Stock Price Prediction AI
The "Feature Matrix" strategy trades only at times when certain technical indicators ("features") show a record of historical profitability. The way it does this is by dividing each of the hundreds of features that we use into bins, and then it backtests trades each of these bin levels to find what level the indicator is most profitable at. It then makes a list of the top 10 of these and turns it into a strategy where it enters the trade when any of these 10 happen.
The end result is that you have a group of highly accurate predictors based on past data, but the question is, do these methods do well on future data? It might be that if you look hard enough at a lot of data, you can always find a list of indicators that do well, but that does not mean they have any predictive ability for the future. Or it might be that these methods really did do well at the time, but the modern highly-computerized market corrects for anomilies such as these, making them useless by the time they are discovered.
To test this theory, we ran a backtest, which means we simulated how much profit it would have made if we had traded it for real using historical data. This is a good indication of how well the Feature Matrix strategy would do in live trading for real.
Results: In our backtest, the AI model's accuracy (up or down prediction) was ______.
The profit/loss backtest had a Profit Factor (PF) of ____ and is shown in the table and chart below. A PF greater than 1 means it would have made a profit. This backtest does not include commissions or the bid/ask spread.
You can run this backtest yourself in our Feature Matrix Colab notebook: Pickstocks.com - Machine Learning. It uses our open-source GitHub repo at Pickstocks GitHub. You can also change the data or AI strategy yourself if you want and then backtest it using our repo.