DD-滴滴

The latest rise in Chinese equities is not being driven by a single policy announcement or a temporary burst of speculative enthusiasm. It reflects the convergence of three powerful forces: an AI-driven memory-chip upcycle, the accelerating localisation of China’s semiconductor supply chain, and a broader reassessment of Chinese assets after years of valuation compression.
This does not mean that every Chinese stock has entered a new bull market. The rally remains highly selective, with capital concentrating in companies positioned to benefit from artificial intelligence, advanced manufacturing, semiconductor self-sufficiency and improving shareholder returns.
Major global investment banks have become increasingly constructive on Chinese assets, particularly in hard technology and large internet platforms. The central argument is that future returns could be supported by improving earnings rather than indiscriminate valuation expansion. After several years of underperformance, many Chinese companies still trade at substantial discounts to their global peers, leaving room for a structural re-rating if profitability continues to recover.
CXMT’s Listing: A Milestone for China’s Memory-Chip Ambitions
The most visible symbol of this reawakening arrived on July 27, 2026, when ChangXin Memory Technologies, better known as CXMT, began trading on Shanghai’s STAR Market under the ticker 688825.
CXMT sold approximately 6.69 billion shares at RMB 8.66 each, raising RMB 57.92 billion. Proceeds could increase to roughly RMB 66.61 billion should the over-allotment option be exercised in full. The transaction became Asia’s largest IPO of 2026 and surpassed Semiconductor Manufacturing International Corporation’s previous fundraising record on the STAR Market.
Its first day of trading was extraordinary even by the standards of China’s retail-driven IPO market. CXMT closed at approximately RMB 49.01, representing a gain of about 466% from its offer price, after climbing by more than 500% at its intraday peak.
Its market capitalisation briefly exceeded RMB 3.3 trillion, making it the most valuable company listed on a mainland Chinese exchange and temporarily overtaking Industrial and Commercial Bank of China.
The scale of the rally was spectacular, but the strategic significance of the listing matters more than its first-day return.
DRAM is a foundational component of modern computing. It is required in smartphones and personal computers, but increasingly also in cloud infrastructure, AI servers, automobiles and other data-intensive systems. CXMT has become China’s largest DRAM producer and one of the world’s leading memory-chip manufacturers.
Its rise demonstrates that China is beginning to establish a meaningful domestic presence in a sector that has historically been dominated by Samsung Electronics, SK Hynix and Micron.
The company’s financial inflection has been equally dramatic. CXMT generated approximately RMB 50.8 billion in revenue during the first quarter of 2026, representing year-on-year growth of more than 700%, while net profit attributable to shareholders reached approximately RMB 24.8 billion.
The improvement reflects rising memory prices, stronger AI-related demand, expanded production and a more favourable product mix. After suffering significant cumulative losses between 2022 and 2024, the company returned to profitability in 2025, with its earnings accelerating sharply in 2026.
In this sense, CXMT is not merely another semiconductor listing. It represents the capital-market expression of China’s effort to develop a more self-sufficient technology ecosystem.
Memory chips have moved from being treated primarily as cyclical electronic components to being regarded as strategic infrastructure for artificial intelligence and national supply-chain security.
CXMT’s listing also illustrates how China’s capital markets are increasingly being used to finance strategically important industries. The speed and scale of the transaction demonstrate the policy priority being given to semiconductor manufacturing, advanced technology and what Chinese policymakers describe as “new productive forces.”
However, the first-day valuation should not be interpreted as a straightforward assessment of CXMT’s long-term intrinsic value.
The unusually large gain was amplified by a conservative offer price, limited immediately tradable supply and the STAR Market’s rule allowing newly listed companies to trade without daily price limits during their first five sessions.
The company’s strategic scarcity value attracted enormous demand, but the resulting share price also incorporated expectations that may take years to fulfil.
CXMT still faces formidable obstacles. The global memory industry remains deeply cyclical, and periods of shortages and rising prices have historically encouraged aggressive capacity expansion, eventually producing oversupply.
The company also remains behind global leaders in certain advanced manufacturing capabilities and faces restricted access to some Western semiconductor equipment. Geopolitical tensions, potential export restrictions and competition from established international manufacturers will remain central risks.
The industry’s current strength may also prove temporary. If global manufacturers expand production too aggressively, or if AI infrastructure spending begins to moderate, memory prices could eventually peak and reverse. A company that appears extraordinarily profitable at the top of the cycle can experience rapid margin compression when supply catches up with demand.
CXMT’s debut should therefore be understood as both a milestone and a stress test: a milestone for China’s semiconductor localisation strategy, but a stress test of whether extraordinary investor expectations can ultimately be supported by technological progress, operating margins and sustained market-share gains.
Hong Kong’s Rally: Technology Rotation Rather Than a Universal Bull Market
The renewed enthusiasm has not been confined to mainland China. Hong Kong equities have also experienced a powerful, though uneven, re-rating.
The Hang Seng Index gained nearly 28% in 2025, while the Hang Seng TECH Index advanced by more than 20%, making it one of Hong Kong’s strongest market years in recent memory. That recovery established a stronger foundation for the more selective technology rallies seen during 2026.
Mainland capital has played an increasingly important role. Southbound investors have continued to purchase Hong Kong-listed securities through Stock Connect, with technology, consumer and financial companies attracting considerable demand.
These flows reflect both a search for undervalued assets and the fact that many of China’s most important internet platforms are listed in Hong Kong rather than on mainland exchanges.
The intensity of the rotation became particularly visible in July. During one notable session, the Hang Seng TECH Index rose by nearly 5%, while Alibaba gained more than 12%. Other major technology platforms and AI-related companies also advanced as investors returned to businesses with improving cloud-computing prospects, potential AI monetisation and valuations that had remained depressed relative to their historical levels.
This rally should not be described as a simple, market-wide surge. It has more closely resembled a reallocation from crowded or highly valued AI hardware positions into comparatively inexpensive Chinese internet platforms, software providers and application-layer businesses.
The distinction is important.
Hardware manufacturers were the earliest and most direct beneficiaries of the AI capital-expenditure cycle. The market is now beginning to ask which companies can convert that infrastructure into commercial applications, recurring revenue and higher margins.
In China, this transition could favour cloud platforms, advertising ecosystems, enterprise software providers, autonomous systems and consumer-facing AI applications.
Hong Kong also offers structural characteristics that mainland markets cannot fully replicate. It provides international and mainland investors with access to major Chinese internet companies, insurers and globally oriented consumer businesses that are underrepresented in domestic benchmarks.
Many of these companies have also introduced larger share-repurchase programmes, higher dividends and more disciplined capital-allocation policies, improving their appeal after several years of regulatory and valuation pressure.
The recovery is therefore not based solely on AI enthusiasm. It also reflects a broader reassessment of Chinese corporate governance and shareholder returns.
For much of the previous decade, investors often criticised Chinese technology companies for prioritising aggressive expansion over profitability. Today, several of the largest platforms are generating stronger cash flow, controlling costs more carefully and returning more capital to shareholders.
This makes the current rally fundamentally different from a purely speculative technology boom.
At the same time, the market remains vulnerable to abrupt reversals. Hong Kong valuations are highly sensitive to global interest-rate expectations, the US dollar, geopolitical developments and changes in mainland investor flows.
The city’s growing pipeline of AI and semiconductor listings may deepen the market, but it may also create additional selling pressure when early investors and cornerstone shareholders become eligible to reduce their positions.
The recent gains therefore represent a valuation repair and a technology rotation—not evidence that fundamental risk has disappeared.
The Deeper Logic: China’s “New Productive Forces” Enter the Equity Market
Viewed together, CXMT’s historic debut and Hong Kong’s technology rally are part of the same broader transition. China’s industrial-policy emphasis on “new productive forces” is beginning to acquire a visible capital-market dimension.
The first pillar is the global imbalance between AI computing demand and the supply of critical components.
Artificial intelligence requires not only processors, but also memory, optical interconnects, data-centre equipment, power infrastructure and advanced packaging. As computing workloads become larger and more complex, these supporting components become increasingly valuable.
The semiconductor opportunity is therefore much broader than GPUs alone. The rise of AI creates demand across an entire industrial chain, from memory and networking equipment to cooling systems and electricity infrastructure.
Companies positioned within these bottlenecks may benefit even if they do not produce the most visible AI products.
The second pillar is localisation.
Export controls and geopolitical tensions have increased the strategic value of Chinese semiconductor manufacturers, equipment suppliers and component producers. For investors, domestic substitution is no longer only a policy slogan.
It is becoming a source of addressable demand, government support, financing access and, in selected cases, genuine earnings growth.
China remains dependent on foreign technology in several important areas, but that dependency itself creates a powerful commercial incentive for domestic alternatives.
Whenever Chinese manufacturers are able to achieve acceptable performance and production scale, they may gain access to a protected and rapidly expanding domestic market.
The third pillar is capital reallocation.
Mainland investors are increasingly using Hong Kong to obtain exposure to technology, insurance and globally oriented companies, while international investors are reconsidering Chinese equities after years of underperformance and valuation compression.
Hong Kong’s active IPO market is simultaneously giving Chinese technology companies access to deeper and more internationally connected pools of capital.
This creates a reinforcing cycle. Strong listings attract investor attention, rising valuations make additional fundraising easier, and the resulting capital can be invested in research, manufacturing and international expansion.
The fourth pillar is valuation.
Chinese equities have spent several years trading at substantial discounts to US and other Asian markets. These discounts were not entirely irrational. They reflected concerns over regulation, property-sector weakness, domestic consumption, geopolitical tensions and uncertainty regarding corporate governance.
However, valuation discounts can become opportunities when expectations are already extremely low.
The current rally suggests that investors are beginning to distinguish between structural problems affecting the broader economy and individual companies capable of delivering strong earnings despite those challenges.
Yet the sustainability of the rally will depend on earnings rather than national strategy alone.
The strongest version of the bullish case is that AI demand, industrial upgrading and improved corporate discipline will produce several years of superior profit growth. Under that scenario, Chinese equities can rise without returning to the extreme valuations seen during previous speculative cycles.
The weaker version is that investors capitalise years of expected growth immediately, while memory prices peak, overseas restrictions intensify and domestic competition compresses margins.
Policy support can provide financing, favourable regulation and strategic demand, but it cannot guarantee commercial success. Industries identified as national priorities may also attract excessive investment, creating duplication, price competition and eventual overcapacity.
This has happened before in sectors ranging from solar manufacturing to electric vehicles. Semiconductor localisation may produce enormous long-term value, but it could also create intense competition among companies pursuing similar markets with similar policy support.
For investors, the implication is not to buy “China” as a single trade. The opportunity is structural, but it is also highly selective.
The most credible beneficiaries are likely to be companies with measurable technological advantages, improving cash flow and defensible positions in memory, advanced packaging, optical communications, semiconductor equipment and AI infrastructure.
At the application layer, the focus should remain on platforms capable of turning AI investment into revenue rather than companies valued primarily on announcements and narratives.
High-dividend companies may meanwhile provide a defensive counterweight to the volatility of technology exposure.
A Structural Opportunity, but Not a Risk-Free One
The comparison between CXMT and Hong Kong’s technology platforms also reveals an important difference between the two sides of the current rally.
CXMT represents scarcity, strategic ambition and the hardware foundation of the AI cycle. Its valuation reflects expectations that China will continue to close the technological gap with global memory leaders.
Hong Kong’s internet platforms represent a different form of opportunity. Many already possess established user bases, substantial revenue and strong cash flow. Their re-rating depends less on technological independence and more on whether they can convert AI into practical products, higher advertising efficiency, cloud demand and new sources of monetisation.
The hardware side may offer faster earnings growth during the current cycle, but it is also more vulnerable to changes in supply and pricing.
The platform side may grow more slowly, but successful companies could produce more durable cash flow if AI becomes embedded in everyday consumer and enterprise services.
A balanced interpretation of the rally must therefore recognise both opportunity and asymmetry.
The strongest companies may be entering a multi-year growth phase, while weaker businesses may simply be benefiting from the temporary expansion of risk appetite.
The challenge is separating companies whose earnings are genuinely changing from those whose narratives are changing faster than their fundamentals.
Conclusion: A Re-Rating That Must Still Be Earned
China’s equity resurgence is more substantial than a short-lived policy rally, but less universal than headline index movements might suggest.
CXMT’s ascent captures the ambition of China’s semiconductor programme and the extraordinary scarcity premium investors are assigning to strategically important technology assets.
Hong Kong’s rally reflects the next stage of the same story: capital rotating toward platforms, applications and companies whose valuations have yet to reflect the possibility of renewed earnings growth.
The central question is no longer whether China can generate exciting technology narratives. It is whether those narratives can be translated into sustainable margins, cash flow and shareholder returns.
Should earnings continue to improve, the combination of technological upgrading, policy support and discounted valuations could support a multi-year re-rating of selected Chinese assets.
Should earnings disappoint, the same market that celebrated CXMT’s historic debut may prove equally unforgiving.
China’s new equity cycle has begun with symbolism and extraordinary momentum. Its durability will ultimately be determined by execution.

Das neueste Modell von Kimi, K3, hat sich diesen Monat an die Spitze mehrerer wichtiger Benchmarks gesetzt.
Das Open-Weight-System mit 2,8 Billionen Parametern von Moonshot AI entspricht inzwischen den führenden geschlossenen Modellen von OpenAI und Anthropic oder übertrifft sie sogar bei Coding- und Long-Context-Aufgaben.
Dieser Durchbruch hat frische Aufmerksamkeit auf den Generative-AI-Rennen in China gelenkt.
Moonshot, 2023 von dem ehemaligen Meta- und Google-Forscher Yang Zhilin gegründet, machte sich mit dem Kimi-Chatbot einen Namen. Das Unternehmen agierte schnell: Es veröffentlichte 2025 die Trillion-Parameter-Reihe K2 und brachte danach im Juli 2026 K3 auf den Markt – mit einem Million-Token-Context-Window und nativer Multimodal-Unterstützung. Die vollständigen Gewichte sollen bis Ende Juli veröffentlicht werden. Erste Nutzer und unabhängige Tests zeigen starke Ergebnisse in Agenten-Workflows und bei komplexem Coding.

羅素2000 $XIWM 指數分析
截至2026年7月中旬
羅素2000指數（Russell 2000）由約2,000家美國小型上市公司組成，是觀察美國小型股表現最常使用的指標之一。市場上最具代表性的相關ETF是 iShares Russell 2000 ETF，代號 IWM。
近期表現
2026年上半年，羅素2000表現相當強勢，累計上漲約19%至22%，創下1991年以來最佳上半年表現。
同期標普500指數約上漲9%，羅素2000明顯跑贏大型股，兩者的上半年報酬差距也是2003年以來最大。
指數在6月底一度來到約3,033至3,046點的歷史高位，之後於7月初出現回檔，目前大致在2,970至3,000點附近震盪。IWM則約在293至300美元區間波動。
這一波上漲的原因
首先是估值修復。
2026年初，美國小型股相對大型股的估值折價，一度接近近30年來最大的水準。當市場資金開始從估值偏高的大型科技股轉向其他板塊，小型股自然成為輪動的主要受益者。
其次，小型公司的營收來源更集中在美國本土。羅素2000成分公司的收入約有77%來自美國，因此相較跨國大型企業，受到美元升值、關稅及海外地緣政治因素的影響較小。
盈利預期上修也是重要推力。市場原先預估羅素2000成分股2026年每股盈餘將增長約23%，之後陸續上調至38%左右。部分增長來自AI基礎設施相關的小型企業，包括晶片供應鏈、電子設備、電力及資料中心周邊公司。
此外，AI行情也開始由大型科技公司向中小型供應鏈擴散。市場不再只追逐少數幾家大型企業，而是逐漸尋找AI資本支出背後的設備、零組件及基礎設施受益者。
美國經濟仍具韌性，加上財政支出及製造業復甦預期，也改善了市場對小型企業營運環境的看法。
不過，6月指數重組後，羅素2000中AI基礎設施相關公司的權重，由約15%降至7%。這代表前期推動指數上漲的重要力量有所減弱，後續能否延續漲勢，仍要看其他產業的盈利能否接上。
投資銀行與機構觀點
高盛：漲勢可能開始降溫
高盛在2026年7月的報告中，對小型股後續表現轉為較謹慎。
策略師Ben Snider認為，羅素2000經過上半年的快速上漲後，估值已不再像年初那麼便宜，預估未來12個月的報酬可能只剩低個位數。
高盛主要擔心幾個問題。
第一，市場已經提前反映不少盈利復甦與估值修復的預期。
第二，美國經濟增長目前接近長期趨勢，未必足以支持小型股持續大幅上漲。
第三，指數重組後，AI相關公司的比重下降，題材帶來的推動效果可能減弱。
另外，羅素2000仍有約四分之一的成分公司處於虧損狀態，未盈利公司的市值占比約23%。一旦盈利不如預期，股價容易出現較大修正。
小型公司也比大型企業更容易受到利率影響。約30%的小型企業債務採用浮動利率，標普500企業則約為7%。假如聯準會維持高利率更久，甚至重新轉向緊縮，小型公司的融資成本會更快上升。
美國銀行：盈利與製造業仍有支撐
美國銀行策略師Jill Hall的看法相對正面。
她認為，羅素2000的領先表現不只是估值反彈，也反映小型企業盈利改善及美國製造業動能回升。
美國銀行同時看好中型股，原因是中型企業通常兼具一定的盈利能力與成長空間，對利率的敏感度又低於小型企業，風險報酬可能更平衡。
其他機構
LPL Financial仍維持羅素2000成分股2026年盈利增長約38%的預估。
Fidelity、T. Rowe Price及Royce等長期研究小型股的機構，整體仍偏向正面。他們認為，小型股相對大型股的估值仍不算昂貴，未來也可能受惠於盈利復甦、併購活動增加及資金重新配置。
從更長的五年至十年來看，小型股仍有機會結束過去長時間落後大型科技股的局面，重新進入相對有利的市場週期。
值得注意的是，高盛及摩根士丹利等機構在2025年底至2026年初時，對小型股的態度也較為樂觀。當時主要看好經濟增長、利率環境改善及市場上漲範圍擴大。隨著羅素2000已經累積可觀漲幅，部分機構才開始調低後續預期。
主要風險
羅素2000目前最大的問題，是成分公司品質差異很大。
部分企業盈利能力穩定，也有不少公司仍在虧損。當市場情緒良好時，投資人通常願意提高風險承受度；但只要經濟或盈利預期轉弱，未盈利企業往往會率先受到拋售。
利率也是不能忽略的風險。小型公司較依賴銀行貸款及浮動利率債務，缺乏大型企業低成本發債的能力，因此高利率環境對小型股的壓力通常更大。
另外，上半年漲幅已經相當可觀，短期出現獲利了結、估值消化或區間整理都很正常。假如美國經濟放緩、通膨重新升溫，或聯準會態度轉鷹，小型股可能比大型股承受更大的下跌壓力。
後續展望
短期來看，羅素2000在快速上漲後，繼續單邊走高的難度已經增加。市場可能需要一段時間整理，等待企業財報及盈利預期確認。
中長期則仍有幾個值得觀察的條件：
美國經濟能否維持溫和增長、企業盈利能否兌現、利率是否逐步回落，以及資金是否繼續由大型科技股流向其他產業。