<aside> ℹ️

About

Michael Mauboussin, Dan Callahan, and Darius Majd wrote this 152-page Credit Suisse Global Financial Strategies report in September 2016. It's the definitive practitioner reference for base rates of corporate performance. The premise is simple: most forecasts are too optimistic and too narrow because forecasters rely on the inside view (the specifics of the case in front of them) and neglect the outside view (what happened historically to a relevant reference class).

The book compiles the outside view. It's the statistical backbone for anyone trying to stress-test a fundamental forecast of sales growth, profitability, margins, earnings growth, or cash returns on investment — and it adds two practitioner-focused sections on what to do when a stock in your portfolio crashes or rips.

Headline insight: when you're modeling a company's future, the first question isn't "what makes this one special?" It's "what does the distribution look like for everything that's been in this situation before?"

</aside>

Part I: The intellectual framework

Inside view vs outside view

The inside view focuses on what's unique about the specific case — the strategy, the management team, the product, the market opportunity. It's how most people naturally forecast. It's also how you end up with an 8-in-10 entrepreneur sample rating their success odds at 70%+ when only 50% of new businesses survive five years.

The outside view (Kahneman and Tversky's framing, later operationalized by Lovallo, Clarke, and Camerer as "reference class forecasting") asks a different question: what happened historically to the reference class this case belongs to? M&A example: insiders talk synergies and strategic fit. Outside view observes that ~60% of deals fail to create value for the acquirer. The inside view wasn't wrong — it was just incomplete.

<aside> 🎯

The central claim of the book: accurate forecasts come from thoughtfully blending inside and outside views. Rule of thumb — if skill dominates outcomes, weight the inside view. If luck plays a large role, weight the outside view.

</aside>

Regression toward the mean — the missing piece

Most investors believe in regression toward the mean. Few understand it properly. The book lays out two rules that matter:

1. The correlation coefficient IS the rate of regression. High correlation (e.g. r = 0.9) means modest regression. Low correlation (e.g. r = 0.1) means rapid regression back to the mean.
2. The arrow of time doesn't matter. Tall sons have tall fathers, but the sons deviate from the mean by a smaller amount than the fathers did. Same relationship runs backwards. This is imperfect correlation, not causality.