MT4 Strategy Tester results are changing over time - why?
You are missing trading opportunities:
- Free trading apps
- Over 8,000 signals for copying
- Economic news for exploring financial markets
Registration
Log in
You agree to website policy and terms of use
If you do not have an account, please register
Maybe someone has experienced the problem that I describe here and can advise what should be done about it.
I noticed for quite some time that the results that I get with Strategy Tester are not consistent. One day I can test an EA, optimize it, be happy with the results and get a printout of the results to document everything. Yet, on a following day I test the EA again using the optimized parameters and the same data only to find that I can’t replicate the same results. It happened not on one but on 2 MT4 platforms that I use to test advisors.
I download tick data from Dukascopy with the Data Manager module of StrategyQuant and I use this software to Export hst and fxt files to the MT4 platform after applying the “MT4 properties” script which adjusts the currency properties specific to the broker to the exported data. The modeling quality obtained with this process should be 99%. The fxt files are read only.
A couple of days ago I analyzed the performance of a certain EA. When I could not find with optimization any parameter that could fix a large drawdown with this EA I restarted the testing platform and began testing another EA. I got a better sustainability on the same data but still far from something that I could use. So I started to optimize the various parameters of this EA, and except for one parameter which improved the performance by just a little, none of the other parameters had influence on the results. For example when I optimized "signal" and "signal period" from 1-100 with step of 1, the balance has not changed at all and remained about +/-$ 800 for a 14 months period. Optimizations of other parameters with increasing ranges basically behaved the same. The Modeling Quality never showed up a value of 99%.
I then needed to go through my weekly comprehensive computer cleaning routine and after doing that I fired up the testing platform and ran H4 strategy tester with the same data that I used before the cleaning and the default setting of the EA. To my surprise this time I got a decent equity curve with 99% modeling quality (see report).
Since everything ”improved” after clearing/cleaning the computer cache, temporary files, junk files, registry etc, I wonder if MT4 generates temporary files during the Strategy Testing and if it does where it stores them and what is the best approach to clean them periodically during the testing. Also does it use cache to store information during the testing that may become sticky and may interfere with the testing itself at a certain point? If this is the case what is the best way to periodically clear it. Any contributing information in this regards would be greatly appreciated.
Thanks in advance, Ron