Based on the recommendation of my friends, I have decided to use StrategyQuant software for creating strategies and expert advisors. Before making a purchase, I started using the Trial version.
Since I had previously used similar software, I was familiar with the general functionality of this type of software. Initially, I loaded the initial data from the broker (Real account) into the software and started generating strategies. Although I have powerful and up-to-date hardware, generating robots still took a considerable amount of time.
I retested the generated robots and eventually optimized them.I transferred all the created robots to Metatrader and started backtesting. The result was unbelievable. The generated robots had opened small trades exactly as I estimated and expected, and the number of trades was exactly according to the settings I had made. However, without exception, all the created robots had drastically reduced the balance from 1000 to 0 with great precision.
Where have I gone wrong in my approach? Shouldn’t I have used broker data? Could there be any specific settings or considerations when entering data into the software?
the generation speed and success rate highly depends on how you configure SQ. If you do incorrectly you can waste a lot of time (wrong markets, templates, indicators, filters). If you filter strategies too strictly for certain markets (forex mainly) it can be very tricky to get decent strategies
As for the results differences it is a common problem usually cause by lower testing precision, incorrect timezone setting etc. Have you exported tick data into MT for backtesting? https://strategyquant.com/doc/quantdatamanager/test-strategy-metatrader-4-tick-precision/
Viewing 1 replies (of 1 total)