Suddenly, “Nvidia” has become a household name for investors. For the uninitiated, “Nvidia is a software and a fabless company that designs and sells products and platforms for gaming, professional visualization, data centers, and automotive.” That company description is brought to you by artificial intelligence, or AI. Said another way, they design semiconductor chips, many of which happen to power AI. (“Fabless” means that the company designs and sells semiconductor chips but outsources their fabrication.)
Nvidia’s stock is up more than 200 percent from a year ago, fueled by the rush of corporate America to join the AI race. According to FactSet, the mention of AI during S&P 500 company earnings calls has spiked.
Investors became so enamored with the possibilities of AI that CEOs were compelled to either announce that they would somehow get into the technology or use it for cost savings and worker productivity. Nvidia found itself at the heart of the AI conversation.
The surge in the stock price of Nvidia has prompted many investors to ask, “How can I invest in the artificial intelligence wave?” There are some easy answers to that question, especially if your sole focus is on Nvidia. For example, if you own an exchange-traded fund (ETF) that tracks the S&P 500, about 3.75 percent of your investment is in Nvidia. There is roughly a 4.5 percent allocation to Nvidia in ETFs tracking the Nasdaq 100.
Owning broad-based ETFs is an easy answer, but it is not the advice inquiring investors want to hear. They want the names of specific companies. You will also have AI exposure if you own stock in companies such as Microsoft, which invested in OpenAI and Copilot, and Alphabet (AKA Google), which has rolled out its Gemini AI tool.
If you prompt an AI tool about how to invest in AI (which I did), you may receive the suggestion to consider Taiwan Semiconductor. Taiwan Semiconductor has a diversified revenue source and supplies products to Nvidia. Also, my AI query noted Samsung Electronics and SK Hynix, which create products that end up in AI semiconductor chips. If you prefer broader exposure to the AI industry, several ETFs exist, including Global X Robotics & Artificial Intelligence (symbol: BOTZ) or iShares Robotics and Artificial Intelligence (symbol: IRBO).
I could have used AI to delve deeper into those options to decide whether to buy or recommend them. But I won’t because I prefer another way to invest in AI (an easier, cheaper, and, yes, wimpier way).
Your margin is my opportunity
Jeff Bezos, founder of Amazon, famously said, “Your margin is my opportunity.” Bezos was referring to the lower prices he could provide consumers by reducing the expenses associated with the traditional intermediaries of the retail supply chain.
For investors, perhaps their best risk-reward opportunity in AI lies not in finding the next Nvidia but in the profit margin of other companies. Investing in AI is as much about margin expansion as it is about revenue expansion.
When it comes to revenue expansion, there are subscription versions of ChatGPT, Copilot, and Gemini, but they are also available for free. There isn’t yet a widely used consumer revenue model for AI companies to benefit from. Some software companies may be able to increase prices or sell more subscriptions if they use AI to improve their products. For example, Adobe is more user-friendly, Tesla is getting closer to self-driving cars, and Facebook is better at targeting advertisements.
Many other companies can use AI to improve their profit margins through dynamic pricing, targeted marketing, risk management, quality control, supply chain optimization, automation, and analyzing data to follow customer behavior. The best risk-reward investment for AI might simply be to buy an S&P 500 index fund.
The price of stocks is closely tied to company profits. If AI can shift the profit margins of S&P 500 companies from 11.1 percent currently to 13 percent, it would not be unreasonable to expect a 20 percent jump in the index. And that is with no multiple expansion.
Nvidia is selling picks and shovels in an AI gold rush
In college, I interned at a brokerage house. That was in the mid-1990s, and the stock market was flying high. Clients would ask me to find the next Microsoft. My line to them was, “If you want the next Microsoft, buy Microsoft.” Today, I have the same response for investors trying to find the next Nvidia. If I had to choose one pure-play investment, it would be Nvidia. (However, later, the dot-com bubble burst, Microsoft’s stock got cut in half and essentially went sideways for a decade. So what do I know?)
Nvidia sells graphics processing units (GPUs). GPUs are used in artificial intelligence to speed up workloads by processing large amounts of data. To better explain the attractiveness of GPUs, let’s contrast them to CPUs (in an oversimplified manner).
People my age will recall their confidence when they bought a desktop computer with the “Intel Inside” sticker affixed to the monitor. Intel is a leading manufacturer of central processing units (CPUs). That Intel sticker assured they could rely on the “guts” of the computer because CPUs are the workhorses of computing power.
Central processing units are designed to take one instruction in, act on it, and create an output. In other words, you ask the computer to do one thing at a time, and it will do one thing at a time—because that’s all it can do. Intel and other semiconductor manufacturers created better and faster chips over the years, but that one-in-one-out factor was limiting.
To accelerate computing power, a chip must take in that one instruction (or many) and then process dozens or hundreds of outputs simultaneously. It turns out that GPUs, traditionally used for gaming, do that well. And Nvidia makes gaming chips.
Nvidia realized that there were limitations to CPU computing. Before Nvidia became a household name, they convinced computer manufacturers to keep using CPUs to perform most of the work but add GPUs for when users wanted to access graphics or video games.
Then people began to realize that the mathematics and process of graphics chips were similar to how artificial intelligence computed output.
Although it could be argued that Nvidia’s stock has gone up way too much to invest in now, the company is growing earnings fast enough that its stock’s trailing price-to-earnings ratio has also fallen.
I wouldn’t fault most investors for buying Nvidia (or Microsoft or Alphabet or any individual stock) so long as it is suitable for them. But keep in mind, according to several studies, investors of individual stocks routinely underperform the broad market. That is true for “civilians” as well as financial advisors. If you are paying someone two percent of your investment portfolio per year in the hope of beating the market, keep in mind that hope is not a strategy. In this instance, it is a way to increase risk. So be careful.
If AI accomplishes what many companies hope for, even non-technology companies will benefit from more significant profits. My preference is to diversify widely. Several broad-index ETFs will give investors direct and indirect AI exposure.
Allen Harris is the owner of Berkshire Money Management in Dalton, Mass., managing more than $700 million of investments. Unless specifically identified as original research or data gathering, some or all of the data cited is attributable to third-party sources. Unless stated otherwise, any mention of specific securities or investments is for illustrative purposes only. Advisor’s clients may or may not hold the securities discussed in their portfolios. Advisor makes no representations that any of the securities discussed have been or will be profitable. Full disclosures here. Direct inquiries to Allen at AHarris@BerkshireMM.com.
