Turkey Problem (Thanksgiving Turkey)

Taleb's parable illustrating the danger of inferring safety from a track record of safety. A turkey is fed every day for 1,000 days, each day confirming its model that 'the farmer is a friend.' On day 1,001—Thanksgiving—the model catastrophically fails. The turkey's confidence was highest at the moment of maximum danger. Every day of feeding was 'evidence' that made the turkey more confident in a model that was about to be violently disproven. In business and investing, long periods of stability can mask building fragility.

When to use it

When a long track record of success is being used to justify continued confidence; when risk metrics show low volatility (which may reflect hidden fragility, not true safety); when evaluating any system that has never been tested by extreme conditions; when 'this has always worked' is the primary justification for a strategy.

How it can help

When a strategy, investment, or system has a long track record of success, ask: am I the turkey? Is this track record evidence of genuine safety, or evidence that I haven't yet encountered the event that will destroy me? The turkey problem is especially dangerous because traditional risk metrics (volatility, variance) look BEST right before catastrophe—the system appears most stable when it's most fragile. Look for hidden fragility beneath surface stability: leverage building up, concentration increasing, or single points of failure accumulating.