Ai drives semiconductor growth as valuations peak

Artificial intelligence continues to drive the semiconductor sector’s surge, but rising valuations and crowded positioning are increasing risk, according to Bernstein’s June 23 quarterly report.

The firm said fundamentals remain strong, with earnings growth underpinning most of the rally, yet warned that expectations are now stretched. Bernstein maintained “outperform” ratings on Nvidia and Broadcom, calling them core to the AI supply chain despite underperformance in their share prices this year.

Semiconductor rally driven by earnings growth

The Philadelphia Semiconductor Index has climbed 155.6% over the past year and 106.6% year-to-date, far outpacing the S&P 500’s 9.2% gain. This has pushed the sector to a 62% premium over the broader market.

According to Bernstein, the surge has been largely supported by fundamentals rather than speculative expansion. Forward earnings per share for the sector have risen 75% since January, with only limited growth in valuation multiples.

However, performance within the sector has diverged sharply. Memory-chip stocks have surged 500% this year, while CPU and optical component makers gained 220%, and GPU and ASIC companies advanced 115%. Demand tied to data centers and fabrication equipment remains strong, though uneven across supply chains.

Valuations climb above broader market

The semiconductor index now trades at a forward price-to-earnings ratio of 34.1, compared with 21 for the S&P 500, highlighting elevated expectations.

Bernstein estimates Nvidia’s adjusted earnings per share could reach $9.19 in 2026 and $12.52 in 2027, implying a multiple of about 25 times at its $315 price target, below the sector average. Broadcom’s $550 target is supported by expectations it could generate $100 billion in AI-related revenue by 2030.

Advanced Micro Devices was upgraded to “outperform,” with Bernstein citing its exposure to both GPU and CPU-driven AI demand. The firm projects AMD could reach $20 in earnings per share by 2028, supported by an early recovery in personal computer chip shipments.

Qualcomm remained at “market-perform” amid weaker smartphone demand, with shipments down 3% year-on-year in the first quarter. Rising memory costs have also pressured handset makers, limiting near-term growth despite Qualcomm’s push into computing markets.

Supply chain risks and crowding concerns

Bernstein maintained positive views on semiconductor equipment firms Applied Materials, Lam Research, and KLA, pointing to continued fabrication expansion. Analog chipmakers Texas Instruments and Analog Devices remain in recovery, with relatively limited data center exposure and valuations of 30 to 40 times earnings.

At the same time, risks are building. Sector crowding has reached record levels, while inventory days have risen above normal ranges. Softer demand in PCs and smartphones could trigger renewed destocking pressures, and any slowdown in server procurement may weaken pricing power for upstream AI chip producers.

Market sensitivity increases as expectations peak

Recent market moves suggest traders are becoming more cautious. U.S. stock futures dropped sharply on June 23, with Nasdaq 100 futures falling nearly 2.6% as crowded AI-related trades began to unwind on valuation concerns.

Bernstein noted that while AI demand remains a powerful driver, much of the expected growth is already reflected in prices. This leaves little room for disappointment, particularly if heavy data center spending fails to translate into anticipated earnings gains.

The firm emphasized that stock selection is becoming more critical than broad sector exposure, as performance gaps widen and risks become more company-specific.

Broader market implications emerge

The semiconductor sector’s outsized gains—more than 100% year-to-date compared with roughly 10% for the S&P 500—underscore a concentration of expectations that could reverse quickly.

Elevated correlations between the Nasdaq and other risk-sensitive assets suggest that any sharp correction in chip stocks may extend beyond the sector. While these relationships can shift, periods of tight linkage have become more frequent, raising the احتمال that weakness in technology sentiment could spill into other markets.

Bernstein concluded that the sector remains fundamentally strong but increasingly fragile, with valuations heavily reliant on sustained AI-driven growth and precise execution across the industry.