Goldman Sachs: AI rally in U.S. stocks is exhausted, shift to defensive sectors in H2, bullish on healthcare and European defense.

In the first half of this year, geopolitical tensions and artificial intelligence (AI) frenzy dominated global markets, but as the second half begins, the macro environment and market logic are undergoing a profound shift.

On July 4, Louis Miller, head of Goldman Sachs' Custom Global Equity Basket business, pointed out in his latest market strategy report that as global growth expectations rebound from a trough and inflationary and interest rate pressures ease, the macro environment in the second half will greatly favor a "selective broadening" of market breadth, with capital flowing out of extremely crowded tech giants to seek new low-valuation opportunities.

Goldman Sachs warns that the AI and momentum trades that dominated the first half have shown signs of exhaustion. Historical seasonal patterns indicate that these popular trades may enter a "summer dormancy period" in July.

In fact, the momentum factor has just experienced its largest drawdown since the first quarter of 2023, with AI concentration in the S&P 500 exceeding 50% and hedge funds' momentum exposure at an extreme 92nd percentile over the past five years. Such extreme crowding means that sharp corrections and liquidation cascades are highly likely in the short term.

Against the backdrop of capital shifting from high to low, Goldman Sachs strongly recommends that investors pivot to defensive sectors and "non-AI related" compounders in the second half. Among them, the healthcare sector is Goldman's top pick, especially bioprocessing and large European pharmaceutical companies with strong merger and acquisition potential. Additionally, the underperforming European defense sector in the first half, due to depressed valuations and reset earnings expectations, presents an excellent entry point.

On the tactical front, Goldman expects lagging cyclical stocks, software, consumer, and real estate sectors to experience catch-up gains. To navigate market volatility from momentum trade unwinding, Goldman advises investors to use its "ex-AI" index basket for hedging, retaining broad market exposure while avoiding short-term downside risks from crowded trades.

AI and Momentum Trades Face "Summer Dormancy," High Crowding Triggers Pullback Risk

Looking back at the first half, AI trading was undoubtedly the market's main theme. U.S. memory chips, new cloud services, AI inference beneficiaries, optical networks, and data center sectors have all posted returns exceeding 100% since the start of the year. Semiconductor and related supply chains in Asia and Europe have also doubled.

However, Goldman Sachs points out that this frenzy is facing a reality check. As the second-quarter earnings season approaches, market focus will shift back to fundamentals, intensifying stock differentiation.

Recently, momentum trades (buying past winners, selling past losers) have suffered their worst two-day selloff since 2022, with consecutive liquidation waves exceeding 5%. This momentum drawdown is causing collateral damage to AI trades, and Goldman Sachs' AI long-short hedging strategy has also recorded its worst single-day performance recently.

Over the past two and a half years, the AI sector has experienced multiple phased corrections. Historical patterns and seasonal factors both point to July as a "breather period."

More concerning is that momentum factor exposure in Goldman Sachs' Prime Book remains at the 92nd percentile over a five-year horizon. Extremely crowded positions mean that once unwound, the impact will be amplified.

AI concentration in the S&P 500 has exceeded 50%, and AI weighting in European markets has doubled from 2023 levels, making systemic risk non-negligible.

In summary, given that summer typically brings weaker U.S. stock market performance, coupled with the need for broader market participation and high factor volatility, Goldman Sachs believes the AI sector faces further selling risk in the near term.

While AI surges ahead, many sectors remain under significant pressure.

Goldman Sachs summarizes several underperforming asset classes in the first half, such as those considered "vulnerable to AI disruption" (e.g., traditional software stocks in the U.S., Asia, Japan, and Europe) and consumer discretionary sectors (e.g., European luxury goods, U.S. low-income consumption).

Goldman believes the most likely outcome for the second half is: lagging cyclical stocks, software, consumer, or real estate sectors experiencing "catch-up rallies." This rotation of capital will support the overall performance of major indices, but will also cause pain for portfolios heavily weighted in momentum tech stocks. Therefore, finding sectors where earnings expectations have been sufficiently downgraded and valuations are attractive for "selective broadening" is the core strategy for the second half.

Healthcare: The Ultimate "Non-AI" Compounder, M&A Cycle Begins

Among the many defensive sectors, Goldman Sachs has high hopes for healthcare, calling it the "ultimate non-AI compounder." This logic is already playing out, with the broad U.S. healthcare sector, European pharmaceuticals, and global drug discovery data segments all rebounding sharply from year-to-date lows.

Goldman emphasizes that the healthcare sector's future excess returns will be driven by two core catalysts:

First, structural growth areas such as global bioprocessing, which is not only one of Goldman's highest-conviction healthcare trades but also a real beneficiary of AI-driven productivity gains.

Second, the M&A cycle of large pharmaceutical companies. European pharmaceutical valuations are currently about 10% below their historical relative market premium. More importantly, large pharma companies have ample balance sheet capacity and face an imminent "patent cliff," which has triggered a strong M&A cycle that will significantly boost sector earnings.

Beyond healthcare, Goldman has also keenly identified a turning point in the European defense sector. Defense stocks underperformed in the first half, but a recent rebound has caught market attention. Goldman believes this rebound is sustainable because current positioning in the sector is very light, earnings expectations have been reset, and depressed valuations set an extremely low bar for second-quarter earnings. The European defense sector is rebounding from year-to-date lows and still has about 12% relative upside to catch up with the broader market.

As for AI-related themes, Goldman is not outright bearish but advocates "buying the dip." In particular, regarding U.S. hyperscalers, Goldman recommends building positions before the earnings season. Meta's recent announcement on cloud business has already triggered the first wave of gains in this sector, and Goldman expects strong second-quarter EPS to drive further upside.

Given the potential for heightened market volatility in the second half, Goldman Sachs advises investors to implement short-term tactical hedging. Due to the excessive AI weighting in the S&P 500, traditional index shorting tools may inadvertently hit "structural winners" with solid long-term fundamentals.

To address this, Goldman has partnered with S&P to launch the SPXXAI (S&P ex-AI index) and recently introduced the "GSXEXXAI" (Ex-AI Europe Market Basket) for the European market, as AI concentration in Europe is also surging (now twice its 2023 index weighting).

These tools provide investors with a highly liquid, lower-volatility hedging alternative, allowing them to safely capture the benefits of broadening market breadth while avoiding the risk of crowded tech stock corrections.

Risk Disclaimer

        Market risk exists; investments should be made with caution. This article does not constitute personal investment advice and does not consider the specific investment objectives, financial situations, or needs of individual users. Users should consider whether any opinions, views, or conclusions in this article are suitable for their specific circumstances. Investment based on this content is at your own risk.