Scenario Probability: Setting the Odds for Bull / Base / Bear / Stress

Where do the "probabilities" in the probability-weighted IRR come from? The engine requires you to assign a probability of occurrence to each of four futures—Bull (optimistic), Base (baseline), Bear (pessimistic), and Stress (extreme stress)—and each must be grounded in evidence, not a convenient fixed value.

Why a Fixed 20/50/25/5 Is Wrong

Many people split the scenarios evenly, or forever apply one fixed set of proportions. That amounts to giving up thinking. A company loaded with debt and facing a regulatory investigation should obviously have higher Bear/Stress probabilities; a company flush with cash and steady demand is the opposite. The probabilities must reflect this company's true odds at this moment.

Where the Evidence Comes From

The engine synthesizes this evidence to set the probabilities, rather than applying a fixed set of proportions.

• Causal reasoning: follow the chain of events—if X happens, it most likely leads to Y.
• Historical base rates: how often events of this kind have occurred historically (for example, the historical frequency of certain capex delays).
• Market-implied signals: probability clues implied by options, credit spreads, and the like, used as corroborating evidence.

“Probabilities are not there to decorate a conclusion; they themselves must be constrained by evidence.”— Institutional Research Principle

Scenarios Are Not Stories Made Up Out of Thin Air

Each scenario's terminal EPS must also be derived from "drivers"—revenue growth, margins, share, capex, dilution, and so on—rather than a growth rate pulled at random. The engine uses a "scenario assumption bridge" to tie each path to official facts or explicit model judgments, keeping scenarios from degenerating into narratives that sound appealing but do not hold up.

How It Shows Up in the Report

The report gives the probability for each of the four scenarios, their respective terminal EPS / multiple / IRR, and the probability-weighted result. When you disagree with a conclusion, it is often because you disagree with one of these probabilities or one of these EPS paths—which is exactly the healthy, discussable point of disagreement.