Defend the decision.
See which geopolitical scenarios move your numbers, how likely they are, and why — before the board asks.
Worked example
Engine selection for a 240 MW behind-the-meter gas plant
A Southeastern power developer · Southeastern United States · values in USD millions
Tariffs hold at current levels
- Probability
- 41%
- Expected value
- $212M
- Downside (P10)
- $118M
- Evidence
- 9 rows
Tariffs escalate; turbine lead times stretch
- Probability
- 27%
- Expected value
- $96M
- Downside (P10)
- −$34M
- Evidence
- 11 rows
Tariffs reverse; OEM backlog clears
- Probability
- 14%
- Expected value
- $268M
- Downside (P10)
- $171M
- Evidence
- 6 rows
Gas regime shifts higher
- Probability
- 18%
- Expected value
- $131M
- Downside (P10)
- $12M
- Evidence
- 8 rows
A real decision, run in Compass. Four scenarios, their probabilities from the evidence table, and the downside through the project's own numbers. As of November 6, 2026.
Who it is for
Power developers and IPPs — tariffs on turbines and transformers, gas regimes, interconnection policy.
Commodity producers — sanctions on buyers, chokepoints, demand shocks.
Data-center developers — chip export controls, power policy, tenant concentration.
Lenders and infra investors — the downside case every IC memo needs.
How it is built and checked
Scenarios.
Your asset, three to five named scenarios, each decomposed into the drivers that set its value.
Evidence.
Each scenario's probability comes from a table of evidence you can read, rate, and override. The AI keeps it current; you decide.
Memo.
One export the board, the IC, or the lender can inspect, with the reasoning attached.