China Securities Industry (Brokerages) In-Depth Research Report

Date: 2026-07-03 Industry: Securities (Brokerages) / Financials Research Archetype: Financials (Brokerage Sub-Sector) Analyst: Loyan AI Industry Research

---

I1. Industry Definition and Scope

This research report focuses on the China securities industry, i.e., securities companies (brokerages) approved by the China Securities Regulatory Commission (CSRC) and engaged in securities business within China. As of the end of 2025, the industry comprised 150 securities companies, with total assets of RMB 14.83 trillion, net assets of RMB 3.34 trillion, and net capital of RMB 2.44 trillion.

Revenue Structure (2025 full-year industry total RMB 541.2 billion): Proprietary investment 34.2%, Brokerage 33.7%, Net interest income 11.9%, Investment banking 7.4%, Asset management 4.4%, Other 8.3%.

Business Chain: Retail/institutional clients → Securities firms (brokerage/investment banking/asset management/proprietary trading/credit) → Exchanges → Clearing and settlement.

Sub-Sector Positioning: Brokerages fall under the financial archetype, with their value drivers centered on "capital market activity × capital scale × license premium." Profitability exhibits significant high-beta characteristics – for every 10% change in ADT, brokerage income changes by approximately RMB 18 billion; for every 10% change in equity markets, proprietary investment income changes by roughly RMB 15-20 billion. Brokerages are leveraged balance sheet businesses: profits can be amplified or instantly wiped out by market conditions, with the greatest risk hiding before a cyclical reversal.

Sources: Securities Association of China (SAC) "Analysis of Securities Company Operations 2025"; Deloitte "China Securities Industry 2025 Review and 2026 Outlook"

---

I2. Demand: Multi-Dimensional Mapping of Capital Market Activity

The "demand side" of brokerages is essentially a function of capital market activity and financing needs. In the first half of 2026, A-share activity set multiple historical records, but beneath the prosperity, structural divergence was acute.

2.1 A-Share Average Daily Turnover (ADT): RMB 2.74 Trillion – "Fire and Ice"

H1 ADT hit a semi-annual record high – but beneath the boom lies extreme divergence: In H1, the median A-share stock return was approximately -14%, with nearly 70% of stocks posting negative returns. The three AI-related sectors – Electronics, Computers, and Telecommunications – contributed 48.6% of turnover, and the technology sector's turnover share reached a record 38% (semiconductor market cap of RMB 14.14 trillion surpassed state-owned big banks). This is not a broad bull market but rather an "absorption effect" of the AI theme – the daily turnover elasticity coefficient was only 0.105 (a 10% increase in turnover drives only a 1% index gain), far below the 2019-2021 bull market average of 0.45.

Sources: Securities Times · Data Treasure "RMB 317.5 Trillion! Total Turnover Nearly Doubled YoY" 2026-07-01; Shanghai Securities Industry Association commission rate data; Caixin

2.2 Margin Trading and Securities Lending (MT&SL) Balance: Breaks RMB 3 Trillion, Approaching Historical Extreme Zone

From 1.66 to 3.06 trillion in two and a half years, an increase of 84%. The margin balance as a share of free-float market cap is about 2.4%, far below the extreme 4.27% in 2015, but the pace of increase has been faster than expected. More critically, regulators have proactively tightened leverage: Since January 2026, the margin financing margin requirement has been raised from 80% to 100% (leverage reduced from 1.25x to 1x), directly cutting new margin buying power by 25%. Current margin financing rates have dropped to 3.0-3.8% (some leading brokerages quoting below 3%), and the net interest spread on MT&SL has narrowed from 2.2-2.5% in 2024 to approximately 2.0-2.2%.

Sources: Southern Finance · East Money MT&SL data; Securities Times

2.3 New Account Openings: Accelerated Retail Inflow, Over 20 Million in H1

Total A-share investors rose from 237 million at end-2024 to approximately 260 million by mid-2026. Penetration rate is about 18.6%, still having room to double when benchmarked against mature markets (US ~50%). However, there is a risk of "hot account openings, cold trading" – many new accounts are short-term speculators attracted by the AI rally.

Sources: Shanghai Stock Exchange monthly account opening data; Caixin; Yicai

2.4 Equity Financing: IPO Recovery + Six New Refinancing Rules Open New Space

On July 3, 2026, the CSRC issued a consultation paper on six new refinancing rules – the most important policy increment in this reporting cycle: Shelf registration system breaks ground, small-and-fast cap raised from RMB 300 million to 600 million, full adoption of market-price issuance, and simplified controlling shareholder private placements. Shelf registration will provide companies with continuous financing channels, opening up incremental investment banking deal flow. However, note that the 2020 refinancing reform only increased industry revenue by 2-4% (Anxin Securities estimate), so the direct elasticity of refinancing reforms on brokerage profitability is limited.

For IPOs, 71-75 companies were listed in H1 (Beijing Stock Exchange 36, accounting for half), and 49 new applications were accepted on the STAR Market (+133%). Giant projects like Changxin Technology, Yangtze Memory Technologies, Kunlun Core, and Unitree Robotics are queuing for listings totaling over RMB 200 billion, with normal approval pace.

Sources: Wind; CSRC six refinancing rules consultation paper (2026-07-03); Shanghai Securities News; Anxin Securities

2.5 Bond Underwriting and Asset Management: Stabilizers Continue to Push

• Bond Underwriting: In H1, brokerages underwrote approximately RMB 4.29 trillion in bonds (including local government bonds), +49% YoY. CITIC Securities RMB 1.19 trillion, Guotai Haitong RMB 891.6 billion, and CSC Financial RMB 821.6 billion ranked top three. Since the launch of the Technology Board bond market in May, sci-tech innovation bond underwriting reached RMB 328.3 billion.
• AM AUM: At end-2025, entrusted assets in management totaled RMB 10.21 trillion (+5.5%), with collective asset management accounting for 33.7%, surpassing single-asset management for the first time, deepening active management transformation. In 2026 Q1, asset management net revenue was RMB 13.5 billion (+27%).

Sources: Wind; 21st Century Business Herald; SAC

2.6 Demand Bridge: 2025→2026E Profit Driver Decomposition

Structural vs Cyclical: Structural growth ~55% (household asset migration + registration-based IPO system + institutionalization), cyclical fluctuation ~45% (ADT volatile with market sentiment + proprietary beta). 2026 is in a phase where structural upward shift and cyclical peak overlap, but the cyclical component cannot be ignored.

---

I3. Supply: Asset Quality and Capital Constraints

The "supply side" of brokerages is essentially the constraint boundary of capital strength and risk control capability – net capital scale determines business ceilings, and risk control indicators set expansion limits.

3.1 Net Capital and Risk Control Indicators: Generally Adequate, Tightening for Leaders

Industry-wide risk control indicators are well above regulatory red lines, but leading brokerages, due to large business scale, have indicators closer to the floor – this is both a sign of "efficient capital utilization" and a source of vulnerability. Tianfeng Securities' net stable funding ratio broke below the warning line, while CITIC Securities' liquidity coverage ratio approached the warning, indicating that bond issuance to expand balance sheets has materially compressed the safety margin of some brokerages.

Leading Brokerages by Net Capital (End 2025): Guotai Haitong outstanding bonds RMB 344.1 billion (industry first), CITIC Securities outstanding bonds RMB 273.2 billion (second); leading brokerages account for 58.7% of industry bond issuance. Industry operating leverage is approximately 4.4x vs. international investment banks 10-15x; the room to increase leverage is a core driver of ROE improvement.

Sources: SAC; Caixin "Bond Blitz of RMB 1.25 Trillion" 2026-06-30; Lianhe Ratings

3.2 Proprietary Investment Exposure: High-Beta Double-Edged Sword – Divergence Intensifies

At end-2026 Q1, the total financial investments of 43 listed brokerages amounted to RMB 7.41 trillion, +4.67% from the beginning of the period. Among them, trading financial assets increased +6.61% from the beginning of the year, and other equity instrument investments +9.24%.

Core Data: Proprietary Investment Income Extremely Divergent – Also the Most Critical Indicator for Assessing Brokerage Asset Quality:

Huatai Securities' trading financial assets at end-Q1 reached RMB 443.8 billion, +17% QoQ, with clear TPL expansion – meaning its profit volatility is directly linked to market direction. CITIC Securities proprietary income was RMB 11.679 billion (+32%), Guotai Haitong RMB 5.839 billion (+46%), GF Securities RMB 5.607 billion, ranking top three, together accounting for over 80% of disclosed data.

Sources: Kaiyuan Securities 2026 Q1 results review; Listed brokerage Q1 reports

3.3 M&A Integration Wave: Scale Jump, ROE Yet to Be Verified

Guotai Haitong's first-year data post-integration is the most important warning: Book ROE 11.1% (including RMB 8.827 billion one-time negative goodwill gain), but ex-one-off ROE was only ~5.75% (ex-one-off net profit RMB 18.982 billion / net equity ~RMB 330.4 billion). Credit impairment losses were RMB 3.863 billion, surging 1,445% YoY – primarily from regular provisions on new financial leasing business added through the absorption of Haitong and accounting treatment requirements for non-identical control business combinations. This reveals that mergers bring not only scale advantages but also full exposure of legacy risk assets. The narrative that "M&A integration boosts ROE" has been falsified by first-year data; integration effects need 2-3 years to materialize.

Sources: Guotai Haitong 2025 Annual Report; Caixin; 21st Century Business Herald

3.4 Talent Structure: Accelerated Shift from "Mass Recruitment" to "Investment Advisors"

CITIC/CICC/CMS broker counts have dropped to zero; Huatai has only 4 remaining. The shift from "mass recruitment tactics" to professional advisor-driven transformation is a long-term structural positive for the industry.

---

I4. Profitability · Asset Quality · Capital (Headline Section)

4.1 Profitability Panorama: From Cyclical Trough to High Cycle, but Extreme Divergence

2025 net profit increased +37.3% YoY, but industry ROE remained only 6.79%. In 2026 Q1, 43 listed brokerages generated net profit of RMB 64.9 billion (+16%), with the top four's annualized weighted ROE of 12.7% already exceeding the 2021 high – but caution: investment income contributed approximately 40-50%, and investment income is highly cyclical.

4.2 Fee-Based vs. Capital-Based Income: Structure Becoming More Balanced but Still Highly Dependent on Markets

2026 Q1 trend continued: fee-based income share rose to ~51% (brokerage +44% leading), proprietary share ~37%. Fee-based and capital-based income have moved from "30-70 split" toward "50-50 split," reducing over-reliance on a single proprietary beta, but proprietary trading remains the largest single profit pole (37.5%) and the core variable for performance divergence.

4.3 DuPont Decomposition: Three Engines and Ceiling for ROE Improvement

2026 Q1 Annualized ROE ~7.7% DuPont Decomposition:

• Net Profit Margin: ~39.6% (2025: 40.55%) – commission rate decline is the main drag
• Asset Turnover: ~4.4% (2025: 3.65%) – driven by doubling ADT
• Equity Multiplier: ~4.44x (2025: 4.44x)

The top four (CITIC/HT/GF/CICC) had better DuPont quality with ROE ~12.7%: net profit margin ~42%, turnover ~5.0%, leverage ~5.5x. But leverage is only 4.4-5.5x, vs. international investment banks 10-15x – this is both a constraint and potential. Leading brokerages' net capital to total liabilities ratios are already near their own risk control ceilings (CITIC 18.51%/HT 18.97%), leaving limited room to expand balance sheets for ROE improvement; international subsidiaries (offshore leverage 7-10x, ROE >20%) are the direction for breakthrough.

4.4 Spread Trends: Commission Rates and MT&SL Spreads Both Under Pressure

Commission rates are declining approximately 15-20% per year. In May, Hebei CSRC bureau reiterated a ban on malicious low-price competition, but the floor is only "no lower than comprehensive cost of about 0.005-0.006%". The "volume offsets price" logic held in 2026 H1 (ADT +95.7% covering commission rate decline), but if ADT mean-reverts to ~RMB 1.5 trillion, price downward pressure will become prominent. The limit of volume offsetting price: if the net spread narrows to 1.5%, MT&SL balance would need to reach RMB 4 trillion to maintain the same interest income.

Sources: Shanghai Securities Industry Association; Securities Daily; Hebei CSRC Bureau

4.5 Earnings Forecast: Three-Scenario Framework

Probability-weighted expected net profit: ~265-270 billion. The base case implies H2 ADT of ~2.46 trillion (a ~10% cooling from H1's 2.74 trillion), already internalizing mean-reversion risk. 2026Q1 net profit of 43 brokers at 64.9 billion, if simply annualized, equals 259.6 billion—so the base forecast is not aggressive but relatively conservative extrapolation.

Source: GF Securities neutral assumption (+27%); Western Securities (+18.3%); Soochow Securities (+36%); three-firm weighted average ~270 billion

---

I5. Competitive Landscape and Profit Pool (Headline Section)

5.1 Concentration: Matthew Effect Accelerating, CR5 Net Profit Captures Half of the Industry

Concentration accelerating upward: 2026Q1 net profit CR5 up +1.7 pct YoY, CR10 up +1.2 pct YoY. Top 10 brokers accounted for 58.7% of bond issuance, exceeding the sum of the bottom 64. M&A integration (Guotai Haitong, Guolian Minsheng) + classified supervision "supporting the strong and limiting the weak" are the core drivers.

5.2 Top Player Profiles: Extreme PB-ROE Dispersion

Key Finding: Traditional top brokers trade at PB of 1.0-1.7x, in the low-to-mid percentile of the past 10 years. Huatai PB 1.07x (38th percentile) vs ROE 9.2%—if ROE is sustainable, valuation indeed appears low; however, its trading financial assets of 443.8 billion (41% of total assets) expose it to TPL volatility, justifying the discount. CICC PB 1.71x (14th percentile) with ROE 11.2%, and PE TTM at the 3rd percentile of 5 years—most pronounced valuation contradiction. East Money PB 3.49x (14th percentile) reflects an internet platform premium but has fallen significantly from historical highs.

Source: Kaiyuan Securities; Shenwan Hongyuan; Western Securities; various broker annual reports/Q1 reports

5.3 Profit Pools: Who Earns Real Money? Who's Chasing Cyclical Mirage?

Real Money (sustainable high ROIC):

1. Client-Driven Proprietary Trading (derivative market-making/cross-border/OTC options) – CITIC 6.79 trillion notional principal, risk-neutral, balance sheet expansion driven, oligopoly among top players
2. Active Asset Management (CITIC AM 12.177 billion—industry's only 10-billion+; GF dual brands) – very high concentration (CR10 87%), relatively stable income
3. ETF Market-Making and Distribution (East Money/Huatai benefit from passive investment trend) – ETF AUM 3.2+ trillion
4. Margin Lending Interest (top brokers have 50-100bp lower funding costs, fatter spreads)

Cyclical Mirage (revenue growth without profit sustainability):

1. Directional Proprietary Trading – Q1: 27 brokers' proprietary income declined QoQ; top players +38% vs SMID -8%; high beta, high volatility. Largest profit source for brokers is also the largest risk exposure.
2. Pure Agency Brokerage – commission rates dropped 20% in one year; new account commissions as low as 0.064 bps; price war without bottom.
3. Sci-Tech Co-Investment Unrealized Gains – 10.5 billion book gain (24-month lock-up); Sci-Tech 50 median PE 128x (99th percentile). Historically, in 2022 during lock-up expiry, over half of co-investments turned to losses. The market is linearly extrapolating book gains into sustainable profits, committing the error of confusing beta with alpha.

5.4 Profit Migration Direction (2026-2028)

• Proprietary: Directional → Market-making/Neutral strategy (top players lead, SMID struggle to participate)
• Wealth Management: Commission from agency → Buyer's advisory allocation fees (distribution revenue +51% already validated)
• Investment Banking: IPO underwriting → Industry enabling + Cross-border + Continuous service under shelf registration
• With rising concentration, profits continue to concentrate among top 5-7 players; SMID broker profit margins are persistently compressed in the price war.

5.5 Return Quality Assessment

Is this business overall easy to make money from? – Below average. Industry ROE center ~6.5-7%, in boom cycles (2025-2026) ~6.8-8.5%, in downturn cycles (2022-2023) only 5.0-5.3%, far below banks (~10%) and insurers (~10-12%), and highly volatile. Only top brokers with capital moats, client stickiness, and business diversification, plus internet brokers, can sustainably generate excess returns.

In an upturn cycle, everyone makes money; in a downturn, who can hold the line? – Only those whose proprietary trading is mainly market-making/neutral, with high fee-based income ratio and strong capital. CITIC Securities and CICC (post-merger) have the strongest counter-cyclical resilience. Huatai's 0.80x forward PB (based on 2026E net assets) precisely reflects the market's reasonable discount for its high TPL exposure.

5.6 Industry Valuation Position

Shenwan Securities II Index PB ~1.12x (as of May 29, 2026), at ~6.4th percentile of the past 10 years. Historical average 1.50x, median 1.45x. Current 1.12x PB corresponds to historical ROE of ~4-5%, while 2026E industry ROE is ~8.1%—a clear mismatch between valuation and fundamentals. However, we must soberly recognize: low PB percentile does not guarantee mean reversion—if the market is pricing "ROE about to mean-revert from 8% to 6%," current PB may be fair rather than undervalued. Focus on top broker PB percentiles: CICC (14%), Huatai (38%), Guotai Haitong (32%) are indeed at historically low ranges, but CITIC (47%) is only mid-low.

---

I6. Policy · Geopolitics · ESG (Elevated to Headline Level)

6.1 Monetary and Interest Rate Environment: Low Rates Persist, July Politburo Meeting is Key Variable

Low-rate environment is triple-positive for brokers: bond proprietary investment gains, lower margin funding costs, and risk-free rate decline lifting equity valuations. The late-July Politburo meeting is the most important near-term macro catalyst—market expects "moderate easing" to continue and possibly counter-cyclical measures to be stepped up; RRR cut probability higher than rate cut.

Source: PBOC LPR quotes; ChinaBond

6.2 Six Refinancing Rules (Consultation Draft July 3) – Most Important Policy Increment This Report Cycle

Policy Assessment: Structurally positive, but limited elasticity. Based on 2020 refinancing reform experience, industry revenue boost is ~2-4%. Shelf registration favors top investment banks (CITIC/CICC/Guotai Haitong) due to stronger project pipeline and continuous service capability.

Source: CSRC "Opinions on Deepening the Reform of Listed Company Refinancing (Consultation Draft)" 2026-07-03

6.3 New Derivative Rules + Risk Control Relaxation + M&A Integration: Triple Institutional Dividends Concentrate on Top Players

• New Derivative Rules (published May, effective Nov 16): First ministry-level regulation on OTC derivatives; net capital ≥0.5 billion, prohibits stock derivatives for major shareholders. Short-term stock adjustment raises costs; medium-to-long term institutionalization benefits top compliant brokers—only 8 primary OTC option dealers
• Risk Control Indicator Relaxation (Jan 2025): AA-rated risk capital reserve coefficient from 0.5 to 0.4; high-quality brokers pilot internal model method
• "Building World-Class Investment Banks": Chairman Wu Qing at SAC Eighth Members' Meeting explicitly supported strong players and restricted weak ones, relaxing risk control indicators for high-quality institutions

6.4 Cross-Border Crackdown + Fee Reduction Pressure: One Bullish, One Bearish

• Eight-Department Illegal Cross-Border Crackdown (May 22): 2-year focused rectification; Futu faces proposed penalty of 1.85 billion, Tiger 0.41 billion → capital flows back to Stock Connect/QDII/Cross-Border Wealth Connect, benefiting licensed Chinese broker cross-border businesses
• Public Fund Fee Reduction Deepening: After three phases, annual profit transfer exceeds 50 billion; commission rate already down to 1.73 bps—the most violent impact of fee cuts is over, subsequent declines will moderate

6.5 Comprehensive Assessment: Net Policy Direction Slightly Positive

---

I7. Core Debates and Scenarios

Debate 1: Is ADT Structurally Shifting to a New 2-Trillion+ Midpoint, or a Cyclical Spike Driven by AI Theme? (Key Decision D1)

Bull: Tech stocks' 38% trading share is an industrial trend—semiconductor/AI market cap has surpassed traditional financials; quantification + institutionalization + household deposit migration (75 trillion in fixed deposits maturing & reallocating) will lift ADT midpoint from 1 trillion to 1.8-2.0 trillion. 2026H1 new accounts 20.16 million (+60%) validates incremental capital inflow.

Bear: H1 ADT of 2.74 trillion saw 70% of stocks decline—this is AI suction, not a broad bull market. Daily volume elasticity coefficient is only 0.105 (vs historical 0.45); massive volume comes from high-frequency churn in tech stocks. Historically, ADT has never stabilized above 2 trillion for more than one quarter (after 2015 peak, 2018 daily average only 369.8 billion). Regulators have proactively tightened leverage (margin requirement 80%→100%); Huijin reduced broad-based ETF holdings 75-89% in Q1; CSFC reduced top broker holdings. Even the bull's own full-year ADT forecast (2.5-2.7 trillion) already implies H2 cooling.

Tracking Indicators: Whether full-market ADT falls below 1.8 trillion for 20 consecutive trading days; whether AI leaders (Zhongji Innolight, Cambricon) miss quarterly earnings; whether Sci-Tech 50 index retreats >20% from peak.

Red Team Verdict: Bear arguments are compelling. The concentration risk of ADT on AI theme and regulatory cooling signals are underestimated by the bull. Structural midpoint shift to 1.8-2.0 trillion is reasonable, but current 2.74 trillion includes significant cyclical premium. Revised ADT midpoint thesis: structural shift to 1.8-2.0 trillion + cyclical spikes to 2.5-3.0 trillion, not simply "new midpoint 2-2.5 trillion."

Debate 2: Are Top Brokers' Low PB a Value Trap or Valuation Discount? (Key Decision D2)

Bull: Shenwan Securities II PB 1.12x (6.4th percentile of 10 years) vs ROE 8.1%; historical PB at this level corresponded to ROE of only 4-5%. CICC PB 1.71x (14th percentile), Huatai PB 1.07x (38th percentile). Historically, similar mismatch windows (end-2018, Oct 2022) were followed by sector excess returns of 20-40% over 12 months. The top four's annualized ROE of 12.7% already exceeds 2021 highs; the valuation mismatch should correct.

Bear: "High ROE + low PB" is not a mismatch but market foresight—current high ROE is driven by investment income (contributing 40-50%), which is highly cyclical. Huijin/CSFC significantly reduced holdings in Q1 (CSFC cut CITIC by 177 million shares -46.3%, GF by 331 million shares -26.1%); "national team" votes with real money. A similar apparent "high ROE + low PB" mismatch occurred in 2015H1, then ROE fell from 15%+ to 6-7%, and PB dropped from 1.5x to below 1.0x. Huatai's trading financial assets 443.8 billion (41% of total assets); once tech stocks correct, unrealized gains become losses, and ROE will quickly fall to 7-8%.

Tracking Indicators: Whether top brokers' 2026 annual ex-nonrecurring ROE exceeds 10%; whether Guotai Haitong's impairment converges; whether Huijin/CSFC continue reducing in Q2.

Red Team Verdict: Bear arguments partially hold. Sector PB 1.12x at historical lows does have room for repair, but no longer an "extreme undervaluation" narrative—low PB includes a reasonable discount for ROE mean reversion. Differentiate: CITIC PB 1.47x (47th percentile) is reasonably low but not extreme; Huatai PB 1.07x (38th percentile) must consider high TPL risk; CICC PB 1.71x (14th percentile, PE 3rd percentile) shows the most significant mismatch.

Debate 3: Can 2026 Full-Year Earnings Growth Be Delivered? Can Six Refinancing Rules + Sci-Tech Investment Take the Baton? (Key Decision D3)

Bull: H1 ADT 2.74 trillion sets a high base; H1 net profit annualized ~280 billion. Six Refinancing Rules (shelf registration breakthrough) open incremental investment banking space for 2027; Sci-Tech investment elasticity (co-investment unrealized gains 10.5 billion + PE holdings 11 billion) is a unique alpha.

Bear: Co-investment unrealized gains of 10.5 billion are book gains (24-month lock-up); Sci-Tech 50 median PE 128x at 99th percentile—once tech stocks correct 30%, gains could shrink to 3-5 billion or turn negative. Refinancing Rules add only 2-4% to industry revenue (based on 2020 experience). 27 brokers already saw QoQ decline in proprietary income in 2026Q1; divergence between top players (+38%) vs SMID (-8%) indicates deteriorating overall earnings quality.

Tracking Indicators: Whether aggregate broker net profit in 2026Q3 falls below 55 billion; whether margin balance falls from 3 trillion to below 2.7 trillion; whether Sci-Tech 50 breaks below 700.

Red Team Verdict: Base forecast of 270 billion is relatively conservative; downside risk is partially priced. The book nature of co-investment gains is the most critical concern—rating must note "unrealized, 24-month lock-up, 99th percentile high valuation." The incremental contribution from Refinancing Rules is limited (<5%); do not over-extrapolate.

Three-Scenario Framework

Catalyst Calendar (2026H2-2027H1)

---

I8. Investment Implications and Beneficiaries

8.1 "Cyclical Direction × Business Quality" Decision Matrix

Warning for Bottom-Right Quadrant – Cyclically Bullish but Poor Return, a Value Trap:

• SMID pure agency brokers: ADT expansion boosts brokerage income, but commission rates continue to decline + excessive directional proprietary exposure; when the market reverses, earnings will collapse. Current high growth is "borrowed."
• Guotai Haitong (painful integration period): Ex-nonrecurring ROE only ~5.75%, credit impairment surged 1,445%; merger brings scale #1 but has not yet translated into shareholder returns.

8.2 Scenario × Target Stock Matrix

8.3 Beneficiary Targets

Top Picks (High Quality + High Cycle Engines):

1. CITIC Securities (600030) – Absolute industry leader, top three in all segments. Derivatives market maker leader (6.79 trillion notional principal), asset management revenue of 12.177 billion leading by a wide margin, international business contributes about 1/3 of total revenue. PB 1.47x (47th percentile) valuation reasonably low, strongest ability to navigate through cycles. Benefits from both upturn and provides defensive cover in downturn, the most certain target in the brokerage sector.

2. East Money (300059) – ROE 16% (annualized) highest in industry, Internet light-asset model with cost-to-income ratio only 25-30%. Dual-platform traffic moat of TianTian Fund + East Money Securities is irreproducible. PB 3.49x (14th percentile) has dropped sharply from historical highs to near 10-year lows. Greatest upside elasticity in a cyclical upturn, but high PB requires market activity support.

3. GF Securities (000776) – Solid asset management moat (GF Fund + E Fund stake), ROE annualized 11.5% entering first tier with steepest improvement slope. PB 1.35x (69th percentile) somewhat expensive, but stable asset management income provides downside protection.

Monitor (Awaiting Catalyst/Verification):

4. CICC (601995) – Strongest alpha target in investment banking brand + internationalization (overseas ROE 15.9%). After merging Dongxing Xinda, net capital will exceed 100 billion, scale leaping to third in industry. PB 1.71x (14th percentile), PE TTM near 3rd percentile over past 5 years – most prominent valuation contradiction. Integration execution risk is the core uncertainty.

5. Huatai Securities (601688) – Most effective in technology empowerment, ZhangLe Wealth Pass leads in monthly active users, wealth management gross margin 55.75%. But the TPL exposure of tradable financial assets of 443.8 billion (41% of total assets) is a double-edged sword. PB 1.07x (38th percentile) low but with reasonable discount element.

Avoid:

• Small pure channel brokers (brokerage revenue share >50% with no differentiation) – commission price war continuously compresses profit margins
• Small and mid-sized brokers with excessively large directional self-operated exposure and weak capital – Q1 self-operated sequential decline of 27 brokers is already a warning
• Guotai Haitong (601211) – PB 1.05x seemingly cheap, but core ROE only ~5.75% + sharp increase in credit impairment + huge interest expense pressure from 344.1 billion bonds. Integration effect requires 2-3 years to verify, not suitable for heavy position currently

8.4 Key Risks

1. ADT Mean Reversion Risk (High): H1 ADT of 2.74 trillion heavily relies on AI theme suction effect. If AI rally fades or quantitative regulation tightens further, ADT could fall back below 1.5 trillion. Elasticity coefficient of 0.105 indicates a small index decline can trigger a sharp shrinkage in turnover. Trigger monitoring: ADT below 1.8 trillion for 20 consecutive trading days.

2. Directional Self-Operated Exposure Risk (High): Self-operated revenue accounts for 37%, directional strategies still dominate. Huatai TPL of 443.8 billion is the largest single stock exposure. A 10% drop in A-shares could drag brokerage profits by 20-30%. 27 brokers already saw sequential decline in Q1, Q2 divergence may intensify.

3. Tech Investment Floating Profit Shrinkage Risk (Medium-High): The 10.5 billion co-investment floating profit is book unrealized gain (lock-up period 24 months). Sci-Tech 50 PE median of 128x is at 99th percentile. Once tech stocks correct 30%, floating profit may shrink to 3-5 billion or even turn negative, dragging Q3-Q4 profits by -10% to -20%.

4. Sustained CSF/Huijin Reduction Risk (Medium): In Q1, CSFC reduced CITIC Securities by 177 million shares (-46.3%), GF Securities by 331 million shares (-26.1%); Huijin reduced broad-based ETFs by 75-89%. "National team" systematically reducing financial exposure; if Q2 continues reductions, it will suppress sector valuation recovery.

5. M&A Integration Underperformance (Medium): Guotai Haitong core ROE of only ~5.75% falsifies the "M&A boosts ROE" thesis; CICC's three-in-one integration challenge is even greater.

6. Diminishing Policy Marginal Effect (Medium-Low): Low rates/registration-based IPO/risk control relaxation already largely implemented; six refinancing measures have limited incremental contribution (<5% boost), room for further surprise policy is limited.

---

Disclaimer: This report is compiled based on publicly available information and data for reference only and does not constitute investment advice. Investment involves risks; please enter the market cautiously. Specific targets mentioned in this report are solely part of industry analysis and do not constitute buy/sell recommendations. Investors should assess risks on their own and consult professional investment advisors when necessary.

Data cut-off date: 2026-07-03 Next update: Suggested after 2026Q3 securities firm earnings season (October 2026)