Key Questions: Will AI Live Up to The Hype?

If you turned on your favorite financial news network or picked up a financial newspaper at any point in 2024, you likely read about an investment theme that has captivated Wall Street and Main Street alike. The artificial intelligence revolution has created huge waves across financial markets as shares of chipmakers have skyrocketed.

Companies outside of technology stocks, such as power producers and electrical component manufacturers, also recorded outsized returns last year. The question that investors debated through 2024 was whether sky-high forward estimates would materialize or whether elevated stock prices signal another early 2000s-esque technology bubble. Our view is that the recent AI revolution is not the same as previous tech bubbles, and so we are cautiously optimistic about AI today. 

AI has wide-ranging potential applications. Primarily, AI is touted as optimizing production and decision-making by analyzing patterns far more quickly than humans or existing computing processes can. The technology is already beginning to be employed by doctors, farmers, industrial operators, educators, and others.

For example, AI is being used to help doctors analyze patient data by identifying patterns and potential diseases more quickly. It’s also being used to help farmers plan crops and to help manufacturers optimize industrial production. Furthermore, AI is being used in everyday life by ordinary consumers: If you use a virtual assistant such as Apple’s Siri or Amazon’s Alexa, you are also using AI.

The range of applications is practically endless, which leads us to acknowledge that the ongoing AI revolution has become something like a digital arms race as companies such as Apple, Microsoft, and Meta seek to develop better AI solutions than those of their competitors. These “hyper-scaler” companies are committing tens of billions in capital expenditures as they build out their AI capabilities to capitalize on developing AI applications.

Critics rightly point out that there is no “critical app” being marketed now. This is a fair criticism today, but we remain cautiously optimistic that the exponential acceleration of progress by AI researchers will flow through into tangible use cases in the years ahead.

Regardless of future use cases for AI, beneficiaries of the AI buildout are reporting notable revenue increases today. Nvidia, the closely followed seller of graphics processing units or GPUs (a critical component in computing), has been a material beneficiary of the broader AI theme. 

More specifically, the company reported quarterly revenues of $18.1 billion in the third quarter of 2023. One year later, owing to explosive growth from AI, it reported quarterly earnings of $19.3 billion. It is significant and unprecedented for a company to report more in quarterly net profit than it did one year prior in quarterly company revenues. In short, demand for AI solutions is real. 

While investors may grow concerned that such levels of growth are reminiscent of a bubble, we note that Nvidia’s price-to-earnings multiple is actually lower now than it was for most of 2023. The stock has, in essence, gotten cheaper as this growth has materialized and the stock has hit all-time highs. This is not indicative of a significant bubble in stock price.

But technology companies are not the only beneficiaries of the AI boom. Independent power producers (companies responsible for generating electricity), natural gas producers, pipeline operators, electrical equipment makers, and commercial HVAC suppliers have also seen notable (albeit, at this point, uneven) increases in their bottom lines and project backlogs as data center construction materializes and opportunities expand. AI data centers consume a lot of electricity and produce a lot of heat, both of which need to be accounted for in a technological buildout that a chipmaker alone cannot provide solutions for.

To reiterate, the AI theme is no longer a hypothetical. GPUs are being ordered, data centers are being built, and electrical components are being sold. The U.S. Census Bureau reported that data center construction is up 60% year over year since October 2023 and 100% since ChatGPT was released in late 2022. The main hyperscalers (Microsoft, Alphabet, and Meta) reported the sixth consecutive year of 25% electricity/energy demand growth, driven by technological advancements. PJM, the largest grid operator in “data center alley," reported power capacity prices rising 10x from the prior year as data centers drive power demand when grid planners were previously expecting low-single-digit electricity demand growth. 

Unlike the internet bubble of the late 1990s, where technology indices significantly outperformed other sectors, performance today is coming from other sectors that will support an infrastructure buildout. From the end of 1998 to the end of 1999, the S&P 500 technology index returned 79%. In comparison, the S&P 500 utility index was down 10% and the S&P 500 industrials index was up 21%. In the past one year (ending 12/2/2024), we see that the same indices are up around the same amount (in the range of 33% to 41%). Again, this does not necessarily indicate a bubble.

AI has upended financial markets as investors scramble to find new potential investment opportunities. Unlike past technology bubbles, valuations are fairly reasonable as orders are materializing on the bottom line for beneficiaries of the AI revolution. While asset prices have surged on the AI theme, our channel checks suggest that we are around the middle innings of it.

To be abundantly clear, we are cautiously optimistic about the AI theme. While there are some exciting data points for AI bulls, we recognize that stock valuations and prices have already moved substantially higher. We are also cognizant of the fact that early commentary from companies thought to be leading the AI charge has been mixed. As with any technological advance that is new and exciting, there remains a risk that AI-linked companies will overspend on the technological buildout today and that future earnings will fail to materialize.

With that being said, current trends lead us to believe that, for the reasons listed above, this is not the early 2000s tech bubble and we are cautiously optimistic there may be more performance ahead for the long-term investor. As always, we endorse prudent risk management and a well-balanced portfolio but suggest that investors may be better rewarded by not missing out completely on the high-flying AI revolution.