I participated in a competition at the U of T Quant Finance Club, where participants were given movements of two energy stocks for the past 365 days, and then had to develop a Python trading bot that managed an initial sum of $100,000 and bought/sold and give a presentation on said bot.
My group used the RSI and mean reversion strategy – we fit the graph to a linear plot, and then bought/sold relative to how far away the current price was from “the mean”. The theory is that despite small fluctuations, the stock price will go back to the mean, and we can sell if it’s above the mean and buy if it’s below the mean.
We were also given examples of qualitative “shocks”, such as changes in energy policy, and I learned how I would fine-tune models to accommodate these shocks.