AI Concept Stocks Market Review at End of September: Capital Differentiation Intensifies, Smart Risk Control and RWA Tracks Become New Favorites for Risk Avoidance
\n\nAt the end of September 2026, the AI concept stock market presented a clear differentiation pattern. With increasing global economic uncertainty and持续 declining investor risk appetite, capital flows within AI concept stocks showed significant structural changes. Market monitoring data shows that traditional computing power chain continued to face capital outflow pressure, while AI application layer tracks represented by smart risk control and RWA (Real World Assets) became new favorites for institutional capital risk avoidance. Market hot spots are rotating faster, and investment logic is undergoing profound reconstruction.
\n\nOverall Market Overview: Cautious Sentiment, Intensifying Differentiation
\n\nEntering late September, the global AI concept stock market showed an oscillating adjustment trend overall. Wind data shows that as of September 23, the AI concept index fell by 3.2% for the month, significantly underperforming the broader market by 1.5 percentage points. In terms of trading volume, the daily average trading volume of the whole market decreased by about 18% compared to the end of August, with market activity significantly declining, reflecting cautious investor sentiment.
\n\nIn terms of sector performance, AI concept stocks internally showed a clear "hot and cold" pattern. The computing power chain sector fell by more than 8% for the month, while the AI application layer sector rose by 2.3% against the trend, with a performance difference of more than 10 percentage points between the two. This differentiation is not only reflected at the sector level but also in individual stock performance, with the valuation gap between leading stocks and small and medium-cap stocks further expanding, and market concentration continuing to increase.
\n\nCapital Flow: Great Migration from Computing Power Chain to Application Layer
\n\nCapital flow monitoring data shows that since September, capital flows in the AI concept stock market have shown significant structural changes. Mainstream capital has continuously withdrawn from the computing power chain sector, with a net outflow of about 8.5 billion yuan for the month, with sub-sectors such as GPU manufacturers and AI chip design companies becoming the hardest hit areas of capital outflow. Meanwhile, the AI application layer sector received about 6.2 billion yuan in net inflow, showing an obvious reverse layout feature.
\n\nLooking at sub-sectors, the smart risk control track has become the most capital-concentrated area, with a net inflow of 2.8 billion yuan for the month, accounting for more than 45% of the net inflow of the AI application layer sector. The RWA on-chain finance track followed with a net inflow of about 1.9 billion yuan, showing that institutional recognition of this emerging track is rapidly increasing. In contrast, pure conceptual sectors such as AI large models and AI content generation faced continuous capital outflow, with a net outflow of more than 3.5 billion yuan for the month.
\n\nNotably, this capital flow is not a short-term phenomenon but has continued since the beginning of August. Analysis shows that this change reflects the market's re-evaluation of AI investment logic, shifting from the early "computing power arms race" to "application value implementation", with capital gradually returning to AI application areas that can generate actual commercial value.
\n\nHot Sector Analysis: Why Smart Risk Control and RWA Tracks Became New Favorites for Risk Avoidance?
\n\nAmong many AI application tracks, why did smart risk control and RWA on-chain finance become new favorites for institutional capital risk avoidance at the end of September? This is due to multiple factors working together.
\n\nSmart Risk Control Track: Perfect Combination of AI + Finance
\n\nThe reason why the smart risk control track has received capital favor is first that its business model has been market-verified. Compared with traditional risk control, AI-driven smart risk control systems have significant advantages such as wide data dimensions, fast response speed, and accurate risk identification, which can effectively reduce the non-performing loan rate of financial institutions and improve risk control efficiency. According to industry data, financial institutions using AI risk control can reduce their non-performing loan rate by 20%-30% on average, while increasing approval efficiency by more than 50%.
\n\nSecond, the smart risk control track has clear regulatory compliance and lower systemic risk. Against the background of increasingly strict global financial regulation, compliance has become an important consideration for financial institutions when choosing technology suppliers. AI risk control systems can not only meet regulatory requirements but also optimize risk control models through continuous learning, forming a virtuous cycle.
\n\nIn addition, leading enterprises in the smart risk control track have established obvious competitive barriers. These companies usually have rich financial data accumulation, professional risk control models, and complete customer service systems, making it difficult for new entrants to form effective competition in the short term. According to market research, the top five smart risk control enterprises have occupied more than 60% of the market share, and industry concentration continues to increase.
\n\nRWA On-chain Finance Track: Innovative Integration of Real Assets and Blockchain
\n\nThe RWA (Real World Assets) on-chain finance track is another major capital inflow hotspot. RWA refers to tokenizing assets in the real world (such as real estate, bonds, art, etc.) through blockchain technology to achieve the digitalization, divisibility, and tradability of assets. The application of AI technology in the RWA field has brought revolutionary changes to this track.
\n\nFirst, AI technology can significantly improve the efficiency and accuracy of RWA asset valuation. Traditionally, the valuation of non-standard assets relies on manual review, which is inefficient and highly subjective. AI technology can conduct comprehensive and multi-angle valuation of assets through big data analysis, image recognition, natural language processing, and other means, greatly improving valuation efficiency and accuracy.
\n\nSecond, AI technology can optimize the risk management of RWA assets. RWA assets usually have characteristics of high value and poor liquidity, making risk management particularly important. AI technology can provide more accurate risk early warning and management solutions for RWA assets through real-time monitoring of market changes and analysis of related risk factors.
\n\nFinally, the RWA on-chain finance track has received active layout from traditional financial giants. Since this year, top financial institutions such as BlackRock and JPMorgan have successively launched RWA-related products, accelerating the commercialization process of this track. The participation of institutional investors has not only brought financial support to the RWA track but also enhanced its market recognition and influence.
\n\nLeading Stock Analysis: Performance Differentiation, the Strong Get Stronger
\n\nAgainst the background of market differentiation, the performance of leading AI concept stock enterprises also showed clear differentiation. Overall, leading enterprises in the smart risk control and RWA tracks showed strong performance, while pure computing power chain enterprises faced greater pressure.
\n\nIn the smart risk control field, leading enterprises such as Zhipu AI and Baikong Cloud showed outstanding performance. Zhipu AI's new-generation smart risk control system released this month improved the risk recognition accuracy to 98.5%, received orders from several large banks, and its stock price rose by 15.2% for the month. Baikong Cloud, with its AI-driven anti-fraud system, continues to expand its market share in the financial technology field, with its stock price rising by 12.8% for the month.
\n\nIn the RWA on-chain finance field, enterprises such as Chainlink and OpenLaw showed bright performance. Chainlink's RWA tokenization platform has successfully connected with several traditional financial institutions, with a cumulative asset processing scale exceeding $5 billion, and its stock price rose by 18.6% for the month. OpenLaw, with its AI-driven smart contract platform, has established obvious advantages in the legal compliance of RWA assets, with its stock price rising by 16.3% for the month.
\n\nIn contrast, leading enterprises in the computing power chain faced greater pressure. NVIDIA and AMD, GPU manufacturers, faced slower market demand, with their stock prices falling by 7.8% and 9.2% respectively for the month. AI chip design enterprises such as Cambrian and Hygon Information also faced similar difficulties, with their stock prices falling by more than 10% for the month.
\n\nInvestment Strategy Suggestions: Focus on Application Layer, Grasp Structural Opportunities
\n\nBased on the current AI concept stock market situation, we believe investors should adopt the following strategies:
\n\n- \n
- Focus on smart risk control and RWA tracks: These two tracks have shown clear commercial value and market recognition, and are the most deterministic investment directions in the current AI concept stocks. It is recommended to pay attention to leading enterprises in these two tracks, such as Zhipu AI, Baikong Cloud, Chainlink, OpenLaw, etc. \n
- Avoid pure computing power chain risks: In the short term, the computing power chain sector still faces capital outflow pressure, and investors should treat it with caution. For investors who have already laid out in the computing power chain, it is recommended to appropriately reduce positions and wait for market stabilization signals. \n
- Pay attention to the integrated application of AI + traditional industries: In addition to the smart risk control and RWA tracks, the application of AI in traditional industries such as healthcare, education, and manufacturing is also worthy of attention. These fields have clear business needs and policy support and are important directions for AI technology implementation. \n
- Grasp structural opportunities: The differentiation of the AI concept stock market is intensifying, but there are still structural opportunities in sub-sectors. Investors should conduct in-depth research on the competitive landscape and business models of various sub-sectors to find enterprises with core competitiveness. \n
Conclusion: Value Reconstruction from Computing Power Faith to Application Implementation
\n\nThe AI concept stock market at the end of September 2026 is experiencing a value reconstruction from "computing power faith" to "application implementation". As AI technology gradually matures, the market is shifting from early concept speculation to attention on actual commercial value. The fact that smart risk control and RWA tracks have become new favorites for capital risk avoidance is a concentrated manifestation of this trend.
\n\nLooking ahead, the AI concept stock market will continue to show a differentiated pattern, but the overall trend is clear: AI application enterprises that can generate actual commercial value and have clear profit models will receive more capital favor, while enterprises engaged in pure concept speculation will gradually be marginalized. For investors, grasping this trend and focusing on the AI application layer will be the key to investing in AI concept stocks in the future.
\n\nAgainst the background of increasing global economic uncertainty, the differentiation of the AI concept stock market may further intensify. Investors should remain rational, pay attention to the fundamentals and long-term value of enterprises, and avoid blind following. Only those enterprises that can truly transform AI technology into commercial value can stand out in the fierce market competition and bring long-term returns to investors.
