Imagine that you have a black box that can predict whether an equity will be binarily up or down in the next three minutes with a 53% accuracy. Given that you will be making many trades with high leverage at the three minute level the bid/ask spread will likely nail you if you only do market orders because the movement on this time window will be so small. What would be the best overall strategy to utilize this black box while not getting killed by slippage/spread and broker fees? Is it even possible?
53% sounds low, but does not take into account expected profit. You need to make more money on the good ones than the bad ones. With that you might be able to drop below 50/50.
You would need to adjust wich market is worth and size =) In example, e-mini would be ok (cheap round trip, high contract size).On the other hand If you'd like to trade micro-euro, probably you would need to increase number of contracts.
The probability of profit means nothing. You could be right 90% of the time and still have a net losing strategy if your losers are 10x greater.