Your CEO asks you mid-committee: "what happens if the peso drops 15% before year-end?" Your team has a base case. You ask for two weeks to model it. The CEO grimaces, says "ok," and your team goes off to process the question.
Here's the moment that defines a CFO: do you deliver the scenario in a week, or do you explain that you already had that question covered because three coherent scenarios live in parallel in your model and you only need to swap one parameter to see it?
Scenario analysis is that difference. It's not building three separate spreadsheets. It's building ONE model where three futures live at the same time, and where any CEO question can be answered live.
By the end of this module you'll have it down. Let's go.
Your CEO asks: "what if the peso drops 15% before year-end?" How do you answer?
In plain language
Before the mechanics, the four basic questions.
Why do this at all?
To make decisions under uncertainty with discipline. A single number (the "base case") hides the truth that the business depends on things you don't control. Three coherent scenarios force you to prepare actions for each. When the world moves, you already know what to do.
Who uses it?
FP&A builds it together with each function's leaders. It's used by: the board for strategic decisions (capex, M&A, capital), the CFO for contingency planning (credit lines, covenants), operating leaders for action thresholds ("if the stress scenario materializes, we freeze hiring").
When does it show up?
When building the annual plan. Before any big decision (investment, launch, acquisition). In every quarterly board review. When a new risk shows up on the horizon (a competitor, a regulatory shift, an FX shock).
What if we skip it?
You operate with a single assumed future. When reality diverges, you improvise. Big decisions (how much capex, how much headcount, what credit line) get made against a central case that no longer applies. You lose months making reactive decisions you could've made proactively.
Meet Andina S.A. (again)
Same Andina from Module 2.6 — LATAM coffee company, $200M revenue, based in Chile. The FP&A team presents the annual plan to the board: base revenue $80M, EBITDA $12M (15% margin).
One director asks: "what if the competitor enters aggressive with 30% lower price?" Another: "what if the peso falls?" A third: "what if the new product launch is delayed?" Three valid questions. Zero concrete answers.
The CFO promises to have scenarios by next session. Two weeks later they show up with three separate spreadsheets: optimistic, base, stress. The numbers don't reconcile across them (each spreadsheet took different shortcuts). The board notices. The session falls apart.
The problem isn't lack of scenarios. It's how they were built. The exit is ONE model, three driver sets, three outputs that fall from the same engine. The visual below shows it.
Three futures, side by side
Three coherent scenarios side by side. Each has 4 explicit drivers (volume, price, launch, COGS) and 3 outputs (revenue, EBITDA, year-end cash).
Tap a scenario to focus on its decision implication. The comparison bar below shows the actual spread between the three futures.
The key: each scenario is a coherent story, not a mix of random numbers. Optimistic = high volume + successful launch. Stress = competitor entry + FX pressure. Base = the central assumption.
Interactive visual
Three futures, one model: Andina S.A.
Tap a scenario to focus on its decision.
What you're looking at
Three coherent scenarios, not three random numbers. Each has explicit drivers, outcomes that fall from the model, and a specific decision to make. This is the exercise a board expects from a mature FP&A team: not a single forecast, but three thought-through futures with their operational implications.
That difference between $15M (optimistic) and $7M (stress) in EBITDA isn't an error margin. It's the real sensitivity of the business to things you DON'T control. If the board approves the plan knowing you live in that range, the decisions they make are different.
Concrete example: if the stress scenario requires opening a $5M credit line by Q3, that decision gets made TODAY (not when the problem shows up). That's what good scenario analysis enables: preventive decisions, not reactive ones.
The mechanics: how to build it well
- Start with the model, not with the scenarios. You need a driver-based model (Module 2.4). Without it, "scenarios" turns into three inconsistent spreadsheets. With it, scenarios are three sets of inputs in the same engine.
- Define scenarios as stories, not as numbers. "Optimistic" isn't "+10% in everything." It's a specific narrative: the competitor doesn't enter, we launch the product in Q2, the peso holds. Each driver in the model receives the value that falls from that story.
- Three scenarios, not seven. Best/base/worst is the standard for a reason: the board can hold three futures in their head. Five or more get confusing. If your business is complex and you need more, add "special stress" for risks the board specifically calls out.
- Assign probabilities. Even if rough (20%/60%/20% is typical), putting probabilities forces the board to think in expected value. Without probabilities, scenarios become a list of possibilities with no relative weight.
- Tie each scenario to a decision. The scenario that doesn't imply a different decision is decoration. Optimistic should trigger A, base should trigger B, stress should trigger C. That's the test that the scenario is useful.
- Sensitivity analysis = what happens if I change ONE driver, holding everything else constant? Useful for understanding which variable matters most. Typical visualization: tornado chart.
- Scenario analysis = three (or more) coherent stories with MULTIPLE drivers changing together. Useful for preparing decisions under different futures. Visualization: side-by-side comparison.
- Monte Carlo = thousands of simulations with drivers that have probability distributions. Useful when the important answer is the distribution (probability of covenant breach, P10/P50/P90 of cash). Overkill for most operational decisions; necessary for major capital allocation decisions.
- Practical rule: most mid-market companies live in scenarios + sensitivity. Monte Carlo enters when the cost of error is high and the uncertainty is genuinely probabilistic (M&A, project finance, large capital allocation).
Adversarial check
Adversarial check
1.The CEO asks you mid-committee: "what's the worst case?" Your model has three scenarios (optimistic, base, stress). The stress shows EBITDA of $7M (vs base $12M). What do you tell them?
2.Your FP&A team built three Excel files for the three board scenarios: optimistic.xlsx, base.xlsx, stress.xlsx. Each with its own assumptions. What's the biggest risk of continuing this way?
3.When is it really necessary to use Monte Carlo instead of scenarios?
Exit checklist
Suggested re-review: 6 months, or when a new risk shows up on the horizon that requires a dedicated scenario.
Optional
Go deeper
Sources and books to dig into the original material