Regular income from stock market using machine learning

Our focus has been to help small investors generate regular income from stock market. After preliminary study of the investors we found they there are 3 key attributes of the potential investors.

1. Lack of knowledge of the stock market. This meant the solution has to be very simple to use.

2. Users always wanted a balance sheet of 1 months i.e. they wanted to see their profit/loss monthly. For us this meant developing short term investment models.

3. Almost all the users knew just one term of the capital market and that was “stock”. This means that building product around fancy instrument of option and future was ruled out.

This finding was the foundation on which we started building our product. These are the 3 objectives with which we started.

Image for post
Image for post
Objective with which we started our Model building

After hand picking the blue-chip stocks on which we will developing the models we started collecting data. Exceedingly early in model building we realized that the magic lies in data wrangling and then feeding to the model. In 4 weeks, we tried to wrangle the data, test its impact on the model. We iterated this till we could not develop any more hypothesis.

When we were building models, we knew that we have to predict less (low recall) but with high accuracy (high precision) to keep small investors confident to invest. Long story short it took us 6 months to achieve this goal. Key traits of this model are

1) Its recommendation rate is not high. We see an average of 5 stocks being held by investors at any given point of point

2) Our %gain/trade is high. This is key to the success because the slippage in market can eat some % from your profit.

3) It has low draw down.

Tentative Growth of ₹100,000 will help you understand the draw down of capital which is quite low. In finance parlance it has a high Sharpe Ratio.

Image for post
Image for post

For users the workflow to use our product is very simple

1) Log in to at 3:00 PM

2) Look at our recommendation and execute the trade from the product itself. It is integrated with broker zerodha. Some users who do not use zerodha manually execute the orders in their broker of choice.

3) On an average user spends < 5 min every day.

4) The holding time of each stock is anywhere between 5 days to 20 days. The average being < 10 days.

We have seen user generate consistent regular income. Most of our user have small account of about ₹ 150,000.

We are motivated to democratize investment in equities for retail investors who have no knowledge of the capital market. We want to make it a mathematical approach so that we have consistency in the long run. Each day is a learning with new inputs coming from users. Some of these inputs even help us improve our model.

Bottom line…If you have clear objective in sight then machine learning can help solve the problem. Having a clear vision of your solution has been the key for us.

Some links that help you better understand about the recommendation engine (Regular Income by short term stock investment) is helping small retail investors generate regular income by short term stock investment using machine learning.

Get the Medium app

A button that says 'Download on the App Store', and if clicked it will lead you to the iOS App store
A button that says 'Get it on, Google Play', and if clicked it will lead you to the Google Play store