Report On Algorithm Performance: 2022-07-29

Hey all, Asher here. I’ll be releasing these in-depth summaries on algorithmic trading performance from here on - ideally, daily. Without further ado, let’s dive right in - here’s how the official HK Trading Algorithms did last Friday.

Performance Rankings

1. The Market: +1.08%

2. Base Algorithm: +0.98%

3. Neural Net: +0.88%

4. Long Term Portfolio: +0.80%

5. Trading Bot: +0.49%

6. Good Vibrations: +0.36%

7. Kokomo: +0.22%

What Happened And Why?

Almost all of our algorithms are based on a machine learning system (the exception is the Long Term Portfolio, which comes from long-term financial modeling - think along the line of DCF’s). To train a machine learning model, you have to set an objective. Ours is Sharpe Ratio: expected annual return over expected annual standard deviation. Or in simpler terms: reward over risk.

Our models are trained to emphasize stable returns, in addition to high ones. It is largely for this reason that on days like this when the market pumps, we expect to be positive, but lag behind a little.

We can see Kokomo and Good Vibrations on the bottom today, albeit positive. Kokomo is designed to be safer than the base algorithm by hedging against the market as a whole (the base algorithm’s market beta is approximately 0.6, whereas Kokomo’s hovers around 0.05); Good Vibrations does the same thing, but with individual sector ETF’s ($XLP, $XLU, $ETC). For this reason, neither was carried by the positive market movements today. In fact, if we had taken either portfolio and ignored their hedges, Kokomo would have returned 0.83% (putting it at #4), and Good Vibrations would have returned 1.00% (putting it at #2).

In summary, this isn’t far off from what we expect to see on a day like today. Lower expected returns - especially on days the market really pumps - is a price we pay to keep Kokomo and Good Vibrations’ returns more consistent than the other algorithms on the list. This is also why we generally advise our community members to utilize those 2 on days they think the market is shaky, or a few sectors will be particularly risky.

As for the trading bot, it made 4 trades on Friday - with all but 1 bringing in positive returns.

While it’s always underwhelming to underperform the market, I won’t complain about a day where everything we publish makes money.

That’s all for today. I know this report was a lot of text - I’ll have more colorful graphs and charts in the next one. Thanks for reading, and happy trading!