Smart AI Investing Isn’t About Chasing the Big One

(Bloomberg Opinion) — How AI will reshape our lives is clouded with uncertainty. Researchers are debating what it means to achieve artificial general intelligence, or AGI, or whether the neural-network systems that drive chatbots like ChatGPT can lead us to that big prize. But smart investors should avoid trying to predict future breakthroughs of a technology that is still at an evolutionary stage.

Rather than participating in OpenAI Inc.’s expensive secondary share sales, they can get exposure to the AI theme using a time-tested approach. The supply chain is long, complex and prone to bottlenecks. At each chokepoint, there are a few economic moats — to use Warren Buffett’s phrase — that deliver outsized earnings a lot faster and surer than ambitious, large language-model developers.

The most well-known constraint occurs in chip design. Demand for Nvidia Corp.’s products is so strong that OpenAI had to find alternative sources, concluding a mega deal with rival designer Advanced Micro Devices Inc. this week for inference functions. Inferencing, or using pre-trained models to generate user content, can make do with less advanced chips.

Dig deeper and there are more obscure but equally interesting opportunities. We are witnessing a data-center boom, especially in the US. In July, construction spending reached $41 billion on an annualized basis, or 47% of the total outlaid on offices. But developers fret over various bottlenecks, from procuring chips to securing power generation.

Try getting your hands on gas turbines, which burn natural gas and turn a generator to produce electricity. It used to take around two years to obtain a new one, now it takes up to five or more.

This is in part due to an oligopoly market structure, dominated by GE Vernova Inc. in the US, Germany’s Siemens Energy AG and Japan’s Mitsubishi Heavy Industries Ltd. After a gas turbine boom and bust in the early 2000s, the three producers have turned cautious and are unwilling to expand capacity, fearing that demand from data centers may not stick. Meanwhile, startups are not trying to edge in either, because the technological barrier to entry is high. In some cases, turbine designs took decades to test and tweak.

The market for high-bandwidth memory, or HBM, essential to support AI systems like those from Nvidia is another good example. South Korea’s SK Hynix Inc. is the industry leader, followed by Samsung Electronics Co. and Micron Technology Inc. The technology is so complex that Huawei Technologies Co. had to use Korean products in some of its leading Ascend AI processors, even though the Chinese conglomerate would prefer to source domestically. Not surprising then that OpenAI also inked strategic partnerships with Samsung and Hynix in an attempt to secure its supply chain and scale up before everyone else. All three stocks are on fire this year.

To be sure, identifying AI winners is not as simple as studying tech entry barriers and market structures. Netherland’s ASML Holding NV has a near-monopoly on the specialized lithographic equipment used to make high-performance chips, but its stock price has missed this year’s AI rally. One key reason is a concentrated customer base. Not being able to sell to the Chinese due to export restrictions, ASML unfortunately relies too much on foundry Taiwan Semiconductor Manufacturing Corp., which balks at purchasing the Dutch company’s ridiculously expensive extreme-ultraviolet-light, or EUV, machines.

So what are the lessons? First, for investors without a computer-science background, identifying supply-chain bottlenecks is perhaps easier than asking if AI research can result in training models that exhibit human-like intelligence. After all, some money managers have already gone through a similar exercise, looking at suppliers around the world during the smartphone era.

Second, the AI-fueled stock boom is not limited to the US or China, because the supply chain is so complex that no country can claim to do all on its own.

Third, there have been a lot of worries about an AI bubble — investors are understandably wary when mega caps from Oracle Corp. to AMD can melt up by more than 20% in a single trading day. But the divergent stock performance of, say, gas-turbine makers and ASML indicates that there’s still rationality in this rally. Not all AI names get to shine, yet.

Most importantly, in this AI race, the bulk of the profit is probably earned upstream, where the biggest bottlenecks are. By 2029, when OpenAI finally starts to break even, boring turbine engineering firms might have made a truckload of money already.