Key Questions: Is Artificial Intelligence (AI) a “Bubble”?

Recently, we have heard much from the financial press about the so-called “AI bubble.” With comparisons made to the late 1990s dot-com bubble, and the associated aftermath, we thought it could be helpful to put market bubbles in perspective and separate fact from fear. In our view, a market bubble, by definition, cannot be as highly debated as the current bubble that we may be experiencing. Said differently, on the fear and greed spectrum, how can one conclude excessive optimism is so present while sentiment, in general, is so depressed? What’s more, market bubbles seem only defined in hindsight, not prospectively. Finally, we would point out that the financial media, and the collective media as a whole, tends to carry a negativity bias, heralding “the end is near” for the purpose of growing viewership.

Historical Parallels

From the Dutch tulip mania in the 17th century to the Mississippi bubble and South Sea bubble in the 18th century, followed by the British railway bubble, Roaring Twenties, Japanese bubble, dot-com bubble, and Housing bubble, which precipitated the global financial crisis (GFC) in 2007–2009, history is replete with examples of financial booms and busts. While times have changed, human psychology (specifically investor greed and fear) – which underpins all manias and crashes – has not changed, in our view. This is why there will always be a next bubble, and we don’t know how or when it will happen. 

We often hear the case of Sir Isaac Newton (arguably one of history’s brightest minds), who was caught up in the events of Tulpenwoerde (tulip madness), which took place in the Dutch United Provinces in 1636 and 1637. Ironically, Sir Newton made a fortune speculating early in the bubble, only to lose it all as he watched his friends become even richer than he. Comparison was the thief of joy for him.

More recently, several mini bubbles occurred over the past 10 years; however, due to the contained fallout, we hear little about them. For example, the cannabis bubble and the special purpose acquisition company (SPAC) bubble both resulted in declines of approximately 90% of affected stocks. Few predicted these mini bubbles. If they had, there likely would have been no dramatic crash within these segments of the market.

This historical context provides us with clues of how bubbles come and go, with numerous false alarms along the way. One common thread is that the bear case is often the most intellectually appealing. Indeed, we recall Morgan Housel’s book "The Psychology of Money," where he sums up this cognitive and behavioral phenomenon quite astutely: “Pessimism just sounds smarter and more plausible than optimism. Tell someone that everything will be great, and they’re likely to either shrug you off or offer a skeptical eye. Tell someone they’re in danger, and you have their undivided attention.” Consequently, we understand the AI skeptics’ fears but are equally skeptical of bears as well as bulls.

Getting Famous for Being Right Once in a Row

We believe that bubbles, by definition, cannot be predicted accurately and reliably because in the short term, markets function as a voting machine (in the words of) Warren Buffett. Said differently, the market prices in the aggregate views of all its participants, and if many participants are calling a market a “bubble,” then price reflects this, and there is likely more upside. In our view, sentiment cannot get overly bullish if bears are taken seriously and both sides are reflected in asset prices.

There are certainly some who break with consensus and herald a bubble. For example, Michael Burry of Scion Capital (made a bit more famous in the movie "The Big Short") made billions on the popping of the U.S. housing bubble that precipitated the GFC. Another example is Joe Granville, who correctly called an equity market top in 1976 and again in 2000. For more than two decades, Mr. Granville had an abysmal batting average. In the words of investor Howard Marks, the investment business is full of people “who get famous for having been right once in a row.”

The key to finding a great investment is to (1) believe something that no one else believes, (2) be right, and then (3) be patient as everyone else comes around to your point of view (also known as reality). This works in both directions, long and short. We watch as AI bears vocally short artificial intelligence when they call it a “bubble,” and we ask ourselves how being short AI can be a great investment when that position does not pass the first point in our framework: that one must “believe something that no one else believes.”

Conclusion

We have no doubt that there will be future bubbles. That said, we do not believe that the AI bubble will result in a spectacular destruction of capital precisely because of so many market participants fearing this specific event. This is how markets work – to outperform you must be willing to do what the average market participant is not willing to do, and then you must be right. Both are equally difficult, but for different reasons. We would encourage clients to view bulls and bears with equal skepticism and to be aware that financial media benefits when we lose our composure.

For more information, please contact your advisor.