At HADE Platform, one project we have been working on intensely for the last year is the prediction of quarterly revenue and the prediction of long-term future performance using machine learning algorithms. When this project started, our goal was to build a program that would beat Wall St by just 10%, because in this industry, a 10% improvement would be an absolute breakthrough, and we have already crushed that goal.
We have worked on these algorithsm for a long time, tested, and continue to improve. We took 15 companies back in the second quarter that we felt strongly about, with the algorithms being used. Below is the average performance for our machine learning algorithms at predicting future quarterly revenue vs the Wall Street investment banks that spend billions of dollars each year to do the same, which plays a large role in their investment decisions.
Over the second and third quarter, HADE Platform’s machine learning algorithms were 57.5% better at predicting quarterly revenue than the consensus of Wall Street analysts. That means our published predictions was closer to the actual revenue of these companies than Wall St.
This figure is determined with the error percentage. As seen above, our error percentage was lower in all 15 instances, and our error percentage exceeded 4% just once. This compares to 7 instances from Wall Street consensus estimates.
In addition to this news, we did test “unrefined” algorithms on 30 additional companies, and it performed 27.5% better than Wall St estimates for the same 30 companies. We consider this to be huge news, given that stock prices are nothing more than a reflection of expected fundamental performance. This further builds the case for why using artificial intelligence and machine learning for investment research is the future, and much cheaper.
HADE Platform 2.0 for individual users will launch within the next two weeks, and all users will have a very clean, easy to use interface. Here, users will be able to access all of content, data visualization and analysis products, quarterly predictions, and long-term outlook predictions from machine learning for FREE!
