Articles on: Getting Started

How to use AI Reports

How to use AI Reports

The AI score constitutes a revolutionary easy way to pick stocks. Usually, stock analysis is a cumbersome process. You could spend days analyzing a business in detail. Even then it is highly uncertain that one would happen to focus on the right combination of parameters without many years of experience.

Enter the world of Artificial Intelligence. We have developed an AI which can analyze a business from a fundamental standpoint and provide a real edge. This is the same basic form of business analysis you would expect from Warren Buffett or a Certified Financial Analyst. Having analyzed the business, the investment proposition is analyzed from a purely mathematical standpoint. Then the technical situation is taken into consideration. And also statistical merits are analyzed. All in all more than 400 data points contributes to an Artificial Intelligence-based evaluation of the merits of any investment.

same basic form of business analysis you would expect from Warren Buffett or a Certified Financial Analyst.
This process is carried out by a supercomputer. It manages to perform a large number of analyses in mere hours, where over a hundred human analysts would take weeks to approximate a similar result.

We refer to the output of this process as the AI score. It presents an effective and simple-to-use shortcut to better decision making.
The AI score is basically a single number. Each stock has an AI score, a number which embodies most of the relevant qualities pertaining to its expected performance, and simply whether or not it makes good sense to invest in this stock.

Imagine then that we divide the stock market into 20 baskets. In each basket, we put 1/20th of the entire stock market, indexed by its AI score. So that in basket #1, we place stocks with an AI score of between -100 to -90, in basket #2 we place all stocks with an AI score between -90 and -80, and so on. Once the entire market has been put into 20 baskets, we compute the performance of all these baskets.

This is the result:

AI score results divided into baskets. Columns = IRR, gray line = win rate

Each bar represents an AI score level. You can, for instance, read from the chart, that an AI score of 50–60, has associated with it an IRR of 17.03% as indicated on the right Y axis. What this means, is that if you were to buy all stocks with an AI score in the interval 50–60, and keep these stocks throughout a very long period, you would, on average obtain a 17.03% return per year. Knowing that the general stock market only offers 6 or 7% pro anno, this is a rather intriguing proposition.

It tells you that stocks with an AI score of 60 for instance, has a rare quality: They outperform the stock market. Of course, many other AI score’s represents outperformance as well. While other lower AI scores levels indicate average or subpar performance.

The gray line on the chart indicates the win rate. This is how many stocks in this basket tends to win, as in produce a positive return at the end of the investment period.

You can observe from the chart that the progression from AI score level to AI score level in terms of both performance and win rate progresses fairly smoothly. From a statistical point of view, this demonstrates the efficacy of the AI score. It actually predicts future returns with an impressive degree of precision. Does this mean, that if you buy a single stock with an AI score of 60, that you will get 17% returns pro anno? Not necessarily. It is a statistical average. Buy 100 such stocks, and keep them through an entire market cycle, and you would have a very good chance to attain a similar number.

The AI score

The score itself is a number between -100 and 100. -100 indicates a toxic stock better left alone (indicated by the color red). Whereas 100 indicates a rather rare combination of positive traits that one would only rarely expect to come across in a single stock (indicated by the color green).

In between is the entire stock market, sorted and ordered by merit from the investor's point of view. It could not be any easier. Just buy green.

Let us have a look at the actual AI score display for a stock.

AI score example of IBM when the price was $107.57

Here you see a dome color-coded from red [DANGER] to green [SAFE]. The indicator points to the score, and it intuitively tells you more about a stocks merits as an investment, than a 12-page investment report would do.

If you would like additional detail, you can observe other scores below such as reward and risk. (more are available)

Reward and risk scores for IBM

Reward tells what the expectancy is from this stock in terms of achieving an investment return above and beyond the market return (alpha). Safety (the inverse of risk, as in “Risk & Reward”) indicates the probability of the stock also delivering on this expected positive return.

But you really only need the total score at the top in order to achieve great results. Only a high implied reward coupled with a low degree of perceived risk could produce a very high AI Score. Risk & reward are to a large degree the most important concepts in the investment process.

The competition.

AI score has been under development at Formula Stocks for 15 years. As pioneers in the field, we deployed it in stock selection in 2009. Our preselected Formula Stocks portfolios use AI score internally too along with several other forms of intelligent analysis. While the portfolios represent stocks preselected by other AI logic, the AI score allows you to select your own stocks relative to your own personal preferences and put together your own personal portfolio.

So for the investor who likes a hands-on approach, wants to select his own stocks, wants to know more about a given investments merits, prefers a certain sector mix, etc., AI score is tough to beat. Whether to use portfolios vs. AI score, is really a question of temperament, or preference for a full-service approach with a pre-constructed portfolio, vs. a portfolio you put together yourself.

Being based on Artifical Intelligence is in no way a hindrance compared to human analysis. On the contrary. The machine is a cool, logical device, not easily misled by a shiny narrative, an illogical bias or a persuasive CEO on a conference call, and possesses logical, mathematical, statistical and memory capacities which vastly exceed that of the human brain by an order of magnitude.

This does not mean that the machine thinks in the way we humans think. But it does mean that it sorts, weighs, remembers, calculates odds and statistics infinitely much better. It learns from mistakes with no ego present and without any complexity constraints. As such it is a brilliant learning machine.

We have come to love AI score. We believe you will too.

Updated on: 23/08/2021

Was this article helpful?

Share your feedback


Thank you!