Testing a New Strategy

alessio · #1 08-12-09, 12:34 PM

When you test trading strategy, it is necessary to understand that if, for example, you enter at the beginning of an hour you will be one or two minutes in the beginning of new hour to enter into the market more often or to expose warrants, and sometimes and it will not be.

Those movements of a following bar which we see in the history, very often occur first minutes of hour or a half an hour, under an exit of the fundamental data or someone's performance. Certainly, - so occurs not always, but it is frequent enough seriously to cut down the profitableness shown by testing.

In general, it is pleasant to see strong movements in correctly predicted party in the history analysis. The problem is that these strong movements occur very quickly. To enter into the market in such situation it is very difficult and dangerous. Indicative quotes which you see, lag behind the market, the real spread extends, quotations are given for very short time, - you can not have time to confirm them and if will be in time, - not the fact that the transaction will take place.

lancehibner · #2 08-22-09, 10:51 AM

It is very useful to us..You have done an great job.Such an informative and well formed post!Thank you for sharing that, it was wonderful of you.

mikeMIKE · #3 01-07-10, 01:48 PM

As far as the success of back-testing is not 100%. It may seem like that, but in back-testing you are eliminating a crucial factor - your psychology.

Forward testing is the only valid way to prove the success of a trading system.

forexexpert · #4 02-02-10, 06:55 AM

The EAs can be testing only through backtesting.

FXBanker · #5 03-06-10, 05:19 AM

Quote:

Originally Posted by alessio

When you test trading strategy, it is necessary to understand that if, for example, you enter at the beginning of an hour you will be one or two minutes in the beginning of new hour to enter into the market more often or to expose warrants, and sometimes and it will not be.

Those movements of a following bar which we see in the history, very often occur first minutes of hour or a half an hour, under an exit of the fundamental data or someone's performance. Certainly, - so occurs not always, but it is frequent enough seriously to cut down the profitableness shown by testing.

In general, it is pleasant to see strong movements in correctly predicted party in the history analysis. The problem is that these strong movements occur very quickly. To enter into the market in such situation it is very difficult and dangerous. Indicative quotes which you see, lag behind the market, the real spread extends, quotations are given for very short time, - you can not have time to confirm them and if will be in time, - not the fact that the transaction will take place.

Thank you for this useful information

Jake Tapper · #6 03-09-10, 08:27 PM

Tell me please - wich difference between real trading and back-testing?
I found new forex robot, and trying to test it.
he may show different results on a real trade?

JamesFX73 · #7 03-25-10, 07:01 PM

alessio,thanks for sharing.Very useful info.

fdfx · #8 03-30-10, 05:45 PM

Back-testing is rubbish. You better test on a demo account.

Currently Active Users Viewing This Thread: 1 (0 members and 1 guests)

Testing a New Strategy

Testing a New Strategy

Currently Active Users Viewing This Thread: 1 (0 members and 1 guests)