About this Topic
▾
The Nasdaq keeps giving investors reasons to believe the next big move has arrived. A rally suggests the technology boom is back. Another retreat revives the argument that artificial intelligence has become a bubble—and that the bubble is breaking.
But six months of repeated reversals tell a more complicated story than a boom simply turning into a bust.
The evidence examined here fits an uneven consolidation: substantial corrections have already happened in individual stocks, while some profitable companies are growing into their valuations. That interpretation has stronger support than treating the technology sector as one speculative trade awaiting a single collapse.
The unusual Nasdaq pattern is the starting point.
Between May 4 and October 29, 2026, the Composite crossed its fixed period average 28 times across 125 trading sessions. It gained about 4.5% overall, while its largest closing decline from a preceding peak was approximately 9.8%.
After the deepest decline, six completed rebounds reached at least 1% above that average before falling below it again. The market repeatedly recovered enough to encourage buyers, then failed to sustain the move.
An exploratory analysis of [daily Nasdaq history dating to 1971](https://fred.stlouisfed.org/series/NASDAQCOM) found no earlier window matching that defined combination of frequent crossings, limited trend, modest drawdown and substantial failed recoveries.
The distinction survives a broader comparison. Even allowing earlier six-month windows declines as large as 15%, none recorded more than 23 average crossings. This period recorded 28 with a smaller decline.
It was not a record on every individual measure. The novelty was the combination: unusually persistent reversals without the deeper damage that accompanied other historically turbulent periods.
That makes it difficult to describe the index as either an uninterrupted speculative surge or an established collapse.
It also makes the familiar question—when will AI stocks finally correct—look late.
A review of 25 selected AI-related stocks found that 19 had already experienced declines of at least 20% during 2026 by early May. By late July, the count reached 23.
Those are falls from preceding peaks, not necessarily losses since January. Some stocks remained ahead for the year. Nevertheless, the corrections were substantial, and many preceded the Nasdaq’s six-month comparison period.
A weighted index can conceal that repricing when stronger companies offset weaker ones. A flat Nasdaq does not mean every component is standing still, or that speculative excess remains untouched.
Some speculative stocks can be suffering a bubble unwind while the index consolidates. Those processes can coexist. Calling the entire Nasdaq a bursting bubble obscures the differences between businesses.
Prices can also become easier to justify without falling.
Nvidia illustrates the mechanism. Between May and October, its shares rose approximately 8%, while reconstructed trailing earnings per share increased about 61%. Its trailing price-to-earnings ratio fell from roughly 40.5 to 27. Operating profit also increased about 52%.
These filing-based calculations compare the previous twelve months of reported earnings available at each observation date. They are not forecasts or single-quarter growth rates.
The implication is straightforward: the business improved faster than its price increased. Investors were paying less for each dollar of reported earnings. That is evidence supporting consolidation, even though it does not establish that Nvidia is cheap.
The argument has limits. AI can transform the economy while associated shares become overpriced. The internet did exactly that; the Nasdaq took about 15 years to regain its 2000 closing record.
Established technology companies also have revenues beyond AI. Those businesses provide support, but share prices still reflect future growth. Disappointing returns on AI spending can hurt shareholders without destroying existing operations.
Nor does the unusual chart establish an imminent rally. Historical tests of selected sideways technology stocks found that about two-thirds rose over the following year—but those stocks usually rose about two-thirds of the time anyway. Exact Nasdaq-pattern analogues were unavailable, and missing delisted stocks limit the comparison.
The repeated reversals have practical costs: options can [lose time value](https://www.optionseducation.org/advancedconcepts/theta) while investors wait, and leveraged futures positions can [require additional collateral](https://www.cmegroup.com/education/courses/understanding-the-benefits-of-futures/the-benefits-of-futures-margins) before a recovery. Those effects explain why consolidation can feel punishing without proving that the technology boom has failed.
October 30’s [2.86% Composite decline](https://markets.ft.markitdigital.com/data/indices/tearsheet/historical?s=COMP%3ANAS) underscored that further losses remain possible. It did not, by itself, resolve the broader debate.
The strongest reading is a market already correcting excess through a mixture of falling prices and improving earnings. The unprecedented six-month combination makes this consolidation worth examining; it does not turn it into protection against another decline.
The test of the AI boom is whether profits and cash generation ultimately justify the prices investors are paying.
But six months of repeated reversals tell a more complicated story than a boom simply turning into a bust.
The evidence examined here fits an uneven consolidation: substantial corrections have already happened in individual stocks, while some profitable companies are growing into their valuations. That interpretation has stronger support than treating the technology sector as one speculative trade awaiting a single collapse.
The unusual Nasdaq pattern is the starting point.
Between May 4 and October 29, 2026, the Composite crossed its fixed period average 28 times across 125 trading sessions. It gained about 4.5% overall, while its largest closing decline from a preceding peak was approximately 9.8%.
After the deepest decline, six completed rebounds reached at least 1% above that average before falling below it again. The market repeatedly recovered enough to encourage buyers, then failed to sustain the move.
An exploratory analysis of [daily Nasdaq history dating to 1971](https://fred.stlouisfed.org/series/NASDAQCOM) found no earlier window matching that defined combination of frequent crossings, limited trend, modest drawdown and substantial failed recoveries.
The distinction survives a broader comparison. Even allowing earlier six-month windows declines as large as 15%, none recorded more than 23 average crossings. This period recorded 28 with a smaller decline.
It was not a record on every individual measure. The novelty was the combination: unusually persistent reversals without the deeper damage that accompanied other historically turbulent periods.
That makes it difficult to describe the index as either an uninterrupted speculative surge or an established collapse.
It also makes the familiar question—when will AI stocks finally correct—look late.
A review of 25 selected AI-related stocks found that 19 had already experienced declines of at least 20% during 2026 by early May. By late July, the count reached 23.
Those are falls from preceding peaks, not necessarily losses since January. Some stocks remained ahead for the year. Nevertheless, the corrections were substantial, and many preceded the Nasdaq’s six-month comparison period.
A weighted index can conceal that repricing when stronger companies offset weaker ones. A flat Nasdaq does not mean every component is standing still, or that speculative excess remains untouched.
Some speculative stocks can be suffering a bubble unwind while the index consolidates. Those processes can coexist. Calling the entire Nasdaq a bursting bubble obscures the differences between businesses.
Prices can also become easier to justify without falling.
Nvidia illustrates the mechanism. Between May and October, its shares rose approximately 8%, while reconstructed trailing earnings per share increased about 61%. Its trailing price-to-earnings ratio fell from roughly 40.5 to 27. Operating profit also increased about 52%.
These filing-based calculations compare the previous twelve months of reported earnings available at each observation date. They are not forecasts or single-quarter growth rates.
The implication is straightforward: the business improved faster than its price increased. Investors were paying less for each dollar of reported earnings. That is evidence supporting consolidation, even though it does not establish that Nvidia is cheap.
The argument has limits. AI can transform the economy while associated shares become overpriced. The internet did exactly that; the Nasdaq took about 15 years to regain its 2000 closing record.
Established technology companies also have revenues beyond AI. Those businesses provide support, but share prices still reflect future growth. Disappointing returns on AI spending can hurt shareholders without destroying existing operations.
Nor does the unusual chart establish an imminent rally. Historical tests of selected sideways technology stocks found that about two-thirds rose over the following year—but those stocks usually rose about two-thirds of the time anyway. Exact Nasdaq-pattern analogues were unavailable, and missing delisted stocks limit the comparison.
The repeated reversals have practical costs: options can [lose time value](https://www.optionseducation.org/advancedconcepts/theta) while investors wait, and leveraged futures positions can [require additional collateral](https://www.cmegroup.com/education/courses/understanding-the-benefits-of-futures/the-benefits-of-futures-margins) before a recovery. Those effects explain why consolidation can feel punishing without proving that the technology boom has failed.
October 30’s [2.86% Composite decline](https://markets.ft.markitdigital.com/data/indices/tearsheet/historical?s=COMP%3ANAS) underscored that further losses remain possible. It did not, by itself, resolve the broader debate.
The strongest reading is a market already correcting excess through a mixture of falling prices and improving earnings. The unprecedented six-month combination makes this consolidation worth examining; it does not turn it into protection against another decline.
The test of the AI boom is whether profits and cash generation ultimately justify the prices investors are paying.