These are the twelve mistakes that will catch every analyst at least once. Memorise them and spot them in other people's models.
Mismatched currencies in cash flows and discount rate
Forecasting cash flows in KES and discounting at a USD WACC. The fix: keep cash flows and discount rate in the same currency. If you must switch, use forward exchange rates, not spot.
Real cash flows discounted at a nominal rate (or vice versa)
If your cash flows do not include inflation, your discount rate must not include inflation either. Mixing the two is a sign you have not thought about which world the model lives in. The standard practice is nominal cash flows and nominal discount rate.
Using net income instead of FCFF
Already covered in module 2. Bears repeating because it remains the most common error in undergraduate models.
Forgetting the mid-year convention
Cash flows arrive throughout the year, not on December 31. The mid-year convention discounts year-1 cash flow at 0.5 years instead of 1.0 — a small adjustment that increases valuation by ~5%. Use it consistently across the explicit period.
Stub period in the first year
If you are valuing as of June 30 but the fiscal year ends December 31, the first 'year' is only 6 months. The cash flow needs to be pro-rated and the discount factor adjusted. Almost every spreadsheet template gets this wrong on first build.
Wrong tax rate
Using historical effective tax rate (which includes one-offs) instead of marginal forward tax rate. Or using the wrong jurisdictional rate for the bulk of the business. Or forgetting to update for the recent tax-rate change in the budget.
Equity vs enterprise confusion
Discounting FCFE at WACC (should be Ke), or discounting FCFF at Ke (should be WACC), or forgetting to subtract net debt to bridge from EV to equity value.
Double-counting the cash
Some models include cash on the balance sheet AND include interest income from that cash AND fail to subtract the cash from net debt. Pick one place to model the cash and stick to it.
Inconsistent capex and depreciation
Forecasting capex of 8% of revenue while depreciation runs at 4%. Over 5 years, the asset base balloons and the model's implicit return on invested capital becomes nonsensical. Reconcile the two.
Terminal-year FCF not in steady state
Year 5 FCF should reflect a steady-state business — growth-rate equal to perpetual growth, margins at maturity, capex equal to maintenance capex grossed up for terminal growth. Many models leave year 5 still 'investing for growth', which inflates TV.
Forgetting the minorities
If the company has minority interests (a partial-ownership stake in a subsidiary), the consolidated EV includes the full subsidiary value but the equity value belongs to all shareholders. Subtract minority interest at fair value to bridge from EV to attributable equity.
Selective comparable-company benchmarking
Choosing the comparables that flatter your conclusion. Discipline: define the comp set before building the model and document why each company is in or out.
Model review checklist
Before you send a DCF to anyone senior, run through these twelve. Print them out. Tick each one. The act of checking is the discipline.
Exercise
You are reviewing a junior's DCF on a Kenyan listed corporate. List four of the most common DCF errors covered in this module and walk through how you would test for each in a 30-minute review. Then describe the kindest way to deliver the criticism to the junior who built the model.