A recent research report outlines its perspective for the second half of 2026, noting that global markets in the first half have moved towards extreme divergence, with AI technology as the sole core theme. This structural market behavior is a direct reflection of the "K-shaped divergence" in the credit cycles of the US and China, characterized by strong technology sectors and weak domestic demand. Looking ahead, whether this K-shaped divergence continues or reverses hinges on two key factors: the sustainability of the tech trend itself and whether any external variables might disrupt it. Regarding asset allocation, technology remains the primary focus, with market crowding not seen as a reason for reversal. US Treasuries offer a favorable balance of probability and potential return, the AI theme has a high probability of success, while the Hang Seng Tech Index offers high potential return. The report provides specific forecasts: 1) The year-end target for the S&P 500 is further raised to 7800-8000, implying a 3%-6% gain from current levels, maintaining a positive stance. 2) US Treasuries still present near-term trading opportunities; if tensions with Iran ease in Q2, expectations for Fed rate cuts could return, corresponding to a long-term yield range of 4%-4.2%, otherwise yields may hold at 4.2%-4.4%. 3) The US dollar is expected to trade within a range (96-98). 4) Gold may need to wait for US Treasury yields to decline in the near term, though its long-term uptrend remains intact, with $5500 being a significant level. 5) Upside for the Hong Kong market indices is seen as limited, with the Hang Seng Index benchmark at 27000-28000, while the Hang Seng Tech Index has room for valuation recovery. 6) Japan is favored over Europe, and emerging markets may continue to diverge internally, with economies benefiting from the AI supply chain remaining attractive but subject to volatility risks.
Root Cause of Extreme Markets: Converging K-Shaped Divergence Between the US and China
The global market in 2026 is characterized by extreme divergence. Setting aside geopolitical disruptions, the core theme is singularly focused on AI technology—either the US-led AI supply chain (involving US stocks, South Korea, Japan, Taiwan, and the offshore-linked segments of mainland China) or China's domestic AI initiatives (semiconductors, equipment substitution, large language models, etc.). Sectors outside this chain, such as domestic consumption in A-shares or the broader Hong Kong market, have significantly underperformed. The underlying logic is straightforward: follow the direction of credit expansion, regardless of consensus; sectors without credit expansion, no matter how long they have been stagnant, do not present a sufficient condition for allocation. The current extreme divergence across global markets, including those in the US and China, is a direct manifestation of the K-shaped divergence in their respective credit cycles. Despite facing different challenges at the end of last year—one with concerns over stagflation or recession, the other with low inflation—both economies have converged towards a K-shaped economic and market structure in 2026: strong technology and weak domestic demand. Technology thrives in the US due to continuous industrial catalysts and accelerating private AI investment acting as a "mid-air refueling," while in China, policy support and fiscal倾斜 also play a role. Weak domestic demand in the US stems from high interest rates, and in China from subdued income expectations and the tapering of subsidies. In aggregate, the strength in technology has masked the weakness in domestic demand, providing robust support for growth and markets. For instance, AI investment contributed over half of the US Q1 GDP growth. While the tech sector propelled the Wind All-A Index to new highs, the consumer sector retreated to pre-924 levels, and household credit growth turned negative for the first time.
In essence, both the US and China are engaged in a "race": a race between strong tech and weak domestic demand, and a race between tech earnings (the numerator) and valuations (the denominator). Assessing the future direction involves determining which force is moving faster: 1) To what extent can strong tech continue to offset weak consumption, otherwise growth pressures may emerge. 2) To what extent can tech profits, revenue, and cash flow continue to offset high valuation multiples, otherwise correction pressures may arise. 3) Whether AI-generated operating cash flow can continue to support expanding capital expenditures, otherwise bubble risks may increase. 4) Whether AI's efficiency gains and growth stimulation outpace its potential job displacement, otherwise long-term concerns may mount. 5) Whether AI's effect on lowering wages and goods prices is more pronounced than its push on electronic component prices, otherwise inflation risks may surface.
Outlook for US and China Cycles: Existing Trends May Widen Divergence
Looking to the second half of the year, the key question is whether this structural market behavior will persist or reverse. This fundamentally depends on two things: first, how far the tech trend itself can go, and second, whether any external variables might interrupt it (such as new growth drivers diverting funds or external shocks disrupting the current trend).
First, regarding the sustainability of the tech trend: it is not yet a full-blown bubble but will advance amidst volatility; focus on penetration rates and cash flow. For the US, the view is that the current AI wave has not reached bubble proportions, resembling more the 1998-1999 period based on demand, investment, capability, and valuation metrics. However, with operating cash flow already depleted by capital expenditures, future investments may rely more on external financing unless further application and monetization breakthroughs occur. This naturally raises market questions, especially since over 90% of US funding comes from the private sector, making returns and cost crucial. Each upcoming earnings season will be a critical test. China's situation differs somewhat in terms of funding sources and cost tolerance. Beyond segments closely linked to US AI, roughly half of domestic funding comes from the government, which can temporarily downplay return requirements and cost pressures in the short term. Therefore, if concerns over the sustainability of overseas AI investment grow, it may反而 highlight the relative resilience of domestic substitution. From an industry perspective, historical parallels with smartphones and new energy suggest penetration rates and free cash flow can serve as indicators for inflection points. First, the main upward浪 for domestic emerging industries in China typically peaks around a 20% penetration rate, with excess returns明显 declining after 25%; current AI penetration in China is estimated at 16-19%. Second, when the ratio of industry free cash flow to revenue falls to around -10%, excess returns tend to conclude; currently, major Chinese头部 companies still maintain positive free cash flow, with operating cash flow sufficient to support current capital expenditure levels.
Second, regarding potential external disruptions: from a US and global perspective, the primary risk stems from global monetary tightening and rising interest rates, which in turn depend on oil prices and the resolution of tensions with Iran. Persistently high oil prices could pressure central banks to hike rates, impacting not only valuations but also capital expenditure by raising financing costs. From China's perspective, with policy acting as a major financier, funding cost is less of a concern. Therefore, under the current growth mix, a significant weakening in external demand might compel policymakers to shift focus from technology to刺激 domestic demand to counter potential growth downside. It's evident that the current scenario (continued tech trend + relatively high oil prices) would likely further widen internal divergence within both the US and China, rather than lead to convergence. High oil prices elevate inflation in both countries but with differing effects: 1) For the US, high nominal rates and rising inflation benefit the high-return tech sector by lowering real financing costs but offer limited relief to sectors like households and real estate with lower returns, thus widening the internal gap. 2) For China, declining nominal rates and rising PPI push real rates for the tech and external demand sectors lower or even negative, entering a reflation phase. However, falling CPI, particularly due to lower pork prices,反而 raises real rates for the consumption端, creating a "quasi-stagflation" effect, exacerbated by already low investment returns in real estate and consumption. This situation is difficult for monetary policy to address, and with strong external demand, the urgency for fiscal policy to significantly ramp up or tilt towards consumption is reduced, thus also widening divergence.
Future variables lie in US Treasury yields and China's external demand, both closely tied to the Iran situation and oil prices, impacting US and China cycles in opposite directions: 1) Higher oil prices are relatively more favorable for China's defensive and domestic demand sectors. A prolonged Iran standoff leading to sustained high oil prices could force US yields higher, impacting US equity valuations and corporate investment, and also weakening China's external demand, creating near-term growth pressure. This might ultimately push Chinese policy towards stimulating domestic demand, while US policy could face stagflationary constraints. After a potential broad-based correction, focus could gradually shift to China's defensive and domestic demand plays. 2) Lower oil prices are relatively more favorable for a US domestic demand recovery and would also support China's cyclical and export sectors. If oil prices retreat, pressure on US yields could at least ease, potentially reviving Fed rate cut expectations. This could lead to温和 improvement in US domestic demand sectors like real estate and consumption. Meanwhile, China's external demand wouldn't face major challenges, likely allowing domestic policy to remain focused on technology. In this scenario, the tech trend might continue alongside valuation volatility, but cyclical and broad external demand sectors could gradually gain attention. Growth stocks suppressed by high US yields, such as Hang Seng Tech and innovative pharmaceuticals, could see valuation recovery. Currently, the base case leans towards the latter scenario evolving over the medium term, with a systemic escalation not being the基准. This is particularly considering the pressure from low approval ratings ahead of the mid-term elections, where inflation, US stocks, and Treasuries are critical for votes. For the Fed, if oil prices fall back to $80-90/barrel in the second half, US inflation could gradually decline below 3% in Q4 due to high base effects, providing some room for rate cuts.
Allocation Strategy: Technology Remains the Core Theme
Based on the analysis of US-China cycle dynamics, under the base case scenario: 1) Technology remains the primary theme in both markets, but attention to timing and position sizing is crucial. Investors can focus on different logic chains within each market and rotate within the tech sector. However, monitoring the rhythm is essential, especially with earnings seasons serving as验证 points. Since 2023, AI-related trends have typically shown strength for two quarters followed by one weaker quarter. Currently, sectors like semiconductor manufacturing, memory, optical modules, and liquid cooling (China) are at high valuation percentiles, while computing chips are around the median, and cloud providers have declined to lower levels since last November. Drawing an analogy to the internet revolution, the current phase is similar to 1998-1999, where hardware led but applications had yet to take off, reflecting a relatively "rational" focus on the盈利确定性强的 hardware infrastructure layer in this AI cycle. 2) If oil prices and US yields decline, it could facilitate a modest recovery in US domestic demand sectors like real estate, boosting cyclical and China's broad external demand sectors. Simultaneously, assets sensitive to US yields, such as gold, non-ferrous metals, Hang Seng Tech, and innovative pharmaceuticals, could achieve valuation recovery. Therefore, a rotation strategy from tech to cyclicals to external demand is suggested. Current market consensus shows crowded positioning in AI themes (e.g., Philadelphia Semiconductor Index, STAR 50, South Korean and Taiwan indices), while US Treasuries and Hong Kong stocks appear less so. However, it is emphasized that crowding alone is not a reason for trend reversal; liquidity and credit cycle momentum are the true drivers. By synthesizing metrics like valuation, micro and macro liquidity, and credit cycles to assess the potential return (赔率) and probability of success (胜率) for various assets, the findings are: US Treasuries offer a favorable balance of both; the S&P 500, Nasdaq, Philadelphia Semiconductor Index, and gold have high probability but lower potential return; the Hang Seng Tech Index offers high potential return but limited near-term probability;原油 and the CSI 500 lack appeal on both fronts.
In terms of specific forecasts: 1) The year-end target for the S&P 500 is further raised to 7800-8000, implying a 3%-6% gain. 2) US Treasuries still present near-term trading opportunities; long-term yields are projected at 4%-4.2% if Fed cut expectations return post-Q2 Iran de-escalation, or 4.2%-4.4% otherwise. 3) The US dollar is expected to trade within a 96-98 range. 4) Gold's near-term performance may depend on a decline in US yields, though its long-term uptrend remains, with $5500 being a key level. 5) Upside for Hong Kong indices is seen as limited (Hang Seng Index 27000-28000), but the Hang Seng Tech Index has valuation recovery potential. 6) Japan is preferred over Europe, emerging markets may see continued internal divergence, and economies benefiting from the AI supply chain remain看好, though volatility risks warrant caution.
Comments