Entering the second half of 2026, the global capital market is standing at a delicate turning point. Recently, Allianz Asset Management, Manulife Asset Management, Lombard Odier Asset Management, Morgan Stanley and other foreign institutions have successively released their latest views on the market for the second half of the year. Combining the views of various institutions, geopolitical risks and AI applications are still the two main themes throughout the year, the global economy as a whole maintains resilience, but the 'K-shaped differentiation' feature is clearly visible in both Chinese and US markets. More notably, against the background of the expected slowdown in AI capital expenditure growth and the re-divergence of global monetary policies, foreign institutions have generally given a 'cautiously positive' judgment on the second-half bond market, equity style switching, and Chinese asset allocation; the market investment logic is quietly shifting from extreme differentiation in the first half of the year to a new era of balanced allocation.
Capital Expenditure Feast Peaking? The 'Shovel Sellers' Era Enters the Second Half
In the past two years, AI has undoubtedly been the strongest engine of the global capital market. Super-large cloud service providers (Hyperscalers) have invested in data centers, chips, and computing infrastructure on an unprecedented scale, driving the expansion of the entire industrial chain such as semiconductors, power equipment, and industrial manufacturing. The Asset Allocation Committee of Lombard Odier stated bluntly in its third-quarter outlook that the AI capital expenditure cycle continues to accelerate, evolving into the core engine driving this macro cycle, and extending the expansion momentum to fields such as public utilities and heavy industry.
However, the surging growth rate is slowing down. According to statistics from Allianz Asset Management, the proportion of capital expenditure of large US technology companies to operating cash flow may reach a high of 98% by the end of this year; the annual growth rate of capital expenditure is expected to slow down from 78% this year to 41% next year and 9% in the following year. Li Changfeng, head of market strategy at Allianz Asset Management, uses the 'gold rush' to describe the current market situation: the market has not yet found the best application method for the AI 'gold mine', but the companies that supply chips, electricity, and cloud computing—the 'shovel sellers'—are the first to profit. However, as the base continues to rise, 'the speed of fundamental upgrades will slow down, this is not a recession, but the upward slope is no longer as steep as before.' In other words, the AI industry trend has not reversed, but the peak of marginal momentum may have passed, which is the core meaning of the 'capital expenditure inflection point'.
K-shaped Differentiation Reaches its Limit, Asset Performance Converges
In the first half of 2026, the global capital market showed distinct structural characteristics: the AI industry chain and technology sector continued to attract capital, while traditional industries and cyclical assets were under overall pressure, and the trends of the two were severely divergent. China Universal Fund recently released its third-quarter allocation ideas, pointing out that the extreme differentiation in the first half of the year is essentially the deep reshaping of the traditional industrial order by the AI technology revolution; entering the third quarter, the core change lies in the correction of liquidity expectations, and asset differentiation is expected to gradually converge towards balance.
Zhu Liang, investment director of Allianz Asset Management, reminded that as of the end of June 2026, the valuation dispersion within A-shares once again reached a historical extreme, 'this kind of valuation divergence is like a rubber band, it cannot be stretched forever'. He further compared the AI industrial structure of China and the US: the proportion of capital expenditure of related companies in the US has risen to about 92% or even higher, while that of related companies in China is only about 50%, which is relatively healthy and more sustainable; China's AI follows a high-cost-performance model route, the capital expenditure expectation is relatively stable, and also leaves room for subsequent valuation recovery.
Comparison of China-US AI Investment Structure
- US: Capital expenditure as a proportion of cash flow is about 92% or higher, with high input intensity and greater pressure of marginal slowdown.
- China: Proportion is about 50%, relatively healthy, following a high-cost-performance route, with sustainability.
- Overall: The annual growth rate of AI capital expenditure is expected to slow from 78% in 2026 to single digits in 2028, with a significant slowdown in expansion speed.
Global Central Bank Policy Divergence, Risk of US Treasury Curve Steepening Rising
Monetary policy is another key variable in the second half of the year. The US Federal Open Market Committee (FOMC) meeting in July maintained the federal funds target rate in the range of 3.5% to 3.75% unchanged, but there were three rare dissenting votes, with three regional Fed presidents all inclined to raise rates by 25 basis points, showing that the Fed's internal concerns about inflation have significantly deepened. Manulife Asset Management analysis pointed out that the new chairman repeatedly reiterated his stance on controlling inflation but refused to provide forward guidance, and the signals released after the meeting were relatively mild, forming a subtle gap with the previous tight pricing in the market.
Looking at the global picture, central bank policies continue to diverge: Europe, Australia and Japan are in an interest rate hike cycle, the US is on hold, and China is偏向宽松. The divergence in interest rate directions is directly reflected in the bond market. Manulife Asset Management observed that the US Treasury yield curve has significantly steepened, with little change in the 2-year variety, and the 30-year yield may break through 5.2%; Allianz Asset Management expects that the US Treasury curve still has room to steepen, and the overall is a bull market of falling yields, but the decline in the short end will be greater than the long end. This means that the tightening of the financial environment comes more from market forces than the Fed's active rate hikes.
Second Half Investment Logic Switching, Value of Balanced Allocation Rising
In the face of the shift from 'building infrastructure capacity' to 'applying AI to reduce costs and increase efficiency' in AI investment, Xing Ziqiang, chief economist of Morgan Stanley China, proposed the 'HALO' asset concept—heavy assets, low elimination rate, advocating for a broader AI ecosystem chain layout in the second half, including energy transformation and other fields. He emphasized that the token cost of China's domestic large models is only about one-tenth of that in the US, and it can lease digital infrastructure to technology enterprises through the national computing power network, hoping to replicate the scale advantage of the mobile internet era.
Overall, the consensus of foreign institutions on the AI track is: there is fundamental support, but volatility will significantly increase in the second half. Lombard Odier maintains an overweight view on global stocks and US large-cap stocks, but 'is tactically cautious and strategically optimistic', and warns that AI stock positions are concentrated and there is a risk of momentum exhaustion. Allianz observed that capital is shifting from overweight tech hardware to underweight, and the recent strong performance of Hong Kong stocks is the result of rebalancing capital inflows.
Key Points for Allocation Observation in the Second Half
- Liquidity expectation correction: Synchronized easing of liquidity in China and the US is expected to support the valuation recovery of non-tech sectors.
- Performance realization capability: The AI investment logic has shifted from 'how much money to spend' to 'how much money to earn', and the progress of converting orders into revenue is the key.
- Interest rate curve trend: If the long-term US Treasury yield continues to rise, it will test the resilience of high-growth growth stocks.
- Chinese asset opportunities: City commercial banks, innovative drugs, precious metals, etc. have allocation value to hedge tech volatility.
At the current point, the K-shaped differentiation of the first half of 2026 may have been fully interpreted by the market. Entering the third quarter, factors such as geopolitical easing, oil price回落, and employment weakness are reshaping liquidity expectations, coupled with the driving forces of new capital inflows and low economic recovery, asset differentiation is expected to converge. In this process, the importance of balanced allocation has significantly increased, and the rebound opportunities of previously lagging assets are worth paying attention to; but AI's core position as a long-term industry trend has not been shaken. Every adjustment and convergence in the market often brews a new round of structural reshaping. Investors should remain clear-headed and patient to seize opportunities in the changing situation.
