Research report interpretation: The semiconductor sector has increased by 155%, Bernstein says NVDA and AVGO are still "absurdly cheap"

Author: Rita

Tide Guide Reading

Bernstein released a quarterly overview of the semiconductor industry on June 23. Core view: AI has become the "only game" in the semiconductor sector, with strong fundamentals, but valuations and crowdedness are at all-time highs. The report also recommends NVDA and AVGO (rated "Outperform the Market"), believing that although they underperformed this year, they are the most core beneficiaries in the AI supply chain, and their current valuations are "absurdly cheap." AMD was upgraded, but QCOM remains cautious due to pressure on its mobile business.

AI Demand Drives Record Gains in the Semiconductor Sector

The Philadelphia Semiconductor Index (SOX) rose 155.6% over the past year, up 106.6% since the beginning of the year. During the same period, the S&P 500 only increased by 9.2%. The premium of SOX over the S&P 500 reached 62%.

This surge is driven by fundamentals, not bubbles. Bernstein's data shows that SOX's forward EPS has increased by 75% from the start of the year to now, while valuation expansion accounts for only a small part.

Within the semiconductor sector, divergence has become exaggerated. From the start of the year to June 22, memory chips rose 500%, CPUs and optical solutions each increased 220%, while GPUs and ASICs only rose 115%. The entire AI supply chain is making money, but the profitability and extent vary. The upstream and downstream of the supply chain benefit the most, as building new production lines requires memory and semiconductor equipment, leading to relatively tight supply. GPUs only increased 115%, despite NVDA holding the vast majority of AI chip market share.

Actual Purchasing Power Under High Valuations

The forward P/E ratio of SOX is now 34.1 times, compared to 21.0 times for the S&P 500, a 62% premium. It sounds expensive, but specific companies tell a different story. NVDA's adjusted EPS forecast for 2026 is $9.19, and for 2027 is $12.52. Based on Bernstein's target price of $315, the P/E in 2027 is 25 times, while the entire sector's forward P/E is 34 times. NVDA is not the most expensive; it is relatively cheap.

Bernstein analyst Stacy Rasgon used a term: "absurdly cheap."

His reasoning is straightforward: NVDA's Blackwell chip series is expected to reach $1 trillion in revenue by 2027. AVGO's situation is similar, with a target price of $550, but if it reaches $100 billion in AI-related revenue by 2030, current valuations look very cheap.

This is why Bernstein rates both companies as "Outperform the Market." Although they underperformed this year, they are the most core links in the AI demand chain. For comparison, Apple’s forward P/E is about 28 times, Microsoft around 30 times, while NVDA is at 25 times. Considering the continuity of Blackwell and Rubin product generations, and AVGO’s monopoly position in switch chips, these valuation discounts seem extremely unreasonable. Capital ignores a core fact: without NVDA and AVGO chips, the entire AI infrastructure cannot operate.

Dual Stories for CPUs, Single Dilemma for QCOM

AMD was recently upgraded by Bernstein to "Outperform the Market." What's the reason? Because AMD not only has opportunities in AI/GPU but also in the proxy AI trend of CPUs. CPU shipments are expected to improve sequentially from Q1 2026, slightly above PC shipments. Bernstein believes AMD's fundamentals are strong enough to support reaching $20 EPS by 2028, with room for the current stock price to rise toward that target.

QCOM, on the other hand, faces a single dilemma. Smartphone shipments declined 3% YoY in Q1 2026, and rising memory chip prices mean higher phone costs, which negatively impacts pricing power for chipset suppliers. Bernstein admits that downgrading QCOM was a "bad decision," but still maintains a "market-perform" rating. The problem is that consumer electronics weakness is now a given, and QCOM will find it hard to find new growth engines. Even if future analyst days reveal new data center stories, compared to AMD’s dual drivers and the structural position of chipmakers, QCOM’s story has limited persuasive power.

Realities of Sub-Sectors

Semiconductor equipment (AMAT, LRCX, KLAC) continues to be favored, with demand for capacity expansion remaining strong. All three are rated "Outperform the Market," with target price increases ranging from 30% to 70%.

The situation for analog chips (ADI, TXN) is more complex. They are indeed in a recovery cycle, with over a year of double-digit growth, but data center business still accounts for only about 10%. TXN and ADI’s P/E ratios are between 30 and 40 times, which is quite expensive. Bernstein gives both a "market-perform" rating, opting to observe.

Crowding and Inventory Risks

Bernstein’s industry sentiment indicator shows that the level of crowdedness in the semiconductor sector is at a historic high. Inventory days have risen again, well above the upper limit of normal historical ranges. Channel inventories have decreased somewhat but remain above average. What does this mean? It indicates that if any signs of downstream demand softening appear, the entire supply chain will face active destocking pressure. PC and consumer segments have already shown weakness, and mobile phones are down YoY. Once inventory pressure spreads to data center procurement, the threat of price wars will become real. Companies near bottlenecks (NVDA, AVGO) will see their pricing power severely weakened.

The strength of AI demand is undeniable, but the current high valuations in the semiconductor sector have already priced in these good news. NVDA and AVGO are relatively cheap, but only if one believes they can meet analyst targets. AMD’s story is attractive but carries execution risks. QCOM has become a forgotten player, with unclear catalysts. Bernstein’s stance is selectively optimistic; at this point, stock picking has become more important than simply betting on the right direction.

Disclaimer

This article is a compilation and interpretation of third-party brokerage research reports by TideResearch. The ratings, target prices, earnings forecasts, and related judgments quoted are solely Bernstein analysts’ opinions, representing only their institutional stance, not TideResearch’s views, nor any investment advice.

Please note three points when reading: 1. Target prices are analysts’ expectations for the next approximately 12 months, predictions rather than commitments, subject to adjustments based on performance and market conditions. 2. Sell-side research reports tend to be optimistic, and some covered companies have investment banking relationships with the broker. 3. The value of the research report lies in its main logic and underlying assumptions, not in any specific target price. Focus on the logic, not just the price.

Markets carry risks; decisions should be made independently. This article should not be used as a basis for buying or selling any securities.

Data sources: Bernstein report (Stacy A. Rasgon et al., June 23, 2026) · Public market data

TideResearch · 2026 June