Every Market Bubble in History Looked Obvious Only After It Popped

If you could actually see a bubble forming in real time, you wouldn't stay in it. Neither would anyone else. That's the core paradox laid out by Lance Roberts of RealInvestmentAdvice.com in a recent analysis published via ZeroHedge: the very fact that bubbles inflate at all proves most people can't spot them until it's too late.

This isn't a new problem. Organized stock trading has existed since the Amsterdam exchange opened in the early 1600s, and speculative manias have followed almost as long.

The Historical Record

The Dutch Tulip Mania of 1636 to 1637 is the case everyone cites first. Tulip bulb prices in the Netherlands rose roughly twentyfold in a matter of months, then collapsed by about 99% in May 1637, according to Roberts.

Less than a century later came the South Sea Bubble. Shares of the South Sea Company went from £128 in January 1720 to £1,050 by June, then crashed back to near where they started by year's end. Isaac Newton, one of the sharpest minds in history, lost a fortune in it. He's credited with the line: "I can calculate the motion of the heavenly bodies, but not the madness of crowds."

The 20th century produced bigger, uglier versions of the same pattern. The Roaring Twenties ended with the 1929 crash and a Dow Jones drawdown of nearly 89% by 1932.

Japan's asset bubble in the late 1980s pushed the Nikkei 225 to 38,915 on December 29, 1989. What followed was a collapse of more than 80%, and the index didn't hit its post-bubble low until October 2008, almost 19 years after the peak.

Then came the dot-com era. Between January 1995 and March 10, 2000, the Nasdaq Composite climbed roughly 572% to a peak of 5,048.62. It then fell 78% by October 2002 and didn't reclaim its 2000 high until April 2015, according to Roberts's analysis.

The 2008 financial crisis broke the mold slightly. Instead of one speculative asset running wild, the excess built up in mortgage credit and spread through the entire global banking system. The S&P 500 still lost 57% peak to trough.

Why This Is Genuinely Hard, Not Just a Cop-Out

There's a real intellectual point buried here, and it deserves to be taken seriously rather than dismissed as market-timing excuse-making. Every one of these episodes looked completely different while it was happening. Tulip bulbs, South Sea shares, 1920s margin lending, Japanese real estate, dot-com stocks, mortgage-backed securities. Different assets, different decades, different countries, different catalysts.

What they share is only visible in hindsight: a period where prices detached from any reasonable measure of underlying value, sustained by the belief that someone else would pay more later. That belief works, right up until it doesn't, and there's no reliable indicator that tells you the exact moment it stops working.

A healthy chunk of financial commentary today is dedicated to declaring the current environment, whether it's AI stocks, crypto, or parts of the housing market, a bubble in progress. Some of those calls will eventually look prescient. Most of them, based on the historical base rate, will be wrong or wildly early, because most bubble calls are.

The honest position is that nobody, left, right, or center, has a working real-time bubble detector. Economists have tried to build one for decades using metrics like price-to-earnings ratios, credit growth, and valuation spreads. None of them reliably called the top before it happened. Alan Greenspan's "irrational exuberance" line came in December 1996, more than three years before the Nasdaq actually peaked in March 2000. He was right eventually. He was also early enough that anyone who acted on it immediately missed a huge chunk of gains.

For investors sitting on cash right now waiting for a top-callable signal in AI stocks or anywhere else, history offers a hard lesson. The ones who correctly call the top usually get there by accident, and the ones who confidently swear it's a bubble months or years before the peak often end up broke or irrelevant before they're proven right.

None of this means bubbles don't matter or that valuations are irrelevant. It means the honest, common-sense takeaway is humility: nobody rings a bell, and the people insisting they can hear one usually can't either. The only reliable pattern across four centuries is that every bubble looks completely unique going up and completely obvious coming down.