No we are not in a bubble yet

Is the U.S. stock market currently in an AI-fueled bubble? Is March 2024 similar to March 2000, the peak of speculative excess driven by wild optimism?

No, not even close. Perhaps we are in the initial stages of a bubble (we surely have all the preconditions), but based on both objective criteria and subjective vibes, we are not there yet. We seem closer to 1995 than 2000.

What is a bubble? Not just overpriced

Defining bubbles

Do bubbles exist? Eugene Fama, one of the towering figures of academic finance and a Nobel laureate in economics, has some thoughts:

I don't even know what a bubble means. These words have become popular. I don't think they have any meaning… Any time prices went up and down—I guess that is what they call a bubble. People have become entirely sloppy.¹

Let us not be entirely sloppy. Here’s my definition: A bubble is a self-sustaining rise in prices over time resulting in the speculative trading of an obviously overvalued asset. The three necessary ingredients are:

• Overvaluation: Prices rise to a level that is widely recognized as overvalued. An unusually large fraction of market participants believe market prices are too high.
• Feedback loop: Initial price increases result in further price increases, due to (a) extrapolation of past increases; (b) the entry of new market participants; (c) social contagion where optimistic ideas and narratives spread through the population.
• Speculative trading: Trading volume is unusually high, as many traders buy the asset despite believing it is overvalued because they intend to sell it quickly at a profit.

For more bubble definitions, you can peruse the references at the end of this post.

Let me differentiate between the words “bubble,” “boom,” and “overvalued.” Overvalued means the price is too high; a bubble is a subset of overvalued. A boom is a rise in stock prices along with improving fundamentals, such as profits and revenue. A typical stock market bubble is the overlap of overvalued and boom. A boom becomes a bubble if the price rises to unreasonable levels with many market participants knowingly purchasing overvalued shares.

Not everything is a bubble

Not everything is a bubble, and not everything that people say is a bubble is a bubble. You can’t just identify a bubble by looking at price alone, because a bubble is defined by multiple non-price criteria.

So here, let me summon my inner Fama² and condemn the sloppy use of the word “bubble.” In my view, the following situations are not bubbles:

• Prices go up and then crash.
  – No, sometimes prices go up and down rationally in response to news.
  – No, bubbles don’t always end with a crash. Neither is it true that all overvalued assets inevitably underperform subsequently; they underperform on average, but in any specific case there is randomness.
• Price mysteriously doubles on low volume. No one knows why.
  – Bubbles are not about inexplicable movements in price. Bubbles are about explicable but unjustified increases in price.
  – Bubbles are social phenomena that operate across time and across people. If you don’t see people communicating with and reacting to each other, you don’t have a bubble.
• Taylor Swift announces she is marrying Travis Kelce, with the wedding to occur at halftime during the Super Bowl.³ Due to the resulting global euphoria and irrational optimism, the S&P 500 index doubles in one day and stays there.⁴ Because everyone is equally euphoric, there is zero trading volume because there are no haters who want to sell. A year later, Taylor Swift announces she is retiring from music to raise a family,⁵ producing a wave of despair across the globe. Due to irrationally negative sentiment, the S&P 500 falls 50% in a single day, with zero volume because everyone is too depressed to trade.
  – Bubbles are not 100% irrational. Bubbles also involve rational speculative trading.
  – A bubble involves a sequence of price rises, not just a one-day price rise.
  – Bubbles are about trading. No volume, no bubble.

Now, lest you think I am a curmudgeonly bubble denialist, let me be clear: while not everything is a bubble, some things are. In my view, the U.S. stock market had a sizeable bubble peaking in 2000 (many, perhaps most, economists would agree) and a smaller one peaking in 2021 (here, my view is less widely held). And in crypto, we’ve had many bubbles, with the two most recent peaking in 2018 and 2021.

My bubble checklist

There are many symptoms of bubbles. Some are objective quantitative metrics that can be constructed with publicly available data from 1926 to today. Some are subjective and impressionistic, such as the plot points from various episodes of The Simpsons (e.g. Homer’s dot-com startup of 1998, Compu-Global-Hyper-Mega-Net, anticipating the market peak two years later). On the impressionistic side, I’ve already given some fanciful scenes from a bubble (many of which were thinly disguised actual events experienced by me in 2000).

In future posts, I’ll be discussing these symptoms (let’s hope the bubble doesn’t come before I’ve had a chance to describe all the many items on my checklist), but for now, let’s talk about my four favorite bubble symptoms.

The four horsemen of the apocalypse

My top four bubble indicators for the U.S. stock market are:

First Horseman, Overvaluation: Are current prices at unreasonably high levels according to historical norms and expert opinion?

Second Horseman, Bubble beliefs: Do an unusually large number of market participants say that prices are too high, but likely to rise further?

Third Horseman, Issuance: Over the past year, have we seen an unusually high level of equity issuance by existing firms and new firms (IPOs), and unusually low levels of repurchases?

Fourth Horseman, Inflows: Do we see an unusually large number of new participants entering the market?

We saw all four horsemen in 2021, which is why, in my judgement, we had a U.S. stock market bubble then. As of March 2024, we may perhaps hear the distant hoofbeats of the First Horseman (overvaluation), who has not traveled far since he last visited us, but there is no sign yet of the other three.

Let me first discuss some conceptual issues about the horsemen and then move on to some specifics of how to recognize them.

On the conceptual side, I want to talk about the issue of dumb vs. smart money. Part of what happens in a bubble is that the smart money sells overpriced equity to the dumb money. When the market is overvalued (First Horseman), opportunistic firms (Third Horseman, issuance) sell equity to eager new buyers (Fourth Horseman, inflows). Indeed, you can’t have inflows into the market (net purchasing by retail investors) without somebody else selling (net selling by firms), so in that sense the Third and Fourth horsemen are yoked together.

However, bubbles are not only about dumb money that doesn’t realize the market is overvalued. Bubbles are also about some of the smart money buying equity while knowing it is overpriced. That is the point of the Second Horseman (bubble beliefs). While firms are able to issue equity to exploit overpricing, other smart money (say, hedge funds) does not have that ability, and may decide to “ride the bubble” and buy.

Do all four horsemen need to be on the scene for a bubble to occur? I’m not sure, but one horseman is surely not enough. Take the Third Horseman, issuance. After the global financial crisis of 2008, many financial institutions cut dividends and repurchases, and issued equity. This behavior was not an opportunistic attempt to sell overpriced equity, but rather was motivated by the need to raise cash (in some cases, forced by regulators). So when we see high issuance but low stock prices (as in 2009), that is not a bubble.

Where are we today?

Here are some specifics of what I need to see before I start thinking we are in a bubble, and where we stand as of March 2024.

First Horseman, Overvaluation. Wise professors of economics or finance declare the market to be overvalued. In the bubble of 2000, we saw many distinguished professors, ranging from Robert Shiller of Yale (a strong believer in bubbles) to Jeremy Siegel of Wharton (a noted bull) declaring the market overvalued. I’m guessing that for this to happen again, stock prices would need to rise significantly higher, perhaps 50% to 100% above their current levels. I am not predicting prices will rise that much, I’m just saying they’d need to rise that far to approach bubble territory.

Second Horseman, Bubble beliefs. The Valuation Confidence Index from the survey of U.S. individual investors conducted by the Yale School of Management falls to historic lows. These surveys, which were started by Robert Shiller, are an often-overlooked resource.⁶ Specifically, I need 65% or more respondents agreeing that “Stock prices in the United States, when compared with measures of true fundamental value or sensible investment value, are too high.” The latest reading on this index is 47%, which is not far from the average value from 1998-2023 and not close to bubble territory.

Third Horseman, Issuance. Over the last twelve months, the dollar amount of cash contributed by shareholders in the form of gross issuance (including issuance from existing firms and from IPOs) exceeds the dollar amount of cash paid to shareholders in the form of dividends, repurchases, and other distributions. In other words, I need to see positive net issuance (or negative net cash distribution) over the past year. Right now, trailing 12-month net issuance is comfortably negative. The U.S. stock market is currently distributing cash to shareholders, not raising cash from shareholders as we see in bubbles.

Fourth Horseman, Inflows. This symptom is less well-defined than the other three, because of the changing nature of stock market participation over time. In the tech stock bubble, you could partly observe retail buying by tracking mutual fund inflows. In 2021, the methods by which retail money entered the stock market were direct purchases of stocks and options (instead of mutual funds). If a new bubble does come, I am not sure how inflows will manifest, but I need to see evidence of new buying activity by individuals or institutions that had previously not participated in the stock market.

Could it be that in the coming months or years, the four horsemen will all arrive? Sure, that’s a plausible scenario given the tremendous (and justified) excitement about AI. If you think we’ve had three U.S. stock market bubbles in the past 100 years (1929, 2000, and 2021) that means that there is a 15% chance that a bubble will occur in any given five-year period. Going forward, it is obvious that today the chance of a bubble is much higher than average. Let’s say the probability has doubled or tripled, implying a 30% to 45% chance of a bubble over the next five years. You might also think that the chances of an ugly decline in stock prices is also higher than usual due to the uncertain impact of AI, among other factors.

References