Circumstance-Based Categorization and Expert Limitations

Key Principle

Investment theory fails when it classifies by static attributes (low P/E, small cap) rather than by the circumstances that determine how cause and effect actually operate. Christensen's theory-building framework shows that mature theories evolve from attribute-based categories to circumstance-based categories. The same stock can be cheap or expensive depending on situational context, not fixed traits.

This categorization error extends to expert cognition itself. In probabilistic, high-degrees-of-freedom domains like markets, individual expertise systematically underperforms collectives. Two cognitive biases form a causal chain: functional fixedness (deep practice locks in one framing) leads to reductive bias (treating complex systems as simple ones), which produces linear thinking applied to nonlinear systems. The antidote is fox-style cognition — holding multiple frameworks loosely and switching between them based on context.

Meanwhile, the profession compounds the problem by dismissing one of the strongest signals of genuine skill: long streaks. Because small base-rate differences amplify exponentially across sequential outcomes, streaks through difficult environments carry far more information about ability than single-period results.

Why This Matters

Attribute-based rules produce false confidence. Using P/E as a market-timing tool over 125 years would have produced poor results because the rule ignores the circumstances that determine whether a given P/E level is actually cheap or expensive. (Ch. 4)
Expert agreement collapses in complex domains. In probabilistic, high-DOF problems like the stock market, expert agreement drops below 20%, and collectives consistently outperform individuals. (Ch. 6)
Functional fixedness is a byproduct of expertise itself. Deliberate practice internalizes domain features, but the same internalization constrains cognitive flexibility. The expert stops considering alternative framings precisely because their practiced framing has historically worked. (Ch. 6)
The profession systematically underweights streak evidence. A 60% vs. 30% base rate is a 2x gap in single attempts but a 32x gap over five consecutive successes. Small skill edges compound exponentially, meaning streaks contain far more information about ability than individual results. (Ch. 7)

Good Examples

Bill Miller's 15-year streak (Ch. 4, 7). Miller beat the S&P 500 for 15 consecutive years using circumstance-based thinking, yet Morningstar criticized him for "straying" from value investing because his portfolio's price-to-book was 178% above the value category average. The criticism itself reveals attribute-based thinking in the critics. His streak probability was 1-in-2.3-million using actual annual rates, far exceeding chance.
Slime mold (Ch. 4). When food is abundant, cells act as independent organisms. When food is scarce, they converge into a collective of tens of thousands. The same entity is "it" or "they" depending on circumstances — demonstrating why fixed classification fails for context-dependent phenomena.
EKG diagnosis (Ch. 6). A computer correctly identified heart attacks 66% of the time vs. a cardiologist's 55% — demonstrating algorithm superiority in rules-based, limited-DOF tasks where expertise adds less value than systematic processing.
MLB 30+ game hitting streaks (Ch. 7). Across 42 players, the average lifetime batting average was .311 vs. league average ~.260. Streaks cluster overwhelmingly among elite hitters, confirming that persistence is a nonlinear skill amplifier rather than a random artifact.

Counterpoints

Circumstance-based thinking is harder to systematize. Investment consultants compel managers to declare an attribute-based style (growth or value) and stick with it, because circumstance-based categories resist easy classification and benchmarking. The industry's measurement infrastructure is built around attributes. (Ch. 4)
Fox-style cognition is organizationally unrewarded. Media fame inversely correlates with calibration — better-known forecasters are less accurate. The information economy rewards hedgehog-style confidence and bold narratives, not the diffident probabilistic reasoning that actually produces better forecasts. (Ch. 6)
Streaks still require extraordinary luck. Even DiMaggio's 56-game hitting streak was less than one-in-a-million given his elite batting average. Skill is necessary but insufficient — it sets the ceiling on what luck can plausibly produce. (Ch. 7)

Key Quotes

"The main message is that much of investment theory is unsound because it is based on poor categorization." (Ch. 4)

"All investors use theory, either wittingly or unwittingly. The lesson from the process of theory building is that sound theories reflect context. Too many investors cling to attribute-based approaches and wring their hands when the market doesn't conform to what they think it should do." (Ch. 4)

"What mattered in predictive ability was not who the people were or what they believed, but rather how they thought." — Philip Tetlock (Ch. 6)

"When we use or think about something in a particular way we have great difficulty in thinking about it in new ways." (Ch. 6)

"Long streaks are, and must be, a matter of extraordinary luck imposed on great skill." — Stephen Jay Gould (Ch. 7)

"The more important issue is that streaks inform us about probabilities. In human endeavors, unlike a fair coin toss, the probabilities of success or failure are not the same for each individual." (Ch. 7)

Rules of Thumb