l5-the-black-swan_bell-curve-persistence

The Gaussian curve persisted because abandoning it would have forced practitioners to relearn their trade, and no alternative theory was readily available. Taleb argues contagion, not validity, determines a theory's fate in social science.

Mandelbrot showed around 1960 that the Gaussian curve did not fit prices, and the 1987 crash was an event that would take billions of lifetimes of the universe under Gaussian assumptions. Paul Cootner admitted Mandelbrot's findings would make most statistical tools obsolete.

The Swedish Academy awarded Nobel Prizes to builders of Platonic models on a Gaussian base, such as Markowitz and Sharpe, without testing the models or consulting practitioners. Software vendors then sold these Nobel-crowned methods for millions.

Sigma, variance, correlation, and the Sharpe ratio all rest on Gaussian assumptions that fail empirically in markets. After the 1987 crash, even practitioners agreed sigmas were charlatan measures, yet the metrics still pervade fund prospectuses.

Gaussian assumptions should not be used when modeling market prices subject to fat tails and extreme events, as demonstrated by the 1987 crash. Taleb's impression argues no empirical observation can justify them, and their use reflects convenience rather than validity.