Markets have fractal, self-similar structure
Price charts look statistically similar across timescales, so market roughness is scale-invariant. This fractal geometry, not smooth randomness, describes real prices.
The (Mis) Behavior of Markets
9 ideas
Modern portfolio theory rests on false premises
Bachelier, Markowitz, and Black-Scholes assume independence and normal distributions that data contradict. Their elegance masks a systematic blindness to catastrophe.
Roughness as a fundamental property
Mandelbrot sees nature and markets alike governed by measurable roughness rather than idealized smoothness. The geometry of the jagged replaces the geometry of the smooth.
Wild randomness versus mild randomness
Mild randomness, like human height, lets extreme values average out, and no single observation moves the total much. Wild randomness, like price changes, has variance so large or undefined that one extreme event can dominate the sum of all the others. Treating a wild process as mild throws away exactly the events that matter most.
Long memory, or the Joseph Effect
Price changes are not independent. Past movements shape future ones over long horizons, so trends and cycles appear without any real cause behind them. Mandelbrot took the idea from Hurst's study of Nile floods, where wet and dry years clustered far more than chance would predict.
Market time is trading time, not clock time
Markets run on a variable internal clock that speeds up during turbulent periods and slows down in quiet ones. Volatility therefore clusters, and calm stretches alternate with violent ones. Rescaling price series to this trading time turns messy, irregular behavior into a self-similar fractal pattern.
Bell-curve finance understates extreme price moves
Under a normal distribution, the index moves seen in August 1998 or October 1987 should almost never happen, at odds of billions or more to one. In real markets they occur many times per century. Portfolio theory, CAPM, and Black-Scholes all assume normality, so they systematically underprice the risk of ruin.
Cotton prices reveal scale-invariant volatility
Mandelbrot studied more than a century of cotton prices. The shape of the price changes looked the same whether he measured daily, monthly, or yearly intervals, and the tails followed a power law instead of a bell curve. This self-similarity across scales was the first empirical evidence that markets are fractal.
The Noah Effect: discontinuous jumps
Prices do not glide smoothly from one value to the next. They leap, skipping over intermediate levels. This breaks the continuity that hedging models depend on, because stop-loss orders and dynamic hedges assume you can trade at every price along the way.