How to invest in the artificial intelligence track through ETF?

Buying an ETF that specializes in investing in artificial intelligence can ensure that its portfolio covers as many different companies as possible and minimize the probability of missing a company.

Which is the hottest investment track in 2024? I'm afraid most readers will choose "artificial intelligence". Since OpenAI announced Sora at the beginning of the year, many of us have been flooded with various videos and short movies automatically generated by artificial intelligence. Of course, how artificial intelligence can improve productivity and how much revenue it brings to related companies seems to be an open question, but companies that "sell shovels" to artificial intelligence gold diggers have made a lot of money. Nvidia, the leader in the GPU market necessary for generative AI, has a market value of more than 2 trillion US dollars, ranking third in the world, second only to Apple and Microsoft. The legendary story of its founder Huang Renxun has also been reported by major media.
As an ordinary investor, how can we get a share of this wave of technological revolution? I believe this is one of the most concerned issues for many readers.
Undoubtedly, the most direct way is to buy Nvidia's company stock a few years ago. But the problem is that we can't turn back time and go back to the past to buy stocks that are sure to rise in the future. In fact, there are thousands of related companies in the field of artificial intelligence. Investors who could have thought of betting on the artificial intelligence track through Nvidia a few years ago were very few, and they might even be considered whimsical.
Therefore, for many individual investors, another way is to buy ETFs that specialize in artificial intelligence themes. These ETFs will not only invest in one or two companies, but will choose to buy stocks of dozens or even more companies. This investment method can ensure that its investment portfolio covers as many different companies as possible and minimize the probability of missing a company.
Of course, even thematic ETFs that only focus on investing in artificial intelligence are very large. This is because artificial intelligence itself is a very broad concept. Since this is the hottest investment theme in the past two years, many companies have chosen to move closer to artificial intelligence and add the word "artificial intelligence" to their products, services or company propaganda. In 2019, a report released by British venture capital company MMC showed that there were 2,830 artificial intelligence start-ups in Europe, of which only 1,580 were actually doing artificial intelligence research and development. In other words, about 45% of the companies that claim to be AI are just trying to take advantage of the popularity of "AI". Secondly, if we dig deeper into the concept of AI, its scope can be further subdivided into: industrial automation, non-industrial automation, 3D printing, language generation, autonomous driving, etc. It is no exaggeration to say that all these sub-sectors are highly related to AI, and the companies in their tracks can be classified as AI themes.
Based on these criteria, we can find about 50 AI ETFs as of March 2024. If we only look at ETFs with a fund management scale of more than US$150 million, the number is reduced to about 18. The investment strategies and focuses of these ETFs are different, but they are all centered on the AI ​​investment theme, with the goal of helping investors obtain investment returns brought by the technological revolution led by AI.
If we analyze the returns of these ETFs carefully, there will be some interesting findings. For example, in the past 12 months, only 4 of the 18 AI ETFs have returned more than the S&P 500 index during the same period. The returns of some of these AI ETFs are terrible. For example, there is an ETF called First Trust Nasdaq Artificial Intelligence and Robotics ETF (ROBT, with a fund size of about $550 million. As the name suggests, this ETF specializes in investing in artificial intelligence and robotics), whose return over the past 12 months has lagged behind the S&P 500 index by 19% over the same period. Another ETF is called WisdomTree U.S. AI Enhanced Value Fund (AIVL), whose return over the past 12 months has lagged behind the S&P 500 index by 13% over the same period.
Many readers must be surprised. I have already found the right track and bet on the hottest and most popular artificial intelligence industry, so why is the return not as good as the average return of the stock market? The main reason behind this is that the returns of artificial intelligence stocks vary greatly. In terms of the percentage of return of company stocks, the biggest winners are Nvidia and Super Micro Computer (SMCI). In addition, the stock prices of large-cap companies such as Amazon, Microsoft and Metaverse Platform have also soared due to the concept of artificial intelligence.
That is to say, whether investors can get a good return on investment in the end depends on whether their investment portfolio includes stocks such as Nvidia, Amazon, and Microsoft, as well as the proportion of these stocks.
As of the end of March 2024, the weights of Nvidia, Amazon, Microsoft, and Metaverse in the S&P 500 index were as high as 5.3%, 3.72%, 7.25%, and 2.55%, respectively. The return of the S&P 500 index in the past 12 months (as of the end of March 2024) was as high as 30%, of which about 1/3 came from the above-mentioned technology giants.
ETFs that specialize in investing in the artificial intelligence track may not be as highly concentrated in the top technology giants in terms of investment weight as the S&P 500 index. This is mainly because the investment target of these ETFs is one track, not one or two companies. If you want to ensure that you buy the entire track, then you need to divide the bullets evenly and spread them as much as possible on multiple companies, rather than concentrating them all on a few companies. This is why many artificial intelligence ETFs limit the maximum weight of any invested stock to 4% or 5%. When the market value weight of a certain stock exceeds this ratio, the ETF fund manager will intervene to rebalance, sell some of the stocks of high-weight companies, and buy some of the stocks of low-weight companies to pull the weight distribution of the portfolio back to a more evenly distributed equilibrium state. This operation can ensure that investors do not miss the entire track, but when encountering "super monster stocks" like Nvidia, it will drag down investors' returns.
What can we learn from the above analysis?
First, it is much more difficult to invest and profit by selecting tracks and industries than many people imagine. Even if you see the right track for artificial intelligence and buy a fund that specializes in investing in the artificial intelligence industry, investors may not be able to beat the market, and may even lag behind the average market return with a high probability. This reminds us that at any time, in our investment portfolio, we should leave a place for index-based passive investment. Since it is difficult to beat passive investment, why not set aside an area in your investment portfolio, such as X% of the total investment amount, specifically for passive investment.
Second, this also shows that the threshold for active investment is very high. For example, some investors have concluded that artificial intelligence is a "trend". But if your knowledge is limited to this and you buy some artificial intelligence theme ETFs, then you are actually far from high-quality active investment and excess returns. Smart investors should be self-aware and understand how much effort is needed to obtain excess returns.
Third, real active investment does not require diversification. Investment legend Buffett has said on many occasions that he does not agree with diversifying his investments too much. For ordinary stockholders, it is enough to buy 8-10 stocks at most and hold them for a long time. The energy and cognition of any investor are limited, and it is impossible to be proficient in everything. Therefore, investors should make choices and make themselves experts in one or two fields, and vertically cultivate in these fields. Let us not forget that the starting point of active investment (stock selection or industry selection) is to pursue excess returns that exceed the market average return based on one's own cognition that is higher than the market. Therefore, on the basis of active investment, diversification is equivalent to diluting one's own cognition that is higher than the market. Of course, the premise for deciding active investment here is that the investor himself does have research and cognition that is higher than the market average. If not, it is better to go back to the first point above and honestly buy a large-cap index ETF for long-term holding.
Fourth, in reality, a more practical approach may be to hold both active and passive investment portfolios. Investors can consider dividing their investable funds into two parts, one part is specifically put into low-cost passive index ETFs for long-term holding; the other part is specifically used for active investment to pursue industries or companies that they are optimistic about. In this way, you can ensure that you get the long-term average return of the market, and at the same time give yourself a chance to prove whether you have the ability to beat the market and get better excess returns.