Strong dollar, rate hike expectations, and AI siphoning

In early May, the dollar strengthened and expectations of interest rate hikes heated up, which precisely corresponded to the acceleration of the K-shaped divergence in global markets. The core issue was the damage to demand in non-AI sectors caused by tightening expectations. As the market has evolved, the K-shaped divergence has reached an extreme stage, with even overseas tech sectors internally shrinking. The pricing of stocks, bonds, commodities, and currencies has already shown early signs of recession trading. If tightening actually occurs, it could further damage carbon-based world demand; conversely, the K-shaped divergence may see a phased convergence. Compared to the relatively volatile overseas markets, A-shares have shown more resilience, with some non-AI sectors already showing signs of left-side capital involvement, and a few undervalued sectors have the foundation for a recovery, merely awaiting a catalyst.

In early May, the dollar strengthened and expectations of interest rate hikes heated up, which precisely corresponded to the acceleration of the K-shaped divergence in global markets

Comparing the liquidity expectations of the dollar with the K-shaped divergence in equity markets, the K-shaped divergence this year can be divided into three phases: Phase 1 was from the beginning of the year to the end of February, when the market still had expectations of Fed easing, the dollar index weakened from 100 to around 96, and K-shaped divergence across markets was not significant. Phase 2 was from the end of February to April, when the market began to reprice hawkishly, the dollar index rebounded, and K-shaped divergence across markets began to emerge. Phase 3 is from May to the present, with further heating of rate hike expectations, the dollar index rising to its year-to-date high, and K-shaped divergence in non-US markets intensifying, creating a clear gap with US stocks. Additionally, Phase 3 corresponds to significant pullbacks in non-AI sectors of the A-share market, with cyclical stocks starting to underperform their overseas counterparts. In the report "A-Share Strategy Focus 20260621 - 'Rate Hikes' Boost Tech?", we analyzed the role of rate hikes in the K-shaped divergence of global markets. The huge disparity in economic/industry outlook leads to the negative impact of rate hikes on non-AI sectors far exceeding that on AI sectors. From this perspective, when the narrative of rate hike expectations and dollar strength cools down, the K-shaped divergence may temporarily ease, and style balancing may follow.

The K-shaped divergence has reached an extreme, and even overseas tech sectors are shrinking internally

Market narratives and capital flows can also push reasonable divergence to an extreme of overpricing. Essentially, as tightening expectations continue to rise, the market's threshold for economic/industry outlook is also constantly increasing. Reflected at the market level, even US stocks show signs of continuous internal shrinking within tech sectors. Since May, when the market perceived that Anthropic's ARR month-over-month growth rate had slowed, coupled with downstream enterprises beginning to control token budgets, the Nasdaq began to consolidate, and the Mag 7 started a sustained correction, falling 12% from their highs as of June 26. The previously high-beta optical communication sector began to oscillate at high levels after mid-May, and the application sector, which had rebounded earlier, came under pressure again in June. The semiconductor sector within US tech stocks performed relatively well early on, but by June, the semiconductor index also began to oscillate at high levels, with a weekly drop of 7.9% this week. Currently, only storage leaders continue to rise due to strong earnings reports.

Overseas risk asset pricing: Tightening may further damage carbon-based world demand

The recent price combination in overseas assets shows a rising dollar, falling US stocks, falling commodities, but also falling oil prices. The decline in oil prices has eased long-term inflation expectations but has not affected short-term inflation stickiness. On June 25, the US Department of Commerce released the May PCE inflation data, with overall PCE rising 4.1% year-on-year, but core PCE also rising 3.4% year-on-year, hitting a new high since 2023, indicating that US inflation is highly sticky and not entirely caused by oil prices. The latest narrative in the capital markets has shifted to AI's strong demand crowding out commodity resources, forcing "carbon-based" entities to pay higher prices, leading to inflation, and compelling the Fed to raise rates. This week, Apple's news of raising prices for consumer electronics products due to memory price increases further strengthened this narrative. The market interprets this as: if the Fed is forced to raise rates in response to supply-driven structural inflation, it may ultimately further damage carbon-based world demand. Meanwhile, the rise in the dollar's real interest rate is seen as high economic growth brought by AI's comparative competitive advantage, making the dollar extremely strong. However, the rapidly flattening yield curve of US Treasuries shows that this narrative is also temporary. Currently, the term spread between 10Y and 2Y US Treasuries is 0.31 percentage points, significantly flatter than the 0.58 percentage points in early March, and at the 28th percentile over the past 20 years. The market may be reflecting that tightening could further damage carbon-based world demand and ultimately hurt long-term economic growth expectations. Overall, the current pricing model for risk assets is extremely contradictory. This may be because the growth effects of AI technological innovation are still confined to a small cycle of hardware manufacturers, model companies, and cloud providers, and have not yet been more broadly integrated into the larger economic cycle. This also means that while the K-shaped divergence has its rationality, it also has its fragility.

A-shares show more resilience, with a few non-AI sectors showing signs of left-side capital involvement

Recently, the technology sector in A-shares has shown significantly stronger resilience compared to overseas, especially the domestic computing power chain, which has exhibited a trend independent of overseas markets. A-share companies in the overseas supply chain follow the same heavy-asset cyclical stock pricing logic as their overseas counterparts, while the domestic chain, under the narrative of domestic substitution and independent control, also exhibits typical growth stock premium behavior. Two different pricing frameworks operate simultaneously. Besides the tech sector, some non-AI sectors, due to their previous adequate adjustments, have begun to show characteristics of left-side capital inflow. Brokerages and chemicals are typical examples. The combined trading volume of brokerages and chemicals once hit a new year-to-date high this week, with significant volume expansion. On some days, they could rise together with AI, and their stock prices have also seen noticeable recovery. For the brokerage sector, its left-side logic is supported by three clues: "low valuation + weakening selling pressure on the capital side + the boom of tech listings." For the chemical sector, the core driver is the widening of spreads after oil prices fell into the "sweet spot," with some companies delivering "expected" strong mid-year performance forecasts, and their stock prices performing very strongly. This may also indicate that market funds have limited holdings and the game is not very intense. Although it is rare for non-AI sectors to rise against the backdrop of capital suction recently, it at least means that there are still many off-market funds in the A-share market continuously watching for opportunities in non-AI sectors, but the marginal changes are not yet sufficient to drive broader entry of off-market funds.

Some non-AI undervalued sectors already have a foundation for recovery, merely awaiting a catalyst

Since May, the non-AI sectors in A-shares have performed weaker compared to their overseas counterparts, fully reflecting many negative expectations such as demand recession, monetary tightening, and twists in Middle East peace talks. They now have a certain cost-effectiveness and foundation for recovery, awaiting some positive changes in their own narratives. This change could come from oil prices falling more than expected after the Strait passage, lowering inflation expectations, or from the synchronized recovery of industrial production and social activities in the global non-AI sector. In a market environment where the dollar is strong and rate hike expectations are heating up, we need to be more precise and patient in selecting varieties. After all, without significant marginal changes, the recovery of weak sectors will not be smooth sailing, and may even see synchronized adjustments due to adjustments in strong sectors. In terms of specific allocation, we still recommend adhering to the structure of AI + energy chemicals. On the AI side, we continue to favor storage, gas turbines, diesel generator sets, semiconductor equipment, and materials. On the energy chemicals side, in the new energy sector, we prefer varieties such as electrolytes and additives, and separators for earnings delivery. In the chemical sector, the decline in the oil price center and volatility brings restocking and operating demand, and the peaking of macro liquidity expectations will be a potential rhythm point later. Currently, we favor varieties with large cost reduction space, relatively rigid demand, and low valuations, such as refrigerants, phosphorus chemicals, spandex, dyes, and large-scale refining. In the metals sector, we recommend computing power metals that have some AI exposure but are temporarily suppressed by the rate hike narrative in valuation, such as tin, copper, and some AI small metals (tungsten). Additionally, we continue to recommend increasing allocation to undervalued brokerages, as the current suppression of liquidity and other flaws may begin to fade in the second half of the year, and mid-year performance forecasts are also catalysts.

Risk Factors

Escalating frictions in technology, trade, and finance between China and the US; domestic policy strength, implementation effects, or economic recovery falling short of expectations; tighter-than-expected macro liquidity domestically and internationally; further escalation of conflicts in Russia-Ukraine and the Middle East; slower-than-expected digestion of real estate inventory in China.

Source: CITIC Securities Research

Risk Warning and Disclaimer

        Markets are risky, and investment requires caution. This article does not constitute personal investment advice and does not consider the specific investment objectives, financial situations, or needs of individual users. Users should consider whether any opinions, views, or conclusions in this article are suitable for their specific circumstances. Investments based on this article are at your own risk.