Broker Check

Is This Another Bubble?

June 02, 2023

Google provides a service called Trends, which graphically displays the popularity of search phrases over time. The chart below is a trend report on “AI,” and it’s pretty clear that Artificial Intelligence (AI) has been top of mind this year1.

It’s not surprising. OpenAI released its AI late last year, and it’s getting smarter by the day. The chart below shows that the most recent upgrade back in March can easily pass tests like the Uniform Bar Exam. If it only took a few months for those green dots to become blue ones, just imagine where OpenAI will be by the end of the year.

There’s an old saying that technology has never moved this fast, and it will never move this slow again. That’s likely why the hype surrounding AI has exploded. But hype has a knack for fueling bubbles, so let’s analyze the current state of AI using a framework designed specifically to look for bubbles before they burst.

Blowing bubbles

William Bernstein’s seminal masterpiece, The Four Pillars of Investing, highlights four necessary conditions for a bubble to develop. The first is a major technological revolution or a shift in financial practice. Think about the impact of the light bulb, automobile, steam engine, etc. These products changed the world and fostered excitement that could not be contained.

The second condition is easy access to credit. If money is cheap and available, then fewer barriers exist for people to put not just their money behind an idea, but, far worse, someone else’s in the form of margin debt (borrowing money to invest).

The third condition is that people must have already forgotten about the last bubble. Since a bubble bursting tends to leave permanent scars, a new crop of investors must emerge that were not around the last time things got really bad.

The fourth is, as Bernstein puts it, “a complete abandonment from time-honored methods of security valuation”2. As bubbles grow, investors lose the ability to value assets on traditional metrics, so they look to new and untested ones to justify the mania.

These four create an environment where speculators buy simply because the price keeps rising. Like a snowball rolling down a hill, the bubble feeds on itself until it loses control and ultimately crashes.  Let’s apply this test to AI to see if a bubble is forming:

Condition 1: There are signs that mimic the early days of other bubbles, like crypto in 2018 and dot-com in 2000. For example, companies in the late 1990s added “.com” to their name to garner attention from investors. In 2017, companies that portrayed a pivot to crypto saw their stock prices soar (this happened again in 2021). Today, registrations of .ai domain names are up 156% year-over-year to 174,000, compared to a 27% increase for .com domains over that same period3.

Simply put, AI qualifies as a major technological revolution, and any stock with even a faint hint of being associated with AI is getting a lot of attention right now.

Condition 2: Despite higher interest rates, access to credit exists for those willing to pay for it. As stocks like Nvidia (ticker: NVDA) triple in a matter of months, there’s certainly an incentive to borrow. For example, Fidelity’s base margin rate is currently 12.075%4. That’s a small price to pay to position for returns like these.

Tangentially, talk of the Federal Reserve pausing interest rate hikes this month could be powering much of the optimism. Low interest rates fueled the tech boom of the last decade, so speculators may be anticipating a return to that regime.

But pausing is not the same as cutting rates to a level that could ignite another boom. If these overly optimistic assumptions are being baked into AI stock prices, then investors could be setting themselves up for a very bumpy ride.

Condition 3: Burning your hand on a hot stove and watching someone else do it doesn’t feel the same. Many investors today were in grade school when the dot-com bubble burst. Much of what happened then would read as a history lesson rather than a reminder of what it feels like to own and watch it get liquidated just 268 days after its IPO5.

Then there are those who missed out on crypto and meme stock mania. They may be viewing this as an opportunity to redeem themselves. If so, it’s hard to think of a more dangerous mindset for an investor than the Fear Of Missing Out (FOMO).

Condition 4: While it’s unlikely that AI is currently being valued on nontraditional metrics, it’s possible this happens if the hype continues to build. That’s what transpired during the dot-com boom, when companies with no earnings or revenue to value resorted to other metrics like the number of website visitors.

But traditional valuation metrics warrant caution. Some AI stocks right now are trading well over 20 times sales. For scale, an expensive growth stock trades at 3-5x sales. Valuations like these set incredibly high expectations that, when not met, can eviscerate a stock price.

The bottom line

Charles Mackay wrote Extraordinary Popular Delusions and the Madness of Crowds in the 19th century, chronicling famous bubbles over time. It’s a fascinating read because of the underlying consistencies throughout each bubble.

The most entertaining of them all was the “tulip mania” that almost destroyed the Dutch economy in the 17th century. Back then, tulips arrived in the Netherlands, and long story short, one species became so popular that it sold for more than ten times the annual income of a skilled craftsman. Yes, tulips. Those flowers that last about a week in a vase.

This bubble eventually burst and destroyed the lives of those who traded their land, life savings, and anything else they could liquidate to get more tulip bulbs. Ultimately, the panic forced the government to step in and support the market, making matters worse. The whole story sounds crazy, but swap tulip bulbs for dot-coms in the late 1990s, and it reads nearly identical. Human behavior fuels bubbles, and that hasn’t changed much over the last several thousand years.

It's hard to say that AI is in a bubble just yet, but based on these four conditions, it could be moving in that direction. If so, investors have two options. First, stay clear of it until we better understand how AI will evolve. Everyone in 2000 knew that the internet would change the world, but it took close to a decade before most companies figured out how to profit from it. AI is moving fast, but our ability to predict tends to be more constrained.

The second option is to be extremely patient and careful about what you own. If an AI bubble were to burst, it could take years before the calamity is behind us. For every Amazon today, hundreds of others didn’t survive. Knowing who will be the next big winner is hard to do, so stay as diversified as possible.

The bottom line is that AI may not be a bubble just yet, but never forget Sir John Templeton’s famous warning:

“The four most dangerous words in investing are: 'this time it's different.'"





2 Bernstein, William. The Four Pillars of Investing. 2003. Print.






This material has been prepared for informational purposes only and should not be construed as a solicitation to effect, or attempt to effect, either transactions in securities or the rendering of personalized investment advice. This material is not intended to provide, and should not be relied on for tax, legal, investment, accounting, or other financial advice. You should consult your own tax, legal, financial, and accounting advisors before engaging in any transaction. Asset allocation and diversification do not guarantee a profit or protect against a loss. All references to potential future developments or outcomes are strictly the views and opinions of Richard W. Paul & Associates and in no way promise, guarantee, or seek to predict with any certainty what may or may not occur in various economies and investment markets. Past performance is not necessarily indicative of future performance.