Who this is for — Anyone using leverage, «normal» models, or backtests on calm periods. Fat tails explain why markets produce black swans more often than expected.
Fat tails (heavy tails) describe distributions where extreme events (± many σ) happen more often than on a normal curve. Financial markets show them systematically — underestimating them = undervalued risk.
In plain terms — «Once in 100 years» that happens every 10 — the bell curve lies.
Why it matters
| Normal assumption | Market reality |
|---|---|
| Extremes very rare | Gaps, crashes, squeezes frequent |
| σ fixes risk | σ understates tails |
| «Smooth» backtest | Live: multiple outliers |
Leverage + mean reversion without stops = exposed to left tail.
Adaptation
- Safety margin on size and daily stop
- Stress test extreme gap/slippage, not only average σ
- Pair with skew and aggregate risk in portfolio
Typical mistake — Calibrating everything on last 12 calm months — fat tail arrives on month 13.
Example — 3× index leverage: +2%/month for 8 months, −18% in one macro day — one day = months of edge.
Summary card
- Rule: assume heavy tails always.
- Action: leverage ↓, mandatory stops, stress tests.
- Review: tag outliers > 3σ in journal.
Gold path — Edge module. Index: Gold path.