The Next Frontier of Artificial Intelligence Predictability in Financial Markets


“This article is written by Yaron Golgher, CEO & Co-Founder I Know First.”

Technology and Automation

In 1837, John Deere invented the steel plow in Grand Detour, Illinois. The tool was designed for farmers to break up tough soil without the soil sticking to it. 48 years later, Karl Benz developed an automobile powered by gasoline, allowing him to drive over 100 km in one trip. 51 years after that, Alan Turing laid the grounds for the future of computing by creating the Turing Machine which was able to solve mathematical problems automatically.

With the steel plow having replaced the necessity of man’s muscle and the automobile having replaced the necessity of horsepower, the last refuge – and one that began to be challenged with the Turing Machine – is man’s mind. We now live in an age of supercomputers that are able to process information and perform operations on the order of hundreds of quadrillions per second. Humans simply cannot compete with these computers.

Disruption in the Modern World

The developments in computing and Artificial Intelligence have coincided with massive disruption in our economic structure, with many jobs being automated out of existence. Soon, the 5 million truck drivers in the U.S. will be left without jobs as autonomous vehicles powered by Artificial Intelligence render their work unnecessary. More advanced jobs are not safe from the development of AI either. For example, machine learning algorithms can now outperform radiologists at identifying diseases from medical imaging.

As the technology of Artificial Intelligence is slowly taking over many fields of work, the financial industry is one in which its effects could perhaps be the most profound.

I Know First

Our company, I Know First, is a leading financial technology company based in Israel, the start-up nation, that is working to formulate stock predictions based on Artificial Intelligence. Alongside Dr. Lipa Roitman, we founded the company with the goal of using advanced techniques to generate stock predictions. Dr. Roitman’s background in chemistry proved useful as he recognized the similarities between the chaotic chemical processes he had studied in the lab and the volatile history of markets worldwide. He noticed that markets tend to follow chaotic patterns, with prices generally reverting to some mean, or trend.

With Dr. Roitman’s algorithmic engineering skills and my experience in financial consulting, we felt a unique dynamic that would help us succeed. Together, we decided to bring our idea to fruition, and founded I Know First. Our mission was to provide institutional and retail clients with insights generated by our predictive algorithm that could help them to make more informed financial decisions. Nowadays, our institutional clients are comprised of several hedge funds, family offices, banks, and wealth management funds.

Since inception, I Know First has significantly grown and consists now of a Research and Development team led by Dr. Roitman and a main business office in Tel Aviv led by myself. Dr. Roitman and his team take an unconventional approach to their research – comprised mainly of PhDs, computer scientists, and mathematicians, they focus on quantitative market models. Although there are financial aspects involved in their work, much of their success can be credited to their strictly scientific approach.

Predictability of Markets

The question of whether or not financial markets are predictable has been a actively debated topic, with some supporting an “Efficient Market Hypothesis” which essentially means that all asset prices reflect available information. Under this theory, it is not possible to consistently outperform the market without illegal means such as insider trading. Over the course of the history of the stock market, this theory seemed to have some legitimacy. Very few were able to consistently achieve outsized returns compared to benchmark indexes, such as S&P 500.

The introduction of computing and artificial intelligence in the financial world seems to have disproved the Efficient Market Hypothesis. Famous hedge funds such as Renaissance Technologies, Two Sigma, and D.E. Shaw all utilize the power of AI to outperform the market. Renaissance Technologies’ Medallion Fund, for example, achieved a 71.8% rate of return from 1994 to 2014. These returns are unheard in the hedge fund world. What allowed them to do so much better than everyone else is their usage of mathematical, statistical, and artificial intelligence powered market modeling in their operations.

The Algorithm

Our algorithm is based on the on artificial intelligence, but what sets it apart from the top hedge funds and other groups is the algorithmic structure engineered by Dr. Roitman. His application of chemical modeling to predict the stock market is unique and not being done anywhere else. Another aspect of our system that sets us apart from others is that it is able to find correlations and relationships across asset classes and across world markets. We currently predict over 10,000 assets in 50 global exchanges, and we are diversified across many sectors and stock universes.

Our I Know First Algorithm not only predicts the movement of stocks, it also incorporates commodities, interest rates, currencies, exchange traded funds (ETFs), and global indices into it. As of now, we forecast these assets across six different time horizons: 3 days, 7 days, 14 days, 1 month, 3 months, and 1 year. The forecasts are then sent daily to our institutional and retail clients. The forecasts we provide not only identify assets that have potential to provide returns, but they also include information about the predictability of the signal associated with the asset. This predictability is a measure of how confident the algorithm is that the predicted asset will be in conformity with the forecast.

We offer customizable solutions to our institutional and retail clients. These include but are not limited to packages targeting biotechnology stocks, S&P 500 stocks, and stocks in a specific country’s exchange, such as Hong Kong. In addition, our algorithm generates daily gold price forecast and forex market predictions to help our clients navigating those markets, especially nowadays.

While we are confident about our forecasting performance, we intend our service to be used with discretion, meaning an investor utilizing these tools should take other information into consideration on top of what I Know First’s forecasts recommend. Doing this, as we have shown from our yearly evaluation reports, can provide clients with a solid investment strategy that will outperform major indexes. In 2019 our algorithm achieved amazing performance by identifying the top S&P 500 stocks that provided significant out-performance against S&P 500 index itself – over the course of the second half of 2019 the outperformance by our top 10 algorithmically identified stocks ranged from some 83% to 150%, on average. This show how our system is clearly effective at outperforming the market through accurate spotting the top stocks to buy, despite the intense overall market volatility, like the world saw in 2019.

The Next Step

Dr. Roitman and I are working every day to make our service even better than it is now. We have been fortunate enough to attend financial technology events around the world and we have won several competitions and awards for being a one of the top fintech companies doing AI-based stock market predictions, according to Geneva WealthTech Forum.

We also have plans for the future, such as expanding our forecasts to other areas such as liquidity and intraday predictions. Our goal is to eventually cover every tradable asset in the world. We are also currently pursuing plans to form a hedge fund using the I Know First algorithm.