A financial model with a single forecast answers a question nobody asked — «what will happen?» — with a confidence nobody has. The forecast's precision is borrowed: the model must produce a number, so it produces one, and the decision meeting then quietly splits into two conversations: the one about the number, and the one everybody actually has, about whether the assumptions behind it deserve the trust the number's neat formatting lends them. Scenario modeling replaces the performance of certainty with the mechanics of judgment: instead of one future, the model carries several, and the decision is made against the set rather than against a point estimate that will be either right or embarrassingly wrong.
The change is smaller than it sounds and deeper than it looks. The model keeps its engine — the same operations, the same drivers — but runs it under two to four coherent futures rather than one. Each scenario is not a random perturbation: it is a named story about the world, with its own logic — the demand-arrives-late case, the competitor-moves-first case, the cost-base-breaks case — and the model translates each story into what it does to cash, runway, and the decisions the company was planning to make. The discipline that makes this real rather than decorative is the same one that separates strategic scenarios from theater: each future must be able to occur, and the difference between them must sit in assumptions the decision-makers actually worry about, not in a mechanical plus-or-minus around a base case that nobody believes anyway.
Three things, all of them decisions rather than numbers. The exposure map. Reading the same plan under three futures shows where the plan is fragile — which single assumption carries the outcome across all scenarios — and that knowledge changes the plan long before any scenario occurs. The trigger table. Each scenario ends with pre-decided responses: what the company does, at which observable indicator, if that future starts arriving. This converts the model from a prophecy into an instrument — the company stops asking the forecast to be right and starts watching for which future is forming. The honest raise. Scenarios change how much to raise: the amount is sized against the hard scenario, not the flattering one, which is why scenario-modeled companies raise amounts that survive their own bad luck instead of amounts that require it to stay away. Lenders and investors, in turn, read a scenario set as evidence about the team rather than about the market: nobody trusts a point forecast, but people trust the people who can show which futures they have considered.
One warning, from practice: scenario sets quietly die by decoration. The three futures are appended to the deck after the decision, as a risk-disclaimer slide; or the scenarios are built so similar that they are one forecast in costume; or the hard scenario is built to lose — loaded with every pessimism at once — so the base case wins the comparison by construction. The test is behavioral: a real scenario set changes at least one live decision — the amount raised, the phasing of a commitment, the definition of the kill-trigger — and if it changed none, it was theater, and the effort would have been better spent arguing honestly about the single assumption everyone was actually worried about.
A forecast is a claim about the future. A scenario set is a plan for several of them — and plans, unlike claims, can be executed.
Building the scenario layer into a financial model — the named futures, the exposure map, the trigger table — is standing work in the project and investment practice, done at the point where the model is still a decision tool and not yet a document of record.
We use cookies to analyse traffic and improve the site. Analytics is enabled only with your consent. Cookie Policy