Goldman Sachs Sees Rotation and Diversification Opportunity After China AI Hard-Tech Pullback
China’s AI hard-tech sector rose an average of 33% in the first half before undergoing a sharp pullback from its June peak.
China’s AI hard-tech sector rose an average of 33% in the first half before undergoing a sharp pullback from its June peak. The STAR50, ChiNext and CSI 1000 indexes have each fallen more than 20% from their highs.
At the same time, H-share internet and soft-tech stocks have significantly underperformed, with the relative-return gap between the two groups exceeding 100%. That has brought the gap to an extreme level last seen during the pandemic-stimulus rebound in early 2021.
Goldman Sachs said the pullback has removed a considerable portion of speculative long positioning, valuation bubbles, leverage and crowded trades. However, it expects market volatility to remain elevated in the short term and highlights considering uncorrelated assets both inside and outside China’s AI ecosystem for diversification.
Key data points include a peak forward P/E of 50 times for the median STAR50 stock and about 26 times for the median ChiNext stock. Outstanding margin financing fell from Rmb3 trillion to Rmb2.6 trillion, but leverage concentration in the AI hard-tech sector is at a record high: Stocks in the top 10% by leverage account for 30% of outstanding margin financing.
Goldman Sachs’ bottom line is that the speculative premium in AI hard tech has been partly cleared, but near-term volatility is unlikely to fall quickly. It points to diversification across H-share internet and soft-tech stocks, policy beneficiaries, domestic-demand-oriented AI stocks, and oversold non-tech stocks with upward earnings revisions.
Potential beneficiaries include leading H-share internet and soft-tech companies such as Tencent and Meituan, stocks benefiting from 15th Five-Year Plan policies, domestic-demand-oriented AI companies, and oversold non-tech stocks with upward earnings revisions. AI hard-tech stocks remain exposed to pressure because absolute valuations are still high and leverage concentration is at a record level.
Current share prices have partly reflected the pullback, but valuation risk is still far from fully discounted or removed; it has not yet been fully priced in.
Potential catalysts include capital-expenditure guidance and AI monetization road maps from large Chinese and U.S. cloud providers during the upcoming earnings season; whether market breadth, return concentration and trading-volume concentration in A shares decline; whether margin financing can continue falling toward historical norms; and whether recent net purchases by China’s national team continue to expand.
At the same time, H-share internet and soft-tech stocks have significantly underperformed, with the relative-return gap between the two groups exceeding 100%. That has brought the gap to an extreme level last seen during the pandemic-stimulus rebound in early 2021.
Goldman Sachs said the pullback has removed a considerable portion of speculative long positioning, valuation bubbles, leverage and crowded trades. However, it expects market volatility to remain elevated in the short term and highlights considering uncorrelated assets both inside and outside China’s AI ecosystem for diversification.
Key data points include a peak forward P/E of 50 times for the median STAR50 stock and about 26 times for the median ChiNext stock. Outstanding margin financing fell from Rmb3 trillion to Rmb2.6 trillion, but leverage concentration in the AI hard-tech sector is at a record high: Stocks in the top 10% by leverage account for 30% of outstanding margin financing.
Goldman Sachs’ bottom line is that the speculative premium in AI hard tech has been partly cleared, but near-term volatility is unlikely to fall quickly. It points to diversification across H-share internet and soft-tech stocks, policy beneficiaries, domestic-demand-oriented AI stocks, and oversold non-tech stocks with upward earnings revisions.
Potential beneficiaries include leading H-share internet and soft-tech companies such as Tencent and Meituan, stocks benefiting from 15th Five-Year Plan policies, domestic-demand-oriented AI companies, and oversold non-tech stocks with upward earnings revisions. AI hard-tech stocks remain exposed to pressure because absolute valuations are still high and leverage concentration is at a record level.
Current share prices have partly reflected the pullback, but valuation risk is still far from fully discounted or removed; it has not yet been fully priced in.
Potential catalysts include capital-expenditure guidance and AI monetization road maps from large Chinese and U.S. cloud providers during the upcoming earnings season; whether market breadth, return concentration and trading-volume concentration in A shares decline; whether margin financing can continue falling toward historical norms; and whether recent net purchases by China’s national team continue to expand.