KlangFool,KlangFool wrote: ↑Fri May 22, 2020 2:17 pmtargetconfusion,targetconfusion wrote: ↑Fri May 22, 2020 2:08 pmThis is the point, I think. Is playing high-reward long shots a good idea? Maybe, depending on how often they occur and how much they cost to play. By Taleb's own description, our sample of market history is too small to reliably estimate that frequency. You'd have to experiment by setting up two competing portfolios and press play for 10k years or however long Taleb thinks it takes to get a representative sample.KlangFool wrote: ↑Fri May 22, 2020 1:47 pmAnd, he had got it right and won big twice. So, there should be some merit in his method.
The kind of very long shot happened in 2008 and 2020. And, the kind of return that we are talking about is 30X to 100X.
<<You'd have to experiment by setting up two competing portfolios and press play for 10k years or however long Taleb thinks it takes to get a representative sample.>>
Why? You can get the best of both worlds by investing 99% the normal way and 1% in the very long shot.
I agree with your thought, and the back of the envelope math - but only if the variables were correct.
For a tail hedge fund to even post 30x returns (highly optimistic, let alone the upper end of your range), it will end up costing more than than the budgeted 1%. To maintain notional exposure, equal to 1% of your total capital, to this type of strategy over a one year period, will end up costing more than 1% per year, often enough.
Not saying the efficacy of the strategy is bad, or no good. Only saying that I think - on average - the cost will be greater than expected, and the upside when markets drawdown is not likely to reach the 30-100x range with confidence.