Since late 2017, financial Websites have been inundated with articles about Artificial Intelligence pureplays, suggesting single stock selections to goose your investment portfolios. Although you will occasionally find a semiconductor equity such as Nvidia (NVDA) included, business writers primarily suggest the FANG stocks. FANG is an acronym CNBC's Jim Cramer coined in the past two years that stands for Facebook (FB), Amazon (AMZN), Netflix (NFLX) and Google (GOOG) (now known at Alphabet). The term recently morphed into FAAANG to include Apple (AAPL) and Alibaba (BABA). With the exception of Apple, all are Internet stocks.
I've written ad nauseam about my personal preference of investing in plain vanilla S&P 500 or Total Market index ETFs, but sector ETFs are an option for investors wanting to generate alpha in their portfolios. Even with elevated expense ratios, these niche ETFs can boost your overall returns if, and only if, you catch them at the right time. They primarily come in two flavors, passively managed Internet ETFs issued by mid tier ETF companies and A.I. specific funds from boutique financial firms.
Internet ETFs
Liquidity and longevity are two critical ingredients when selecting subsector ETFs. In the Internet space, First Trust Dow Jones Internet Index Fund (FDN) and PowerShares NASDAQ Internet Portfolio (PNQI) are the two largest with track records going back at least a decade. In fact, since the market crash of 2008-2009, they are two of the best performing ETFs in the financial universe. Domestically focused FDN is the preferred investment vehicle for traders because of its liquidity, but PNQI has performed equally as well with an international leaning. FDN has 9 times the daily volume as PNQI. Expense ratios for both are steep, 0.54% for FDN and 0.60% for PNQI.
Nevertheless, with market-cap-weighted holdings greatly exposed to the FANG stocks, both ETFs have kicked in the afterburners where performance is concerned. Average gains are roughly 23% per year the past five years outpacing the S&P 500. Besides the liquidity, the big difference between the two funds are geographic allocation and number of holdings. PNQI includes overseas equities with a significant exposure to Chinese Internet companies with the exception of Alibaba. The exclusion of Alibaba continues to perplex me, but organizations like Baidu (BIDU) and JD.com (JD) are under its umbrella. Another heavy hitter in the fold is British travel company Priceline Group (PCLN). PNQI holds 88 securities, over double the amount as FDN.
A.I. ETFs
- Global X Robotics & Artificial Intelligence ETF (BOTZ): News travels fast in the financial world. BOTZ had its inception date on 9/12/16, and in a short period of time, it became one of the best non-leveraged ETFs of 2017 gaining 49%. Some of this performance may have to do with an overweight position in Nvidia which constitutes almost 10% of holdings. With only 28 stocks, it's a concentrated portfolio, but that hasn't stopped speculators from bidding it up. Like all of these specialty ETFs, the expense ratio is high, 0.68%. No FANG stocks in its top 10 holdings, so you're getting more of a robotics story here.
- ROBO Global Robotics and Automation Index ETF (ROBO): ROBO has been trading three years longer than BOTZ, but has a much higher expense ratio of 0.95%. You expect some alpha generation with those fees, and although it gained 36% last year, it pales compared to the performance of BOTZ. ROBO holds 89 equities which gives you some global diversification. Like its brother BOTZ, there is not a lot of liquidity with this fund, so you are best suited to use limit orders. The top holding only constitutes 2% of the portfolio.
- ARK Industrial Innovation ETF (ARKQ): ARKQ is another high flyer from 2017 gaining approximately 45%. It's also another low volume, lofty expense ratio ETF charging 0.75% annually. International in scope and actively managed, I found it interesting that Tesla (TSLA) comprised almost 10% of the portfolio of 43 equities. Inception date was 9/30/14, so it's got some history behind it, but don't confuse brains with a bull market. There are thousands of ETFs issued worldwide and any significant downturn in the market could put any of these thematic ETFs in jeopardy.
Conclusion
An oligopoly has formed in the A.I. arena. The same Internet companies we've come to depend on for our everyday technology needs devour smaller startups. The Google and Amazon of 15 years ago are no longer mom and pop shops. They are the IBM and AT&T of the 1950's, if not the Standard Oil of New Jersey in 1900. The Gilded Age redux. That is why if I were to purchase an Artificial Intelligence ETF, I would select either PNQI or FDN with heavier leanings toward the Internet. Unless the European Union or the United States Government breaks them up, there is plenty of room to run.
FDN receives my most favored nation status only because of its relatively high volume. In an up-to-the-second connected world, PNQI, BOTZ, ROBO and ARKQ may frustrate investors or traders with quotes delayed as much as 5-10 minutes. Sometimes as much as a half an hour. That is why I can't stress enough the importance of using limit orders. Although your broker can probably provide you with CFRA reports powered by S&P Global on these smaller exchange traded funds, I find ETF.com is the best source for statistical analytics. It's free. Just go to their Website and register.