The artificial intelligence rally is now showing a factor structure and market concentration with a striking resemblance to the 1999 dot-com bubble, according to a Morgan Stanley report.
The AI-driven surge in equity markets has reached historic proportions, with the Philadelphia Semiconductor Index (SOX) peaking at a price 62 percent higher than its 200-day moving average. This spread is more than double the margin seen before the 1987 Black Monday crash and significantly exceeds the lead-up to Black Tuesday in 1929, according to a Bank of America note. The current market structure shows the highest similarity to the dot-com bubble, which saw the Nasdaq's spread hit 55 percent before it burst in March 2000.
"There is a risk that markets are running too hot on the AI story," Michel Lerner, head of UBS's HOLT division, said in a research note. He noted that the equity price action in April represented a nearly three-standard-deviation event over the last 25 years, cautioning that markets are assuming AI firms are "immune to normal competitive dynamics."
The parallels to 1999 are stark. A May 14 report from Morgan Stanley MUFG Securities found that factor returns in April 2026—with momentum factors dramatically outperforming value, low-volatility, and small-cap factors—most closely mirrored the period from June 1999 to February 2000. This is reinforced by extreme market concentration, with the Nikkei-to-TOPIX ratio hitting a record high above 16 on April 24, reflecting buying pressure narrowing to a few high-beta, large-cap stocks.
While the rally has been powerful, some analysts see it as a classic bubble, forcing investors to chase returns despite mounting risks. Morgan Stanley's report concluded that investors have "no choice" but to participate in what could be a "once-in-a-generation" bull market. It suggests the next six months could be a critical window, referencing the roughly six-month final phase of the internet bubble from October 1999 to March 2000.
Intel Loses 950 Basis Points of Server Market Share
Beyond broad market concerns, competitive pressures are intensifying for legacy chipmakers. Intel is steadily losing its footing in the lucrative server CPU market, with rivals AMD and Arm directly benefiting, according to fresh data from UBS. In the first quarter of 2026, Intel's market share fell to 54.9 percent, a year-over-year decline of 950 basis points.
AMD's share grew by 330 basis points to 27.4 percent over the same period, while Arm-based processors seized another 620 basis points to capture 17.7 percent of the market. The data confirms a persistent trend of Intel ceding ground in the data center, a segment critical for AI infrastructure growth. The competitive shift prompted a selloff, with Intel shares falling 6.43 percent to $108.48 in Friday's trading. Advanced Micro Devices and Arm Holdings also retreated 3.4 percent and 4.4 percent, respectively, as investors took profits.
Valuations Signal Caution Despite Index Headroom
A key difference between the current environment and the 1999 peak is index-level valuation. As of May 10, the Nasdaq 100 and SOX index traded at forward price-to-earnings ratios of 24.9 and 21.5, respectively. While not cheap, these levels are far from the extreme valuations seen in 2000, suggesting that broad indices may still have room to run.
However, the picture for individual stocks is different. A basket of "AI Enablers" defined by Morgan Stanley shows median and average forward P/E ratios climbing steadily and already well above 2025 levels. This indicates that valuation pressure and crowding risk are building in more concentrated, pure-play AI names rather than in the mega-cap-weighted indices.
For Intel, the UBS analysis offered some potential bright spots. The firm noted that Intel's upcoming Coral Rapids processor line could help stabilize its market position. Furthermore, analysts see an opportunity for Intel to "benefit on the PC side as locally run agentic workloads drive demand over the medium term." Still, the report underscores the accelerating momentum held by AMD and Arm, concluding that Intel needs compelling product innovations to halt the competitive erosion.
This article is for informational purposes only and does not constitute investment advice.
