Which setups are more attractive?

The problem is that markets are not static, plus on a lower win % system, you lose on commissions.

So in reality, I would not trade either system since over time both would lose money.
 
Quote from kut2k2:

SPS == k*W

k(A) = .55/407 - .45/433 = .0003121

SPS(A) = k(A)*433 = 0.135

k(B) = .10/50 - .90/1000 = .0011

SPS(B) = k(B)*1000 = 1.1

System B is much more attractive.
More...

According to SPS, lol, system C (see above) is even more attractive with a SPS of 1.12.

But tbh, i wouldn't want to trade a system where the prob of a winner is 0.1%.
 
The ideal system for me has the following:

win rate of at least 30%
profit factor of at least 2
expected value per dollar risked (or SPS lol) of $1 or more
 
Quote from Visaria:

Took me a few minutes admittedly to figure out, but this "SPS", lol, is just the expected value per dollar risked!!!!!

:D
More...
That's not the general case, just the special case of exactly two outcomes (only one win amount and only one loss amount). For the general case of multiple outcomes, I prefer the formula I wrote (k*W) using the exact value for k, because the Kelly approximation (= p/|L| - q/W) is always an overestimation, which means you're overtrading. Not good to overtrade.
 
Quote from Visaria:

According to SPS, lol, system C (see above) is even more attractive with a SPS of 1.12.

But tbh, i wouldn't want to trade a system where the prob of a winner is 0.1%.
More...
SPS(C) = 1.221 actually.

Despite the low winrate, the payoff makes the system best of the three. Would you refuse to play one of the multi-state lotteries even if the jackpot reached a billion dollars? Lol
 
Quote from Visaria:

The ideal system for me has the following:

win rate of at least 30%
profit factor of at least 2
expected value per dollar risked (or SPS lol) of $1 or more
More...
PF = pW/qL = .3W/.7L >= 2 : W >~ 5L

SPS >~ .3*5-.7 = .8 < 1

Looks like SPS is your bottomline factor. Lol
 
Quote from kut2k2:

... , because the Kelly approximation (= p/|L| - q/W) is always an overestimation, which means you're overtrading. Not good to overtrade.
More...

kut2k2,

I made an equivalent claim in:

http://www.elitetrader.com/vb/showthread.php?s=&postid=3028937#post3028937

Any references to and/or proofs of the claim would be greatly appreciated.

The proof I have uses Jensen's Inequality for conditional expectations which I believe to be beyond the mathematical level of most traders.

Thanks,
Jim Murphy
 
You must log in or register to reply here.
Back
Top