The AI sector experienced significant volatility; diversified allocation helps manage market uncertainty.

Market focus has recently shifted clearly toward the sustainability of capital expenditures in the artificial intelligence (AI) industry and the rotation of capital among different investment styles. The semiconductor sector has underperformed, as investors express concern over whether high capital expenditures can translate into stable returns. The latest earnings reports from major tech companies show that $Alphabet-C (GOOG.US)$ 、 $Microsoft (MSFT.US)$ 、 $Amazon (AMZN.US)$ and $Meta Platforms (META.US)$ they continue to maintain high levels of investment in AI infrastructure, but free cash flow has come under significant pressure—Google’s free cash flow turned negative, Meta’s quarterly free cash flow nearly dropped to zero, and Amazon’s cash flow also continued to contract sharply. Meanwhile, the Federal Reserve decided at its July meeting to keep interest rates at 3.50%–3.75%, but saw an unusually high level of dissent for recent years, with three votes opposing the decision and advocating for an immediate rate hike. The new Chair’s comments also leaned hawkish, leading markets to raise expectations for further policy tightening. Under the dual pressures of elevated capital spending and potential interest rate hikes, some highly valued, leveraged AI-related positions have undergone deleveraging, while capital has rotated toward quality tech stocks with strong free cash flow and pricing power, resulting in a clear style rotation.

According to Bloomberg data, US equities in July $S&P 500 Index (.SPX.US)$ 、 $NASDAQ 100 Index (.NDX.US)$ and $Dow Jones Industrial Average (.DJI.US)$ delivered mixed performance, while Hong Kong stocks $Hang Seng Index (800000.HK)$ 、 $Hang Seng TECH Index (800700.HK)$ and $HKEX Tech 100 Index (800666.HK)$ also showed divergence, with selling pressure concentrated primarily in AI supply chain-related sectors. As of the US trading session on July 30, the memory and semiconductor sectors rebounded across the board, with $PHLX Semiconductor Index (.SOX.US)$ surging 8.2%, marking its biggest single-day gain since April 9, 2025; $SK hynix (SKHY.US)$ memory stocks performed even more prominently, $Micron Technology (MU.US)$ jumping 17.5%, $SanDisk (SNDK.US)$ soaring 18.4%, and

rising nearly 26.0%. The sharp rebound in the sector may be linked to deleveraging. JPMorgan released a new report titled 'Korea Equity Strategy Update on the De-leveraging Process in Korea (2),' outlining three key points:

1. Regulatory floor: On July 29, the Korea Exchange initiated an internal assessment to confirmthe technical feasibility of a temporary short-selling ban and narrowing daily trading limits to ±30%. On the same day, the Ministry of Finance convened an emergency meeting and activated a cross-agency 24-hour market monitoring system. Historically, markets have often seen phased rebounds following short-selling bans.

2. Leverage unwinding: JPMorgan’s report on July 29 noted thatleveraged ETF assets under management declined from $50 billion to $17 billion, indicating that forced liquidations are largely complete; hedge funds’ long/short ratio dropped from 5.7x to 3.2x, suggesting approximately 90% completion of deleveraging.

3. Corporate capital entry: On July 30, SK Group Chairman Chey Tae-won personally purchased 3,620 shares directly for the first time in history, $SK Hynix (000660.KR)$ amounting to KRW 4.8 billion—deliberately staying just below the mandatory disclosure threshold (purchases exceeding KRW 5 billion require a 30-day advance filing)—and explicitly stated, “Memory will continue to be needed.”

The report primarily highlights that leverage unwinding is nearing completion, with regulators and industrial capital simultaneously signaling a market bottom, potentially opening a window for a phased rebound in Korean equities. However, it remains uncertain whether the semiconductor cycle has peaked. Earnings reports from hyperscalers—including Google, Amazon, Microsoft, and Meta—indicate sustained capital expenditures, with the AI compute infrastructure race still ongoing; thus, declaring a cyclical peak in the sector may be premature.

Nevertheless, it is still too early to conclude whether current market conditions have truly emerged from the recent correction and re-established an uptrend. It remains to be seen whether this rebound is merely a technical recovery following a sharp decline or reflects genuine fundamental improvements in the sector—further data will be needed for confirmation. In the near term, investors should closely monitor the following key events and trends.

1. Changes in interest rates and yields

(Source: CME FedWatch, as of July 31)

As of July 31, according to CME FedWatch data, interest rate futures indicated that over 60% of investors expected the Federal Reserve to raise rates at its September 16 policy meeting. Key reasons include: three FOMC members already voted in favor of a rate hike at this month’s meeting, signaling growing hawkish sentiment within the committee; additionally, the new Chair’s recent comments have been more hawkish, leading markets to increase expectations for further policy tightening in the second half of the year.

Beyond the benchmark rate, $U.S. 10-Year Treasury Notes Yield (US10Y.BD)$ the movement of longer-end yields deserves closer attention, as changes in long-end rates often more directly impact market risk appetite and valuation levels. In fact, heavy capital expenditures by the AI sector and rising prices of key components such as memory storage are already exerting some upward pressure on inflation from the cost side. If this cost-push inflation persists, it could increase the likelihood that the Fed maintains a restrictive stance—or even hikes rates further.

The yield spread between long-term and short-term bonds will be a key indicator to watch。

2. Tech giants’ investment intensity in AI

Chart: Overseas CSPs revising capital expenditure plans upward

(Source: Bloomberg, JPMorgan, as of July 28, 2026)

Market views on the AI sector have recently diverged significantly. Following earnings announcements from Google and Meta, their share prices came under pressure, with investors focusing on the sharp contraction in free cash flow amid high capital expenditures,Concerns about the timing and certainty of return on investment。

In contrast, Amazon and Microsoft have received more optimistic assessments, primarily due to theircloud businesses (AWS and Azure) demonstrating stronger growth momentum, with the market believing their AI infrastructure investments are more likely to translate into sustainable revenue growth. Even companies with extremely strong cash flows $Apple (AAPL.US)$ have not been entirely immune—pressure from rising storage prices on costs and margins has still been reflected in post-earnings stock performance.

Overall, hyperscalers continue to ramp up capital expenditures, providing near-term support for demand in AI infrastructure and semiconductors. However, the market is increasingly focused on the sustainability of the entire sector. Beyond the scale of capital spending, recurring revenue (ARR) performance among upstream AI supply chain companies is gradually becoming a new metric investors use to assess long-term value. In simple terms, whether AI-related investments can ultimately generate a reasonable return on invested capital (ROIC) will be a key reference point for the market in evaluating long-term value.

Figure: Static ROIC for AI inference workloads is already quite attractive; Microsoft estimates ROIC at 25%–50%

(Source: Morgan Stanley Research Estimates, as of July 27)

3. Deleveraging Progress and Shifts in Market Sentiment

Another key area to closely monitor in the near term is the pace of market deleveraging and shifts in speculative sentiment. After an extended period of AI-themed trading, some high-valuation, highly leveraged positions have recently shown signs of stress during the recent pullback, prompting capital to rotate from high-beta semiconductor and AI-related names toward higher-quality tech stocks with more robust cash flows. It remains to be seen whether this style rotation will deepen further or reverse as sentiment recovers and funds flow back into thematic assets. Additionally, speculative activity in the options market, retail investor flows, and changes in volatility levels will all help determine whether the current rebound is merely a technical recovery or signals a renewed rise in speculative sentiment. If the deleveraging process has not yet run its full course, market volatility may remain elevated.

Brief Summary

Market volatility has recently intensified noticeably, with some of the pullback potentially linked to accumulated leverage and speculative activity in prior periods. Regarding the AI sector, market focus is shifting from pure thematic enthusiasm toward greater scrutiny of the efficiency and sustainability of capital expenditure returns. Semiconductor demand continues to receive support from sustained investments by major cloud service providers, while the memory segment faces relatively higher uncertainty; fundamentals in other chip segments have not changed significantly.

Looking ahead,AI-related themes present both opportunities and risks, and market performance may lean more toward structural rotation rather than broad-based, one-sided moves.. In addition to focusing on AI infrastructure, investors may also consider companies benefiting from the gradual commercialization of AI applications. In a high-volatility environment,moderate diversification can help manage market uncertainty.Furthermore, certain non-pure-AI thematic sectors with strong free cash flow may serve as a balancing component in portfolios, mitigating the impact of volatility tied to a single theme.

E Fund Hong Kong has established an ETF product lineup covering multiple themes and investment styles, allowing investors to select options aligned with their risk tolerance and allocation objectives:

– $E Fund HKEX Tech 100 ETF (03456.HK)$ : Primarily invests in Hong Kong-listed technology stocks, with diversified exposure across six key areas: artificial intelligence, biotechnology and pharmaceuticals, internet, information technology, electric vehicles and autonomous driving, and robotics. It offers representative exposure to Hong Kong tech stocks and is suitable as a core holding for Hong Kong technology allocations.

– $E Fund (HK) FTSE AI Select Index ETF (03489.HK)$ : Offers balanced exposure to global AI themes, covering both Hong Kong and U.S. equities—not solely concentrated in hard tech—which helps diversify risk during market rotations.

– $EFund A SEMICON ETF (03486.HK)$ : Focuses on Asian semiconductor manufacturing and memory segments. Suitable for capturing cyclical recovery opportunities in the semiconductor supply chain following significant corrections. Its advantage lies in not being persistently concentrated in the memory sector alone, but rather benefiting from the broader semiconductor industry’s capital expenditure cycle—such as sustained infrastructure investments by major cloud service providers driving demand.

– $EFund Biophar ETF (03186.HK)$ ：Its equal-weight design helps reduce idiosyncratic risk from any single biotech stock. The investment universe spans global biopharmaceutical companies listed in both Hong Kong and U.S. markets, offering growth-oriented diversification beyond technology themes.

– $E FUND (HK) MSCI Asia Pacific Select High Dividend Index ETF (03483.HK)$ : Uses relatively high dividend yield as the core screening criterion, investing in high-dividend equities across the Asia-Pacific region, primarily diversified across high-yield sectors such as energy and financials in markets including China, Japan, and Macau. It can provide a source of dividend income for the portfolio and is suitable as part of a cash flow allocation during volatile market conditions.

– $E Fund (HK) CSI Liquor Index ETF (03189.HK)$ : Focuses on leading and major companies in the baijiu (Chinese liquor) sector. It has performed relatively well recently amid heightened volatility in tech stocks, reflecting the resilience and independence of traditional consumer sectors. Signs of cyclical improvement in the baijiu industry are gradually emerging, with companies maintaining strong cash flows overall. Given its low correlation with AI-related themes, it offers valuable non-tech diversification for portfolios.

Through a combination of different themes and investment styles, investors can respond more flexibly to current market rotation and uncertainty. AI remains a long-term structural theme, but short-term volatility and style rotation have intensified; diversified allocation helps manage risk. Investing involves risk—past performance does not guarantee future results—and investors should carefully assess their own circumstances before making decisions.

Important Notice

This material is issued by E Fund Asset Management (Hong Kong) Company Limited. This material is for reference only and does not constitute an offer or recommendation to invest in fund units. This material is for display purposes only and must not be shown to any person in any jurisdiction where such display would be unlawful. Investing involves risks, and you may lose a substantial portion of your principal. Prior to investing, investors should carefully read the relevant investment risks described in the offering documents (including the 'Risk Factors' section) of the fund. This material has not been reviewed by the Securities and Futures Commission of Hong Kong.

For detailed important notices and disclaimers regarding the above fund, please visit E Fund (Hong Kong)'s website:

https://www.efunds.com.hk/tc/etfDisclaimers/

521K Views