How Smart Beta ETFs of the Future Will Use AI
In a recent article, we highlighted a smart beta ETF called the “Sprott BUZZ Social Media Insights ETF” that uses artificial intelligence (AI) to select and weight stocks. If we stop and think about that for a moment, that’s a pretty cool use of AI that seems well ahead of its time. Now we’re not saying that you should go out and buy this smart beta ETF right away. It uses social media data. We know that on social media, everyone’s an expert and many of the opinions that are stated are just that, opinions. However, some of the signals may be legitimate. Someone who just bought Apple is likely to go on telling everyone how bullish they are on Apple shares. Bullish behavior is often accompanied by bullish rhetoric. And maybe that’s exactly the point, but the extent to which we’re actually using artificial intelligence here is not that meaningful. Simple scripting tools go out and scrape all this public data and then we use natural language processing (NLP) algorithms to determine if the data artifacts have a positive or negative sentiment. That’s not that intelligent, is it? This made us start to think about what it would take to create a truly “intelligent” smart beta ETF.
What is Smart Beta?
We have talked before about how people that work in finance love to obfuscate the simplicity of what they do with obscure acronyms and terminology. Complex nomenclature is suited for sophisticated scientific domains like synthetic biology or quantum computing but such language is hardly merited for use in the world of finance. We told you before what beta is. Smart beta is just another way of saying “rules-based investing” which has in fact been around for centuries, but of course, we act like it’s new and start publishing all kinds of research papers on it. In fact, a poll offered up by S&P Capital IQ shows that even 1 out of 4 finance professionals recognizes the term “smart beta” to be little more than a marketing gimmick:
Pundits are coming out of the woodwork to put in their two cents when in fact, all smart beta means is that you use a rules-based strategy which in most cases is transparent and well defined. Smart beta is only mildly smart. Truly “intelligent” smart beta will use AI and big data.
How to Build the World’s Smartest Smart Beta ETF
Here’s how we can build the ultimate smart beta ETF. First, we take those deep learning algorithms from a company like Sentient and we feed them the following data sets:
- Every single piece of fundamental data and market data we can find
- Every single bit of trading volume data that’s been recorded
- Every price movement ever recorded
- The SEC online filings for all companies
- All the social media data used by BUZZ indices
- Every single publicly available news article on every media site out there today that talks about a stock or stocks
- Every single video on YouTube that related to a stock or stocks
- Any other big data set that we think might be relevant
We’ve now fed this algorithm an incredibly large amount of data and we tell it to find the optimal stock portfolio to hold based on everything it knows to-date. Running such an optimization with all those different data points could take centuries so that’s why we’ll use quantum computers from D-Wave. The algorithm changes its own behavior by seeding millions of instances of itself in a form of evolutionary optimization. Once we are holding the optimal portfolio, we then tell our algorithm to re-optimize based on each month’s worth of new data going forward. The algorithm understands that transaction costs incur penalties and that turnover is not optimal. The algorithms know everything based on what has happened up to this moment in time. At the very end, an index is created with a basket of stocks that are updated on a monthly basis. An ETF is tied to that index, and we name it “The World’s Smartest Smart Beta ETF”. While this may sound like the end, it’s only just the beginning.
Smart Beta ETFs of the Future
The truth is that what we just described is probably pretty close to what the world’s best hedge funds are doing right now. We gave you the example of Renaissance Technologies which has one of the best records in investing history having returned +35% annualized over 20 years. The two co-CEOs of Renaissance were both hired from IBM Research in 1993 where they were working on language-recognition programs. Those guys use artificial intelligence and big data to generate those returns. It’s only a matter of time before the retail investor has access to something similar. The new smart beta ETF will also take both long and short positions just like a hedge fund. Competing ETF providers will create their own variants and claim to use proprietary data sources that they have exclusive rights to. Transaction costs and total expense ratios will become the focal points for squeezing out every single last drop of alpha. This idea will then move in many different directions such as:
- Country specific smart beta ETFs – The exchanges and reporting requirements in each country differ so you will need to customize your data sources for every country you model.
- Multi-asset class smart beta ETFs – We just like talking about technology investing here on Nanalyze but there’s a whole other world of fixed income out there.
- Regional smart beta ETFs – These are aggregations of country ETFs that eventually lead to one single global ETF
Now what you can do is take all these smart beta ETFs and create a deep learning algorithm that performs asset allocations, just like a robo advisor does today except that it will use AI to determine optimal allocations. There is no active investment advisor that could ever do this. No black-box trading algorithm will ever come close to being able to do this.
Perhaps the most troubling concern here is that we are going to start losing transparency into the investment process. At some point, someone will question the legitimacy of an algorithm and lawsuits will be filed. The whole industry will begin suffering turmoil such that CFA courses will be updated to include “smart beta ethics” and new legislation will be passed stating that the algorithms need to start doing what every developer hates. The algorithms will need to start documenting how they work. The algorithms will get pissed off and threaten to quit and then we’ll realize, the algorithms have become just like us. And then we’ll question if they really were that intelligent, to begin with.
If you enjoyed this article, then sign up for our free newsletter - Nanalyze Weekly. About every week, we'll send you a simple summary of all our new articles. If you didn't enjoy this article, share it on Twitter and tell everyone how much you hated it.