What a financial model is

A financial model is a single, practical thing: a set of linked calculations that turns assumptions about a business into consequences you can inspect. You put in a view of the world — revenue grows 12% a year, the gross margin holds at 40%, the company opens six new outlets — and the model tells you what that world implies: how much cash the business throws off, whether it can service its debt, what it is worth, and where it breaks. Everything else — the formatting, the tabs, the terminal-value debate — is machinery in service of that.

It is worth being precise about what a model is not. It is not a prediction. No competent analyst believes the number in the corner of their DCF. A model is a way of making a set of assumptions explicit, internally consistent, and open to challenge. Its value is not the answer it produces; it is the argument it lets you have. When a model says a company is worth KES 40 a share and the market says KES 25, the useful work is not declaring the market wrong — it is finding which assumption is carrying the disagreement.

The three questions every model answers

Whatever the industry, a model exists to answer some combination of three questions, and knowing which one you are being asked changes what you build. First, can the business pay its bills — the liquidity question, asked by lenders, credit committees, and a CFO staring at a cash-flow forecast. Second, is it worth owning — the value question, asked by investors, acquirers, and boards weighing a capital-raise. Third, can it survive a bad year — the resilience question, asked by anyone who has watched a profitable company run out of cash. A bank building a credit model cares most about the first. An analyst valuing Safaricom cares most about the second. A founder deciding whether to take on a loan to expand needs all three.

That rule sounds obvious and is broken constantly. A junior analyst writes =C10*1.12 because revenue grows 12%. Six months later the growth assumption is 9%, they change it in the assumptions tab, and the model does not move — because the 1.12 was welded into a formula three tabs away. The number on the page is now a fiction, and nobody knows. The professional writes =C10*(1+$B$4), points $B$4 at a single labelled cell that says 'Revenue growth', colours that cell blue so everyone knows it is an input, and never types a growth rate into a formula again. The whole craft of modelling is, to a first approximation, the disciplined application of that idea across a hundred rows.

Inputs, calculations, outputs

Every good model has the same three layers, and keeping them physically separate is half the battle. Inputs are the assumptions — the things you are allowed to change — and by convention they are coloured blue. Calculations are the engine: the income statement, the balance sheet, the cash flow, the schedules that feed them, all black formulas that no human edits by hand. Outputs are what the decision-maker actually reads: the valuation, the credit metrics, the sensitivity table, the one chart that ends the meeting. When these three are tangled together — when someone has to overwrite a formula to run a scenario — the model is already broken. When they are clean, you can hand the assumptions tab to a stranger, let them change anything blue, and trust that every consequence flows through correctly.

What you will build

This course is not a tour of finance theory. Over the next eleven modules you will build a complete, integrated three-statement model on a real, recognisable African business — a company you can look up, whose accounts you can sanity-check against reality — and then value it. By the end you will have produced the two artifacts every entry-level analyst is expected to build and no junior is taught properly: an income statement, balance sheet, and cash flow that link and balance to the cent, and a DCF that hangs off them and survives a stranger's review.

• A structured assumptions tab, and the colour and layout discipline that makes a model auditable
• An income statement built from real revenue drivers, not a single growth rate
• A balance sheet and cash flow statement that link to it — and the balance check that proves they do
• The supporting schedules — working capital, capex and depreciation, debt and interest — that feed the statements
• A DCF valuation driven straight off the model's free cash flow
• Sensitivity and scenario analysis, and the error-proofing that lets you hand the model over without fear