Are we in an AI investment bubble?

In recent weeks, I've been struck by the amount of media coverage focused on whether investors' enthusiasm for artificial intelligence (AI) is pushing the market towards an AI bubble.

As I struggle with this question myself, I thought it would be helpful to share my perspective, not only as an investor in AI stocks, but also as a former telecom analyst who witnessed the bursting of the tech bubble in the late 1990s. I still remember that experience vividly, and it taught me lessons that I believe apply to today's environment.

The important question today is whether the year 2000 is an appropriate analogy. The year 2000 means we are in a bubble that is about to burst. or 1998 – this means we have not yet seen a major turning point and there may be more room in the market.

I know bubbles are difficult to foresee and evaluate, but I believe we are closer to 1998 than 2000. As some of you may remember, 1998 was the year that Global Crossing, a company that defined the Fiberbuild era and one of the cornerstone stocks of the late '90s bubble, went public at $19 per share. Nine months later, the company's stock was trading at $64 per share. According to Worldcom's famous report, Internet traffic is doubling every 90 days.

Of course, both companies later went bankrupt, but they went on to do spectacular business. From the time Global Crossing went public on August 14, 1998 until the market peaked on March 10, 2000, the tech-heavy Nasdaq 100 index rose more than 245%.

For an experienced portfolio manager who had been left out of the market rally and skeptical of a growing bubble, that period of time off was extremely painful. They were eventually proven right after 2000, but the journey was an unpleasant one.

Will history repeat itself?

In the future, I think it's quite likely that there will be a bubble at some point, followed by a root-breaking correction. But I don't think we're there yet. As Mark Twain once said, history may rhyme, but it does not repeat itself exactly.

Today, companies making AI-related investments are even more powerful. The giant hyperscalers, internet and cloud platform providers such as Amazon, Microsoft, and Alphabet, can support capital expenditures far better than the telecom startups of the late 90s.

Some economists said spending on AI-related investments, including buying advanced chips and building huge data centers, has helped shield the U.S. economy from recession this year.

According to some estimates, the AI ​​spending cycle is so large that it accounts for approximately 7% of the U.S. gross domestic product, or more than $2 trillion. This massive spending is likely necessary if Big Tech incumbents are to avoid being disrupted by young emerging AI companies.

I believe that as long as companies like Amazon, Microsoft, and Alphabet view AI spending as existential, they will continue to invest. And that will continue to fuel the AI ​​boom.

The missing piece: The pending IPO boom

Another big difference today is that OpenAI, the leading company of this era, has not yet gone public. OpenAI kicked off the latest round of excitement in November 2022 with the launch of ChatGPT, an AI-powered chatbot. ChatGPT quickly became the most downloaded app of all time at the time.

Other innovative startups such as Anthropic, Cohere, Mistral AI, and xAI also remain private for now. While we haven't yet reached a global moment, I believe it's only a matter of time before these startups enter the next phase of growth through the initial public offering (IPO) process.

One of the factors that inflated and sustained the 1990s tech bubble was the promise of accelerated revenue growth and subsequent profitability. These pre-IPO companies are the modern equivalent. Once a company goes public and investors have greater visibility into its financials, higher growth rates are likely to be rewarded by the market.

Additionally, it is worth noting that the US Federal Reserve is currently on an interest rate cutting cycle. Easy monetary policy could be a tailwind for high-value technology stocks. In 1998, after the bankruptcy of Long Term Capital Management, Fed officials began cutting interest rates very aggressively. Amid growing concerns about the Y2K issue, interest rates were maintained at low levels.

Today, some would argue that tariffs and weakening labor markets are equal concerns prompting the Fed to act. In any case, there was and is still a large amount of liquidity in the system, which tends to stimulate the animal spirits of investors.

What happens when the AI ​​bubble is about to burst?

Another lesson I've learned from 30 years of investing is that the market can sometimes humble you. It's entirely possible that I'm wrong about the scope and timing of the AI ​​bubble.

In my portfolio, I invest in a 1998 or 1999-like situation with the goal of fully participating in the powerful AI trends that continue to unfold among dynamic, growth-oriented companies. But I'm also on the defensive and trying to add some balance to my portfolio.

From that perspective, I'm actively looking for companies that may not have a great reputation right now, but could do relatively well once the AI ​​bubble bursts. In my opinion, energy and cable companies fall into this category.

Both sectors are trading near historically low valuations. And both contain carefully selected companies with decent returns, valuable long-term assets, and the potential for surprises.

For example, the energy sector currently makes up about 2.8% of the S&P 500 index. This is only slightly higher than at the height of the coronavirus crisis, when oil prices briefly fell below zero.

Looking at the index data, it has never been this low. This area of ​​the market was effectively left for dead, and I think the level of pessimism had gone too far.

Similarly, the rapid decline of cable TV has made cable stocks unpopular for a long time. But for investors who want to take a closer look at this sector, I think there are some overlooked gems with growing businesses and healthy cash flow that are trading at very low multiples. Investors don't often have the opportunity to invest in a growing company that can earn up to 6x earnings.

Not to sound any alarm bells, but this is how I hedge AI-related risks in my portfolio.

Considering the bubble problem

My comments should not be interpreted as suggesting that I am unconvinced by the rapid progress of AI and its potential to be an incredibly transformative technology.

I'm not an AI skeptic. I believe it will change the world, just like the internet. I believe it will set the stage for the birth of new, innovative and disruptive companies, just as the advent of the Internet paved the way for Amazon, Alphabet, Meta, and Netflix.

But I think it's also important to assess where we are on the AI ​​adoption journey, investor enthusiasm, and the potential for problems down the road. If we are moving towards bubble territory, where we are on that timeline will be very important. I would argue that we are nearing the early stages.

And if you look at the history of the tech bubble in the late 90s, you might come to the same conclusion as I did. This means that it is probably too early to leave the attractive opportunities presented by this powerful new technology to the risk of a bubble problem.

The author is an equity portfolio manager at Capital Group.