Unit 3 · Unit 3: Revenues
The top line drives everything below it. Learn to build revenue bottom-up, top-down, or by growth rates — and to fade it honestly to a rate that can last forever.
Revenue sits at the top of the waterfall, so its error propagates into every line below it.
Revenue is the first line of the income statement and the foundation of the entire forecast. Costs are frequently modelled as a percentage of revenue, working capital scales with revenue, and capex is often sized to support a revenue level. Get the top line wrong and every downstream number inherits the error — margins, free cash flow, and ultimately the valuation. This is why analysts spend more time on revenue than on any other single assumption.
It is also the line you know least about. Margins and tax rates are relatively stable and mean-reverting; revenue depends on demand, competition, pricing power, and the macro cycle — none of which you control. A disciplined modeller treats the revenue forecast as a range of defensible outcomes, not a point estimate, and is explicit about the assumptions driving it.
Garbage in, garbage everywhere Because so many other lines are anchored to revenue, a 2% error in your growth rate is never just a 2% error in revenue — it compounds across the horizon and is magnified in the terminal value, where most of a DCF's worth usually sits.
Two independent routes to the same number — the best forecasts reconcile both.
Bottom-up forecasting builds revenue from the operational drivers of the business. You forecast the quantity sold and the price, then multiply and sum across products or segments. It is the most credible approach when you understand the unit economics — units shipped, subscriber counts, same-store sales, average revenue per user — because each driver can be defended on its own.
Revenue = Σ (Volume_i × Price_i) for each product or segment iForecast volume and price separately for each line of business, then sum. This separates a pricing thesis from a volume thesis.
Top-down forecasting works in the opposite direction: estimate the size of the total addressable market (TAM), assume how it grows, and apply a market-share assumption to land on the company's revenue. It is most useful for early-stage or fast-changing markets where unit economics are unstable, and as an independent cross-check on a bottom-up build.
Revenue = Total Addressable Market × Market ShareIf your top-down and bottom-up numbers disagree wildly, you have learned something — go find out which assumption is wrong.
| Approach | Best when | Watch out for |
|---|---|---|
| Bottom-up | You understand unit economics (units, price, segments) | Spurious precision; many small errors compounding |
| Top-down | Early-stage or fast-changing markets; cross-checks | An implausible market-share path; soft TAM estimates |
| Growth-rate | Mature, stable businesses; quick first pass | A flat rate that ignores fade and competition |
Triangulate, don't pick The strongest revenue forecasts are built one way and checked the other. If a bottom-up build implies a market share the top-down lens says is impossible, the model is telling you to revisit your drivers before you defend the number.
No company grows above the economy forever — fade your opening rate to a sustainable terminal rate.
The growth-rate method applies a percentage increase to the prior period's revenue. Its great danger is holding a high rate flat across the whole forecast. Competition, market saturation, and the law of large numbers all pull growth down over time. A credible forecast therefore fades the opening growth rate toward a sustainable terminal rate — a rate the company could in principle grow at forever.
Revenue_t = Revenue_(t−1) × (1 + g_t)g_t is the growth rate for period t — and g_t should generally decline as t increases.
g_t = g_start − (g_start − g_terminal) × (t / N)Fade the opening rate g_start to the terminal rate g_terminal linearly across the N-year horizon. Other shapes (exponential decay) are common too.
The terminal rate is a hard ceiling A terminal growth rate above long-run nominal GDP (roughly 2–4% in developed markets) implies the company eventually becomes larger than the whole economy. That is impossible, and reviewers will flag it instantly. Keep the terminal rate at or below long-run nominal GDP.
Multiple fitted growth models, an automatic terminal fade, and consensus plus event factors layered on top.
In The Cash Flow Projector you don't start from a blank growth assumption. When a model is seeded from SEC EDGAR history, the engine fits multiple revenue-growth models to that history and applies a terminal fade automatically, producing a default trajectory you can inspect and override on the Assumptions tab. You are choosing between defensible, fitted paths rather than typing a single hopeful number.
Consensus and catalysts, built in The Factors tab and Market Intel triangulate consensus from several sources and overlay AI-clustered event factors onto projected revenue and EPS. This is the external, real-world check the Excel CFA course can't give you — your top line is tested against what the market and the news flow actually expect.
Override with intent The fitted default is a starting point, not a verdict. The skill the course is building is knowing when the history is a good guide and when a structural change — a new product cycle, a pricing reset, a recession — means you should override it.