Dozens of AI Trading Strategies, All With Backtests
For over 40 years I have been trying to beat the stock market. This site is a free collection of all my strategies.
A typical old-fashioned trading system might be something like "buy AAPL if it goes down for 5 days in a row". Candlestick patterns (from stock charts) are also a common strategy. Using AI is more complex though. You can't simply feed the raw stock price data into the AI and expect it to be profitable. At a minimum, you need to add technical indicators such as moving averages, oscillators, and trends. This site handles all of that.
There are also many ways to create an AI model. Dozens of different machine learning algorithms (Logistic Regression, XGBoost, Random Forest, CNN, LSTM, AutoML, etc.) can be modeled to see which is the most accurate. And the data can be prepared in time-series format, regression, or up/down categorical. Then there are data issues to deal with such as missing rows, outliers, binning, and scaling. All of this is tested here on this site.
Even once you generate a good prediction, when do you exit the trade? You could close out at the end of the day, wait until you get a signal from the AI saying the stock is going to move in the opposite direction, or use reinforcement learning which tells you exactly when to buy and sell but some days does no trades at all. And should you use daily data, 1-hour data, 1-minute data? Should you trade an index (like Dow30 or SPY), a specific stock (like Tesla or Google), Forex (FX - Foreign Currencies), futures (like gold, crude oil, or soybeans), or crypto (such as Bitcoin)? Or should you do pair trading where you buy a good stock and sell short a bad stock? Or maybe trade based on news, rumors, or social media postings? This site has the answers to all of these questions.
To get started, choose a trading category: