view my tester values

 

Hello All

I am new to Forex and MetaTrader4. I have spent many sleepless nights studying and developing my own EAs. I feel that Im getting very close to going live with real money. I would ask that you view my tester values below and suggest any concerns and things you feel I should look at to improve my stats. thank you in advance for your help: Greg

The below are the ONE YEAR $3000 balance MAX 1 POSITION BACKTEST RESULTS for eur/usd Jul2010 jul 2011

Bars in test 2619
Ticks modelled 13953681
Modelling quality n/a
Mismatched charts errors 1632
Initial deposit 3000
Total net profit 62801.26
Gross profit 107012.85
Gross loss -44211.59
Profit factor 2.42
Expected payoff 11.17
Absolute drawdown 426
Maximal drawdown 7015.00 (10.06%)
Relative drawdown 26.02% (1237.40)
Total trades 5623
Short positions (won %) 2811 (3.52%)
Long positions (won %) 2812 (4.02%)
Profit trades (% of total) 212 (3.77%)
Loss trades (% of total) 5411 (96.23%)
Largest
profit trade 6732.8
loss trade -346
Average
profit trade 504.78
loss trade -8.17
Maximum
consecutive wins (profit in money) 2 (6741.80)
consecutive losses (loss in money) 278 (-2777.00)
Maximal
consecutive profit (count of wins) 6741.80 (2)
consecutive loss (count of losses) -2777.00 (278)
Average
consecutive wins 1
consecutive losses 27

 

From the statistical side it looks good, the biggest difference to 'normal' ea's is that you have only ~4% winning trades, this means that you are using a very tight stoploss (which suggests also the average loss/trade) or a extreme high risk revard ratio. The downside (from what i can say watching this statistic) is that the strategy might NEED low spread. What would happend if the spread gets increased to 2.5Pips and sometimes reaches even 20 to 30 pips?

I hope that you are aware of the tester limitations and the difference from a test scenario to a demo or even a live account.

 

I agree 278 consecutive losses is a bit much.

Instead, optimize on one time period (09-10,) and see how those values work on a different period (10-11) If it works there, THEN you have something.

See also What is backtesting? | Invest Press



 
zzuegg:

From the statistical side it looks good, the biggest difference to 'normal' ea's is that you have only ~4% winning trades, this means that you are using a very tight stoploss (which suggests also the average loss/trade) or a extreme high risk revard ratio. The downside (from what i can say watching this statistic) is that the strategy might NEED low spread. What would happend if the spread gets increased to 2.5Pips and sometimes reaches even 20 to 30 pips?

I hope that you are aware of the tester limitations and the difference from a test scenario to a demo or even a live account.

Thanks for the reply. Now im going to have to reveal how newbee I really am. I am just begining to study "money management" so I have difficulty understanding how you figured out my issues.

"~4% winning trades, this means that you are using a very tight stoploss"

Im not using any stop loss. only a moving avarage closes a BUY.

"or a extreme high risk revard ratio"

I have not declare a RR so you are correct I bet the whole farm in the backtest.

"strategy might NEED low spread"

this statement I find most interesting as I noticed that this stategy is only successful when the BUY/SELL price ratio is very close. I suspected that this had something to do with ask/bid "spread" and began researching "spread" last night. I dont really understand how these two relate but concluded that the BUy/SELL spread must be low in order to trade successfuly with this strategy. i hope you can see my ignorance and advise me. I really am very new even thou Ive learned how to create what appears to be a very high profit.

"What would happend if the spread gets increased to 2.5Pips "

The strategy doesnt work on wide spread. Ill need to understand why. Im hoping someone can advise how.

"difference from a test scenario to a demo or even a live account"

My plan is to download good hst history to improve back testing and have heard that the MBTrading HST im receiving may noy be adaquate. Please advise a site. I hear ALPARI hst is one option.

 
WHRoeder:

I agree 278 consecutive losses is a bit much.

Instead, optimize on one time period (09-10,) and see how those values work on a different period (10-11) If it works there, THEN you have something.

See also What is backtesting? | Invest Press




Please forgive my ignorance but I dont understand what you mean by "optimize on one time period" (09-10,) or (10-11). I run optimize on whatever period the strategy is designed for. I am possibly not using he optimizer to its full potential seeing that I dont understand your advice. Please clarify.
 
  1. trueresource:
    Please forgive my ignorance but I dont understand what you mean by "optimize on one time period" (09-10,) or (10-11). I run optimize on whatever period the strategy is designed for. I am possibly not using he optimizer to its full potential seeing that I dont understand your advice. Please clarify.

    A strategy can not be designed for any time. You don't know the future.

    Check the Use Date box (just above Visual Mode box) Optimize over one year, then run it with the optimized settings on another year. If that's profitable you may have something.

  2. I dont really understand how these two relate but concluded that the BUy/SELL spread must be low in order to trade successfuly with this strategy. i hope you can see my ignorance and advise me
    When a trade looses, you pay the spread+SL, The larger the spread the more each loosing trade costs.
 
trueresource:

Thanks for the reply. Now im going to have to reveal how newbee I really am. I am just begining to study "money management" so I have difficulty understanding how you figured out my issues.

"~4% winning trades, this means that you are using a very tight stoploss"

Im not using any stop loss. only a moving avarage closes a BUY.

This is quite the same as a stoploss, if you set the stoploss at the price of the moving average you will get the same result but without requote issues on forward trading

"or a extreme high risk revard ratio"

I have not declare a RR so you are correct I bet the whole farm in the backtest.

I meant that you take many small losses for the chance of a big win, your resulting balance curve most likely look like a stair with slightly downward's periods followed by a big move up

"strategy might NEED low spread"

this statement I find most interesting as I noticed that this stategy is only successful when the BUY/SELL price ratio is very close. I suspected that this had something to do with ask/bid "spread" and began researching "spread" last night. I dont really understand how these two relate but concluded that the BUy/SELL spread must be low in order to trade successfuly with this strategy. i hope you can see my ignorance and advise me. I really am very new even thou Ive learned how to create what appears to be a very high profit.

Yes, spread is the difference between buy/sell price. On tester you have always the same spread, while on forwardtrading the spread increases and decreases during the day depending on the current volume beeing traded all over the world.

"What would happend if the spread gets increased to 2.5Pips "

The strategy doesnt work on wide spread. Ill need to understand why. Im hoping someone can advise how.

On a Alpari standart account typical spread is 1.6 on eurusd, but during the night it will be bigger, (current spread at close of market is 3.6Pips) this might be a issue you have to solve

"difference from a test scenario to a demo or even a live account"

My plan is to download good hst history to improve back testing and have heard that the MBTrading HST im receiving may noy be adaquate. Please advise a site. I hear ALPARI hst is one option.

MT4 is very limited regarding the backtests, as more tick and spread sensitive your strategy is as worser the Backtest/realty ratio gets. No variable spread, no delay in execution ecc..

 
zzuegg:

Thanks for the time you spent with this. Can you recomend a more appropriate back test solution?
 
WHRoeder:
  1. trueresource:
    Please forgive my ignorance but I dont understand what you mean by "optimize on one time period" (09-10,) or (10-11). I run optimize on whatever period the strategy is designed for. I am possibly not using he optimizer to its full potential seeing that I dont understand your advice. Please clarify.

    A strategy can not be designed for any time. You don't know the future.

    Check the Use Date box (just above Visual Mode box) Optimize over one year, then run it with the optimized settings on another year. If that's profitable you may have something.

  2. I dont really understand how these two relate but concluded that the BUy/SELL spread must be low in order to trade successfuly with this strategy. i hope you can see my ignorance and advise me
    When a trade looses, you pay the spread+SL, The larger the spread the more each loosing trade costs.


Ok, I thought you meant periods as in M1, h4 d1 etc. I didnt realize 09 was 2009 etc. You are correct, the strategy did not fair well in the 2009 to 2010 year test.

What Ive learned here today is very helpful. Im thinking about studying for the series 34 forex exam to help me understand what Im doing. Im really flying Blind. If you or anyone can suggest study material (outside of series 34 education) I would certainly appreciate it. Thanks again.

 

to zzuegg or anyone else that can answer:

Im still trying to figure out how zzuegg (way above) could tell that my strategy required low spread? Also how did he come up with the 2.5 pip question. He certainly openned up a new world for me as Ive been studying spreads every since his comment. You see I had already figure out that my strategy would only trade eurusd and usddkk successfully. The only thing I could see that the pairs had in common were the spread. I just dont understand many of the statistic parameters. So if someone can enlighten me. my goal is to learn how to adjust my strategy to changing spreads and/or to limit trading during high spread.

 

Hi,

You have a winning rate of 3.77%: what does this mean?

You are generally go out of the trade pretty soon if the price goes against you. Also the average loss/trade supports that idea. You can't trade high lotsizes since your system survived 278 losses in a row.

This can suggest that your system might be very sensitive to small movements of the price (generally the tester is not very reliable when it comes to tick sensitive trading systems). Also a spread increase of a few points might make your system hit the stop more often. (In a worst case scenario also a part of the winning trades).

The 2.5 pip questions had no specific reason, but i assumed that your general spread was lower and from personal (kind of expensive) experience i learned that spread is all other than similar with a standart testing environment. Fluctuations up to 30pips can and do happen.

But you could give us a few more informations, do you exit by stoploss or based on some indicators? Sorry, that might be all we can tell based on the informations you gave us. All we can see is a statistical overview of your strategy.

At last i suggest to get some good history. You have a few mismatched chart errors. This can also be a killer already