Fat Tail Risk
The higher-than-normal probability of extreme investment outcomes, caused by return distributions with heavier tails than a normal curve.
What is Fat Tail Risk?
Fat tail risk (also called heavy tail risk or leptokurtosis risk) refers to the phenomenon where financial asset returns exhibit a higher probability of extreme outcomes — large positive or negative moves — than a normal (Gaussian) distribution would predict. A normal distribution has thin tails, assigning near-zero probability to moves beyond three standard deviations. Real financial data has 'fat tails': market crashes, flash crashes, and sharp rallies occur far more frequently than normal models suggest. Measuring fat tail risk uses metrics like kurtosis (the fourth moment of the distribution; values above 3 indicate fat tails), Conditional Value at Risk (CVaR/Expected Shortfall), and extreme value theory. Standard risk management tools like Value at Risk (VaR) using normal assumptions systematically underestimate the probability and magnitude of tail events, contributing to catastrophic failures such as Long-Term Capital Management (1998) and the 2008 financial crisis.
Example
A daily equity return model assuming normal distribution predicts a 5% single-day loss will occur once every 3.5 years. In reality, the S&P 500 has experienced daily losses exceeding 5% multiple times per decade — far more frequently than the model predicts. This fat tail means that VaR estimates based on normality chronically understate extreme risk.
Source: Nassim Nicholas Taleb — The Black Swan (Random House, 2007)