Ex-Ante

Information timing matters: a manager preparing a project forecast before approval has to use assumptions about demand, cost and implementation. Those assumptions can be reasonable without turning the forecast into a fact already established.

Ex-ante analysis helps choose among options by comparing expected benefits, costs and risks before money or resources are committed, so the question is whether the assumptions and alternatives are useful, not whether every forecast will prove exact. A forecast should identify the information cutoff, because later knowledge should not quietly be inserted into a supposedly original prediction.

Preserving the version and date allows a fair comparison with what the decision-maker could actually know at the time. The OECD's regulatory impact analysis material discusses assessment before policy decisions and its relationship with later evaluation, emphasising structured comparison and evidence rather than treating forecasts as automatic instructions.

Expected and realised returns are different, so an investor may expect a positive return but experience a loss after conditions change. The observed outcome alone does not establish that the initial analysis was unreasonable, though it can reveal assumptions that need improvement.

Probability-weighted estimates make uncertainty explicit by combining several possible outcomes, but the probabilities are themselves assumptions that need explanation and cannot be made reliable merely by entering them into a spreadsheet. Sensitivity analysis asks how the result changes when an input changes, showing whether a project depends heavily on a sales price, interest rate or implementation date.

A forecast with fragile economics should not be described as strong simply because its base case is positive. Scenario analysis examines combinations of conditions, as a recession could change demand, costs and customer credit together, and correlated changes can matter more than the separate sensitivities considered in isolation.

Ex-post review compares outcomes with the original plan, separating execution differences from changes in external conditions while recognising that some causes are difficult to isolate. The aim is learning and accountability, not automatically blaming whoever made the earlier forecast.

The choice of comparison matters too, because a policy outcome may improve even if the policy had little effect, so a meaningful assessment needs a baseline or counterfactual rather than assuming every change after an intervention was caused by it. A business should distinguish an ex-ante target from an expected forecast, since a target expresses an intended result while a forecast estimates what is likely under stated assumptions.

Decision records should preserve assumptions, alternatives and reasons for choosing, so that when outcomes differ the record helps identify whether the problem was information, judgment, implementation or a new shock. For a non-finance manager, label projections as ex-ante, state their date, explain the range of outcomes and review realised results against the actual original analysis, using hindsight to improve future decisions without pretending it was available before the event.