Retail investors contributing to professional's returns

AI Machine Learning
1

So I am pretty sure that I have done this. Or at least it is a question I should ask myself.

If I kept investing in strategies while they were doing well and then abandoned them when they had a big drawdown what would happen?

I think the short answer would be that I would be giving money to the professionals that had a strategy that they could stick with.

This is not a new idea. It is the idea of gamble’s ruin. And is clearly part of why casinos stay in business. In addition to having the odds in their favor with each spin of the roulette wheel a gambler will leave the table after he has lost all of his chips—clearly a net win for the casino (and not requiring a proof, I hope). The casino winning almost every time because of the bet size and the casino’s larger bankroll. And not just because of the odds.

Gambler’s run adds that the casino wins even if the odds are with the gambler with each play—like with a good blackjack card counter. If her bets are too big she still gives up all of her money to the casino even if her card counting is flawless and the odds were truly with her on each play. And I would only add that she loses if she gives up the first time she is behind as well as when she faces complete loss of her original bankroll. This is actually the gambler’s ruin guarantee: “A persistent gambler who raises his or her bet to a fixed fraction of the gambler’s bankroll after a win, but does not reduce it after a loss, will eventually and inevitably go broke, even if each bet has a positive expected value.” Source: Gambler’s ruin

I am copying this idea. It has been a while and I forget which book it was but the author basically thought value strategies work because a disciplined investor could eventually go home with the money that the average retail investor leaves on the table after they closed out their brokerage (and possibly P123) account.

Jim

2

This work only for long strategies that are based on valuations.

On strategies that are not based on valuations and the drawdown is not caused by market selloff, the method of abandoning strategy after is stops working for statistically significant period is a good idea.

I have read few books and seen few interviews with hedgefund managers and traders using shadow patterns and trading strategies making millions. If those patterns are discovered and get known, first an antipattern appears and then the pattern looses its profitability, so it is better to abbandon the strategy.

3

I think I agree with this and cannot think of an exception. The key being drawing “statistically significant” inferences versus just not being able to stand the pain (and perhaps not being certain of the statistics). I.e., the state of many retail investors and I do not claim to be anything other than an average retail investor at times.

And I think this is—at least somewhat—captured in my idea of using Bayesian model averaging which gradually adjusts ones confidence in a model rather than suddenly going-in in a big way or suddenly going out of a model according to an arbitrary p < 0.05 cut-off.

Well said I think! Maybe a different perspective but I think I fully agree.