Why model
A model turns your view of a business into numbers: revenue, costs, cash, debt and value. It forces you to be explicit about assumptions and lets you test how much the answer depends on each one.
Revenue
- Top-down: start from the economy or market. Forecast industry size from GDP growth or another driver, then apply the company’s market share.
- Bottom-up: build from the company’s own drivers: units × price, number of stores × sales per store, subscribers × revenue per subscriber, capacity × utilization.
- Hybrid: combine the two, using each as a check on the other.
Try a simple model. Each line follows from the drivers you set:
Costs and margins
- Cost of sales: often a share of revenue, but adjust for input prices, hedging and mix.
- SG&A: has fixed and variable parts; it often grows more slowly than revenue.
- Operating leverage: a company with mostly fixed costs sees profit swing more than revenue. Examples: software, airlines, steel.
- Financing costs: interest = average debt × interest rate.
- Taxes: use the effective rate, adjusted for known changes and one-off items.
The balance sheet and cash flow
- Working capital: forecast receivables, inventory and payables from DSO, DOH and days of payables.
- Capital spending: split into maintenance (roughly related to depreciation) and growth. A capital-intensive company growing fast needs a lot of cash.
- Financing: debt, buybacks and dividends follow management’s policy and the cash available.
Make the statements balance: cash or a revolving credit line usually acts as the plug.
Competition, inflation and technology
Porter’s five forces help judge whether margins can last: threat of new entrants, threat of substitutes, bargaining power of buyers, bargaining power of suppliers, and rivalry among existing firms.
Inflation or deflation affects margins through pricing power, input costs and the timing of price changes.
Technological change can shrink an old market or create a new one. Scenario analysis, for example a base case and a disruption case, captures the range.
Behavioural biases
- Overconfidence: ranges that are too narrow.
- Illusion of control: believing more data or detail means a better forecast.
- Conservatism: underreacting to new information.
- Representativeness: extrapolating a pattern from a few observations.
- Confirmation bias: seeking evidence that supports your view.
Remedies: compare forecasts with base rates, seek opposing views, track your forecasting record and keep models simple enough to understand.
Long-term forecasts
Choose a horizon long enough for the business to reach a normal state. Use normalized earnings for cyclical companies, adjust for discontinued businesses and acquisitions, and make sure the terminal growth rate is believable relative to the economy.
Exam focus
Expect a short forecast calculation, identification of a bias, the meaning of operating leverage, and how competitive forces or inflation would change your assumptions.