Bubbles-in-the-sun

Nothing new under the sun, part 2

As I mentioned in Part 1, the present is more like the past than we think. There’s nothing new under the sun, especially when it comes to investor behavior. And that gives us a chance to learn from the past.

2021’s everything bubble shares many parallels with the irrational exuberance of the late 1990s U.S. tech bubble and the soaring markets of the Roaring Twenties (the 1920s).

Cliff Asness, co-founder of AQR, published Bubble Logic in August 2000. The NASDAQ and S&P 500 were just off their tech bubble peaks at the time. As we now know, the NASDAQ would plunge 78% and the S&P would drop 49%. But that hadn’t happened yet. Investors still wanted to believe and had many ways of justifying the record-high valuations of U.S. equities and especially tech stocks at the time.

Cliff Asness critically examined these justifications, which he labeled “bubble logic:”

“Equities Always Win over the Long-term”

Yes, equities have tended to outperform other asset classes over long periods of time (although later research reveals this is not as consistent as we may have first thought). Defenders of the high valuations during the tech bubble claimed that since equities had won (in the past), they must win (in the future).

There is no requirement that the “stock market must do well…” Yes, equities should offer a risk premium but there’s no law that says they will deliver one. As Asness noted, equities were cheaper over the years typically captured by long-run analysis of historical returns. Counter to conventional wisdom, a long horizon doesn’t rescue you from high valuations. Extreme valuations lock in low or even negative returns and lower the odds of winning (outperforming other asset classes) over the next 10 or 20 years.

If you were counting on the S&P 500 and NASDAQ to win in 1999 or early 2000, you would have been deeply disappointed. The S&P peaked in 2000. And it would not surpass its peak in real terms (adjusting for inflation) until 2014 (2013 with dividends reinvested). The NASDAQ would take a bit longer.

“It Is Silly to Compare Today’s P/Es to Those before the Great Depression”

1929 was the prior high watermark for U.S. equity valuations. Although 70 years had passed, there were several parallels between 1929 and 1999. The rise of a new technology – the radio in the 1920s, the internet in the 1990s. A belief that record earnings growth would continue – “Apparently, then [in 1929] as now [late 1990s], investors were looking at recent growth, and pricing stocks as if this growth would go on forever.” And a view that the Fed would rescue the economy if the markets “went on the rocks.” (The Greenspan put is nearly as old as the Fed!)

Investors in the 1990s knew how the 1920s had ended – with “a tremendous crash and bear market…and a full 20 years of effectively zero real return on stocks.” Accepting the similarities meant accepting, at least in part, that the good times might end. And that end, when it came, could be ugly. The tech bubble did end, the most speculative stocks declined the most, and U.S. equities experienced a lost decade.

This isn’t to say there is a right valuation level or that P/Es from the 1920s were 100% comparable. But the valuations in the late 1990s were also higher than they had been in any recent decade.

“If You Had Worried about X’s Valuation…You Would Have Missed Making a Fortune”

“…this logic is precisely the same as pointing to the winner of a lottery and declaring that lotteries are a good investment.” We remember the “super expensive growth stock that was really worth the price.”

We like the potential of a high payoff. But so do others and, as a result, lottery ticket stocks tend to be a terrible bet (as a group). Yes, there will be the occasional stock that returns 50x or 100x. And these winners are incredibly salient. But they are rare (as we know from Bessembinder’s research). They often seem obvious in hindsight, but they are hard to recognize beforehand (otherwise, we’d all be rich!).

“You Will Be OK, Just Stick to Buying Great Companies”

This is another way of saying that valuation doesn’t matter – as long as we buy great companies. “Recognizing a great company before the market does [emphasis added] is clearly a way to get rich.” But this only works if we can identify great companies before others or detect greatness that is not already priced in. Valuation still matters! We tend to forget this, especially in the later years of a long bull market.

Cisco Systems was exhibit A for this in the late 1990s. Cisco was the classic seller of picks and shovels in the dotcom gold rush and was, briefly, the most valuable company in the world. From 1999 through 2022, revenue would quadruple, and profits would quintuple. Even so, the company never lived up to the lofty expectations priced in during the tech bubble. At its peak, Cisco was worth more than $500 billion or ~30 times trailing sales. Although the company subsequently performed well, investors who bought at or near Cisco’s peak valuation would regret it – even if they held on for the next 10 or 20 years.

Closing thoughts

“…we, like everyone throughout history, have the hubris to think that the present is radically different than the past.” We tend to discount the past and, as a result, learn less from it than we could.

What are some lessons for today?

  1. The Fed is not all powerful – stock markets can decline when the Fed is cutting rates.
  2. With world-changing technologies, it can be hard to predict the eventual winners. Investors, caught up in the hype, usually overpay for companies connected to the new new thing.
  3. Equities are not a sure thing – they are riskiest when they feel the safest (after years of strong performance).
  4. Markets can stay expensive for years but valuations matter…eventually.

Although U.S. equities are down from their 2021 highs, they are still very much in bubble territory. This is especially true for growth stocks and the tech sector. And many of the justifications used in the late 1990s and prior bubbles have resurfaced. The markets are up this year but It’s Not Over Yet.


There’s much more in the article (PDF) by Cliff Asness.

Share this post
Scroll to Top