10/19 Prediction INTC - close lower AMD - close higher FB - close higher MSFT - close higher The cloud server is down and it takes me forever to run those simulation in my home pc.
Congrats, you are at coin toss levels. Without listing the cumulative net $$ gain/loss, the win rate itself does not have a lot of meaning. Don't get me wrong 48% is fine as long as your cumulative net profit vs loss is skewed in your favor. With just a 60% up/down prediction rate you still don't have tradeable information.
I ran some numbers for you that make more sense than a win loss ratio. I started with 10K, 2500 for each stock. If the prediction was for a lower close the next day I sold short. If the prediction was for a higher close I bought. I closed and reversed a position when the prediction changed. (So for AMD I sold at the close on Oct 9 , Reversed my position on Oct 15 and Oct 21 and am now short.) I didn't use stops or profit targets. It kind of looks like if I did the loss would have been greater. As it turned out if you cashed in tonight you would be down 2.26%. If you had just bought and held SPY you would only be down 1.19%