Machine learning & Indian Equity Markets

Young engineers with amazing coding skills have a tremendous opportunity in the Indian equity market. However, many of them get carried away by the allure of "algo trading", which is the process of automating a trading strategy and connecting it to a trading account using technical tools such as API keys and bridge software.

While back testing an algo trading strategy may show low drawdown (loss) and high returns, there is a risk of slippage when the strategy is deployed in the market. Slippage occurs when the stop loss doesn't get triggered and the funds in the account get blown away, which can discourage programmers from continuing to trade.Many of the "algo babas" on YouTube sell their software for this reason - they don't have the courage to deploy all their capital using their own strategies.

Instead, programmers with coding skills should use their skills to crunch data and find hidden insights from the vast amount of data that is now freely available or at a low cost. Some sources of free stock price and volume data include Interactive Brokers, while historical options price (premium) and volume data can be purchased from Accel Pix at a monthly subscription of Rs 1600.

Once data is made available, it may be beneficial for programmers to work with experienced traders in the market to deploy machine learning techniques using Java or Python. If a good and predictable pattern emerges, a scanner can be created to filter out trades. To test a strategy, a paper trading account with Interactive Brokers or Ninja Trader can be used.

Keep an eye out for a follow-up blog post on this topic next weekend for more information.