We are going to value Safaricom (NSE: SCOM) end to end. The choice is deliberate: large, mature, predictable cash flows, listed comparables (regional and global telcos), and Kenyan-currency reporting that lets us avoid the FX layering. By the end of this walkthrough, you will have a base-case valuation, a sensitivity table, and a defensible answer to 'is it cheap?'.
Step 1 — gather the inputs
Pull the latest annual report (FY24, period ended 31 March 2024). Note: revenue ~KES 335bn, EBITDA margin ~50%, capex ~KES 50bn (incl. Ethiopia spend), net debt ~KES 100bn. Pull Bloomberg or Reuters for share price and shares outstanding to compute market cap.
Step 2 — forecast revenue, 5-year explicit
Three segments: Mobile (M-PESA + voice + data), Fixed (enterprise + home fibre), Ethiopia (loss-making, growing). For Mobile Kenya, model 5-year revenue growth at 7-9% nominal (in line with 2-year trailing); for Fixed, 12-14%; for Ethiopia, model toward EBITDA breakeven by year 4. Aggregate forecast revenue: roughly KES 350bn → KES 470bn over 5 years.
Step 3 — costs and margin
Hold operating margin flat at ~30%, with Ethiopia drag declining as the unit scales. Tax rate at 30% marginal. Capex tapers from KES 55bn down to KES 35bn as Ethiopia rollout completes. Working capital flat at 2% of revenue change.
Step 4 — WACC build
Rf (KES 10y) ≈ 13.5% (late 2024). Beta (regression vs NSE 25, 5y weekly) ≈ 0.85; comparable telco beta ~0.7-0.9. ERP build: developed-market 6% + Kenya CRP 5% = 11%. Ke = 13.5% + 0.85 × 11% = 22.85%. Cost of debt: KES corporate yield ~14%, after-tax 9.8%. Capital structure ~85% equity / 15% debt at market values. WACC = 0.85 × 22.85% + 0.15 × 9.8% = 20.9%.
Step 5 — terminal value
Gordon growth at 5% (above long-run inflation expectation of 5-6%, below long-run nominal GDP growth of 8-9%). TV/EBITDA implied multiple ≈ 6.4x. Cross-check against Kenyan comparables and global emerging-market telcos: regional comps trade at 4-7x EV/EBITDA, so 6.4x is at the upper end of the band but defensible for a market-leader, M-PESA-monetising business.
Step 6 — discount and assemble
Discount each year's FCFF at WACC, plus TV at the year-5 discount factor. Sum to enterprise value. Subtract net debt and minority interests to get equity value. Divide by shares outstanding to get implied share price. Compare to current market price.
Step 7 — sensitivity
Build the WACC × g table. With WACC ranging 19-23% and g ranging 4-6%, the implied price range is wide but bounded — typically a ±25% range from base.
Step 8 — defend the answer
The verbal defence has three parts: (1) Why the forecast is reasonable — anchored in the historical 5y, with a story about why the next 5y differs (or doesn't). (2) Why the WACC is right — the Rf, beta, and ERP choices each defended with one sentence. (3) Why the TV is right — both Gordon and exit-multiple cross-checks land in similar places.
The whole point of a DCF
Not to produce a number, but to produce a structured argument about what a business is worth and why. The model is the artefact of the argument; the spreadsheet is the calculator. The senior analyst's reputation rests on the argument, not the spreadsheet.
Exercise
Imagine you have just completed an end-to-end DCF on a Kenyan listed bank (KCB or Equity, your choice). The base case lands at a per-share value of KES 60 against a current market price of KES 45. The CIO sits down with you and asks: 'Convince me — in 5 minutes — that this is a real investment idea and that I should size it as a 3% position.' Write the verbal pitch. Then list three specific questions you should expect from the CIO and the answer you would give to each.