Normal Distribution vs Real Market Returns

The bell curve is elegant. The market is not.

The normal (Gaussian) distribution is the most-used model in quantitative finance. Option pricing, VaR, and risk management all lean on it — yet real market returns violate it regularly. Understanding the gap is what separates a trader who survives a crash from one who is ruined by one.

The bell curve

A normal distribution is fully described by its mean (μ) and standard deviation (σ):

P(x) = (1 ÷ (σ·√(2π))) · exp(−(x − μ)² ÷ (2σ²))

Key properties traders use:

• ~68% of values fall within ±1σ
• ~95% within ±2σ
• ~99.7% within ±3σ

This is where statements like "a 4σ move" come from — under a normal model, that should happen roughly once every 30,000 days.

Why the model is useful

Even though markets aren't perfectly normal, daily returns are roughly bell-shaped. That makes the model a useful approximation for:

• Quick risk sizing (treat 2σ as a typical "bad day")
• Z-scoring indicators
• Back-of-envelope probability estimates

Where it breaks: fat tails

Real returns have fat tails — extreme moves happen far more often than the bell curve predicts. Equity index "4σ" and "5σ" crashes occur multiple times per decade, not once per century.

Other violations:

• Negative skew — big down-moves outnumber big up-moves
• Volatility clustering — calm and stormy periods cluster, contradicting the model's independence assumption
• Excess kurtosis — more density in the tails and the peak than a normal curve has

The practical danger

Risk tools built on the normal model underprice tail risk. A model that says a 2008-style drawdown is "impossible" will size positions far too aggressively. This is the core lesson of Long-Term Capital Management and every VaR failure since.

How to adapt

1. Treat the normal model as a first approximation, not truth
2. Use historical simulation or t-distributions for tail-sensitive risk
3. Watch skew and kurtosis (covered in the next article) to detect when the model is dangerous
4. Always stress-test against actual crashes, not just 3σ
5. Size for the tails you've seen, not the tails the curve predicts

Summary

The normal distribution is a useful baseline for everyday volatility sizing. But markets have fat, negatively skewed tails. Trust the bell curve for routine sizing — and distrust it the moment you start thinking about crashes, where it consistently underestimates how bad things can get.