Newsletter commentary May 2026

Market Commentary

In May, the A-share market remained highly polarized. Apart from the Communications and Electronics sectors, which delivered strong performance, most industries underperformed, The strength in Communications was largely driven by the continued rally in optical module companies, supported by growing overseas demand for increasingly sophisticated optical connectivity solutions. Meanwhile, the Electronics sector benefited from ongoing optimism surrounding domestic semiconductor development, driven by technological progress, expectations for the listing of the two major memory manufacturers, and the expanding opportunities arising from future large-scale capacity expansion.

The current market environment reflects the unique complexity and diversity of China's equity market. Globally, markets such as Japan and Korea have been significant beneficiaries of AI-related hardware demand originating from the United States. With limited trade barriers and relatively slow industry capacity expansion, these markets have enjoyed both volume growth and pricing power, supported by a compelling long-term growth narrative. In contrast, many economies less directly linked to U.S. AI demand have remained largely influenced by the inflation-interest rate narrative, resulting in more moderate market performance.

China occupies a unique position. Certain sectors are directly exposed to U.S. AI-related demand, although trade barriers limit the breadth of the benefits. For those directly involved companies, earnings growth has been substantial, yet the spillover effect to the broader market remains limited. At the same time, China has its own AI development story. Continuous technological progress and capacity expansion are creating recurring investment opportunities. While the earnings visibility resulting from domestic AI development may be less predictable than export-oriented AI beneficiaries, the long-term direction remains clear. As a result, share price behavior may also differ, with potentially greater volatility.

Beyond AI-related sectors, China's market continues to contain a large number of companies linked to consumption, real estate, chemicals, and non-ferrous metals. These sectors have faced pressure from soft domestic demand, while fluctuations in expectations surrounding the Strait of Hormuz have introduced additional volatility, particularly for commodity-related industries.

Globally, market concentration has become increasingly extreme. In many markets, both trading activity and market capitalization have reached historically elevated levels in a narrow group of sectors and companies. Investors appear increasingly willing to extrapolate short-term earnings trends far into the future and apply terminal-value-based valuation frameworks. This shift is evident across a wide range of research reports. Discussions around long-term competitive dynamics, industry structure, and sustainability have become less prominent. At present, asking too many questions about competitive positioning may hurt investment returns; however, there may also come a time when not asking such questions becomes equally costly. The challenge, of course, is determining when that transition occurs.

The market also appears increasingly insensitive to interest rates. In China, liquidity remains abundant and interest rates remain low. Yet in many overseas markets, elevated rates have not meaningfully constrained valuations. One possible explanation is that current earnings growth is so strong that a modest increase in discount rates appears insignificant by comparison. The gravitational pull of interest rates may still constrain traditional industries, but its influence on rapidly growing sectors appears diminished. That said, changes in liquidity conditions tend to exert their influence with a lag, and their impact should not be underestimated.

Daily turnover in China's equity market has remained close to USD 450 billion, and market behavior continues to become increasingly momentum-driven. Several episodes during the month demonstrated heightened market sensitivity to short-term developments. Persistent outflows from broad-based index ETFs have also affected market style. Traditional sectors with weaker earnings trends have continued to face capital outflows, which in turn have pushed more capital toward popular growth sectors. Meanwhile, incremental off-exchange capital inflows appear limited, leaving the market increasingly dependent on margin financing and the emergence of new thematic opportunities. Broad ETF outflows may, in some sense, also act as a balancing mechanism by limiting excessively rapid market appreciation.

Globally, economic recovery has become a more prominent investment narrative this year. As markets increasingly conclude that ongoing geopolitical conflicts are unlikely to significantly derail growth, discussions surrounding overseas economic recovery have gained traction. In China, however, domestic demand has weakened since April, with both consumption and investment remaining soft. This has been negative for cyclical sectors and consumer-related industries. The combination of high savings, weak investment demand, and abundant liquidity has resulted in financial markets outperforming the real economy. Weak fundamentals combined with loose liquidity conditions have further contributed to increasingly extreme market behavior.

Looking ahead, the sustainability of the global AI investment cycle remains the most important driver for capital markets. Based on various industry conferences and product releases, the path of technological advancement appears increasingly clear. The key question is whether large language models can continue to evolve into systems capable of performing a broader range of tasks, effectively replacing or augmenting human physical and cognitive labor through ever-greater deployment of computing power and electricity. Until the resulting productivity gains encounter constraints arising from existing economic and social structures, the productivity narrative may continue to dominate. Predicting when such constraints may emerge is extremely difficult; for now, observation remains our most reliable guide.

On the demand side, there are also emerging developments worth monitoring. Recent reports have highlighted concerns over the cost of token consumption, while Google has begun large-scale fundraising efforts to support future capital expenditures, despite being one of the strongest cash-generating cloud service providers globally. Its decision to raise capital may not be an isolated case. At the same time, DeepSeek, Tencent and other large model providers have significantly reduced pricing, in some cases by as much as 90%. For model providers whose marginal costs are not close to zero, the future economics of tokens will depend on whether competitive advantages come from cost leadership, performance leadership, or a combination of both. The industry remains far from reaching a stable equilibrium and continues to evolve rapidly.

Forecasting remains inherently difficult. Our approach is therefore to observe carefully and focus on opportunities that we can understand and explain. As long as model capabilities continue to advance rapidly and the underlying economic structure has yet to undergo significant adjustment, the broader AI value chain appears relatively well supported. Within that framework, identifying areas where market expectations and fundamentals diverge significantly may offer particularly attractive risk-reward opportunities.

At the same time, we continue to emphasize portfolio diversification. In an environment where capital is increasingly concentrated in a small number of favored themes, diversification can at times feel like a headwind. However, without a strong informational advantage that would justify highly concentrated positioning, maintaining a diversified portfolio remains our preferred approach. Over time, we believe the intrinsic value of companies in less favored industries will continue to be recognized.

From a longer-term perspective, the AI revolution is reshaping the value of many industries. AI increasingly replaces or enhances both human physical and intellectual labor. Yet the fundamental nature of the economy remains unchanged. Economic activity is ultimately a circular system of production, income, and consumption. One person's consumption remains another person's income. While AI can dramatically improve productivity, how the benefits of that productivity are distributed across society may ultimately become one of the most important questions for investors in the years ahead.