I like your angle on things, Amahrix, re staying statistically passive. As you pointed out, uncertainty is always there, both on the payoff and win ratio, so the best one can do is manage outcomes intelligently. Reducing size if/when the win ratio takes a hit is one way to buy some time to assess things with a clear head
Thanks for your reply. And yes, Kelly Criterion. You reduce the size of the next bet relative to portfolio value so that the law of large numbers comes around in your favor, which it should if you have a positive statistical expectation. You’ll have more losers than winners, net. But your account will grow, over time, on average.
Not necessarily. Edit: In some strategies, you want to accumulate securities that are deemed cheap before buyers rush in, if ever, so that you can sell when everyone rushes into the door demanding your product, and you can sell high, very high, than where you bought it originally. Not buy at the same time as others as this will increase the price and decrease the edge. I’m talking options, stocks aren’t really multidimensional and surely aren’t nonlinear.
Ideally, you will have more losing trades relative to winning trades, but your account value will grow over time, on average. Why? Because payoff of winning trades pays for losers and more. Similar to Venture Capital model of Silicon Valley... invest in 100 businesses, and 1 becomes an Uber. Rinse and repeat. This is simplified but I am trying to showcase the concept.
Often overlooked is the key role that trade frequency plays right? If one's method is low-output in terms of the # of signals, it's much harder to deal with the equity curve while waiting for the average/expectation to bear out. Think Ed Thorpe at the blackjack table...
Can you rephrase and elaborate? I want to understand correctly. Edit: I think I understand what you’re saying but want to make sure. Edit2: And I meant to say, regarding Kelly Criterion, that you only reduce bet size as portfolio decreases in value, but increase bet size as portfolio gains in value.
Yeah sure, basically more signals per timeframe traded allows for a smoother ride to the trading method's expected value. That's one of the things Ed Thorpe mentioned once, that when he was at the blackjack table it was easier for him to manage a big drawdown (he was down $1700 on a $2500 bankroll at one point or something like that) because of the number of hands he could play/number of bets available to make.
That is Taleb's barbell strategy right? Take no/low risk and take enormous risk at the same time. For example, put your cash in t-bills and use the interest income to fund your high payoff strategies. btw, to be a bit quantitative for the newer folks: payoff = probability*(reward/risk)
There are many ways. That’s one example. Edit: I recall reading somewhere, maybe Nassim's twitter, that he admitted to making the mistake in assuming T-Bills were safe securities. I tried locating and if I find, I'll reference.
they are fantastic for those who issue them they pay less than the inflation rate so the issuer can pay out interest with a lot cheaper money