.
Tax rate: 23%. The effective rate was 22.3%. Using something near the recent effective rate is standard. Using the 21% federal statutory rate would ignore state taxes.
D&A 3% of revenue, 3.4%.Capex above means the asset base grows slightly, which is consistent with a retailer still refitting stores. If you set capex below D&A you are modeling a shrinking business, and you should say so on purpose rather than by accident.
Working capital: 5% of the change in revenue. A simplification. The lesson argues for forecasting days instead, and for a retailer with a negative the effect here is small either way.
: 7.5%. Built in the cost of capital lesson from a 4.2% , a relevered near 0.9, and a 4.5% , weighted against about 20% debt.
ConventionMid-year convention is on. Each forecast year is discounted from its midpoint, and the terminal value is still discounted the full five years because it sits at the end of year five. Turning the convention off lowers the answer by roughly 3.8% on the explicit years.
What it produces
Build it yourself
Everything above is preloaded below. Change one assumption at a time and watch which ones move the answer. The sensitivity grid at the bottom is the honest output of a , more so than the single share price above it.
The order to build in
- Revenue, from drivers if you have them and from a growth rate if you do not.
- Operating margin, and therefore .
- Tax on EBIT at the effective rate, giving .
- Add D&A, subtract capex, subtract the increase in working capital.
- Discount each year. Decide and state the .
- , both methods, and check each against the other.
- Sum to , bridge to , divide by .
- Sensitivity across WACC and the terminal assumption. Report the range.
Test yourself
01Walk me through a DCF.
Project unlevered free cash flow for five to ten years: revenue, operating margin, tax on EBIT, add back D&A, subtract capex and the increase in working capital. Discount each year at WACC. Calculate a terminal value at the end, either by growing the final year forever or by applying an exit multiple, and discount that back too. Add them for enterprise value. Subtract net debt and other claims for equity value, then divide by diluted shares.
Then say the part most people leave out: check what share of the value is terminal, and say what the model is most sensitive to. That is what separates someone who has built one from someone who has read about one.
02What are the two or three assumptions that matter most?
WACC and the terminal assumption, by a distance, because both act on the largest and most distant component. Then the operating margin, because it scales everything. Revenue growth in years two through five matters much less than people expect, which is uncomfortable given that it is where most of the modeling effort goes. Run the sensitivity grid and you can see the ordering rather than assert it.
03Your DCF says $143 and the stock trades at $151. What do you conclude?
That the model and the market are within about 5%, which for a DCF is indistinguishable from agreement. The honest read is that the market is pricing slightly better fundamentals than my assumptions, and I could ask what growth or margin the current price implies. Reverse-engineering the assumption the price requires is usually more informative than defending my own.
A tutor that knows this lesson. It asks before it explains, and it will not tell you what to do with your own money.
Every figure in this lesson that names a company comes from Target Corporation's Form 10-K for Fiscal 2025, the year ended 2026-01-31, filed 2026-03-11. Educational material, not investment advice.