The HaiKhuu Trading Algorithm: Updates And Our Future Road Map
Updates And Announcements
To begin, I’d like to announce the bigger changes we’re making. First off, we are discontinuing the daily HaiKhuu algorithm. In its place, we’ll be moving our focus to the new Day Trade Bot in our Discord Server. We will continue publishing our Long Term Portfolio, along with stats on it. We’ll go into more detail on our future plans later in this report. In the meantime, I’d like to go over some reasons we’re making this move.
We feel that, in the long run, our quantitative analysis will be more beneficial to us on a bot that works on 5-minute candles than on an algorithm which runs on daily candles. I went over some reasons for this when we began our internal betas (see this report), but I’d like to go through them again here.
Greater Control Over Our Positions. Our new model uses 5 minute candles instead of daily candles. This means that we can afford to be pickier with the trades we take. We’re currently aiming for 100 notifications per day, or 50 positions (1 notification to enter, and 1 to exit each position). And we usually want to take a good number of trades. The more we can get the law of large numbers to work for us, the more consistent returns we can get. Our long term portfolio is 100 tickers right now. Meaning that, with 1 daily candle per ticker per day, a daily algorithm would need a 50% trade rate in order to get 50 positions per day.
With the same number of tickers, and 78 5m candles per ticker per day, an intraday system only needs a 0.64% trade rate to get the same 50 positions. When you consider that our daily algorithm is long only, and the Day Trade Bot isn’t, this is an even bigger divide - with the daily algorithm only hitting those 50 positions if it has a 100% trade rate, and even that assumes that it’ll go long at least half of the time. The bot can afford to be more than 150 times pickier with its trades, and get roughly the same benefit from the law of large numbers.
For this reason, we’re able to use different forms of data analysis which, in backtests and paper trades, have shown more promise.Less Influence From Fundamental Factors. All of our trading algorithms exclusively use numerical patterns and technical signals - no news or fundamental analysis. Holding each position for an entire trading day leaves us exposed to a lot of these factors. There’s a higher risk of news impacting a position coming out over the course of a 6.5 hour trading day than just a 2 hour hold time. This will let our technical analysis shine through, and should further help us to bring in consistent returns.
Greater Ease Of Use For Users. A problem we’ve had with the daily algorithm is getting into the positions. Since every trade assumes we buy at market open, and sell at market close, that’s dozens of trades we want to make at once. Regardless of how good the backtests are, there’s no denying that’s a really difficult experience for a casual user. When we were running our pilot account, we actually had to make an automated tool to get these entries and exits for us.
The Day Trade Bot calls positions throughout the day. This means that if you want to consider its trades, you have much more time to examine each position.More Information Per Trade. Our daily algorithm only outputs an allocation per ticker - that’s it. Our Day Trade Bot includes a stop loss, target exit price, maximum hold time, and can be configured to include an expected exit time. This gives a good idea of how much potential downside and upside we see in each position, in addition to how long we expect that capital commitment to last. If you’re a daytrader, and watching the bot, you’re given much more information about each trade.
At time of writing, our Day Trade Bot is most of the way through its open beta. This is a live run, going throughout the month of January. Should it end up positive at the end of the month, we’ll move into a full release. Our plan has been to discontinue the daily algorithm once we did a full release of the Day Trade Bot - however, we’ve made the choice to take this step a bit early.
Performance Of The Daily Algorithm
This week’s performance ratings are:
Overall Market: +1.56%
Long Term Portfolio: +1.39%
Variable Sector Neutral: +0.6%
Base Algorithm: +0.41%
Variable Market Neutral: -0.19%
Sector Neutral: -0.38%
Market Neutral: -0.84%
This marks a second week in a row of major under-performance by the algorithm.
Looking at sector and ticker breakdowns, we can see why. The algorithm is going deeper into energy, and away from tech. With tech making great returns, and energy falling near the bottom, the algorithm misses all of the portfolio’s biggest winners, while failing to dodge the losers. It’s a similar pattern to what we saw last week.
In the long run, we expect the algorithm to be profitable, but given our plan to switch away from it anyways, we don’t think it has a long term available. Using the daily algorithm for a year or more is a relatively low-risk, moderate reward move. Using it for another 2 weeks (or even 4-6), is much higher risk and much lower reward. Therefore, given the short usage horizon and relatively poor performance in the immediate term, we’re making the decision to discontinue this algorithm earlier than expected.
As mentioned, we will continue publishing our Long Term Portfolio, as well as reports on its performance. We believe our fundamental analysis will remain strong, and given that a portion of our user base is interested in longer term moves, we don’t see a downside here.
Performance Of The Day Trade Bot
Overall, the Day Trade Bot is in the green! Since the beginning of the month, it’s up just about 32 bps. That doesn’t sound like a lot, but keeping in mind that most of its capital is unused, most of the day, gives an indication that this is solid performance.
Since launching the open beta on January 3rd, the bot has taken 531 simulated positions - an average of about 40 per day. Its win/loss ratio is 1.15. It’s average winning trade brings in a return of 0.8%, while an average loss costs it 0.46%. Of the 531 positions it’s taken, 213 of them (about 40% of its trades) have been profitable. All in all, its performance is solid and satisfactory so far. If it continues performing like this, I’ll be extremely happy with it.
We’ll go more in-depth on these numbers next week (with more visual aids), but in the meantime, I’d like to highlight some its best (and worst) trades of the last week.
Our most profitable trade of the week was a short position on BXP. The bot went short at 10:20 AM (EST), and got out of the position at 3:15 PM. The stock never got close to its stop loss or target exit prices, but the bot responded to changing technical signals to exit. All in all, this trade netted a simulated return of 2.1%.
Our most profitable long position (which has been exited) was KIM, on Friday. The bot called a buy at $19.86, and a few hours later it hit its target exit price at $20.215, for a profit of 1.79%.
Disclaimer: we have another long position on NXPI, which is currently up 1.83%, but that position has not been closed at this time. While KIM isn’t our most profitable long trade of the week, it is the most profitable to have been completed.
Lastly, in the interest of transparency, let’s look at the worst trade of the week. This was a short position on IFF that lost us 1.26%. No stop loss hit, but the bot chose to exit due to unfavorable technical signals.
That’s all I have for you this week. Thank you for reading, and best of luck trading this week!
