Methods to back test a strategy - (manually)

Have a question about trading? Any questions are welcome. Go ahead!

Methods to back test a strategy - (manually)

Postby RonLow » Sun Aug 30, 2009 7:20 am

Dear all

Can someone advice me on how to back test a strategy manually? The reason is because i don't know much about coding.

1) I would like to back test a strategy by running through for 1 year. However, i always have this problem of getting to that specific date to start with. Is there a way that i can zoom straight to a specific date so that i can continue? Ie, 01-01-2008?
2) Is there an easier method to test a strategy without running through the EA?

Appreciate your advice

Thanks

Ron
RonLow
 
Posts: 2
Joined: Sun Aug 30, 2009 7:07 am

Re: Methods to back test a strategy - (manually)

Postby Edward Revy » Sat Sep 05, 2009 5:48 pm

Hi Ron,

1) When you run a strategy tester for an EA (the Tester window is opened with Ctrl+R), you can specify the exact date to start backtesting from as well as the exact date to finish the test.

2) Without an EA the only way to test a strategy is to do a manual forward testing from the required date.
Attachments
mt4backtest.png
mt4backtest.png (6.88 KiB) Viewed 8414 times
User avatar
Edward Revy
Site Admin
 
Posts: 209
Joined: Wed Jul 01, 2009 6:20 pm

Re: Methods to back test a strategy - (manually)

Postby RonLow » Sun Sep 06, 2009 5:18 am

Hi Edward

Thanks for your reply. Seem like i may have to start learning how to do EA.. :(

Btw, can i ask you another question?

I started exploring Forex based on USD/YEN but i found it to be too slow. I am currently on EUR/YEN.

If i were to practice my money management based on 1:3 win/loss ratio, it would means that i would need a TP of 60pips if i were to set my SL at 20pips. My current strategy is on 5mins chart using 3 EMA lines (8,14,21) and may start to execute a trade when EMA 8 crosses EMA 21. I am also using Stoch (5,3,3), MACD (12,26,9) and Support & Resistance as references.

My challenges would often be when to take profit! Most of the time, it may not reach 60pips, or i might closed off the trade when the EMA crosses again but later to realised that it reversed to hit my TP, or when sometimes, i will just close the trade when it reversed on me to to hit 0pips profit so as not to avoid any losses. I started with a $2000 account and now i am down to $600. I was doing fine at first but started losing when began to punt. I would extend my SL knowing (/hoping) that my trade would reversed but end up losing more. Many a times, my wins cannot cover my losses even i have a 60~70% win ratio.

I have stop trading and would like to back test again my strategy and some of those you have shared to find something that is suitable for me. Definitely my confidence has been affected!:(

Would really appreciate some guidance from you on my strategy and money management as i want to improve and start making some money.

Hope to hear from you soon or any traders that have experienced these and have a great week ahead.

Cheers.

Ron
RonLow
 
Posts: 2
Joined: Sun Aug 30, 2009 7:07 am

Re: Methods to back test a strategy - (manually)

Postby Dennis » Wed Sep 09, 2009 9:56 pm

I have a trade simulator that might help.

Dennis
Attachments
Trading simulator.zip
(490.98 KiB) Downloaded 712 times
Trade_Simulator_Installer.zip
(119.96 KiB) Downloaded 686 times
Dennis
 
Posts: 1
Joined: Mon Sep 07, 2009 10:31 pm

Re: Methods to back test a strategy - (manually)

Postby Edward Revy » Wed Oct 07, 2009 7:07 pm

Hi Ron,

as I see it, there is still a lot to be done before you'll be able to make your system follow any desired risk:reward rule.

Having a strategy that was never designed around risk:reward ratio in the first place is one thing, adding that risk:reward rule afterward is another, totally different thing.

The problem is, you and me and other traders, we're all on the same boat: we cannot just wish for any strategy to follow our requirement of, say, 60 pips TP and 20 SL; some systems will never be able to adopt that.
I've tried to address this question here: forex-risk-reward-folly-t114.html

In your particular case, we have an "unpredictable" system when it comes to taking profits: Moving averages, MACD, Stoch and many other indicators don't advise on future profit targets upfront. This immediately means that we cannot implement risk:reward ratio just by setting/choosing approximate or desired numbers.

When it comes to risk:reward, all numbers must be justified, e.g. setting a profit target of 100 pips for 1 min charts is unreasonable, same as setting a stop loss of 10 pips on daily charts. We know it from practice and experience. But then, why we allow guessing on other time frames? What would be a practical reason to believe that a profit target of 60 pips for EURUSD on hourly charts using certain entry method is a reasonable target? True answers can be found only after thorough evaluation, basically days and moths of dedicated testing. All trading strategies that cannot advise on profit targets upfront require such evaluation.

What strategies can advise on profit targets upfront?
These are: Support/resistance based strategies, channels, trend lines, pivot points, Fibonacci, Elliott etc - the fundamental pillars of technical analysis.

Regards,
Edward
User avatar
Edward Revy
Site Admin
 
Posts: 209
Joined: Wed Jul 01, 2009 6:20 pm


Return to Newbie questions & answers



cron