Goldman Sachs Interview: Reassessing HALO Value, Anchoring AI Investment New Cycle | Investor Insights

Ask AI · How does the AI lobster farming boom reflect China’s unique advantages in AI commercialization?

Jinjin Liu Chief China Equity Strategist at Goldman Sachs

Graduated from the London School of Economics with a Master’s in Finance, CFA, and CPA. Joined Goldman Sachs Asia-Pacific Strategy Research team in 2004, responsible for research in Greater China and North Asia.

“Lobster Farming” marks an important milestone in AI application China has more opportunities in AI implementation

First Financial: Recently, the “lobster farming” craze driven by OpenClaw has spread from industry circles to the capital markets. From your perspective, do you see this as a new paradigm or turning point in AI execution, or just a short-term, phase-specific hype?

Jinjin Liu: I think both. Looking back, the turning point associated with DeepSeek over a year ago proved that Chinese tech companies are capable of developing highly competitive models on a global scale. But this time, the “lobster farming” boom is a significant milestone in AI application and commercialization. From this perspective, it’s a very important inflection point for AI investment, which can boost confidence in many related stocks within the AI ecosystem and overall landscape. In the coming period, investors’ confidence in their commercialization and monetization will likely increase substantially.

More importantly, I believe China’s AI landscape, ecosystem, and our understanding of AI development differ greatly from overseas. Abroad, their business models tend to be high-investment, high-margin; domestically, many models are open-source and low-cost, so we see more opportunities for monetization and deployment in China.

Optimistic about investment opportunities in AI-related power and infrastructure stocks

First Financial: Indeed, the “lobster farming” trend has sparked a rally in A-shares related to computing power, chips, and related applications. Which sectors or segments do you think have sustainable performance potential, and which might just be emotional hype?

Jinjin Liu: We all know the entire AI supply chain is a very long ecosystem. Recently, in our reports, we’ve spent considerable time exploring this. Briefly, we conducted a major study mapping about 3,000 AI-related companies worldwide, categorized by region—Europe, America, China, and North Asia—and by sector, roughly aligned with Nvidia CEO Jensen Huang’s five major segments.

The top layer includes power and semiconductors; the middle involves models and foundational monitoring; the bottom layer is applications. Based on this framework, we believe that Chinese listed AI-related stocks in power and infrastructure are highly competitive globally. Therefore, from profit growth and current market valuation perspectives, these sectors may offer more investment opportunities.

The role of some “old economy” stocks in the AI ecosystem is significant

First Financial: As you mentioned, AI development relies heavily on infrastructure. Recently, Goldman Sachs proposed the concept of HALO assets (Heavy Assets Low Obsolescence)—simply put, physical assets that AI cannot replace and that are also indispensable. This concept has sparked Wall Street’s investment enthusiasm. Can you explain how HALO assets differ fundamentally from traditional value stocks, cyclical stocks, and dividend assets?

Jinjin Liu: The HALO concept has gained popularity among investors. I think there’s a misconception that AI is mainly about intangible assets, but they overlook heavy-asset companies like power and infrastructure. In our HALO portfolio, we include many stocks considered part of the old economy, yet they are crucial components of the AI ecosystem.

For example, power transmission and storage are vital parts of the supply chain. Our HALO portfolio contains traditional industries that still play a key role in AI revenue and development.

Key resource minerals and some manufacturing sectors also hold HALO value

First Financial: Are the value and opportunities of HALO assets long-term? Which industries might truly withstand cycles?

Jinjin Liu: Currently, global energy security and supply are major concerns, so our HALO portfolio includes energy-related stocks. Additionally, resource stocks, including upstream critical minerals, are part of the portfolio.

Another main theme of HALO is heavy-asset manufacturing. We’ve selected relevant stocks in cyclical sectors and manufacturing for inclusion. Overall, AI remains a very important investment theme in the near future, and we can explore related stocks from multiple angles.

High valuation and unfulfilled profits are potential risks for HALO

First Financial: Are there any potential risks associated with HALO assets? When should investors reduce their allocations?

Jinjin Liu: We see three main risks. First, HALO stocks performed very well over 3-6 months, significantly outperforming the broader market and some AI stocks, so a correction risk exists. Second, the market’s profit expectations for HALO-related stocks are already quite high. In the coming weeks, as companies release last year’s annual reports and Q1 results, if actual earnings growth falls short, that could be a major risk. Third, valuations for HALO sectors are above historical averages, which warrants caution. However, we remain optimistic about the earnings growth of A-shares, HALO assets, and the broader AI theme. We expect the overall market profit growth to reach about 13-14% this year. If companies meet these expectations, the upward or slow bull trend in A-shares should continue.

HALO and AI are complementary themes in investment portfolios

First Financial: For investors, is HALO—heavy assets and solid foundation—complementary, hedging, or cyclical relative to the “lobster chain” new paradigm of AI execution? Do you prefer embracing HALO’s certainty or participating in the growth potential of the “lobster chain”?

Jinjin Liu: From a fundamental perspective, it’s like many investors asking whether to choose A-shares or Hong Kong stocks. I believe it’s not a binary choice. We’ve always been optimistic about both markets, as they offer different allocation options and flexibility.

Regarding HALO and AI themes, I see them as complementary. For diversification, both are important. We particularly favor power and infrastructure stocks within AI, and certain sectors within HALO. From a diversification and risk-spreading standpoint, both are key components of a balanced portfolio.

AI proliferation could boost listed companies’ profits by about 3% annually

First Financial: We’ve observed that since the DeepSeek moment in 2025, the stock prices of Chinese AI-related listed companies have increased by roughly 50%. AI has thus changed our capital market investment narrative. Looking ahead, as AI further spreads and deepens, what specific impacts will it have on the profitability, capital inflows, and valuation levels of A-shares?

Jinjin Liu: We’ve conducted a mid- to long-term projection. As AI develops, it will enhance productivity and spawn new business models. We believe this will lead to about 3% annual profit growth for listed companies over the next decade. Currently, AI-related companies account for roughly 35% of total market profits; we expect this to rise to about 45% in ten years. So, AI is a significant driver of profit growth for listed firms.

Regarding valuation levels, we benchmarked against the US market. Since ChatGPT’s launch, the average valuation of the US AI ecosystem has increased by 15-20%. We believe that as AI further takes root and develops in China, Chinese listed companies will also see valuation increases of about 15-20%.

Goldman Sachs Disclaimer:

This content is for professional investors only. Markets carry risks; invest cautiously. Under no circumstances does the information or opinions herein constitute investment advice to any individual.

Author: Zou Ting, Zhao Yiwen, Kong Fantian

Editor on duty: Su Xiao