Arjun Bhalla

Learning how to trade

Back

June 18, 2018

For a while now, the idea of the stock market has intrigued me, ever since I watched the Big Short a few years ago and started to understand how exactly the 2008 financial crisis came about. I’ve studied economics for a few years, and am fairly mathematically inclined, but have never quite had the courage to make the leap into exploring the markets.

All that changed with two things - first, my discovery of Robinhood: a commission free trading app. Second, my rapidly growing interest in Machine Learning & Artificial Intelligence. From what I knew of the stock market, swing trading seemed like the perfect challenge for applying reinforcement learning techniques. However, jumping into the process, and allowing an untrained computer to gamble away my money didn’t really seem like the best course of action.


Initial Steps

After waiting about a week and emailing back and forth with the Robinhood support staff, I finally got myself ready to start trading. It was overwhelming at first - positions, options, puts, limit orders, stop loss, ETF’s, and so on. As a result, I chose to take a step back and to start reading about all of these terms, as well as basic investing strategies on Investopedia.

Later, I put $500 into my account and began trading. I put the majority of my cash (admittedly little for the trading world) into the $DIA ETF, and spent a little more buying shares in companies who I thought would soon improve upon their current positions (e.g. $SNAP).

Finally, I made my first profitable trade! With a stop loss of 2-3% and an expected profit margin of 5-10%, I evaluated $SNAP’s position and sold it for a 2% profit ($0.50!). Not much, but it led to greater confidence and interest in the market, and less than a week later I found all my trades being positive, selling 13 shares of $AMD for a 6% ($11.00) profit. Baby steps, of course - but progress, to be sure.


Future Plans

The main reason that the stock market captivated my interest, as I touched on before, was due to the underlying patterns behind short-term price movement, and the fact that it seemed like a challenging, intriguing problem to which I could apply my knowledge of Machine Learning.