DCF model checks: 8 mistakes that change the value

Most DCF errors come from eight places: a terminal growth rate at or above the discount rate, a discount rate that does not match the cash flow, missing tax or reinvestment, wrong timing of cash flows, working capital moving the wrong way, double-counted cash or debt, the wrong share count, and a terminal value that is most of the answer without anyone noticing.

Checked 3 November 2026 · sources below

The eight checks

CheckWhat to look forWhy it matters
Terminal growthLong-run growth is below the discount rate and close to long-run inflation plus real growthAt growth = discount rate the Gordon formula divides by zero
Discount rate matchFree cash flow to the firm uses WACC; cash flow to equity uses cost of equityMixing them over- or under-values the business
Tax and reinvestmentCash flow is after tax, capex and working capitalProfit is not cash
TimingYear-end or mid-year discounting is chosen on purpose and used everywhereMid-year adds roughly half a year of discounting back
Working capital signA rise in receivables or stock reduces cashA sign flip turns a cash drain into a gain
Bridge to equityNet debt, leases and minorities are taken off onceDouble counting moves value per share
Share countDiluted shares, including options in the moneyValue per share is wrong otherwise
Terminal value shareShow what % of value comes from the terminal valueOften 60-80%: the answer rests on one assumption

A worked example

Year-5 free cash flow 100, discount rate 9%, terminal growth 3%: terminal value = 100 x 1.03 / (0.09 - 0.03) = 1,717. Change growth to 4% and it becomes 2,080 - over 20% more from one input. That is why the sensitivity table (growth across, discount rate down) belongs next to every DCF answer.

Quick self-test

Set growth to 0% and the discount rate up 2 points. If the value per share barely moves, a link is probably broken. If the balance sheet stops balancing, the DCF is reading from the wrong place.

Questions

What terminal growth rate is normal?
Many analysts use something near long-run inflation, often 2-3% for developed markets. It must stay below the discount rate.
Should I use mid-year discounting?
It assumes cash arrives evenly through the year. Either choice can be fine; what matters is using one choice everywhere and saying so.
Can a DCF be checked automatically?
The mechanics can: signs, links, growth vs discount rate. Whether the forecast is realistic needs a person.

Sources

See the Excel model check

We check DCF and three-statement models cell by cell and list each issue with the fix.

See the Excel model check