On 12 June 2026, the National Bank of Belgium quietly cut its growth forecast for the year to 0.6%, roughly half of what had been projected only months earlier. The trigger wasn’t a spreadsheet error, it was the war in Iran, an event no economist could have priced in with certainty a year ago. Alongside its official baseline, the ECB had already prepared a set of alternative scenarios for exactly this kind of shock, because betting an entire monetary policy on one forecast would be reckless.
That discipline isn’t reserved for central banks. Any business making decisions about staffing, stock, pricing or financing for the year ahead needs the same underlying exercise. It’s called scenario modelling, and unlike a lot of finance vocabulary, it’s a genuinely practical tool that scales down to a five-person company as easily as it scales up to the Eurosystem.
In this article, we cover what scenario modelling actually is, why businesses should bother with it, what to prepare and which data to gather, how to keep it alive rather than let it gather dust, and, just as importantly, which conclusions you’re entitled to draw from it and which you’re not.
Key Takeaways
- Scenario modelling is not forecasting. A forecast gives you one likely path. Scenario modelling deliberately builds several plausible, internally consistent futures so you can see how your business holds up across all of them.
- The output is a decision tool, not a prophecy. A scenario passing the test tells you your business can withstand that shock. It says nothing about the next one, and it should never be read as a profit projection.
- Good scenario modelling depends more on discipline than software. A handful of well-chosen variables, tracked consistently and revisited often, will outperform an elaborate model built once and never touched again.
What Is Scenario Modelling?
Scenario modelling is a planning method that defines several distinct, plausible future environments and tests how a business’s finances would hold up in each of them. It’s often confused with three related but narrower tools, and the differences matter:
| Method | What it changes | What it answers |
|---|---|---|
| Forecasting | Extrapolates current trends forward | “What will most likely happen?” |
| Sensitivity analysis | One variable at a time | “What happens if just the price moves by 5%?” |
| Stress testing | One severe, often single-cause shock | “Would we survive this specific extreme event?” |
| Scenario modelling | Several correlated variables together, within a coherent narrative | “What are the different worlds we might operate in, and how do we perform in each?” |
The distinction that matters most for a founder: a forecast produces one number to plan around. Scenario modelling produces several defensible stories, none of which is “the” answer, each of which changes how you’d act.
It’s worth stressing what scenario modelling is not trying to do. The method that dominates professional and academic practice, known as “intuitive logics,” deliberately avoids assigning probabilities to scenarios. Each scenario is meant to be equally plausible. The value isn’t in picking the most likely one; it’s in stress-testing your own assumptions against several coherent alternatives so that surprises become “we’ve thought about this” rather than “we never saw this coming.”
A Short History: Betting Against the Single Story
Scenario planning as a business discipline traces back to Royal Dutch Shell in the early 1970s, under a planner named Pierre Wack. Rather than producing a single-point forecast for oil markets, Wack’s team built multiple, internally consistent stories about how the future might unfold, one of which involved OPEC nations gaining enough leverage to trigger a sudden, sharp oil price shock.
Wack was explicit that these scenarios were not predictions. He called them rehearsals for the mind. When the 1973 oil embargo actually hit, Shell was the only major oil company that had already imagined, discussed, and planned around that world.
It didn’t make Shell psychic. It made Shell fast, because the decision had effectively already been rehearsed before the shock arrived. That’s the entire point of the exercise, being prepared: scenario modelling doesn’t improve the accuracy of your predictions. It improves the quality and speed of your decisions once reality picks one of the paths you’ve already thought through.
Preparing Your Model
Before opening a spreadsheet, a scenario exercise needs some structure, or it collapses into a vague brainstorm. The standard approach follows a few sequential steps:
- Define the focal issue. Not “the future of our business” in general, but a specific decision: should we hire two more people, extend our credit terms, take on a loan, enter a new market?
- List the forces that affect that issue. Cast a wide net across political, economic, social, technological, environmental and legal drivers, then group related forces into a handful of clusters.
- Split predetermined elements from genuine uncertainties. Some things you can reasonably assume (a signed contract, a known tax change already legislated). Others you genuinely cannot know (a client’s payment behaviour next year, energy prices, a competitor’s next move).
- Select your critical uncertainties. Pick the two forces that combine the highest impact on your focal issue with the highest genuine uncertainty. These become the axes of your scenario space, typically producing four distinct, named futures.
CFOrent Tip: start smaller than you think you need to
Founders often try to model everything at once and stall before finishing. Pick one focal issue (e.g. “can we afford to hire this quarter?”), two critical uncertainties (e.g. “client payment speed” and “energy costs”), and build from there. A simple model that gets updated monthly beats a comprehensive one that gets built once and abandoned.
What Data You Need
Scenario modelling for a business’s finances draws on a mix of internal records and external reference points. Neither is optional. Internal data grounds the model in reality, external data keeps your assumptions honest rather than optimistic.
| Data type | Source | Why it matters |
|---|---|---|
| 12–24 months of revenue and cost history | Internal accounting | Establishes a credible base case, not a guess |
| Cash flow drivers: sales volume, pricing, collection period (DSO), inventory levels, major fixed and variable costs | Internal | These are the levers you’ll actually flex between scenarios |
| A rolling forecast structure | Internal | A 13-week rolling forecast for short-term liquidity, paired with a 12-month view for strategic planning, is a common and workable SME setup |
| Macroeconomic reference points (GDP growth, inflation, interest rates) | External: NBB, European Commission, Eurostat | Prevents your “worst case” from being either fantasy or needlessly panicked |
| Industry or peer benchmarks | External: sector associations, bank sector reports | Sanity-checks whether your assumed ranges are realistic for your market |
A Worked Example
Take a business with monthly revenue of €100,000, cost of goods sold at 45% of revenue, fixed operating costs of €48,000 per month, and a cash buffer of €60,000.
Analysing Cash Flow Across Scenarios
Monthly cash flow = Revenue × (1 − 0.45) − €48,000 = 0.55 × Revenue − €48,000
Base case (flat revenue): 100,000 × 0.55 − 48,000 = +€7,000/month. Cash buffer grows.
Adverse case (a 20% demand shock, echoing a scenario the NBB flagged for 2026): 80,000 × 0.55 − 48,000 = −€4,000/month. At that burn rate, the €60,000 buffer would last roughly 15 months, so not an immediate emergency, but a clear early warning worth tracking monthly.
Reverse stress test: instead of picking a shock and measuring the damage, start from an outcome you can’t afford (say, exhausting your buffer within 6 months) and work backwards. Solving for the revenue level that produces a €10,000/month burn (60,000 ÷ 6 months) gives a required revenue of roughly €69,000, a sustained drop of about 31% from base. That’s your trigger line: if sustained revenue decline approaches that mark, action needs to happen well before the buffer runs out, not after.
For illustrative purposes only. A tailored model would reflect your actual cost structure and financing options.
Keep It Running: Follow-Up and Iteration
A scenario model that’s built once and filed away is close to worthless. It needs a frequent updates:
- Update cadence should match how volatile your situation is. A stable business might revisit its scenarios quarterly. A business with tight liquidity should be re-checking weekly, since early warning only works if it arrives early.
- Keep the exercise collaborative, not delegated. One of the most consistently cited failure modes in scenario planning is leadership handing the exercise to junior staff or outside consultants and only reappearing for the final presentation. If the people making decisions weren’t part of building the scenarios, they tend not to act on them when the moment actually arrives.
- Attach a trigger and a named action to each scenario, not just a number. “If DSO stretches past 45 days, we draw down the credit line” is a plan. “DSO might get worse” is an observation.
CFOrent Tip: pair scenario modelling with a reverse stress test
Forward scenarios ask “what happens if X occurs?” A reverse stress test flips the question: “what would have to happen for us to be in real trouble?” Running both gives you a fuller picture, since forward scenarios can miss combinations of smaller shocks that reverse stress testing surfaces directly.
Do’s and Don’ts
| Do | Don’t |
|---|---|
| Keep the number of scenarios small (2–4) and clearly defined | Build a dozen barely-distinct variations that dilute focus |
| Ground every assumption in either your own data or a named external source | Invent a “worst case” number that simply feels appropriately dramatic |
| Make each scenario internally consistent | Combine contradictory assumptions in the same scenario without explaining the link |
| Involve the people who will actually make the decisions | Delegate the exercise entirely and review only the final slide |
| Revisit and update on a set cadence | Treat the model as finished once it’s built |
| Attach a concrete trigger and action to each scenario | Leave scenarios as descriptions with no decision attached |
Two traps are worth naming explicitly, because they undermine the very tool meant to counter them. Anchoring on today’s conditions and quietly building every scenario as a minor variation of the present. And the planning fallacy, where optimism creeps back in and the “adverse” scenario ends up looking suspiciously mild. Both are well documented in strategic planning research, and both are best countered by simply asking the team out loud: “what are we assuming will continue that might not?”
Conclusions (not) to make
You’re entitled to conclude
- Which of your cost or revenue lines carries the most exposure, and roughly by how much, under a plausible shock.
- Whether your current cash buffer, credit lines, or covenants are adequate for the range of outcomes you’ve modelled.
- Which specific trigger points deserve a pre-arranged response, arranged calmly now rather than improvised under pressure later.
You are not entitled to conclude
- That your business’s actual future profit will land inside the range you modelled. Even EU banking regulators are explicit that adverse scenario results should not be read as profitability projections, precisely because the scenario is hypothetical by construction.
- That surviving one scenario proves resilience to all future shocks. A model that holds up against a demand shock says nothing about, say, a supplier failure or a currency shock it was never built to test.
- That the assumptions baked into your model (a static cost base, a stable client mix, a specific time horizon) don’t have blind spots. Every model has a boundary. Know where yours sits before trusting the output too far.
Conclusion
Scenario modelling won’t tell you what’s going to happen next year. What it does is make sure that when reality picks one of the paths you’ve considered, your business isn’t rehearsing its decision for the first time under pressure. The exercise scales from a national central bank down to a five-person company: define the question that actually matters, gather the data honestly, build a handful of distinct and internally consistent futures, and revisit them often enough that they stay useful.
At CFOrent, we help founders and SMEs build financial models that hold up, whether that’s a straightforward cash flow scenario or a full review of your business’s exposure to the year ahead. Feel free to get in touch if you’d like a tailor-made approach to scenario modelling for your business.
Sources
- National Bank of Belgium. Macroeconomic projections for Belgium, published 12 June 2026. nbb.be
- European Central Bank / European Banking Authority. EU-wide stress test methodology and results, 2025. bankingsupervision.europa.eu
- European Banking Authority. Guidelines on Institutions’ Stress Testing (EBA/GL/2018/04), including reverse stress testing requirements.
- Wack, P. (1985). Scenarios: Uncharted Waters Ahead and Scenarios: Shooting the Rapids. Harvard Business Review.
- Wright, G., Bradfield, R. & Cairns, G. Does the intuitive logics method produce “effective” scenarios? — foundational academic treatment of the intuitive logics scenario method.
- McKinsey & Company. Overcoming obstacles to effective scenario planning (2015).
Related articles
- Business Processes, Metrics
