Value investing. Step to the next level

If you are interested in value investing, you probably know that fundamental analysis is the most effective and at the same time, the most complicated way to determine the intrinsic value of a company. It allows you to make informed investment decisions by purchasing promising companies at a fair price. But this approach takes a lot of time and effort, limiting the investor’s horizons.

To solve this problem, I started the COVANN project. Today, it is a cascade of artificial intelligence models whose task is to analyze publicly listed companies and determine their intrinsic value by using fundamental analysis (P&L, BS, CF and economic data). In this article, I would like to share the results achieved. I have not seen such projects on the internet and have not seen them in articles on arXiv.org. I hope it would be interesting for you.

Important note: this article does not contain recommendations about buying or selling shares. Make investment decisions deliberately and independently.

Human-level or another vision?

Value investing, as well as evaluating the value of a company, is more often referred to as an art than a science. Why? For example, if you give two qualified specialists the task of estimating the value of a company, you will get approximately similar values in absolute terms, but they will not be the same.

Simplistically, this is because everything in this world is relative. In more detail, there are different methods for evaluating companies, analysts may have different sources of financial information, different assumptions may be used in calculations, and a large number of other reasons why the estimated values of companies will be different.

During the development of COVANN, I focused on my calculations and vision of the value of companies. As a result, a model was developed with the results of which I agree in the vast majority of cases, but nevertheless, it is artificial intelligence and its understanding of the intrinsic value of companies.

In order for you to evaluate the results of COVANN’s work, I have prepared a small sample of companies, which you can see below. COVANN analyzes the intrinsic value of companies at the date of publication of the report, so you can easily compare its results with your calculations.

AAPL fair value
TSLA fair value
Ford fair value
GPOR fair value

The value investor knows that the market is inefficient, and the charts shown above demonstrate this well. Therefore, all attempts to directly predict stock prices by quotes or any other means, in my opinion, are doomed to failure.

At the same time, it is this market behaviour that creates good investment opportunities, provided that the intrinsic value of the company is understood.

Where’s the money?

I am convinced that a person should make investment decisions. It was a fundamental idea in the development of COVANN. The goal of this AI is to simplify the search and monitoring of companies that are attractive and promising for investment, regularly analyzing thousands of companies listed on the market.

The concept of using COVANN in a simplified form is as follows:

  1. COVANN – assessment of the intrinsic value of companies. By purchasing shares at a price close to or below fair value, we reduce our risks and increase our potential profit.
  2. COVANN – forecast of fundamental data to identify promising companies.
  3. COVANN – the ranking of companies by the degree of attractiveness for investment. Significantly simplifies the work of the investor in finding and monitoring companies for investment.
  4. Investor – analysis of companies from the ranked list and making investment decisions. It is necessary to make an informed decision and avoid possible AI errors (data errors, etc.)
  5. Profit

And that’s all?

This article presents the results of the first version of AI COVANN, which I accepted for testing as the most promising. It took more than a year to create, many model variants were tested, and terabytes of data were processed, thousands of GPU hours were spent, and there is still a lot of work to develop this project.

If you find this project effective and useful in making investment decisions, please support it with a repost and your opinion. It will help me prepare this project for open testing as close as possible to the needs of the investment community. You can always contact me via the contact form on my blog or LinkedIn profile.

Leave A Comment

Your email address will not be published.