Will Artificial Intelligence Eat ETFs for Lunch?

How generative AI could disrupt the entire ETF industry and what ETF issuers need to do.

|
Reviewed by: Lisa Barr
,
Edited by: Lisa Barr

Recent innovations in artificial intelligence might allow customers, in search of a perfect expression of their investment preferences, to bypass ETFs entirely. What can exchange-traded fund issuers do in response?

Since its inception, the ETF industry has provided investors with a cost-effective way to gain exposure to two different investment opportunities.

The first is asset class exposure. Think of broad-based trackers like the S&P 500 ETF Trust SPY and the iShares U.S. Treasury Bond ETF (GOVT). These index trackers are the backbone of asset allocation solutions for smaller customers; even someone who invests $500 per month can build a reasonably diversified portfolio with ETFs.

The second is thematic exposure, including anything from sector ETFs to baskets of robotics companies. Active traders tend to make up the majority of investments in this space, sacrificing diversification for the sake of thematic alpha.

Generative AI

However, by combining generative AI with fractional share trading, investors could potentially bypass ETFs and instantly create a custom basket of securities that meet their hyperspecific investment needs. The combination of generative AI and fractional share trading has the potential to completely upend the ETF industry.

For example, an investor might want to trade the S&P 500 but exclude energy companies and prefer a value approach. Oh, and they want a 4% dividend yield. With generative AI, the investor can simply provide their investment criteria, and it will generate a custom basket of securities that meets those criteria.

Any financial analyst can work that out with Excel and Bloomberg. It takes 15 minutes, tops—five if you know what you are doing. Until yesterday, retail brokers had to fiddle with finicky filtering systems and would not necessarily get exactly what they wanted. 

But today? Today we asked generative AI or ChatGPT4 this:

“Build a basket of the top 500 largest U.S. stocks excluding energy and expensive companies, and with a minimum 3% dividend yield. Then show me the historical performance,” and once the investor is satisfied, they could simply send in the request via chat and say, “Buy $1,000 of this basket,” and it’s in their portfolio. 

Poof. Done. Everybody can have their own customized basket within seconds to write, using only plain English and generative AI tools. And with the right integration onto their brokerage platform, they may never have to leave the chat interface to execute their trades.

While the ETF industry is still likely to have a place in the investment landscape, it will need to adapt to this new reality if it hopes to remain relevant. Issuers need to embrace this emerging technology, whether that’s developing their own data science tools (which can be costly and time consuming), or by carefully vetting third-party AI providers.

By creating hyperspecific baskets and offering more options for consumers, they’re taking an extra step of work out of the equation for investors, who are used to customization and often have a unique set of requirements in mind.

Giuseppe Sette has been co-founder and president of Toggle AI since its founding in 2019. Prior to that, he was co-CIO of global macro at asset manager Lombard Odier. Before that, Sette invested at Brevan Howard and Davidson Kempner, and ranked top of his consultant class at Bain Italy. He holds an MBA from Wharton and an M.S. in electronical engineering from La Sapienza University, as well as a patent in acoustic propulsion via ultrasound.