How reliable is backtests?
forex_trader_202879
Member Since Aug 07, 2014
406 posts
Sep 23, 2014 at 20:22
Back testing doesn't only have to be used for an EA. If you trade manually you will be TESTING the system which you would use in live. None the less, back testing is a very effective way to tell if a system works or not. The problem is most people back test, and come up with very good rules, but once they begin trading live they forget all that they have tested, which leads one to believe that back testing is useless, but it is the actual trader who isn't consistent enough to trade as they back test.
forex_trader_202879
Member Since Aug 07, 2014
406 posts
Sep 23, 2014 at 20:22
theHand posted:
Probabaly already made you throw some very good systems in the dustbin and made you try losing ones.
Look do the test, take your EA run it for a week, then back test the same week. Compare the results. If you can run it on a second broker.
You'll end up with three sets of results.
You just can't build a future on that. And you're wasting your time trying.
My advice to you. Test live, you see very quickly if you like the system or not.
WHat are you talking about? It depends on the system you are actually running. If your using the silly support and resistance strategy, and brokers have different quotes (due to spreads) then of course you will get different results! You guys are making opened statements and have yet to provide proof of what you are claiming.
Backtesting not working is like saying ¨studying before an exam doesn´t make sense¨.
forex_trader_202879
Member Since Aug 07, 2014
406 posts
Sep 23, 2014 at 20:22
The only way to reverse engineer anything is to BACKTEST IT!
Member Since Nov 21, 2011
1718 posts
Sep 23, 2014 at 21:36
Cholipop posted:theHand posted:
Probabaly already made you throw some very good systems in the dustbin and made you try losing ones.
Look do the test, take your EA run it for a week, then back test the same week. Compare the results. If you can run it on a second broker.
You'll end up with three sets of results.
You just can't build a future on that. And you're wasting your time trying.
My advice to you. Test live, you see very quickly if you like the system or not.
WHat are you talking about? It depends on the system you are actually running. If your using the silly support and resistance strategy, and brokers have different quotes (due to spreads) then of course you will get different results! You guys are making opened statements and have yet to provide proof of what you are claiming.
Backtesting not working is like saying ¨studying before an exam doesn´t make sense¨.
Well if you believe using support & reistance is silly thing it means only one thing... you don't know how to use it.
Sorry to say TheHand is right as I have already experienced it. Backtesting is some case ins't able to reproduce what have done the EA for the same day.
Now a simple test with both EA included into any MT4 broker with 3 differents brokers will generate 3 different results. That's it.
Sep 24, 2014 at 06:41
@Cholipop
I've been at this a decade. My statement is based on hard won lessons over the years. You can back test all you want, I strongly suggest that people who have any intention of getting somewhere don't trust it at all. There's just to many factors that prohibit an accurate result, with MT anyway...
Even testing will still give a different result to live account. There is only one way to really test a system, that's live with a funded account.
I've been at this a decade. My statement is based on hard won lessons over the years. You can back test all you want, I strongly suggest that people who have any intention of getting somewhere don't trust it at all. There's just to many factors that prohibit an accurate result, with MT anyway...
Even testing will still give a different result to live account. There is only one way to really test a system, that's live with a funded account.
forex_trader_202879
Member Since Aug 07, 2014
406 posts
Sep 24, 2014 at 06:47
CrazyTrader posted:Cholipop posted:theHand posted:
Probabaly already made you throw some very good systems in the dustbin and made you try losing ones.
Look do the test, take your EA run it for a week, then back test the same week. Compare the results. If you can run it on a second broker.
You'll end up with three sets of results.
You just can't build a future on that. And you're wasting your time trying.
My advice to you. Test live, you see very quickly if you like the system or not.
WHat are you talking about? It depends on the system you are actually running. If your using the silly support and resistance strategy, and brokers have different quotes (due to spreads) then of course you will get different results! You guys are making opened statements and have yet to provide proof of what you are claiming.
Backtesting not working is like saying ¨studying before an exam doesn´t make sense¨.
Well if you believe using support & reistance is silly thing it means only one thing... you don't know how to use it.
Sorry to say TheHand is right as I have already experienced it. Backtesting is some case ins't able to reproduce what have done the EA for the same day.
Now a simple test with both EA included into any MT4 broker with 3 differents brokers will generate 3 different results. That's it.
Ok Crazy Trader. Now since you say that three different brokers will give three different results, why not go one step further to identify the reason as to WHY! Once you understand the reason why it will then lead to the answer as to why. It isn´t fair to say it DOESN´T work when it HASN´T worked for you. For the most part lack of constant signals, or different quotes to brokers will result in the change. Yet, i take it neither of you actually programmed the so called EA which you use. Which of course explains your pessimitic view of back testing not working.
Member Since Nov 21, 2011
1718 posts
Sep 24, 2014 at 07:20
I have already explained Backtesting is designed to check if EA is well coded.
Example, You have the time to check it on week end, Market is closed, but you can use backtest to get an idea how it would have worked. Check if EA behaves as it should do. Then you know if it worth to try it demo once you make sure you don't have any bugs.
There is no need to understand why it doesn't produce the same result for each test. You guys are using it in wrong way. Simply because you are fooled by tons of scamers that present amazing backtests. They show backtest because they can't show Live account. Backtesting is far away from reality.
Example, You have the time to check it on week end, Market is closed, but you can use backtest to get an idea how it would have worked. Check if EA behaves as it should do. Then you know if it worth to try it demo once you make sure you don't have any bugs.
There is no need to understand why it doesn't produce the same result for each test. You guys are using it in wrong way. Simply because you are fooled by tons of scamers that present amazing backtests. They show backtest because they can't show Live account. Backtesting is far away from reality.
Member Since Nov 21, 2011
1718 posts
Sep 24, 2014 at 07:23
Ps: Backtesting file can be edited and modified before uploaded!
Oct 01, 2014 at 15:12
RSTrading posted:
I suppose you meant 'not even at 100%'?
So does that extra 9% make a huge difference. Should one rather go that route?
I know backtesting doesn't factor in stuff like spread and commission, but lets say you set the spread to a ridiculous figure and know what your commission is per trade, will that do?
What do you guys then do to get accurate backtesting?
Actually when you do backtest you can set spread.
So, i mean if modelling quality is more than 90% and also spread is calculated, then it's good.
Stable 20% per month without manual intervention...
forex_trader_139412
Member Since Jul 16, 2013
385 posts
Oct 07, 2014 at 21:01
So if I get good results on 90% modeling quality It won't perform like that on real?
Member Since Apr 20, 2010
814 posts
Oct 08, 2014 at 06:57
It didn't perform well as in backtest is because you did not enough back test to see the bad performance might happened.
You need backtest year by year start from 2000 - 2014.
After it pass all these year, you might need try random start date for 3 month test, if stop by close you need increase longer period.
From backtest, you need to figure out what is the save margin Account to play with minimum lot and what is the Risk Reward payout.
Of course there are exception for some EA like the : fast scalping less than 1 pips within split second where backtest can able perform the operation while real it get re-quote and spread factor result scalp Loss instead of Profit.
You need backtest year by year start from 2000 - 2014.
After it pass all these year, you might need try random start date for 3 month test, if stop by close you need increase longer period.
From backtest, you need to figure out what is the save margin Account to play with minimum lot and what is the Risk Reward payout.
Of course there are exception for some EA like the : fast scalping less than 1 pips within split second where backtest can able perform the operation while real it get re-quote and spread factor result scalp Loss instead of Profit.
Information is Gold when come to organised.
*Commercial use and spam will not be tolerated, and may result in account termination.
Tip: Posting an image/youtube url will automatically embed it in your post!
Tip: Type the @ sign to auto complete a username participating in this discussion.
