im not sure why people keep talking about black swans. i simply want to know how someone can think they have a statistical edge when they don't know all the variable in the market. i was just pointing out 1987 as an 18SD event which would disprove anyones statistical evidence of an actual edge.
What I find kinda amusing is, people say "but the distribution isn't normal" as if it discounted the severity and frequency of crashes observed.
crashing is good for trading if your liquid or long convexity at the end of the day.. bring it on.. haha jk
i'd try to propose a simple answer here. perhaps you're a good fundamental trader who makes sound decisions based on various market variables. most technical traders (who do have statistical edges) simply Not! just 2 cents!
The price change distributions of most financial instruments are platykurtotic (ie fat tailed.) Some of the more sophisticated econometric models do attempt to account for kurtosis, and skew. Skew and kurtosis exist because of swan events, fear, greed, insider knowledge, fraud, etc Markets are far more complex than even Nobel Prize winners are able to accurately model. True edge doesn't necessarily come from a model, but from things like knowledge of order flow and insider knowledge.
you're kinda right - his strategy was to put a majority (like 90%) in t bills and use the rest for playing black swans. another point - now you get no interest from t-bills and it's going to be that way for a long time (at least the next 2 years).