(2008-09-16) Taleb Limits Of Statistics

Nassim Taleb on on the limits of Statistics. Examples: Long Term Capital Management, Credit Crisis 2008. Close to 99% of the variations, over the span of 20 years, will be represented in 1 single day. As I show in the appendix, this is typical with ANY socio-economic (Social System) variable (commodity prices, currencies, inflation numbers, GDP, company performance, etc. ). No known econometric statistical method can capture the probability of the (Black Swan) event with any remotely acceptable accuracy... This absence of "typical" event in Extremistan is what makes Prediction Market-s ludicrous, as they make events look binary. "A war" is meaningless: you need to estimate its damage - and no damage is typical... The "WisdomOfCrowds" might work in the first three quadrants; but it certainly fails (and has failed) in the fourth.

He defines the Fourth Quadrant as cases where

  • heavy and/or unknown tails in distribution
  • complex payoffs (non-linear outcomes)

Phronetic Rules: What Is Wise To Do (Or Not Do) In The Fourth Quadrant (Decision Making)

  1. Avoid Optimization, Learn to Love Redundancy (Slack). The only weak point I know of financial markets is their ability to drive people & companies to "efficiency" (to please a stock analyst's earnings target) against risks of extreme events.

  2. Avoid prediction of remote payoffs

  3. Beware the "atypicality" of remote events.

  4. Time. It takes much, much longer for a times series in the Fourth Quadrant to reveal its property.

  5. Beware Moral Hazard.

  6. Metrics. Conventional metrics based on type 1 randomness don't work.

  7. Where is the skewness?

  8. Do not confuse absence of volatility with absence of risks.

  9. Beware presentations of risk numbers.

So how do you make decisions and weigh trade-offs?


Edited: |

blog comments powered by Disqus