Hello! Question about Win Rates in systems trading

ragzROI

Newbie
Messages
5
Likes
0
As my first post, I just wanted to say Hi, Ive been lurking here for a bit, and am excited to start getting involved.

Now Im not looking to beat a dead horse about asking about Win Rates or anything so Im rather looking for someone to tell me if I'm doing something wrong.

Well Ive been meticulously developing a system for the past month or so, and since numbers (Mac's excel equivalent) is bogging down because of the thousands of simultaneous calulations I'm running I decided to write a compile able program to sim a bunch of stocks that I can load into it...

Now onto my question. So far Im under the assumption that Win Rate = Winning trades/total trades.

How is it anybody is achieving over 60-70% Win Rate, when more or less half your trades are coming from entries?

My system right now is a long only system dealing with US Equities, and I keep track of total trades to factor in commissions in various frequency ranges for the various discounts brokers offer. My win rates are averaging to about 24-30%, with many tight stop losses. My profit factor is averaging between 2.00-3.5, as well as my average profit per trade being anywhere from $20 to $150 depending on trades.

Total trades 363 (including entering positions which I dont count as losses or gains only on sells)
2013-03-21 12:39:18.830 BackTest[17618:303] TOTAL TRADES: 363
2013-03-21 12:39:18.831 BackTest[17618:303] TOTAL TRADES PER QTR: 18
2013-03-21 12:39:18.831 BackTest[17618:303] PPT: $72.33
2013-03-21 12:39:18.831 BackTest[17618:303] Winning Percentage Of Total Trades: 23.69%
2013-03-21 12:39:18.832 BackTest[17618:303] Sharpe Ratio: 2.12
2013-03-21 12:39:18.832 BackTest[17618:303] Max Drawdown: 16.59%
2013-03-21 12:39:18.832 BackTest[17618:303] Most Consecutive losses: 6
2013-03-21 12:39:18.833 BackTest[17618:303] Most Consecutive wins: 5
2013-03-21 12:39:18.833 BackTest[17618:303] Probable Wins: 86
2013-03-21 12:39:18.833 BackTest[17618:303] Probable Losses: 95
2013-03-21 12:39:18.834 BackTest[17618:303] AVG Win: $193.71
2013-03-21 12:39:18.834 BackTest[17618:303] AVG Loss: $-91.38
2013-03-21 12:39:18.834 BackTest[17618:303] AVG TIME IN TRADE: 5 DAYS

2013-03-21 12:39:18.835 BackTest[17618:303] APPT: $115.63

Am I calculating this win rate thing wrong? I mean I know its not the be all and end all of everything...but I seem to be seeing a recurrence of people disliking lower win rates, and claiming 70%..but how are the trades entries factored in then?

IE out of 363 trades 182, are coming from entering the position.

(If it helps this is over a 5 year backtest, with EOD trading) (Not intraday)
 
Last edited:
How is it anybody is achieving over 60-70% Win Rate, when more or less half your trades are coming from entries?

I don't understand this question.

Also bit confused by your results, since I don't know what probable wins are. It's a backtest, so they won or they didn't I guess. Lets assume 86 wins our of 363 trades, that comes to 23.69%. But you only have 95 losses, so where are the other trades? Still open? Breakeven? Calculating the %winners you really need to have all trades closed.

Perhaps you've counted double...? I.e.2*(86+95)=362. Well then you are calculating win% incorrectly. You can't win an entry. A trade is an entry and an exit.
 
I don't understand this question.

Also bit confused by your results, since I don't know what probable wins are. It's a backtest, so they won or they didn't I guess. Lets assume 86 wins our of 363 trades, that comes to 23.69%. But you only have 95 losses, so where are the other trades? Still open? Breakeven? Calculating the %winners you really need to have all trades closed.

Perhaps you've counted double...? I.e.2*(86+95)=362. Well then you are calculating win% incorrectly. You can't win an entry. A trade is an entry and an exit.

Like I said in my post, my strat is long only.
Therefore, I count all trades including when I buy my shares for ex. When factoring in win rate percentage.

So I only count winners/losers when I sell my position and refer it back to whether I made money or lost.

I also say "Probable" because nothing is for certain. Backtesting is just that. Statistical. Probable wins in a vacuum therefor is what was calulated. I cant account for limit orders not being filled etc..

PS You did technically answer my question :) I think my brain farted, and I *did* calculate double, since one 'trade' with a broker that charges commission is technically half of a proper trade, and I forgot about that.
 
Last edited:
Like I said in my post, my strat is long only.
Also, I count all trades including when I buy my shares for ex.

So I only count winners/losers when I sell my position and refer it back to whether I made money or lost.

I also say "Probable" because nothing is for certain. Backtesting is just that. Statistical. Probable wins in a vacuum therefor is what was calulated. I cant account for limit orders not being filled etc..

Right. But you've counted it twice haven't you? You go long, you count that as one trade, you exit, your count is now 2.

This will skew your win % on a trade. Count one trade as an exit and an entry.
 
Right. But you've counted it twice haven't you? You go long, you count that as one trade, you exit, your count is now 2.

This will skew your win % on a trade. Count one trade as an exit and an entry.

Thanks mate, I guess I dont have to throw my system in the garbage just yet! (y)
 
Thanks mate, I guess I dont have to throw my system in the garbage just yet! (y)

Look into: Expectation and Drawdown. The win rate is somewhat important, but not nearly as important as the other two.
 
Look into: Expectation and Drawdown. The win rate is somewhat important, but not nearly as important as the other two.

Yea Ive read/heard some different things. Too much is bad, but "below 30% is not risking enough" Or something like that.

My DDs are averaging between 15-20%

As for expectancy, thats something that I dont think ive come across yet. I will google
 
Thanks mate, I guess I dont have to throw my system in the garbage just yet! (y)

Forget about win rates. They are a product of your stops and take profit levels. The tighter the stops the worse your win rate will be. There's nothing wrong with having a system that losses 3 out of every 4 trades if at the end of the day it still makes a profit.

I can guarantee you that most of the systems you see with 70%+ win rates do so by running wide stops.
 
Top