An Example of Why "Genius" Fails at Predicting Markets

This is from David Swensen's incredible book mentioned earlier (his words in italics):

On October 19, 1987, the U.S. stock market, as measured by the S&P 500, declined by 23 percent. The unprecedented decline nearly defies probabilistic description. Based on historical daily volatility, the one-day collapse represented a 25-standard deviation event...

A one-standard deviation event happens one third of the time, two-standard deviation events 1/20th of the time, three-standard deviation events 1% of the time, and so on. Ever hear of Six-Sigma and Jack Welch? That's where all this mumbo-jumbo comes from. Sigma is the Greek letter used to represent standard deviations in statistics. Anyhow, Swensen then throws out this gem to describe how special a 25-standard deviation event is:

Based on a 250-day year, an eight-standard deviation event occurs once every three trillion years. Twenty-five-standard deviation events should not happen.

Wow. So the market is not at all normal (or '80s investors were collossally unlucky), and it seems almost certainly left-sided. (I don't think a 25-sigma uptick could happen as there's no real equivalent of a "positive bank run.")