Would You Trade This System?

Quote from intradaybill:

No problem. To buy 2500 SPY you currently need about $450K. Much above 4x margin in your case. This means that you may have used hypothetical profits in your backtest to size positions. This is highly unrealistic backtest. I would suggest you try again with a fixed 100 shares positions. What do you get then for win rate, avg. trade, etc?
More...


Well, im using prop leverage to trade and not retail accounts with just 4x.

You can see that worst drawdown was just $17k and this was in 2008 crazy times. Normally it gets no where near that, but even still on a $100k bank i feel a 17% drawdown is not so bad. So i dont feel im using hypothetical size in my backtest as systems bank size is more than adequate to cope with that size of drawdown.

As for trading 2500 shares, i would only require $450k if i was trading with a retail broker. My example above shows that the intial bank is more than enough for the drawdowns so why use only 4x, why not use prop leverage and then you can trade more effientley making better use of your capital. Whats the point of 4x if it does not put your capital to maximum work for you?...........And Im still well within my risk management.

I can get you the stats for 10x less if you like and post them here, a size of just 250 shares but it will be exactly the same results just 10x less overall net profit. Win rate and profit factors will be the same.
 
here is the same system with just 250 fixed share size. I have not looked yet but im almost sure it will be the same profit factors and other stats should be relative. I can post any share size you wish but it will all be relative. It is very much a scalable system.


Total Net Profit $41,073.50
Profit Factor 1.27


Total Number of Trades 6032
Percent Profitable 63.71%
Winning Trades 3843
Losing Trades 2189
Even Trades 0

Avg. Trade Net Profit $6.81
Avg. Winning Trade $49.63
Avg. Losing Trade ($68.37)
Ratio Avg. Win:Avg. Loss 0.73
Largest Winning Trade $420.50
Largest Losing Trade ($789.50)


Max. Consecutive Winning Trades 17
Max. Consecutive Losing Trades 9


Max. Shares/Contracts Held 250
Total Shares/Contracts Held 1508000
Account Size Required $1,744.00
Total Commission $12,064.00


Trading Period 9 Yrs, 11 Mths, 22 Dys, 5 Hrs, 37 Mins
Percent of Time in the Market 3.90%
Time in the Market 4 Mths, 20 Dys, 5 Hrs, 40 Mins
Longest Flat Period 5 Dys, 1 Hr, 26 Mins

Max. Equity Run-up $42,239.00
Date of Max. Equity Run-up 11/09/11 18:08


Max. Drawdown (Intra-day Peak to Valley)
Value ($2,306.50)
Date 10/23/08 19:22


Max. Drawdown (Trade Close to Trade Close)
Value ($1,744.00)
Date 10/23/08 18:13


Max. Trade Drawdown ($847.50)



01/01/2012 $4,041.00
01/01/2011 $4,267.50
01/01/2010 $3,204.00
01/01/2009 $5,825.50
01/01/2008 $8,047.50
01/01/2007 $3,069.50
01/01/2006 $2,274.00
01/01/2005 $1,687.00
01/01/2004 $2,564.00
01/01/2003 $2,981.50
01/01/2002 $7,216.00






Created by TradeStation: 07/02/2012 21:35:54, Copyright (c) 1991-2004 TradeStation Technologies, Inc.
 
With a Percent Profitable of 63.71% and a Ratio Avg. Win:Avg. Loss of 0.73, you will have almost no margin for error.


Unless you can avoid ever making mistakes, this system is not workable without further refinement.
 
Quote from monti1a:

With a Percent Profitable of 63.71% and a Ratio Avg. Win:Avg. Loss of 0.73, you will have almost no margin for error.


Unless you can avoid ever making mistakes, this system is not workable without further refinement.
More...

thanks for feedback, this is exactly what im looking for.


yes i agree things are tight. But my personality is not trend chasing, i have no patience for that. Smaller targets are easier to get. Im not mircro scalping for a couple of ticks though either, my average is around 20 - 25 cents profit target. Most RTM type systems are 1-1 or less. I can post you improved win/loss avg if i increase my tragets to 2-1, but beleive it or not when i do that the return and smoothness is worse. This is probably because my entry could be improved but im not sure perfect entry exists all of the time, with noise and randomness we get in the markets.


I can post the improved win/loss with a 2-1 if you like but as i say the return is lower and not as smooth as the system that gives room and takes profits quickly.
 
here is the monthly net profit...............

Yes im with you that the profit factors and win/loss needs improving. Any views or opinions on how i can do this and ill post results.
 

Attachments

here is the same signal trading google, I get similar performance on several stocks and still testing them individually which is taking forever..............please hurry up tradestation with the portfolio software.



Imo this shows robustness, but i agree the edge is borderline noise trading and needs improving and im all ears. Alot of pro's here and i need your help.


Google 500 fixed shares

Total Net Profit $506,761.00
Gross Profit $1,960,350.00
Gross Loss ($1,453,589.00)
Profit Factor 1.35



Total Number of Trades 4811
Percent Profitable 65.58%
Winning Trades 3155
Losing Trades 1656
Even Trades 0 0 0

Avg. Trade Net Profit $105.33
Avg. Winning Trade $621.35
Avg. Losing Trade ($877.77)
Ratio Avg. Win:Avg. Loss 0.71
Largest Winning Trade $4,516.00
Largest Losing Trade ($9,229.00)


Max. Consecutive Winning Trades 16
Max. Consecutive Losing Trades 9
Avg. Bars in Total Trades 32.44

Max. Shares/Contracts Held 500
Total Shares/Contracts Held 2405500
Account Size Required $18,131.00
Total Commission $19,244.00

Return on Initial Capital 506.76%
Annual Rate of Return 24.20%
Buy & Hold Return 460.56%
Return on Account 2795.00%
Avg. Monthly Return $4,458.62
Std. Deviation of Monthly Return $6,493.02

Return Retracement Ratio 0.36
RINA Index 13785.08
Sharpe Ratio 0.75
K-Ratio n/a

Trading Period 7 Yrs, 5 Mths, 11 Dys, 3 Hrs, 9 Mins
Percent of Time in the Market 3.93%
Time in the Market 3 Mths, 15 Dys, 18 Mins
Longest Flat Period 5 Dys, 53 Mins

Max. Equity Run-up $509,952.00
Date of Max. Equity Run-up 01/31/12 16:40
Max. Equity Run-up as % of Initial Capital 509.95%



yearly profits

Last 12 month $79,831.00
01/01/2011 $80,310.00
01/01/2010 $40,016.00
01/01/2009 $49,902.00
01/01/2008 $154,136.00
01/01/2007 $67,605.00
01/01/2006 $71,628.00
01/01/2005 $37,335.00
01/01/2004 $3,536.00
 
i agree that it does seem like noise trading here. But is it possible that an edge can exsist in noise trading?. I mean using volatility to capture small targets?

I have posted 2 different symbols now and can post many more, all using the same signal and one un changed set of rules for a 10 year period. Can these results be luck or conincedent?, even when both are net profitable year in year out going back 10 years?.....................If it was truely random and no edge was there, would i not get losing years in between?

Please can someone clarify?

I have posted fair tests as i dont want to waste anyones time hear.....Like i say i can post up now a completely different set of results if all i do is change entry and exits to fill when price is just touched. I have not done that and the results you see have all had entries and exits moved through price before being considered in results, please bear this in mind. Alot of backtesters would not have done this, but they are not being realistic.
 
It seems as if you're relying too much on statistics. Actual trading rarely plays out like the backtested statistics for any sustained period of time.

Another approach would be to use your signals only as a guideline for entering the market. Then, use your discretion to determine whether to take the signal and when to exit.

I know this defeats the purpose of trading a system only methodology but I fear that you will be solely disappointed when trying to trade this system in a mechanical fashion.
 
Your equity curve hints at possible "look ahead" in your backtest. This is either a heavily curve-fitted system or a system that exploits some looking ahead hole in backtest.

If none of the above, then you may have found an equivalent of HFT in lower frequency mode.
 
You must log in or register to reply here.
Back
Top