General

Posted On: September 7, 2023

AI driven investment management is possible?

A very heavy topic for the day I guess…

Let us define what we mean by AI driven investment decisions and where are we investing and what is investing.

AI driven investment decisions are just following an algorithm that is powered by neural networks and AI platforms to invest in stock market.

Investment management industry has been experimenting with multiple versions of this concept starting from simple rule-based investing strategies through quantamental models for signal generation to various Flavors of it namely find alpha, find theta and various other Greek alphabets.

So, what is investment management?

Well, it all boils down to few basic fundamental concepts:

1. Identify the stock to invest in
2. Identify how much of this stock to hold in a portfolio.
3. Determine when to buy and hold and exit the stock.
4. Constantly monitor the portfolio and adjust for changes in the market.

So, can all this be done by an AI? is AI so powerful that we don’t need investment managers anymore?

The answer is Yes and No (yes for the first part and No for the second part)

Let us dive deeper beyond this theoretical rambling…

1. Identifying stocks: this is done through deep analysis of fundamental, technical, chart, macro data, market view and consensus of fellow analysts.

Do machines help in this step:  yes, many software packages are freely available today that help you identify a stock based on the various parameters you determine.

Can this be further enhanced to auto-decision?  Yes, auto-decisioning is possible.

• How much of stock hold in a portfolio?

Do machines help in this step: yes, back testing a portfolio is a basic feature in most investment management software today.  Infact, if it’s not available, the software is not favoured by the investment management community.

Can this be further enhanced to auto-decision?  Yes, auto-decisioning is possible with a bit more algorithms and machine learning.

• When to buy or hold or exit a stock?

Do machines help in this step:  yes, it is a basic feature in most stockbroker and investment management software today.

Can this be further enhanced to auto-decision?  Yes, auto-decisioning is possible with a bit more algorithms and machine learning.

• Portfolio monitoring

Do I need to say more?

Now the big question?  If all this is automatable and machines can handle it, what is the need for a human being?

Well, most of the airplanes fly by wire but still have pilots in them, why?   The expertise of the pilots are needed to handle the occasional turbulence, planning the approach and the take off and the flight paths.  Similarly, the investment managers will use an arsenal of tools to help smoothen the entire investment management process for a safer and smoother investment journey.