The risk of extreme, low-probability outcomes that occur in the "tails" of a return distribution — events far outside what standard risk models predict.
Deeper Explanation
Most risk models assume normally distributed returns — but investment returns have "fat tails": extreme outcomes occur far more frequently than a normal distribution would predict. The 2008 financial crisis, the 1987 crash, and the 1998 LTCM collapse were all events that standard models suggested should occur once in thousands of years — but in practice occur every decade or two. Marks argues that investors systematically underestimate tail risks because they calibrate risk based on recent experience rather than historical extremes. Explicit tail-risk hedges — owning assets that perform well in extreme scenarios (gold, long-dated government bonds, put options) — provide portfolio insurance against the scenarios that most investors are not pricing.
Continue Learning
Go deeper into the Macro school — frameworks, case studies, and decision systems.