Sorry to Burst Your Bubble, but This Isn’t 1999

Asset bubbles occur when speculation and emotion drive prices far beyond their fundamental value.

With the S&P 500 delivering double-digit gains in 2023, 2024, and 2025, it’s easy to see why some assume we’re in one today. 

While every bubble has its own backstory, they all follow a familiar pattern. Here are some of the key attributes to look for.

Abandoning traditional valuation metrics

In the late 1990s, traditional valuation metrics such as a price-to-earnings ratio or free cash flow yield were substituted away. Instead, new internet startups with negative earnings began to invent non-financial metrics like “chasing eyeballs”. For example, website traffic was looked at to justify skyrocketing stock charts, even with no path to future profitability.

Today, investors are still focused on earnings, cash flow, and margins. A good example is the Magnificent 7 stocks (Alphabet, Amazon, Apple, Meta Platforms, Microsoft, Nvidia, and Tesla), which are only up 3% collectively since the start of 2026 because of the near $800 billion that the group will spend this year to build out their AI infrastructure. Because investors want to see evidence of a profitable return on investment, these stocks have underperformed the overall market.

Our concern over abandoning traditional valuation metrics: 1/10.

Extreme valuations

During the tech bubble, the S&P 500’s price-to-earnings ratio peaked around 34 times in the spring of 1999. To put this valuation into perspective, the market historically trades at around 15-16 times earnings. 

Today, the forward multiple sits at around 19.6 times. While this is elevated compared to the historical average, earnings are growing at a remarkable 47% year-over-year (the strongest rate since Q2-2021) and net margins are running at 16.7% (the strongest since at least 2009). When earnings are growing and margins are high, the valuation multiple deserves to trade at a premium relative to the historical average.

Our concern over extreme valuations: 3/10.

Euphoria

Bubbles take root when healthy market confidence mutates into collective euphoria. Driven by media hype, social pressure, and the “fear of missing out”, speculators buy stocks simply because prices are going up.

Unlike the dot-com era, today’s AI expansion is grounded in tangible infrastructure construction rather than pure speculation. North America currently has roughly 700 to 830 data centers actively under construction and another 1,500 to 1,760 planned. The $1 trillion being spent this year is set to grow to $2-3 trillion by the end of the decade.

Before the tech bubble burst in 2000, the S&P 500 delivered nine consecutive years of positive total returns, compounding almost  21% annually. Over the last nine years, two years produced negative returns (2018 and 2022) while the market has posted a 15% annualized total return.

Our concern over euphoria: 5/10.

Leverage

Buying on “margin” is when a trader borrows money to buy stocks. While the dot com bubble was about overvalued equity, the 2008 financial crisis was about excessive debt. Yet, both bubbles exhibited a clear increase in margin buying.

In 1999, margin debt rose by 61% compared to the year before to almost $230 billion. In 2006 and then 2007, margin debt rose 23% and 21%, respectively, year-over-year to hit $340 billion the year before Lehman Brothers collapsed. 

Today, margin debt sits at $1.2 trillion, and it has grown by over 30% year-over-year in both 2024 and 2025. Rising margin debt represents a primary vulnerability for today’s bull market. If prices drop, traders face margin calls requiring immediate cash deposits. Those who can’t cover the difference will see their stocks forcibly sold by brokers, transforming a routine pullback into a more severe sell-off. 

Our concern over leverage: 9/10.

Flooded offerings with new IPOs

A massive increase in initial public offerings (where private companies go public by selling shares to everyday investors) often signifies a bubble. 

During the dot-com bubble, hundreds of unprofitable tech startups rushed to public markets every year. In fact, there were 477 operating companies that went public in 1999. Of those companies, 74% had negative earnings and the average return on the first day of trading was +71%.

Since 2022, fewer than one hundred operating companies have listed each year. In the first half of 2026, only 48 operating companies have so far gone public. Furthermore, full-year IPO activity is tracking well below the 10-year historical average for public market debuts.

Our concern over flooded offerings: 2/10.

Overall Conclusion

While the rise in margin buying bears watching, today’s rally is backed by rising profits, tangible investments in AI infrastructure, and fairly priced valuation multiples.

-written by Jeff Pollock