Discrepancy Between Backtest and Live Trading Results on Exness Broker with MQL5 Bot

Hello everyone,

I'm encountering an issue with my trading strategy that I've developed using MQL5 on the Exness broker platform. The strategy performs exceptionally well in backtests using historical data provided by Exness (from100$ to 10000$ in 6months). However, when I deploy the same strategy in a live trading or demo environment with Exness, the results are unexpected and significantly differ from the backtest results.

I've taken several steps to investigate the issue, including:


I'm still unable to pinpoint the exact cause of the discrepancy between the backtest and live/demo trading results. I'm reaching out to the community to seek advice and guidance on potential reasons for this inconsistency and possible solutions.

If anyone has encountered similar issues or has insights into factors that could affect the performance of a trading strategy between backtests and live/demo trading on the Exness platform, I would greatly appreciate your input.

Thank you in advance for your help!

Backtesting is suitable for testing to see if the strategy or EA works. Historical data is no guarantee that the strategy will work similarly in the future. Additionally, many backtesting platforms have an issue including real-world situations like latency, spread gaps, slippages and errors. In simple terms, the tugs and jerks you see in live situations do not exist in backtesting situations or are hard to replicate. This gives you smooth backtesting results which are not realistic.

If you like, we can discuss more in private. I do a lot of backtesting for people to see if a system works but not to data dredge or fit the curve.

Does your backtesting based on live tick data + include commission and slippage on raw trading data?
If it is a scalping method then the above is pretty much the reason why backtest vs live such a difference, given if they are under the same brokers.

When you are running live trades, run in another terminal in a demo account and compare two trades and you will exactly know the reason.

As the previous commented pointed out. Live trading will include spread, latency and slippages. Depending on the strategy entry point is also a factor. does your strategy enters at candle open? candle close? etc
Ideally you would run a paper trading and live trading at same time to compare entry points and see if the problem is commission/spread/fees or entry point.

depends mostly on the strategy, strongly recommend using only mql5(metatrader 5) algos

use "Every tick (based on real ticks)" as the data source

backtest its only backtest, forward test its the real validation (always let it run on demo for whatever time it needs to confirm that the trades are as close as possible to the backtest), also cross validate on the end of week (after 1 week on realtime forward testing on demo, run an backtest on the same period and compare the results)

for example I have an algo did by myself which does not behave like in the backtest but still its profitable in realtime (that as an example of keeping your expectations grounded, you test it to know how it behaves, if you know how it behaves then you can know what to expect )