Fat-tail distributions model the reality that market crashes and booms happen far more often than a normal bell curve predicts. Ignoring fat tails in retirement simulations underestimates risk — plans that look safe under normal assumptions can be 5-10 percentage points less likely to succeed in the real world.
A fat-tail distribution is any probability distribution where extreme events — values far from the mean — occur more frequently than a normal (Gaussian) distribution would predict. In financial markets, this means crashes like 2008 (-37%) and single-day drops like Black Monday 1987 (-22%) are not freak anomalies but statistically expected features of how markets behave.
How It Works
The normal distribution assigns vanishingly small probabilities to extreme events. A daily move of 5+ standard deviations should occur roughly once every 14,000 years under normal assumptions — yet markets experience such moves every few years.
Fat-tail distributions fix this by placing more probability mass in the tails. The most common model is the Student's t-distribution, where a degrees of freedom (DOF) parameter controls tail thickness. Lower DOF = fatter tails = more frequent extremes.
Interactive chart: fat-tail-comparison
Normal distribution vs. Student-t (DOF=5) — notice the heavier tails
Coming soon
Why It Matters for Retirement Planning
Retirement portfolios are asymmetrically sensitive to tail events. A retiree withdrawing from a portfolio is hurt more by a crash than they benefit from a boom of equal magnitude — because withdrawals lock in losses during downturns (sequence-of-returns risk).
This asymmetry means that underestimating tail risk systematically overestimates plan safety. Retirement Lab addresses this by using the Fernandez-Steel skewed Student-t distribution, which captures both the frequency of extremes (kurtosis) and their asymmetry (skewness) — producing more honest success rate estimates.
Frequently Asked Questions
- How much do fat tails affect retirement planning?
- Switching from a normal distribution to a fat-tailed Student-t distribution typically reduces Monte Carlo success rates by 5-10 percentage points. This means a plan that appears 92% safe under normal assumptions may actually be closer to 83-87% — a meaningful difference when your retirement security is at stake.
- What causes fat tails in financial markets?
- Fat tails arise from herding behavior, leverage, liquidity crises, and feedback loops in financial markets. When many investors sell simultaneously (panic), prices fall further than random models predict. These structural features of markets make extreme events inherently more frequent than a bell curve assumes.