By April Ye & Connor Anderson - Algorithmic Trading Team

All trades and graphs below were made in Python

Algorithmic Trading: executing trade orders using automated pre-programmed trading instructions accounting for variables such as time, price, and volume. There are 2 main types of trading strategies: Mean Reversion and Momentum. We used a momentum strategy.

Our strategy: we relied on Heikin-Ashi candles to calculate indicator date, which we then ran through our momentum trading strategy to determine whether or not to buy or sell. Our momentum strategy relied on Heikin-Ashi candles as well, using the 3 days before an indicator date to determine the current direction of the market.

Example Buy & Sell Patterns (Based on CVS stock):

Our goal was to create a momentum trading strategy that performed positively on 4 chosen stocks:

• CVS
• Tesla
• Twitter
• Walmart

To calculate indicator dates, we looked for Heikin-Ashi candles with "wicks" more than 3 time the length of the candle on both the upper and lower shadows.

After determining indicator dates, we looked at the previous 3 days' worth of candles to determine if a clear shift in momentum could be concluded. This entailed looking for 3 negative days prior for a buy indicator, and 3 positive days prior for a sell indicator.

We created a visual of indicator dates on our Heikin-Ashi Plot (fig 1), then compared the Heikin-Ashi Plot to a regular candlestick plot (fig 2).

Figure 1: Heikin-Ashi Plot

Figure 2: Regular Candlestick Plot

While trading with $1,000,000 and no stock broker fees, simulating how our algorithm would have performed this past year, our results were as follows:

• CVS -> 1.004%
• TSLA -> 1.021 %
• TWTR -> -0.013%
• WMT -> 1.009%

Had we actually traded this past year, we would have made $18,000. While these returns are narrow, this makes sense since we were conservative with when we bought stock:

Had we traded on the Dow Jones Industrial AVerage this past year, we would have had 1.008% positive returns (trading using $100,000,000).

Our momentum trading strategy thrives on small gains that compound over time. Coming into this, we had the perception that we would be making lots of money with every trade. After studying this strategy for a semester, we now realize that you can't be perfect with your trading strategy. Instead, the goal is to turn a long-term positive profit margin, no matter how small that margin is.

Using multiple strategies is almost a necessity. We had to combine our overarching Heikin-Ashi strategy with concepts surrounding bearish and bullish markets, and regular candlestick trading theory.

Having worked on a momentum trading strategy this semester, the next logical step is for us to tackle a mean reversion trading strategy.

We would also like to further explore momentum trading, possibly with a more complex algorithm with more modern theory. The Heikin-Ashi candle trading theory is not well flushed out, and was a very open-ended project. It would be interesting to tackle a more advanced strategy that has been well researched by modern-day algorithmic traders.