China securities app MAU falls 1.87% MoM to 183M in Aug; analyst calls it temporary correction

Reporter: Sun Yongle Editor: Jiang Shiqiang, Xiao Jia Source: 21st Century Business Herald

Securities App MAU Declines Month-on-Month

In August 2026, stock market trading volume retreated, leading to a decline in active users of securities-related apps.

According to the latest data from Analysys Qianfan, the overall monthly active users (MAU) of securities service applications in August stood at approximately 183 million, down 1.87% month-on-month. This marks the first sequential decline since the Spring Festival, though it still represents a 5.78% year-on-year increase.

Yang Ling, a securities expert at Analysys Qianfan, believes this is more of a phased correction following a period of high activity, rather than an exodus of low-frequency users or a loss of core existing users.

Contrast with Short-Term MAU Fluctuations: Accelerated AI Investment

In contrast to the short-term fluctuations in app MAU, brokerages are accelerating their investment and competition in the field of artificial intelligence (AI).

On one hand, brokerages' proprietary apps are iterating with a focus on optimizing trading tools, transforming wealth management, and implementing AI capabilities. On the other hand, investment in online customer acquisition continues to increase, with top-tier brokerages now allocating budgets in the tens of millions of yuan. More notably, brokerages' AI strategies are shifting from private deployment to open ecosystems, leveraging third-party AI platforms to output professional capabilities and compete for new traffic gateways in the AI era.

MAU Trends in 2026

Looking at the trend for the year:

• January: ~184 million MAU
• February: Fell to a year-to-date low of 170 million due to the Spring Festival
• March to July: Rose for five consecutive months, reaching a year-to-date high of 187 million in July
• August: First decline of the year

Yang Ling explained to a reporter from 21st Century Business Herald that the simultaneous decline in MAU and daily active users (DAU) in August was mainly due to decreased market trading activity and lower participation from incremental capital. However, the MAU scale remains at a relatively high level, with a nationwide penetration rate of 16.72%, indicating a stable user base.

She further noted that compared to July's high-volatility environment, which triggered concentrated demand for monitoring, rebalancing, and risk management, investor high-frequency interactions cooled in August, leading to the decline in both MAU and DAU. Therefore, August is better viewed as a phased correction after a period of high activity, rather than a loss of users.

New Account Data Supports the View

This assessment is corroborated by new account opening data for August, which reflects short-term fluctuations but a strong cumulative growth trend for the year.

• August 2026: 2.3973 million new A-share accounts were opened, a 9.55% year-on-year decrease and a 9.72% month-on-month decrease, marking the lowest single-month level for the year.
• Cumulative (Jan-Aug): Total new A-share accounts on the Shanghai Stock Exchange reached 25.214 million, a 46.49% increase from 17.2118 million in the same period last year. This figure is equivalent to 91.90% of the total for the entire year of 2025 (27.4369 million).
• Projection: If the current average pace of 3.15 million new accounts per month continues, the total for the year could approach 38 million, surpassing the 2025 total of 27.43 million and setting a new historical record. This suggests that while short-term data fluctuates, the overall trend of investors entering the A-share market remains unchanged.

Margin Trading (Two-Finance) New Accounts

• August 2026: 107,400 new margin trading accounts were opened, a 41.31% year-on-year decrease and a 22.96% month-on-month decrease, hitting the lowest single-month level for the year.
• Cumulative (Jan-Aug): Total new margin trading accounts reached 1.2074 million, a significant 31.42% year-on-year increase.
• End of August: Total margin trading accounts stood at 16.6754 million.

Brokerage Proprietary Apps Focus on AI Implementation

At the level of brokerages' proprietary apps, the MAU of the top 20 all declined month-on-month.

• Huatai Securities (ZhangLe Wealth Connect): 12.7241 million MAU, continuing to lead, down 1.60% month-on-month.
• Guotai Haitong (JunHong): 11.1689 million MAU, down 1.35% month-on-month. These are the only two platforms with over 10 million MAU.

Ranking Changes (August vs. July): Most positions remained unchanged, with only three swaps:

• CITIC Securities (Xin e Tou) rose to 4th place, pushing China Merchants Securities to 5th.
• Xiao Fang rose to 10th place, pushing Zhongtai Qifutong to 11th.
• Everbright Securities (Jin Yangguang) rose to 14th place, pushing SDIC Securities to 15th.

Third-Party Platforms

The "Big Three" platforms—Tonghuashun, East Money, and Dazhihui—continued to dominate the top three:

• Tonghuashun: 37.8639 million MAU, down 2.10% month-on-month, maintaining a significant lead.
• East Money: 18.9414 million MAU, down 2.74% month-on-month.
• Dazhihui: 13.7448 million MAU, down 1.43% month-on-month.

AI-Focused Iterations

The MAU scale reflects not only the quality of brokerage app development and online customer acquisition but also indirectly indicates the ranking and trends in AI competition among brokerages.

In August, version iterations of mainstream brokerage proprietary apps focused on:

• Optimizing trading tools
• Transforming wealth management
• Implementing AI capabilities

Examples:

• Huatai Securities (ZhangLe Wealth Connect): Optimized intelligent account diagnostics and conditional order strategies.
• Guotai Haitong (JunHong): Advanced intelligent customer service and personalized news push.
• CITIC Securities (Xin e Tou) & China Merchants Securities: Enhanced integration of sector capital flow and top trader data.
• CSC (Qingting Diangjin) & GF Securities (Yi Taojin): Focused on iterating their wealth management marketplaces.

Yang Ling categorizes the current AI iterations by brokerages into three directions:

1. Trading and Decision Support: Focused on high-frequency scenarios like natural language stock selection, intelligent monitoring, and conditional orders. Representative products include Huatai's AI ZhangLe and GF's Yi Taojin AI Lens.
2. Portfolio and Wealth Management Services: AI enters areas such as account diagnostics, portfolio analysis, asset allocation, and investment advisory support. Guotai Haitong's Lingxi App 3.0 is a typical example.
3. Downward Deployment of Professional Analysis Capabilities: CSC's client terminal launched AI Stock Diagnosis, transforming professional capabilities like individual stock analysis into intelligent services for clients.

Competition among brokerage proprietary apps is shifting from a single trading channel to a comprehensive capability contest centered on intelligent decision-making, wealth management, and ecosystem services.

Brokerages Increase Online Customer Acquisition; Top Budgets Reach Tens of Millions

In 2026, while app MAU continues to face pressure, brokerages are increasing their investment in online customer acquisition.

On September 15, Tencent's official marketing WeChat account disclosed that in 2026, the company and brokerages had established a full "account opening and customer acquisition" chain. Over 40 traditional brokerage headquarters have invested in customer acquisition this year, with top-tier brokerages allocating budgets in the tens of millions of yuan. Branch offices and sales departments are also accelerating their entry. The cooperation model has evolved from basic advertising to deep business development involving customized mini-programs, account opening, investment advisory, and live streaming.

The online customer acquisition strategy for brokerage account openings unfolds across four stages:

1. Cognitive Reach: Using large-scale content on platforms like Video Accounts and Moments.
2. Interest and Trust: Building trust through authoritative media endorsements and the "licensed institution" gold label on mini-programs.
3. Active Search: Intercepting high-intent traffic with search ads.
4. Conversion and Closure: Using Enterprise WeChat and mini-programs to facilitate account opening.

With the help of tools like AIGC and digital humans, brokerages can produce account opening materials in bulk at a lower cost.

This shift is driven by the customer acquisition window created by the 25.214 million new A-share accounts opened from January to August 2026, a year-on-year increase of approximately 45%. Brokerages have entered a heavy-investment phase in online customer acquisition, with most doubling their spending on business promotion and customer acquisition fees.

Industry insiders generally believe that with the implementation of the "Measures for the Management of Online Marketing of Financial Products" on September 30, brokerages' online customer acquisition will place greater emphasis on headquarters-level compliance coordination. The customer acquisition chain is shifting from decentralized branch-level investments to headquarters-coordinated, full-chain operations.

Brokerage AI Enters the Era of Intelligent Agents

Notably, beyond the rankings, the financial sector, where brokerages operate, has seen a flurry of recent activity, positioning it as the next key battleground for AI implementation.

Since September, global leading AI companies have made intensive moves in the financial industry:

• September 10: OpenAI released ChatGPT for Financial Services.
• September 14: Anthropic launched Claude for Financial Advisors.
• September 17: Moonshot AI (Kimi) released its financial industry solution.

From overseas to domestic, top model companies have almost simultaneously turned their attention to the financial sector.

Focusing on the brokerage industry, most institutions have not separately disclosed their specific AI investments, making it more appropriate to observe from a resource allocation perspective.

A recent research report from Dongwu Securities suggests that "AI + Finance" is moving from the large model stage to the Agent intelligent agent era. B-end financial institutions like brokerages have taken the lead in initiating technology deployment and scenario implementation. The current mainstream path is to use general-purpose open-source large models as a base, combined with internal databases, knowledge bases, and business systems for private deployment, balancing data security and business adaptability.

DeepSeek, with its open-source and low-cost advantages, has become the preferred choice for local deployment by brokerages. By March 2025, over 20 brokerages had completed local deployment of DeepSeek-R1. After the release of DeepSeek-V4 in April 2026, several top-tier brokerages, including Guotai Haitong, Industrial Securities, SDIC Securities, Zhongtai Securities, Sinolink Securities, and Shanxi Securities, completed upgrade deployments. Huatai Securities, Guotai Haitong, GF Securities, CICC, and China Merchants Securities have all completed varying degrees of deployment and application, with the technical foundation largely in place.

Relying on private deployment bases, brokerages have implemented Agent capabilities in core business areas such as intelligent Q&A, industry research, investment advisory, risk control, and institutional services. In the investment research field, intelligent agents can handle repetitive foundational tasks like data queries, information sorting, and report preparation, significantly freeing up analysts' time and driving the intelligent upgrade of B-end investment research models.

From Internal Efficiency to Open Ecosystems

As private deployment matures, brokerages' AI capabilities are moving from internal efficiency improvement to external ecosystem openness.

A previous report by 21st Century Business Herald ("Brokerages Flock to AI Platforms like WorkBuddy, Vying for the Next Traffic Gateway") noted that recently, multiple brokerages have been intensively entering third-party AI ecosystems. For example, they are launching Skills packages or Buddy application zones on Tencent WorkBuddy, and conversational financial agents on Alibaba's Qianwen app. Previously, brokerages' self-developed AI investment research and AI investment advisory functions were mostly deployed within their own apps. Now, brokerages are moving their professional capabilities into external platforms, a trend driven by the competition for new traffic gateways in the AI era.

"Top-tier brokerages are concentrating resources on large model bases, computing power, financial knowledge bases, intelligent agent platforms, and business system upgrades. Small and medium-sized brokerages are more likely to adopt third-party model access and lightweight development, prioritizing the resolution of specific scenario implementation issues. AI is accelerating its integration into investment research, investment advisory, trading, and wealth management processes, while simultaneously opening up capabilities to external AI ecosystems through Skills, Agents, and other methods," Yang Ling pointed out.

SFC Produced by: 21 Finance Client, 21st Century Business Herald Editor: Zhang Jiayu