Budgeting and Forecasting: Drivers, Seasonality and Variance Analysis
Budgeting & Forecasting
A budget is a financial plan for a defined period, usually annual and fixed; a forecast is a rolling estimate updated as actuals arrive; a long-range plan covers three to five years. Questions often turn on process: top-down (management targets, fast, low buy-in) versus bottom-up (departments build it, higher ownership, risk of sandbagging), and zero-based budgeting (every cost justified from zero, strong overhead discipline) versus incremental (last year plus a percentage).
Build forecasts on drivers rather than a single historical growth rate: units x price, headcount x cost per head, stores x sales per store. Driver-based models make sensitivity analysis meaningful and expose weak assumptions. Model seasonality explicitly - dividing an annual plan by twelve will misstate peak cash needs in a seasonal business.
Traps: conflating cash with accruals (a budget showing profit but omitting capex and working capital overstates cash); treating sunk costs as relevant; double counting inflation by inflating both volumes and prices; and ignoring that a favourable variance in one line may be offset elsewhere.
Variance decomposition: total variance = volume effect + price/rate effect. For revenue, volume effect = (actual volume - budget volume) x budget price; price effect = (actual price - budget price) x actual volume.
Worked example: budget 1,000 units at 10 (=10,000); actual 900 units at 11 (=9,900); variance -100. Volume effect = (900-1,000) x 10 = -1,000. Price effect = (11-10) x 900 = +900. Total -100 - a serious volume miss almost fully masked by pricing, which deserves its own line in the commentary.