Backtesting
i just wondering, is the backtest just a dream or can we keep it for strategy plan that we do? your coment please....
I agree we should put some efforts to make backtest a reality, that we can trust on, instead of wasting time on forward testing, we should make backtesting more accurate, there are so many ea's that perform great in backtest but fails in forward test.
We have plateform, computer power even some guys arrange tick data also, now some solution for real backtest.
I agree we should put some efforts to make backtest a reality, that we can trust on, instead of wasting time on forward testing, we should make backtesting more accurate, there are so many ea's that perform great in backtest but fails in forward test. We have plateform, computer power even some guys arrange tick data also, now some solution for real backtest.
I think accurate backtesting is still a dream!
The accuracy/results of MST is a very misguided and the only guide is the forward testing.
It's just a humble opinion from one who trusted the MST to the farthest!
Well, another thing , if the MST accuracy/results are 100% accurate for the history of the market? what does it mean for the future of the price? (The same issue for indicators ploting the history)
I think accurate backtesting is still a dream!
The accuracy/results of MST is a very misguided and the only guide is the forward testing.
It's just a humble opinion from one who trusted the MST to the farthest!
Well, another thing , if the MST accuracy/results are 100% accurate for the history of the market? what does it mean for the future of the price? (The same issue for indicators ploting the history)I agree backtest doesn't gurantee future results. But at leaset back test should match forward test. It all mathmatical if some EA has produced some result in forward testing, it should produce same result on back test on same data?
Plus if there could be some playback feature can be added, sometime its much easier to see how EA would have placed a traded in market in playback mode or how indicator perform in runing tick by tick mode.
I really like Strategy tester that is under developemnt by Canadian programmer, that has playback featyure.
I agree backtest doesn't gurantee future results. But at leaset back test should match forward test. It all mathmatical if some EA has produced some result in forward testing, it should produce same result on back test on same data?
Plus if there could be some playback feature can be added, sometime its much easier to see how EA would have placed a traded in market in playback mode or how indicator perform in runing tick by tick mode.
I really like Strategy tester that is under developemnt by Canadian programmer, that has playback featyure.Where can i find the playback strategy tester thet you mentioned? can you attaching it on this forum? thanks....
There is nobody to blame but MT4 programmers who didn't care about every tick to be stored and retrieved from the history properly or not at all.
That says it all!
My scalping EA(1min chart) shows different backtesting and forwardtesting on both demo & live accounts(different servers) of interbankfx. That is 4 different result. Plus when currencies begin to bounce like 20-30 pips on a 5min bar for several minutes - during this time backtesting for many month(entire history period actually) is different from a regular backtesting.
I get similar profit result during backtesting and forwartesting though. Major difference is different open time and I have more trades on live account than demo, both forward and backtesting.
There is nobody to blame but MT4 programmers who didn't care about every tick to be stored and retrieved from the history properly or not at all.
That says it all!
Okay keris, I'm lost here. Can you help me by explaining what asmdev mean on his last statement ? Thank you in advance
Those who wrote MetaTrader did lousy job working on history. They still fixing bugs but not the bugs we need to be fixed.
Ex: Recalculate option - most of the time you need this to be ON in build 184; in build 192 if this option is checked,unckecked many times I have to close MT4 otherwise "start" button doesn't work. I read somewehere on their website that they included automatic recalculate in later builds. And how much time was it since build 184 - many months
Those who wrote MetaTrader did lousy job working on history. They still fixing bugs but not the bugs we need to be fixed. Ex: Recalculate option - most of the time you need this to be ON in build 184; in build 192 if this option is checked,unckecked many times I have to close MT4 otherwise "start" button doesn't work. I read somewehere on their website that they included automatic recalculate in later builds. And how much time was it since build 184 - many months
Hi asmdev, thank you for the explanation. All the best for you
There is nobody to blame but MT4 programmers who didn't care about every tick to be stored and retrieved from the history properly or not at all.
Please everyone, report found MT bugs to
https://www.metaquotes.net/bugtrack/
This is the direct link to developers of MT.
- Free trading apps
- Over 8,000 signals for copying
- Economic news for exploring financial markets
You agree to website policy and terms of use
i just wondering, is the backtest just a dream or can we keep it for strategy plan that we do?
your coment please....