Hello, I would just like to make sure my math is correct for how much I should be making with my EA. I predict the next bar on a 15M chart. The system has been getting around 62% accuracy in backtest/demo account. So my calculations were

3 pips * 5 dollars (pip value) * .12 (accuracy) * 24 (hours in a day) * 4 (4 15M candles in an hour) = 172.8 dollars/day.

The currency is EURUSD.

Does my math look correct for profit/day?

Does your stategy predict *each* next bar?

As for me, most calculation basod on the consistency of something that will take place each day are closer to assumtions. Another point is how your system deals with closing trades. It seems that the rules prescribe to close position after the 3 pip movement, am I right?

The accuracy you`ve metioned is great, and with the right risk-reward ratio the strategy could be very pofitable. What is an average risk-reward ratio an how the system operates with stops?

By the way, which software do you use for backtesting? Most of such tools will provide you with an answer on that question.

Yes, it predicts each next bar. It also closes the trade at 15 minutes after opening the position without taking profit/loss into account.

The three pips is an assumption of mine. I am simply assuming that every 15 minutes, the market will move by 3 pips in some direction.

The R:R is around 1:1, however it varies in different market conditions. It comes out to about 1:1 once averaged.

Unfortunately I have to manually backtest because the strategy involves both hardcoded logic (candlestick patterns/basic cycle analysis) and a bagging method neural network which does not backtest in MQL4. The numbers for the accuracy is simply the neural net accuracy on the testing data.

In conclusion, it opens a trade on every next bar regardless of market conditions, has 1:1 seemingly, and I assume after 15 minutes, the market will move 3 pips.

Also, disregard spread/commission in this calculations. I get about .3 spread on EURUSD average with $2 round trip, so that is rather insignificant.