Decision-making under deep uncertainty
No one can promise to get the future right. Every big decision rests on assumptions that may turn out to be wrong. We stress-test your strategy across thousands of plausible futures and map how each decision holds up across them, under which conditions it breaks, and what to watch for before it does.
We tell you under which futures your decision breaks. The decision is yours.
Why it's different
This is not a replacement for your DCF or your investment model — it's an exploratory layer wrapped around it. Same decision, same model logic. The difference is how many futures you look at before you commit.
The standard approach
A well-built model, best-estimate assumptions, a confident answer. It tells one story — usually the story everyone in the room already believes. What it cannot tell you is how often that story survives contact with reality.
The exploratory layer
The same model, run across the full range of plausible assumptions. Now the question changes: not "what is the expected IRR?" but "in what share of futures does this deal fail — and what do those futures have in common?"
The analysis
Drawn from our Meridiaan Groep case dossier — a synthetic Dutch family operating company weighing four strategies. All data below is fully synthetic; the analytical machinery is exactly what runs on your numbers.
Hover or tap a strategy to isolate its futures
The base-case winner is not the robust winner. "Expand now" has the highest expected value — and underperforms doing nothing in roughly a third of futures. Ranked by survival rather than by the average, the ordering inverts.
Scenario discovery (PRIM) searches all 5,000 futures for the region where failures cluster. Here, failure isn't spread randomly — it concentrates where cost pass-through drops below 62% while input inflation runs above 3.8%.
That's not the variable this board debated most. It rarely is. The conditions that actually break a strategy are usually mundane, measurable — and monitorable.
Becomes a board tripwire:
"Pass-through below 62% while trailing input inflation exceeds 3.8% — tripwire fires; the board's pre-agreed response applies."
All figures are drawn from fully synthetic case data and are illustrative of the deliverable format, not of any real company.
The method
Every engagement is structured around the XLRM framework: a transparent, inspectable map of what you can't control, what you can, and how the two interact. Glass-box models with documented equations — no black boxes, ever.
Demand, rates, input costs, regulation — every factor outside your control, sampled systematically across its plausible range rather than fixed at a guess.
The candidate decisions on the table — expand, stage, hold, divest — each tested against the same set of futures so comparisons are honest.
A transparent quantitative model connecting uncertainties and levers to outcomes. Every equation documented and inspectable by your team.
Outcomes in your vocabulary — IRR, MOIC, and downside frequency for funds; continuity, liquidity, and generational wealth preservation for families.
Case dossiers
Each dossier is a complete, board-ready analysis built on fully synthetic data. One is public — download it and judge the work directly. The other two are available on request.
Single-deal robustness stress test
A synthetic battery-recycling growth equity deal, stress-tested across 5,000 futures. The dossier maps exactly which combinations of price, ramp-up, and cost conditions drive the deal below hurdle — before the IC signs.
Download the full dossier (PDF)Strategy robustness for a family operating company
Four candidate strategies for a synthetic Dutch family business. The base-case winner inverts once uncertainty is taken seriously — and the variable that decides survival is pricing power, not the ones boards usually debate.
Whole-balance-sheet vulnerability scan
Operating company, real estate, and securities portfolio scanned together. The signature finding: failure concentrates where four mundane macro conditions coincide — and sleeve-level reporting actively hides it. Diversification on paper is not diversification under stress.
All case data is fully synthetic and clearly labeled as such. No client information appears in any dossier.
After the analysis
Every quantified failure condition converts into a concrete threshold your board tracks quarterly. When a tripwire fires, you act early and deliberately — instead of debating whether "the environment has changed."
Optionally continued as a quarterly monitoring retainer: we re-run the ensemble against your updated numbers and report which tripwires moved.
Engagement formats
One deal or one decision, stress-tested end to end. A board-ready dossier with failure mapping and tripwires. The fastest way to see the method on your own numbers.
Multiple candidate strategies compared under the full uncertainty ensemble — including staged and hybrid options — with a robustness scorecard and regret analysis.
Your whole position — operating assets, real estate, portfolio — scanned for shared exposures that sleeve-level reporting hides. Optionally continued as a quarterly tripwire retainer.
Principal-level engagement on every project. Open-source, inspectable tooling. Methodology from the RAND / TU Delft decision-science tradition — built for decisions where being wrong is expensive.
Tell us the decision you're facing, or request one of the case dossiers. We reply personally within two business days.