Three Years of 72% Gains: Market Streak Echoes Patterns From Decades Past

The S&P 500 has historically delivered an average annual return of about 10% over the past century. The past three years, though, have been anything but typical: the index has climbed 72% over that span — roughly double the long-run pace — fueled largely by investor enthusiasm for generative artificial intelligence.

That kind of outperformance raises a question familiar to market historians. Extended stretches of above-average returns have rarely persisted indefinitely, and previous eras of rapid gains have at times been followed by years of muted — or even negative — returns. Analysts caution that the current environment bears similarities to earlier speculative technology cycles.

One frequently cited parallel is the dot-com era of the late 1990s, when internet optimism drove a string of outsized annual gains: 33% in 1997, 28.5% in 1998, and 21% in 1999. That run culminated in the S&P 500 touching an all-time intraday peak of 1,552.87 on March 24, 2000, before the bubble deflated. The current AI-driven rally has drawn comparisons to that period, given the concentration of gains in technology companies positioned as beneficiaries of the new computing paradigm.

The differences are also notable. Today’s leaders include chipmakers, cloud providers, and hyperscalers generating substantial revenue from AI infrastructure — unlike many late-1990s firms whose business models remained unproven. Still, valuations across parts of the tech sector have expanded rapidly, and historical precedent suggests elevated growth phases tend to mean-revert over longer horizons.

Individual stocks in the AI supply chain have seen volatile sessions as investors weigh the sustainability of the boom. Pattern AI Holdings (PTRN), for instance, traded at $20.20 in recent action, down 1.89% from its prior close of $20.59, giving the company a market capitalization of roughly $3.58 billion.

Whether the market is primed for a slowdown, a correction, or merely a period of normalization remains a matter of debate among strategists. What history does show is that century-long averages are built on both boom years and bust years — and that extended streaks of double-the-average returns have been rare occurrences, appearing only a handful of times over the past several decades.

What to watch

• Upcoming quarterly earnings from major AI infrastructure and semiconductor companies, which will test whether revenue growth supports elevated valuations.
• Forward guidance from S&P 500 constituents for signs of spending discipline or acceleration in AI-related capital expenditure.
• Whether the index sustains its multi-year streak of above-average returns or begins to converge toward the historical 10% annual norm.

Source: original release