We are in the process of training our artificial intelligence (AI) stock valuation model. It takes time and a lot of data. Initial results are encouraging, and we plan to share them with you shortly.

Artificial intelligence is a very broad term that covers a lot of very different techniques, even as simple as a straight line fitting a scattered data set. We go much further that this through use of deep learning and neural networks. The power of these advance AI techniques is truly mind-boggling. Quite often the result looks like a miracle.

It is important to note here that we are not using AI to recognize past stock price patterns and use those to predict future stock prices. What we use AI for is to discover deep interrelationships between company’s finical data and its underlying or intrinsic value. We assume that eventually the market capitalization and the company’s intrinsic value do converge.

Contrary of our intrinsic stock valuation model, where we calculate the intrinsic value directly, in our AI stock valuation model we track changes in the company’s intrinsic value (as it is reflected in its stock price) and changes in the company’s financial performance (as reflected in its financial statements) that cause such changes.

So, in effect, the stock price target generated by our AI stock valuation model could be considered as “relative”, i.e. it is the current stock price plus the price change that our AI model stipulates as justified by the recent change in the company’s financial performance. Whereas our intrinsic valuation model calculates the “absolute” price target, as is not based on or influence by the current stock price.

Please also keep in mind that the AI stock price targets are “market-neutral” in the sense that they assume no change in the overall market prices, which, of course, is almost never the case in real life. The actual stock price performance usually is very strongly correlated with the overall market performance, so the performance of our stock targets (and ratings) should be measured in relation to the performance of the overall stock market.