Shenwan Hongyuan Group: Crossing the "J curve" in the short term, focusing on liquidity risk pressure, and continuing to lay out the AI industry trend in the medium term.

Shenwan Hongyuan Group released a research report stating that in the second half of 2026, the market is in a critical period of transition from the left side to the right side of the "J-curve." In the short term, one should be cautious of the volatility risks brought about by geopolitical and US dollar liquidity disturbances; in the medium term, one should focus on the trend of the diffusion of AI industry dividends, as well as the opportunities for improving the cost-effectiveness of Chinese assets in global allocation. Strategically, a combination of tactical flexibility and strategic determination is required to navigate the cycle and prepare for the future. On the tactical allocation front, considering the risks of oil prices and US dollar liquidity disturbances, the analysis suggests: In the short term (1-3 months), increase cash allocation, maintain hedges for commodities and energy positions, continue to underweight US bonds, and reduce equity allocations in the Korean market that experienced significant gains in the previous period and in the Southeast Asian market where fundamentals are under pressure from oil prices. In the medium term (3-6 months), opportunistically increase allocation to high-quality equity assets in Chinese and American technology stocks based on changes in US dollar liquidity. From a strategic allocation perspective, commodities, high-quality Chinese assets, and opportunities in the AI industry trend are still favored. The latest edition of the quantitative asset allocation "Five-Star Model" analysis suggests overweighting commodities, Chinese equities, and cash in Q3 2026; maintaining a standard allocation to US equities and Chinese bonds; and underweighting US bonds and other emerging markets. The main views of Shenwan Hongyuan Group are as follows: In the first half of 2026, under the macro disturbances such as geopolitical risks, rising inflation, and fluctuating interest rates, global asset prices have shown dramatic fluctuations and differentiation. Global stock markets: AI hardware stocks and certain markets that benefit (such as the Korean, Japanese, and US tech stocks) have performed strongly, while other sectors and markets are under pressure; Global bond markets: diverging interest rate trends at home and abroad, with US bond rates sharply rising, while Chinese bond rates are marginally decreasing. Commodities: Crude oil has surged, industrial metals have remained resilient, and precious metals have undergone significant adjustments. Foreign exchange market: the US dollar is relatively strong, while the Chinese Renminbi is steadily appreciating. Looking back at the second quarter of 2026, our global quantitative asset allocation "Five-Star Model" has achieved excess returns by overweighting commodities and equities, with a cumulative return of 3.79% since March 2026. Global asset differentiation pattern on the left side of the productivity "J-curve": In 2026, global stock markets are moving towards extremes and differentiation, reflecting on several levels: 1) market level: funds have continued to flow into the US market since April, flowing out of non-US markets; 2) style level: the performance gap between tech and non-tech stocks in the first half of 2026 has further widened compared to 2025; 3) within tech stocks: further differentiation in the performance of AI hardware-related companies and software stocks. The current global market differentiation pattern is a typical feature of the left side of the productivity "J-curve": in the early stages of the introduction of AI technology, corporate capital expenditures precede productivity improvements, thus the market's upward momentum is concentrated in direct areas such as tech hardware, resources, and capital goods, with other sectors relatively lagging. The market is expected to transition from the left side of the "J-curve" to the right side in two ways. Optimistic path (breadth recovery): if the AI productivity dividend successfully spills over to more industries, driving demand improvement and inflation abatement, the market will achieve comprehensive growth. Pessimistic path (muddled waters): before the productivity dividend is realized, capital costs (rates and risk premiums) may rise significantly first, which could disrupt the AI investment cycle, leading to stagnation in growth and a general market decline. Macro and US dollar liquidity: In the short term (next 1-3 months), US dollar liquidity disturbances are ongoing, and tactical volatility risks need to be monitored. Firstly, geopolitical risks and concerns about oil prices are likely to fluctuate, particularly with the Strait of Hormuz being blocked, oil inventories continually being used up, and oil prices being more prone to rise than fall. Secondly, the policy attitude of the new Federal Reserve Chairman Walsh has entered an observation period. Lastly, with mid-term elections approaching in the fourth quarter, US policy uncertainty may increase. Key areas to focus on are the changes in the US bond yield curve and volatility, with a significant increase in volatility and "bear flattening" posing pressure on the US stock market. In the medium term (next 3-6 months), the Federal Reserve's considerable rate hikes are facing constraints, and US dollar liquidity may lack a trend-based tightening foundation. On one hand, the US economy is showing a split pattern of "weak consumption, strong investment," driven by investments in AI-related equipment and intellectual property, but declining consumer savings rates, decreasing consumer confidence, slow growth in real wage income, and a tightening labor market are limiting factors. On the other hand, high US fiscal pressures and structural tensions in the reserve system will limit the Fed's ability to significantly raise rates or engage in trend-based "balance sheet reduction." If the Fed places more emphasis on indicators like the "tail mean" PCE, or further raises the threshold for rate hikes, looking ahead to the second half of the year, the 10-year US bond yield is expected to hover at high levels, with the core volatility range estimated at 4.2% to 4.7%, potentially peaking before falling again. In a downside scenario (geopolitical risks easing, labor market weakening), yields may fall back to 3.7%; in an upside scenario (continuing conflicts, resonance between AI investments and fiscal deficits), they may break through 4.7%. A trend towards a stronger US dollar usually occurs during periods of significant Fed rate hikes or deep recession in the US, with the core volatility range expected to be between 97-103 in the second half of 2026. Global stock markets: As the "J-curve" transitions from the left side to the right side, will global stock markets exhibit a "breadth recovery" or a "muddled waters" scenario? From a fundamental perspective, there is still significant room for improvement in the AI tech industry trend and adoption rate, with the AI fundamentals trend yet to be systematically falsified. AI adoption is expanding from the tech sector to areas such as finance and professional services, with cloud business revenue and profits growing in sync, confirming the industry trend. However, as capital expenditures increase, US AI cloud manufacturers' reliance on debt financing is rising, necessitating attention to operational pressures and impending risks of capital expenditure fluctuations for some companies. From a valuation and risk perspective, valuations have not yet reached bubble levels, but there are tactical volatility risks to be wary of in the third quarter: the PE valuations of the S&P 500 and Nasdaq 100 have not reached extreme bubble levels, but the PB percentile is high. Short-term risks to be cautious of include: surging US bond volatility (MOVE index) suppressing the stock market (particularly tech stocks), heightened policy uncertainty in election years, and the historical trend of poor post-IPO market performance for large companies. There is also a need to be vigilant of the spillover effects of liquidity volatility in markets like South Korea. From a structural perspective, after short- to medium-term macro liquidity disturbances, there is still an opportunity for the AI semiconductor market to expand into other tech sectors and styles. Currently, the relative excess returns for semiconductors compared to other industries are approaching the highs of 2000, but the overall relative excess returns for the tech industry as a whole are still some distance away from the highs of 2000, and the growth of large-cap tech stocks compared to small-cap value stocks is significantly lower than during the dotcom bubble period. China stock and bonds: Strategically, one should continue to look at the opportunities for Renminbi asset allocation under the backdrop of Renminbi internationalization. Unlike the widespread pressure on interest rates and significant market volatility in overseas markets, liquidity in Chinese assets has shown marginal improvement since 2026, further highlighting the resilience of Chinese assets under global turmoil. In the bond market, Renminbi appreciation through trade surplus inflows acting as a "passive reserve requirement cut" has provided some support for the bond market. The central range for the 10-year government bond yield is expected to fluctuate between 1.60% and 1.80% in the second half of 2026, supported by factors such as expectations of a stronger yuan, relatively controllable imported inflation pressures, continued loose monetary policy, and strong institutional allocation demand. In terms of the equity market, the overall upward trend of A-shares since 2025 has been primarily driven by domestic incremental fund pricing, with external fund allocations and direct impacts from US dollar liquidity having minor effects in the short term. As domestic AI computing power and high-end manufacturing industries see a resurgence, there is still room for structural market gains. Hong Kong stocks' sensitivity to US bond rates has increased this year, with short-term foreign liquidity risks still exerting pressure. Considering significant adjustments in earnings expectations already absorbed, and negative factors to a certain extent already factored in, there is room for further market pricing of the value of Chinese manufacturing and the AI industry trend. Global funds' allocation to the Chinese stock market still remains at historically low percentiles, and from a global stock market perspective on PB-ROE ratios, A-shares and Hong Kong stocks are positioned favorably in terms of cost-effectiveness. In the medium term, with expectations of underlying Renminbi asset recovery, there is still significant room for global funds to increase their allocation to Renminbi assets. Strategically, there is continued optimism for commodities, with a focus on opportunities for Shenzhen Agricultural Power Group to catch up: driven by global fiscal deficits, the recovery of the manufacturing cycle, geopolitical tensions and resource control, and demand resonance from AI capital expenditure. Crude oil: under tight supply-demand balance, oil prices are more likely to rise than fall, with hedging value against geopolitical risks. Industrial metals (copper, aluminum): constrained supply due to mine accidents and rising costs, coupled with demand from AI computing power infrastructure and energy transition, suggests the upward trend may not have ended. Short-term volatility in gold prices: gold prices in 2026 are primarily driven by changes in interest rate expectations triggered by US-Iran conflicts. Gold ETF holdings (such as SPDR) have seen significant outflows following the rise in actual US bond rates. Long-term logic remains solid, with high US fiscal deficits and central bank gold purchases under the global trend of "de-dollarization" providing long-term support. With implied volatility at low levels, it is advised to patiently wait for allocation opportunities following a reversal in US dollar liquidity expectations. Shenzhen Agricultural Power Group: the allocation window has emerged. Global grain stocks are low, while rising energy prices are pushing up planting costs and demand for biofuels. Historically, in commodity cycles driven by supply shocks, Shenzhen Agricultural Power Group has often been the "last to rise." Additionally, the possibility of an El Nio event forming in 2026-2027 may further pressure Shenzhen Agricultural Power Group prices upwards. Risk considerations: escalation of geopolitical conflicts, nonlinear weakening of the global economy, and significant declines in US AI tech stocks.