The Goldman Sachs Data-Driven World ETF (GDAT) tracks a bespoke index that seeks exposure to artificial intelligence, big data, cybersecurity, the internet of things and data infrastructure. Goldman’s marketing materials say that GDAT provides “exposure to the beneficiaries of technological innovation, regardless of sector, geography or market capitalization.” The investments are selected and weighted “by a function of ‘thematic’ beta.’” Top holdings include NVIDIA, Amazon, Microsoft, Apple, and Google-parent Alphabet Inc.
GDAT’s management fee is high compared to ultra-low-cost plain-vanilla index funds, but it isn’t outrageous for smart-beta funds or niche products. Still, for the cost-conscious, there’s significantly cheaper technology index ETFs out there, like the Vanguard Information Technology ETF (VGT). For investors who want innovation and don’t mind paying a bit extra for it, there are both indexed and actively-managed alternatives. The ARK Innovation ETF (ARKK), run by the stock-picking team at ARK Invest, absolutely trounced GDAT’s performance in the first five months of 2020, a tumultuous period marked by pandemic shutdowns and social unrest. There are also other specialized index funds that cover same of the same themes, like the Global X Cloud Computing ETF (CLOU) or cybersecurity ETFs like HACK and CIBR.
GDAT is one of five ETFs that Goldman launched in 2019 with indexes designed by Motif Investing Inc. The indexes selected stocks using artificial intelligence and machine learning. Goldman worked with Motif to design indexes that aim to identify companies that fit with broad themes that Goldman’s money managers thought would drive growth. (The “Motif” moniker was removed from the fund name in May 2020 when Motif dropped out as the index provider.) The funds were yet another example of Wall Street’s increased use of sophisticated automation. Computers read faster than human analysts, quickly scouring regulatory filings, company reports and news stories for information on a company’s prospects. Using AI and machine learning isn’t that much of a divergence — after all, most factor funds and smart-beta ETFs are really just attempts to codify investing strategies employed by human stock pickers. Whether it’s a plain-vanilla index or something more sophisticated, index methodology is basically just programmed code that tells the index what to buy and sell.