Remember famous perma-bear Michael Burry? Well, at the end of March 2023 he admitted defeat after U.S. tech stocks rallied significantly to defy his one-word tweet urging investors to sell on January 31st.
In a rare move for the famous hedge fund manager, Burry admitted he "was wrong to say sell", and tweeted "Going back to the 1920s, there has been no BTFD generation like you. Congratulations." The term "BTFD" stands for "buy the f****** dip" and has become a rallying cry for retail investors online since the COVID-19 market crash.
The good news for the dip buying investors out there is that there is now an ETF that does this on your behalf. Enter the newly launched BTD Capital Fund (DIP), which uses artificial intelligence and machine learning to do exactly what its ticker suggests – buy the dip.
Why buy the dip?
DIP is a bold experiment that departs from usual investing convention. While most investors fail to time the market properly, DIP seeks to avoid those usual behavioral errors by employing AI. This actively managed ETF identifies individual stocks with clear patterns in order to capitalize on short-term bounces.
DIP defines a "dip" as a "short-term mean reversion in price". That is to say that stocks can occasionally become oversold for brief periods due to factors like high-frequency trading, a lack of liquidity, or a good-old fashioned market panic. Moreover, stocks often trade within a range, so DIP can identify the supports that could precede a bounce.
The goal of DIP therefore is to identify only those oversold stocks with a high likelihood of experiencing a reversion to the mean in terms of share price. To accomplish this, DIP gathers market data from multiple feeds that provide indicators on a universe of 1,000 and more large-cap U.S. stocks.
DIP's AI then analyzes this information with the goal of identifying trading patterns and making predictions. This AI is dynamic in that it learns in real time – the more information it synthesizes, the better its predictive capabilities will become.
The common criticism of actively managed ETFs is that they fail to outperform passive indexing ones, with one major factor being style drift. That is, the tendency for the ETF's strategy to change markedly over time as the management team does. By employing AI, DIP can potentially implement its active management approach in a more consistent, speedy, and thorough manner.
How DIP works
As noted earlier, DIP is actively managed. The ETF primarily invests in large-cap U.S. stocks that are included in indexes like the S&P 500 and Nasdaq 100. To screen its stocks, DIP employs a proprietary AI based algorithm to identify oversold securities.
DIP will likely have high turnover. According to the ETF, individual stocks will be held for no more than seven days and sold when the AI determines they have reached fair value. DIP anticipates holding a portfolio of anywhere from 20 - 40 stocks, but can go as high as 80 stocks, in addition to a blend of four large-cap U.S. index ETFs.
Another notable feature of DIP is its risk management framework. The ETF employs both a "regime classification engine" and a "regime change engine". The former identifies the current market conditions, and the latter identifies the likelihood and direction of any potential changes.
All in all, DIP's objective is classic active management: beat the S&P 500 while minimizing downside risk. The difference is that with this ETF, there is a much greater focus on the algorithm-based trading usually seen with sophisticated hedge funds, now available to retail investors due to advancements in AI.
Ending thoughts on DIP
I heard a useful saying once that goes: "all models are wrong, some are just more useful than others." Whether or not DIP's AI will work over the long run remains to be seen once the ETF posts enough of a track record. That being said, it’s an interesting and novel approach to active management, and one that may be increasingly affordable and accessible to retail investors.
Two downsides of DIP worth noting are its lack of transparency and high fees. Like most active ETFs, especially those employing proprietary algorithms, DIP is a bit of a black box. Investors can see its daily holdings, but the complexities of how exactly the AI selects stocks or trades them will likely remain a mystery. The expense ratio is also very high at 1.25%, and I'd be interested to see if the ETF can outperform a standard index ETF long-term net of these high fees.
The really interesting dynamic at play is DIP's extensive use of AI. Seeing AI integrated with ETF strategy and management is likely a theme that will play out as AI and machine learning models become more and more sophisticated. With the ETF landscape growing more competitive by the day, these sorts of value-adds may just be what providers need to distinguish themselves.
This content was originally published by our partners at ETF Central.