Rating: Cautiously Bullish | Target Price Range: HKD 9.0–9.8 (base-case fair value) | Current Price: HKD 7.82 (close of 2026-09-02) | Margin of Safety: approx. +15% (computed by the system as "(base-case fair value lower bound − current price) / current price") | Probability-Weighted Expected Return: approx. +16% | Time Horizon: 12 months
FX basis: 1 HKD = 0.85683 RMB (2026-09-03). Unless otherwise stated, all financial data in this report are in RMB; all valuations are anchored to the H-share close of HKD 7.82 on 2026-09-02.
| Dimension | Previous (2026-06-05) | Current (2026-09-03) | Drivers of Change |
|---|---|---|---|
| Stance | Bullish | Cautiously Bullish | Direction maintained but confidence downgraded: volume retreat shifted from "risk scenario" to "happening now"; Central Huijin consolidation option squeezed out |
| Confidence | 0.70 | 0.58 | Two core catalysts failed/were falsified within three months: the new five-year strategy lacks quantitative targets, and Galaxy is absent from the current round of Huijin-system consolidation; Q3 volume fell 14% QoQ |
| Target price range | HKD 8.76–11.10 | HKD 9.0–9.8 | Upper bound lowered: consolidation option and bull-market extrapolation removed; lower bound raised: H1 2026 net asset accretion (BVPS ex-perpetual bonds RMB 11.04) + interim earnings delivery |
| Valuation judgment | Undervalued (margin of safety +37%) | Undervalued (margin of safety approx. +15%) | PB methodology corrected: last time 0.49x on an including-perpetuals basis, described as an "extreme 5-year low percentile"; this time standardized to 0.61x ex-perpetuals — 43rd percentile over 3 years, 66th over 5 years. Absolute discount remains deep, but the percentile is not extreme; the degree of undervaluation narrows vs. the prior assessment |
| Delivery record | — | Interim dividend of RMB 0.15/share delivered (+20%); market volume delivered in H1 (ADT RMB 3.07 trillion, +57.3%) drove record interim results | Both dividend and volume theses delivered, but the stock is down 25% YTD — earnings and valuation diverging |
One-sentence summary of this update: Fundamentals delivered the previous bullish thesis (net profit +20%, a record high; dividend paid out), but the market has re-imposed "cyclical peak" pricing — the debate is no longer "is Galaxy worth anything," but "when, and on what catalyst, will the H-share broker discount converge."
China Galaxy's H1 2026 net profit attributable to shareholders of RMB 7.797 billion (+20.18%) hit a record high for the period, but underperformed the listed-broker industry average (+49%), rooted in low elasticity of proprietary investment; at the current H-share price of HKD 7.82, the stock trades at 0.61x PB (ex-perpetual bonds, 2026-09-02), the lowest among six leading brokers' H shares, with a PB/ROE ratio of 0.063 also the lowest in the group; the implied perpetual ROE backed out of the current price is only about 6.2%–7.3%, below the company's 2023 bear-market low (7.52%) — the market is pricing Galaxy as if "profitability stays forever at the bottom of the last cycle." Zero-growth earnings power value (EPV) is about HKD 10.5, roughly 34% above the current price. But clarity is required: about half the discount is a sector-wide structural phenomenon of H shares (the concurrent A-share 1.13x PB implies a 10.9% ROE, already consistent with fundamentals); realization depends on catalysts such as volume stabilization, a payout-ratio hike, or southbound inflows, and may never converge. Market volume retreated in Q3 (ADT RMB 2.45 trillion; below RMB 2 trillion in late August) — the biggest constraint; negative FY2027 attributable profit growth is this report's base case, not a risk scenario. Cautiously bullish; participate from the left, controlling drawdown via the dual floors of dividends (FY2026E approx. 5.7% gross; approx. 4.5% after tax for Stock Connect individual investors) and valuation; re-assess if the price breaks HKD 6.2 (bear-case floor).
Two honest qualifications: first, the low PB has percentile context — the 43rd 3-year and 66th 5-year percentiles coexist because H-share brokers traded at 0.3–0.5x for long stretches in 2021–22 (years of weaker fundamentals); 0.61x is low in absolute terms but not a historical extreme. Second, the concurrent A-share 1.13x PB implies an ROE of about 10.9%, already consistent with fundamentals, showing roughly half the H-share discount is a sector-wide structural phenomenon of southbound pricing power and liquidity discount, not pure mispricing — realizing the mismatch requires catalysts and cannot be assumed to auto-converge. The bear view also holds that the A/H premium stems from investor-structure and dividend-tax differences, and may widen rather than converge at the index level over time. This report's stance: the structural discount explains "why it is cheap" but does not overturn the mismatch fact that "implied ROE is below realized bear-market levels."
Boundary conditions must be clear: share resilience ≠ revenue resilience. A +0.09pp margin share gain (+1.7% relative) is far from enough to offset a shrinking market — industry margin balance has fallen 11.1% from the June peak of RMB 3.0 trillion, for a net effect of roughly -9.5% on Galaxy's margin-related revenue; wealth management revenue is a lagged confirmation of H1's RMB 3.07 trillion ADT, and Q3's -14.3% ADT retreat will show up in the Q3–Q4 reports. This thesis asserts "cyclical-chip characteristics with an intact base," not short-term defensiveness. Note: market reports (Eastmoney, 2026-09-01) put Galaxy's net-profit ranking "down to seventh" after the interim results, differing from this report's ranking of leading brokers (5th) due to sample-methodology differences (treatment of ties, inclusion of unlisted brokers); the directional conclusion of ranking decline (the mirror of underperforming the industry's +49%) is consistent.
Three qualifications: ① Split methodologies — the H-share segment figure of RMB 3.047 billion (including interest and investment income from the Hong Kong + Southeast Asia platforms) differs by nearly 2x from the RMB 1.688 billion (+53.52%) disclosed for Galaxy International Holdings in A-share filings; this report uses the H-share segment basis and notes it here. ② Segment revenue includes beta from the HK bull market, not purely operating income. ③ Evidence for the "second curve" remains thin — the 6 IPOs are a list of deals from 2026's big year; IPO underwriting revenue is highly cyclical, and 2027 pipeline plus Southeast Asia retail revenue/profit margins have no itemized disclosure. At a FY2025 segment margin of only 13%, international business currently contributes less than 10% of profit; quality awaits annual-report verification.
Honest qualification: part of the market discussion lifts FY2026 DPS expectations to RMB 0.42–0.48 (implying payout ratios of 33.8%–38.7%), but the company's 30.5% payout inertia has never wavered in three years, the interim payout ratio is fixed at about 21%, and no management guidance supports an increase — this report uses RMB 0.38 (30.5% inertia) as the base, treating a payout hike only as an option (if raised to 35%, yield approx. 6.5% gross). The dividend floor's fragility must also be stated: if the volume-collapse scenario materializes (attributable profit RMB 11.0–12.5 billion), DPS falls to RMB 0.31–0.35, after-tax yield approx. 3.7%–4.2%, and with rising rates the spread protection narrows significantly — the dividend is a "gross soft floor," not a hard floor.
This item does not contradict C1: C1 says "price is below the bear-market pricing of profitability"; C5 says "current earnings are at a cyclical peak and cannot be linearly extrapolated" — the former supports the PB-based odds, the latter constrains the false safety of the PE basis. Combined operational implication: buy the normalized earnings power at 0.61x PB plus the dividend floor, not the 5.3x PE "cheapness."
The two-way asymmetry must be presented honestly: consolidation is not a reason to hold the position (no official progress; the company has clarified in writing twice); but quantifying re-ignition probability to 0.1 is as falsely precise as claiming "falsified" — 14 months after the November 2023 first joint clarification, a February 2025 Reuters rumor drove A shares limit-up and H shares up over 18% intraday. The correct framing: elimination from this round is confirmed; the residual option value is two-way and asymmetric (single-day elasticity from rumors re-igniting far exceeds the squeeze-out magnitude); treat as event-driven, exclude from the valuation center.
| Metric | FY2023 | FY2024 | FY2025 | 2025H1 | 2026H1 (Latest) |
|---|---|---|---|---|---|
| Operating revenue (RMB bn) | 30.257 (restated) | 34.048 (restated) | 38.338 | 18.798 | 22.323 (+18.75%) |
| Net profit attributable to parent (RMB bn) | 7.879 | 10.031 | 12.520 | 6.488 | 7.797 (+20.18%) |
| Recurring net profit (this report's estimate, RMB bn)¹ | 7.564 | 9.830 | 12.246 | 6.340 | 7.624 |
| Weighted average ROE (%) | 7.52 | 8.30 | 9.84 | 5.16 (half-year) | 6.00 (half-year; ~12% annualized) |
| Total assets (RMB bn, period-end) | 663.205 | 737.471 | 855.745 | — | 1,060.474 (+23.92% vs. start of year) |
| Net assets attributable to parent (RMB bn, period-end) | 130.466 | 140.481 (estimated) | 147.780 | — | 155.703 (+5.36% vs. start of year) |
| NAV per share (RMB, period-end, ex-perpetuals) | 9.20 (estimated) | 10.12 (estimated) | 10.78 | — | 11.04 |
| Operating leverage (x, own assets/equity attributable to parent) | 4.16 (estimated) | 4.07 | 4.28 | — | 4.66 |
| Margin financing balance (RMB bn, period-end) | — | — | 136.2 | — | 164.757 (+21.0% vs. start of year; 5.45% market share) |
| DPS (RMB, incl. interim) | 0.22 | 0.28 | 0.35 | 0.125 (interim) | 0.15 (interim, +20%) |
| Payout ratio (%) | 30.53 | ~30.5 | 30.57 | — | Interim payout ratio 21.0 |
¹ Recurring net profit is this report's own measure: net profit attributable to parent − other income and gains (non-recurring items such as disposals); exclusions for FY2023–2026H1 were RMB 0.315 / 0.201 / 0.274 / 0.173 bn respectively, all <4% of net profit, so the gap between reported and recurring is small. The broker's operating cash flow has been persistently negative (FY2023–2026H1: −30.2 / −6.5 / −25.5 / −35.6 bn), which reflects the balance-sheet nature of "financing client advances + trading financial assets + resale agreements" (attributed in the cash flow statements of each reporting period) and does not negate earnings quality; hence no OCF/net profit quality line is set. Gross margin / free cash flow / net debt-EBITDA lines are structurally not applicable to the financial industry.
Reasons for metric changes (line items with YoY ≥ ±20%, all per management explanations in the 2026 interim report):
Volume: Revenue of RMB 22.323 bn (+18.75%) and net profit attributable to parent of RMB 7.797 bn (+20.18%), both record highs for the period; weighted ROE 6.00% for the half year, +0.84pp YoY (~12% annualized). Structure: Growth was driven by brokerage and margin financing (net commission income +47.91%), while the investment & trading segment fell −22.66% — low elasticity of proprietary trading is the direct reason it "underperformed the industry's +49%" (industry net profit +49% in 2026H1, based on 43 listed brokers); but it also means that if markets pull back, the damage will be smaller than for high-beta peers. Balance sheet: Total assets exceeded RMB 1 tn for the first time (RMB 1,060.474 bn, +23.92% vs. start of year), with client-funds-related items (payables to clients +50.38) being liability-side accumulation; own-account operating leverage rose to 4.66x (+0.38), and expansion eroded the liquidity buffer — LCR fell to 246.04% (down 130pp over two years), and in 2026H1 the company issued another RMB 15 bn of perpetual subordinated bonds, lifting net capital to RMB 125.811 bn (+9.2%). Comparative judgment: Management gives no quantitative earnings guidance (the "15th Five-Year" strategy remains mostly qualitative); the market's read of the interim results is "strong earnings, weak stock" — earnings estimates were nudged up but valuations compressed; the H shares fell −3.3% on the day after the interim close (after 8/28), leading declines among top brokers — the market has begun re-pricing for an H2 slowdown in trading volumes.
Business model in brief: A full-license state-owned broker controlled by Huijin, blending a "capital-heavy retail brokerage + margin financing" model with "asset-light wealth management" — 453 branches form one of the industry's largest offline networks, with 20.5 million clients and RMB 287 bn in financial products under custody forming the retail base; proprietary financial investments of RMB 451.3 bn (62.2% of own assets) are fixed-income-heavy and absolute-return oriented; international business (Hong Kong + Southeast Asia) is the second growth curve. Revenue has no subscription/recurring concept and fluctuates with market volume and price cycles.
Cash content of earnings and the "recurring earnings" test: A broker's OCF includes changes in client funds and trading assets, so four consecutive negative periods are a business characteristic (see financial table notes) and OCF/net profit is not used for judgment; instead, revenue mix tests "cycle exposure": FY2025 investment income and P&L 34.21% + interest income 35.35% ≈ 69.6% market-beta-type revenue; commission and fee income of 29.72% is relatively stable, but within it investment banking + asset management together account for only 12.8% of net commission income — stable revenue is under 15% of the total. This is the structural root of "earnings quality tied to market conditions," and also why ROE rose from 7.52% to 9.84% over three years alongside a market recovery. The gap between recurring and reported net profit is <4% (no one-off embellishment), but effective tax rates of just 3.1%/4.6% in FY2023–24 (normalizing to 13.1% in FY2025) systematically inflated earlier reported profits (see Chapter 8).
Return on capital: Weighted ROE was 7.52%→8.30%→9.84% from 2023–2025, and 6.00% for 2026H1 (half-year); versus industry leaders (CMB International-listed peers: China Merchants 9.94%, CITIC 10.59, GF 10.16%), this is upper-middle. Normalized (full-cycle) ROE is set at 8% (between the weak 2023 at 7.52% and the bull-year 2025 at 9.84%, cross-checked against FY2023–25 recurring averages) — above the 6.2–7.3% implied by the H-share price, but this is not an elite moat, rather an above-average return from "scale + SOE credit."
Capital consumption and regulatory metrics (in lieu of maintenance CapEx scrutiny): Brokers have no CapEx/depreciation concept; instead we track "balance sheet expansion's drain on regulatory buffers": risk coverage ratio improved from 242.95%→252.72%, but LCR fell 376.07%→246.04% (−130pp over two years) and capital leverage ratio 14.92%→13.54% — expansion and liquidity safety margins trade off; capital replenishment relies on perpetual subordinated bonds (7 outstanding, RMB 40.8 bn; NAV per share differs by RMB 3.20 with/without perpetuals), interest-bearing debt of RMB 479.5 bn in 2026H1, half-year interest expense of RMB 4.83 bn. Verdict: expansion path is smooth (SOE credit; short-term paper at only 1.53–1.6% coupon), but the true shareholder return must be assessed on an ex-perpetual basis.
Moat and quality red flags:
Words-vs-deeds consistency: partially delivered. ① The 2023–2025 three-year strategic plan pledged "building a first-class domestic, internationally excellent modern investment bank"; the 2025 annual report claimed "successful completion of the three-year strategic plan" — yet weighted ROE of 9.84% remains below 2021's 12.29%, and industry rankings in investment banking (12th in equity underwriting) and asset management (Galaxy Jinhui with RMB 90.7 bn AUM) showed no leap; the "delivery" was in completing actions, not in achieving standing. ② The 2025 annual report projected "the securities industry entering a healthy upward earnings cycle" — delivered: 2026H1 revenue +18.75%, net profit +20.18%. ③ The new five-year strategy (2026–2030) still had no quantitative targets as of the interim report and early-September public channels; "three investment banking leads, three-化 synergy empowerment" is a qualitative framework — the quantitative targets hoped for in the last report (original K4) failed to materialize, one basis for downgrading confidence this time.
Shareholder friendliness: neutral-to-friendly (dividends), but no buybacks and prior dilution. Payout ratio ~30.5% for three consecutive years (FY2025: RMB 3.827 bn, 30.57%), three consecutive years of interim dividends, 2026 interim of RMB 1.64 bn (+20%) paid on schedule; by contrast, the 2022 RMB 7.8 bn convertible bond fully converted by end-2023, diluting share capital by ~7.9% (total shares rose to 10.934 bn), with no further equity issuance since — capital has instead been replenished via perpetual subordinated bonds (interest bypasses P&L; another RMB 15 bn issued in 2026H1) — true returns to ordinary shareholders must be computed on an ex-perpetual basis. No buybacks in 2026H1 (the company confirmed no purchase/sale/redemption of listed securities during the period), and no company-level initiative on the H-share discount.
Risk signals: ① Compliance culture needs strengthening — 7 issues surfaced in 2024 (including two warning letters from the Beijing CSRC bureau: imprudent private fund distribution, imprudent OTC options and stock pledge business; a warning letter to subsidiary Galaxy Jinhui for evading lock-up rules), down to 1 in 2025, and 2 branch-level warning letters in 2026H1 (improper competition via commission rebates, excessive incentives in fund sales); branch internal-control problems recur; ② Elevated personnel turnover: CRO change (March) within 2026, executive director Qu Yanping retiring at term (8/10, ~half a year in office), independent director rotation (Vitus Lo completing term and exiting, Li Weihong elected 8/25) — all routine changes with confirmed board consensus and no evidence of governance conflict, but combined with the Huijin-system executive shuffles (CICC–Galaxy "three leadership changes in 100 days"), management stability depends on group-level personnel arrangements; ③ Related-party transactions: the October 2025 supplementary agreement with controlling shareholder Galaxy Financial Holdings substantially raised the annual cap on related-party services (RMB 0.944 bn for 2026); actual usage in 2026H1 was only RMB 0.047 bn — the cap is ample but signals deepening intra-group dealings.
| Segment (FY2025) | Revenue share | Segment margin | YoY | 2026H1 dynamics | One-line business logic |
|---|---|---|---|---|---|
| Wealth management | 42.75% | 46.7% | +18.8% | +32.07% (RMB 9.784 bn) | Brokerage commissions + margin interest, driven by clients and turnover; 53.06% of segment profit, the profit engine |
| Investment & trading | 31.06% | 54.6% | −0.6% | −22.66% (RMB 5.000 bn) | Proprietary FICC + equities, absolute-return oriented but high beta; highest margin, currently a drag |
| International business | 11.93% | 13.0% | +6.3% | +48.46% (RMB 3.047 bn) | Hong Kong + Southeast Asia, mixed light/heavy; fastest growth, lowest margin |
| Parent-subsidiary integration | 8.57% | 35.2% | +2.9% | +39.65% | Futures (Galaxy Futures net profit RMB 0.599 bn) + asset management + alternatives synergies |
| Institutional business | 4.67% | 79.0% | +434.8% (low base) | +79.52% (RMB 1.950 bn) | PB/custody/market-making/derivatives, light capital, high fees |
| Investment banking | 1.79% | 13.2% | +17.8% | RMB 0.306 bn (weak) | Strong in bonds (6th in industry), weak in equities (12th in underwriting) — a structural shortcoming |
Profit engine verdict: Wealth management, with 42.75% of revenue × 46.7% margin, contributes 53.06% of segment profit — the largest revenue segment is also the profit engine; investment & trading has a higher margin (54.6%) but turned into a drag in 2026H1. Margin structure differences (segment margins 13.0%–79.0%, a 66pp spread) reflect fundamentally different businesses: institutional business is a capital-intermediation fee business, international is a heavily invested growth-stage platform, and investment banking is suppressed by weak equity markets — one cannot apply a single "broker" label uniformly. Core tension: the profit engine's (wealth management) fortunes move with market volumes, the second profit source (investment & trading) is in a downcycle, and the growth engines (international/institutional) are still small (together under 20% of profit contribution) — the segment structure confirms that "the cyclical nature of earnings has not yet been diversified away."
Accounting red flags (5 items, pattern + severity + textual evidence):
Cross-period consistency (4 items): ① Effective tax rate 3.1%→4.6%→13.1%, unexplained by the company (inconsistent); ② OCF negative for four consecutive periods against positive earnings, each period attributed to client funds and trading assets — a business characteristic (consistent); ③ Interest income share 42.14% (FY2021) →47.75%→35.35% trending down, attributed to falling deposit rates, directionally consistent with the rate environment (consistent); ④ Impairment losses 0.439→0.769 bn (+75.16%) → 2026H1 credit impairments again +238.74%, continuously escalating; the company vaguely cites "rise in expected credit risk" but does not disclose counterparty detail for the fastest-growing other-receivables impairment (partially unexplained).
Overall judgment: no systematic accounting manipulation detected (no revenue inflation, no direct evidence of profit shifting across periods), but the combination of "low-tax years + OCI hiding unrealized losses + perpetual bond basis" implies: the historical ROE series carries an upward bias; true earning power, viewed on an ex-perpetual basis at a normalized tax rate, is lower than the headline reading — consistent with the conservative treatment in the "business model" chapter of using normalized ROE of 8% (rather than the apparent 9.84%).
H shares at HK$7.82; total market cap HK$85.5 bn (10.934 bn shares × H price, full share count); A shares at RMB 12.47 (2026-09-01 close), H/A ratio 0.537, H discount to A of 46.3% (H shares of brokers generally trade at deep discounts; 46% is sector-normal, not Galaxy-specific). PB (ex-perpetuals) 0.61x — 17th percentile over 1 year / 43% over 3 years / 66% over 5 years / 62% over 10 years: the percentile readings are highly window-sensitive (pre-bull-market H shares long traded at 0.3–0.5x on weaker fundamentals); the absolute discount is deep, the historical percentile neutral — both presented side by side, without forcing a single narrative. PE(TTM) ~5.3x (TTM net profit attributable to parent RMB 13.829 bn) — earnings are at a cyclical high (2026H1 industry net profit +49%, ADT at RMB 3 tn level), the low PE is peak-cycle math; "percentiles for reference only" — not a headline cheapness signal. Dividend yield: FY2025 basis 5.21%; FY2026E (30.5% payout, DPS RMB 0.379) ~5.66%, ~4.53% after tax for Stock Connect individual investors. H-share average daily turnover ~HK$127 mn (last 30 days), 0.149% of market cap — above the 0.1% low-liquidity warning line, but RMB-100 mn-scale position building would need to be executed over multiple days; the liquidity discount is one reason the H shares are cheap.
Peer comparison (H shares, 2026-09-02, same-source basis):
| Company | PB (x) | PE (x) | FY2025 ROE (%) | 2026H1 net profit growth | Notes |
|---|---|---|---|---|---|
| CITIC Securities H | 1.21 | 9.03 | 10.59 | +50% | Full-service investment banking leader |
| China Merchants Securities H | 0.92 | 6.52 | 9.94 | ~+15–20% (high annualized ROE) | Steady SOE type; PB/ROE ratio 0.093 |
| CICC H | 0.84 | 6.41 | 9.39 | — | Rank moved up after absorbing Dongxing Xinda |
| Huatai Securities H | 0.77 | 6.73 | 9.20 | +54.9% | Digital wealth management |
| Guotai Haitong H | 0.69 | 7.08 | 9.78 | Top of industry | First tier by scale post-merger |
| China Galaxy H | 0.62 | 5.3 | 9.84 | +20.18% | Lowest PB, 4th in ROE; PB/ROE ratio of 0.063, lowest in group |
Reverse-engineering via the Gordon model (r = 9–10%, g = 2–3%): the current price of 0.61x P/B implies a perpetual ROE of roughly 6.2%–7.3%. In plain terms, "what does the current price demand of the company": it demands that Galaxy's profitability stays forever at the 2022–23 bear-market level, below even the 7.52% actually achieved in 2023. The reality: ROE has risen for three consecutive years (7.52 → 8.30 → 9.84), annualizes to about 10–11% in 2026, margin finance share is edging up, and client numbers are at record highs. Meanwhile, the A-share price of 1.13x P/B implies an ROE of about 10.9%—A-share pricing is already consistent with fundamentals; roughly half of the H-share mispricing is a structural phenomenon of southbound pricing power and liquidity discount (its closure depends on catalysts rather than automatic convergence). On the sell side: Investing aggregates 6 H-share analysts—all Buy, 0 Sell—with a mean target of 11.06, a low of 8.80, and a high of 13.70; two named downgrades year-to-date (CICC cut to 11.2 on 4-08; J.P. Morgan cut from 13.0 to 11.8 on 4-22 while trimming 2026–28E net profit by 5% each)—all target prices predate the interim results and the Q3 volume-decline data, so the anchors are broadly stale. Should volume collapse, downward revision of consensus becomes the source of the bear scenario (which is why this report does not directly adopt the sell-side mean as its target price).
| Scenario | Probability | Fair Range (HKD) | Midpoint | vs. Current Price | Key Driver |
|---|---|---|---|---|---|
| Bear | 30% | 6.2–7.0 | 6.6 | -20.7% ~ -10.5% | ADT collapses to CNY 1.4–1.6tn + bond market correction: ROE 6.5–7%, P/B 0.47–0.53x; dividend floor falls with earnings |
| Base | 50% | 9.0–9.8 | 9.4 | +15.1% ~ +25.3% | ADT midpoint CNY 2.0–2.3tn (down 25–35% from H1 but no collapse): full-year ROE 8.5–9%, P/B 0.68–0.73x × 2026E BVPS CNY 11.42 |
| Bull | 20% | 11.5–12.5 | 12.0 | +47.1% ~ +59.8% | Bull market restart on incremental capital (ADT CNY 2.8tn+): ROE 10%+, P/B 0.87–0.95x; Galaxy's low-beta prop trading lags high-beta peers, hence a cap below the sell-side high of 13.70 |
Bear-case anchor check: the bear midpoint of HKD 6.6 is about 25% below the most pessimistic published target (aggregated low of 8.80)—justified because all sell-side targets predate the Q3 volume decline; the bear case assumes "double kill" of earnings and multiples and is cross-checked against stress tests (net profit attributable of CNY 7.7–9.5bn and ROE 4.9–5.3% under volume collapse; plus widening-pure-discount cases of 6.6–7.0 and an extreme stock-bond double whammy of 5.4–6.2), covering the extreme extension of J.P. Morgan's 5% downgrade direction; the base range (9.0–9.8) falls between the sell-side low of 8.80 and high of 13.70. Anchoring basis for the base-case exit multiple (multiple crossed anchors, not the company's own price as self-proof): ① forward Gordon: ROE 8.5–9%/r 10%/g 3% → fair P/B 0.79–0.86, cut to 0.68–0.73 after the H-share liquidity discount; ② peer H-share P/B median 0.84 (CMB Intl—no, CITIC peers: China Merchants at 0.92 with comparable ROE, Huatai at 0.77); ③ the company's own 3-year P/B median of 0.64 as a lower bound—three anchors converge on the 0.68–0.73 range.
| Period | Revenue (CNY 100mn) | Net Profit Attributable (CNY 100mn) | Key Assumptions |
|---|---|---|---|
| FY2026E | 405–425 | 132–141 (neutral 136) | H2 ADT CNY 2.3–2.6tn (Q3 already 2.45); brokerage commission rate stable with slight decline, margin finance share 5.4–5.5%; with low-beta prop trading, full-year net margin ~44%; DPS ~CNY 0.379 (30.5% payout ratio) |
| FY2027E | 380–410 | 118–132 (neutral 125) | ADT midpoint recedes to CNY 2.0–2.3tn; investment banking partially hedged as IPOs normalize; -8%~-2% YoY on a high base |
Management gives no quantitative guidance (the "15th Five-Year Plan" contains no quantitative targets); on the sell side, J.P. Morgan cut 2026–28E net profit by 5% each in April (direction consistent with the volume decline)—three-way comparison: this report's FY2026 neutral figure of CNY 13.6bn sits mid-range of pre-upgrade sell-side estimates, and the FY2027 turn-to-negative call is more conservative than sell-side consensus.
Quality: leading central-SOE broker, ROE rising for three years, solid market share, low-volatility prop trading—upper-middle but not top-tier (investment banking weaker than the "two C's plus H" of CITIC/CICC/Haitong; stable-income share <15%). Price: 0.61x P/B / implied ROE 6–7% / dividend yield 5.7% / 34% EPV discount—favorable odds structure but roughly half the discount is structural. Verdict: undervalued; target price range HKD 9.0–9.8 (base case); margin of safety ~+15% (base fair-value low of 9.0 vs. current price of 7.82). Probability-weighted expectation ~HKD 9.08 (+16.1%). Low P/E alone is not a buy case (C5); the H-share discount lacks a company buyback catalyst, and the timing of convergence is unknowable—this is the full meaning of "prudent" rather than "bullish."
China's securities industry (about 150 licensed brokers, per the Securities Association of China) posted FY2025 operating revenue of CNY 5,411.71bn and net profit of CNY 2,194.39bn; at end-2025 total assets were CNY 14.83tn, net assets CNY 3.34tn, net capital CNY 2.44tn, and assets under discretionary management CNY 9.53tn (SAC release dated 2026-03-27). On the market side, A-share total market capitalization broke CNY 100tn for the first time in 2025. The industry's core revenue driver is not "industry scale expansion" but the market's volume-price cycle: 2025 A-share average daily turnover (ADT) was CNY 1.73tn, +61.7% YoY (full-year turnover CNY 420.69tn per exchange data, cross-checked against independent calculation of CNY 1.70tn/day); margin finance balance reached CNY 2,540.7bn at end-2025, +36.3% YoY. H1 2026 saw the cycle peak: 43 A-share listed brokers posted 2026H1 revenue of CNY 364.71bn (+45.87%) and net profit attributable of CNY 155.37bn (+49.01%) (Securities Daily tally); the quarterly ADT series ran Q1 CNY 2.56tn (+71%) → Q2 2.86tn (+132%) → Q3 through 9/2 ~2.45tn (+17.5% YoY, -14.5% QoQ)—the slope of the decline from peak volume is the sector's first-order variable for earnings and valuation. Structurally, in the first 8 months of 2026 there were 104 IPOs raising CNY 191.878bn, +55%/+192% YoY, already 45.6% above full-year 2025 (Cailianshe tally)—the investment banking line is on a recovery path; public fund AUM was CNY 39.11tn (end-July 2026, AMAC) and equity ETFs CNY 3.51tn (2026-08-28, 1,551 funds)—massive indexation expansion, but ~CNY 1.27tn of net outflows year-to-date (a signal of profit-taking at highs). No industry authority publishes a quantified CAGR—the sector's "size" is essentially a function of market volume and price and should not be expressed as a linear CAGR.
The value chain: exchanges and clearing/depository (infrastructure, member-based conduit) → brokers (five lines: brokerage / investment banking / asset management / credit / prop trading) → retail customers (150mn+ investors, client margin deposits of CNY 3.24tn) and issuers; wealth management extends downstream to public/private fund managers and distribution chains. Brokers occupy the pivotal position in the "exchange–broker–AM/public funds–distribution" chain: both the trading conduit and the node for product issuance and sales. The secular shift in value distribution: channel commissions keep deflating (industry average net commission rate has fallen below 0.02%, per Securities Daily, 2026-09); after the three-phase public fund fee reform, the blended non-money-market fee rate fell from 1.41% in 2022 to 0.93% in 2025 (Phase 3 forgoes ~CNY 30bn/year, cumulative over CNY 50bn/year across three phases), and distribution is shifting from "transaction-driven" to "holdings-driven"—the profit pool is migrating from channel fees to "holdings + services" (fund advisory, ETF market-making, margin spreads) and capital businesses (prop trading, market-making, derivatives). Galaxy's position: the representative franchise on the retail chain (453 branches, CNY 287bn in financial product holdings, 4,400+ investment advisors, PB business of CNY 481.4bn, +38.3%); bargaining power vs. upstream (exchanges/funding markets) strengthens with scale (short-term financing paper at 1.53–1.6% on central-SOE credit); bargaining power vs. downstream retail is eroded by internet platforms (East Money, etc.) and bank wealth management—overall bargaining power neutral-to-weak, offset by scale and capital.
Supply side = licensed service capacity of 150 brokers + balance-sheet expansion; no capacity constraints (high supply elasticity), with tightness expressed in "capital vs. business opportunity." The 2024–2026 M&A wave is clearing out mid/small licenses and raising head concentration: Guotai Junan's absorption of Haitong closed (2025, creating a mega-broker with CNY 63.1bn revenue / CNY 27.8bn net profit, topping the industry in 2026H1 net profit); Guolian + Minsheng, Zheshang + Guodu, Western + Guorong, Guosen + Wanhe successively closed; CICC's merger of Dongxing Securities + Cinda Securities passed review in 2026-08, and Soochow Securities' CNY 10bn acquisition of Donghai Securities was accepted for review in 2026-09—broker count is shrinking while per-entity capital steps up. Concentration: 2026H1 listed brokers by net profit: CR5 = 52.1%, CR10 = 74.3% (Qichacha/Ziyu Tong statistic)—marked head concentration, still rising on M&A. Competitive intensity: brokerage price war is real (net commission <0.02% and still deflating); on the investment banking side, the "two C's plus H" (CITIC, CICC, Haitong system) took ~40% of 2026H1 investment banking revenue, intensifying head-end oligopoly; margin/market-making/derivatives capital businesses are the arena for heads. Substitution threats: internet brokers/third-party platforms siphoning retail trading, bank wealth management substituting distribution, ETF indexation substituting active advisory. Entry barriers: licenses essentially frozen (access via M&A only), net capital regulation, branch networks and client stickiness, classification ratings (Galaxy: cultural-category AA for four consecutive years).
Current cycle position: "mid-bull-market correction, volume-price falling from highs"—the 2025 bull run (Shanghai Composite 3,320 → 3,969, a ten-year high) → 2026H1 peak (ADT Q2 CNY 2.86tn, margin balance ~CNY 3.02tn at end-June near historical peak, public funds topping CNY 39tn at end-April) → July–August turning to correction (Q3 ADT -14.5% QoQ, margin -11.1% from peak, CSI 300 ~10% off its 52-week high, broker H-shares leading declines). Not a bear market, but beta is pointing down. Leading indicators: ADT (monthly), margin balance, equity ETF flows (CNY 1.27tn net outflow YTD), IPO acceptance pace, monthly change in public fund AUM (July -56.3bn, ending three straight increases), broker H/A discount. Regulatory direction: neutral, favoring heads—① support for M&A (the Guotai-Haitong-scale merger won regulatory blessing; Galaxy, part of the Huijin system, was not drawn in but benefits from industry consolidation); ② three-phase public fund fee reform (compressing distribution margins short term, forcing a shift to holdings long term); ③ capital market activation and long-term capital inflows (high-quality ETF development, expansion of personal pensions—Galaxy has opened 230k+ pension accounts); ④ institutional initiatives such as trading system external-access rules and the consultation draft on comprehensive broker account management (2026-09). Net implication for Galaxy: fee reform and commission cuts suppress channel revenue long term, but M&A easing and capital business expansion favor Huijin-system head platforms.
| Company | 2025 Revenue (CNY 100mn) | 2025 Revenue Growth | 2025 Net Profit Attributable (CNY 100mn) | Weighted ROE (%) | 2026H1 Profit Growth | Key Difference vs. Galaxy |
|---|---|---|---|---|---|---|
| CITIC Securities | 748.54 | +28.79% | 309.9 (net margin 41.4%) | 10.59 | +50% | Full-service investment bank leader; leading across institutions/IB/derivatives |
| Guotai Haitong | 631.07 | +87.4% (consolidated) | 278.09 | 9.78 | Industry No.1 | Top tier by scale post-merger; Galaxy did not participate in this round of consolidation |
| Huatai Securities | 358.10 | +6.83% | 116.92 (2026H1, +54.9%) | 9.20 | +54.9% | Benchmark in digital wealth management; "low commission + online" model |
| GF Securities | 354.93 | +34.33% | 137.02 | 10.16 | — | Dual public fund licenses (GF Fund / stake in E Fund); no H shares |
| CICC | 284.81 | +33.50% | 97.91 | 9.39 | — | Top two in IB/institutions; rank rises after absorbing Dongxing and Cinda |
| China Galaxy | 283.02 (restated) | +24.34% | 125.20 | 9.84 | +20.18% | Retail/margin franchise + unique SEA footprint; weak in equity IB |
| China Merchants Securities | 249.72 | +19.53% | 123.50 | 9.94 | — | Steady central-SOE type; highest net margin in group at 49.3% |
(GF's 2025 net profit of CNY 13.702bn is per East Money F10; CITIC's attributable profit is back-calculated from net margin, for scale reference only.)
Scale ranking (FY2025 basis): revenue of CNY 28.302bn ranks 6th among listed brokers (just 0.6% behind CICC's 28.481bn); net profit attributable of CNY 12.520bn ranks 5th (ahead of China Merchants' 12.350bn, behind GF's 13.702bn); ROE of 9.84% ranks 4th among 7 comparable heads. Core share: margin finance balance CNY 164.757bn, 5.45% market share (2026H1, up slightly from 5.36% at end-2025, industry No. 4–5); bond underwriting CNY 682.575bn, industry No. 6 (Wind basis, No. 5 in local government bonds); No. 4 in both STAR Market and BSE market-making; equity underwriting No. 12 (FY2025, CNY 8.697bn)—a clear IB structural weakness: strong in bonds, weak in equity. Brokerage: the annual report only states that "core business market rankings have long remained in the industry's top tier" without disclosing specific stock-fund turnover share—2026H1 brokerage net revenue of CNY 6.273bn (+47.5%) equals ~5–6% of industry interim brokerage net revenue, consistent with the No. 2–4 tier (estimated basis). Caveat: the "24.57% ETF holdings market share, industry No. 1" cited in the previous report (2026-06) could not be verified in this round's annual/interim report text and is treated as unverified and downweighted; verifiable retail hard metrics are 20.5 million clients, financial product holdings of CNY 287bn (+13.9%), and 700k+ buy-side advisory clients. Share trends: margin/brokerage share edging up, IB ranking flat-to-weak, international share rising (first top-10 entry as a HK IPO sponsor, share in four SEA countries). Moat sources: scale network (453 branches + client deposits), central-SOE credit (low-cost funding), localized SEA licenses—positioned as the "Huijin-system retail wealth platform," forming a division of labor with CICC (IB/cross-border) within the group rather than head-on competition; relative weaknesses: scale gap passively widening in the M&A wave (Guotai Haitong is over 2x Galaxy's size), and lagging elasticity rankings in bull markets (2026H1 +20% vs. industry +49%).
Securities brokers' valuation and earnings are not determined by "industry TAM," but by three quantifiable decisive factors: volume (ADT × margin financing), price (commission rate), and direction (equity/bond allocation in proprietary trading). This chapter quantifies the transmission of all three to Galaxy, serving as the demand-side basis for valuation scenarios.
① ADT Sensitivity (Primary Driver): In 2026H1, average daily stock-and-fund turnover of RMB 3.07 trillion (+57.3%) corresponded to the company's net commission income of RMB 6.94 billion (+47.91%). Elasticity estimate: for every RMB 500 billion decline in ADT, net brokerage commission income falls by approximately RMB 1.13 billion per half-year; incorporating the linked margin financing effect, net profit attributable to shareholders declines by approximately RMB 1.07 billion per half-year, or roughly RMB 2.1 billion annualized—equivalent to -13.8% of the 2026 annualized net profit (approx. RMB 15.6 billion). Against the current baseline: Q3 ADT has already fallen about 20% from the H1 average (2.45 vs 3.07 trillion), implying that if Q4 holds at this level, FY2026 net profit attributable will land in the lower-middle portion of this report's forecast range (RMB 13.2-14.1 billion). Trigger lines: ADT below RMB 2.0 trillion for 20 consecutive trading days (the single-day RMB 1.81 trillion on August 26 has already approached this several times), and total market margin financing balance below RMB 2.4 trillion.
② Margin Financing: Share Gains Fail to Offset the Shrinking Pie: The company's margin financing balance stood at RMB 164.757 billion with a 5.45% market share (up slightly +0.09pp), but the market balance is down -11.1% from its peak—statically estimated, the net effect on the company's margin-financing-related income is approximately -9.5%. With a maintenance ratio of 281.2% and pending buyback of pledged shares of RMB 17.779 billion (collateral coverage of 342.4%), the current safety cushion is ample; the risk trigger line is a further 15-20% market decline, which would bring the maintenance ratio down to 239%-225% (still above the 130% forced-liquidation line, but tail accounts would face pressure). The 2026H1 credit impairment of RMB 0.33 billion (+238.74%) is already an early signal of prudent provisioning.
③ Proprietary Direction and Interest Rate Exposure (Second Decisive Factor): Financial investments of RMB 451.325 billion (62.2% of own assets) are dominated by fixed income—FVTPL bonds of RMB 130.4 billion + FVOCI bonds of RMB 106.1 billion + FVOCI equities of RMB 55.6 billion (high-dividend strategy with unrealized losses in OCI) + FVTPL equities of RMB 25.2 billion (only 11% of FVTPL). Per the interest rate sensitivity disclosed in the interim report: rates +100bp → pre-tax profit -RMB 2.762 billion and other comprehensive income -RMB 5.825 billion (i.e., +50bp ≈ profit -RMB 1.38 billion + net assets -RMB 2.9 billion)—the income statement impact is milder than a total-caliber duration estimate, but the net asset (the denominator of PB valuation) channel cannot be ignored. VaR (1D, 95%) is RMB 0.339 billion. 2026H1 investment & trading segment revenue fell -22.66% (high base from the H1 2025 bond bull market), validating the low bull-market elasticity of the "fixed income-led + absolute return" strategy; the symmetric implication is that losses in downturns will also be smaller than for high-beta peers.
④ Stress Test (Consistent with the Bear Scenario Caliber in the Valuation Chapter):
⑤ Management's Cycle Discipline: During this upcycle, capital actions have mainly been "pro-cyclical balance sheet expansion + debt issuance for capital replenishment" (total assets +23.92%, leverage 4.28x→4.66x, another RMB 15 billion of perpetual subordinated debt issued), with no counter-cyclical buybacks; the payout ratio has been stable at 30.5% but has not been raised despite the industry upswing—cycle discipline is neutral: no aggressive equity expansion at the peak (a plus), but also no enhanced shareholder returns at the bottom (a regret).
Overall assessment: Cautiously bullish, confidence 0.58, time horizon 12 months. Odds structure: the current price of HKD 7.82 is below the base fair value of 9.0-9.8 (+15% margin of safety, probability-weighted expected return of approx. +16%); the bear case of 6.6 (-15.6%) and bull case of 12.0 (+53.5%) form an approximate 1:1.3 upside/downside shape, plus a 5.7% dividend yield for total holding return. Certainty is suppressed by triple constraints—volume is declining in real time (FY2027 turning negative is the base case), the structural nature of the H-share discount (half of it is not mispricing), and reliance on external catalysts (payout ratio upgrades and southbound fund flows are not fully within the company's control). Strategic implications: participate in tranches on the left side, use dividends and the 0.55x PB floor (approx. HKD 7.1) to manage drawdown expectations; above the upper bound of 9.8, the margin of safety is exhausted; below the lower bound of 6.2, the revaluation thesis re-engages.
Risk Warnings (Ranked by Impact): ① Volume collapse (ADT persistently below RMB 1.8-2.0 trillion): FY2026/27 earnings double whammy, fair value center dropping to around the HKD 8 level; ② Bond market correction (10Y holding above 2.1%/2.3%): dual-channel hit to income statement and net assets (+50bp ≈ profit -RMB 1.4 billion / net assets -RMB 2.9 billion); ③ Margin financing/pledged shares credit risk (a further 15-20% market decline, impairment elasticity of RMB 0.7-1.7 billion per half-year); ④ H-share liquidity and discount widening (pure discount scenario: another 10-15% underperformance vs A shares); ⑤ Deterioration of competitive structure (further commission rate cuts, widening gap among top players post-M&A); ⑥ Two-way volatility in the integration option (pulse elasticity from reignited rumors coexisting with crowding-out risk).
Monitoring Checklist (Points for Falsification and Validation): Monthly ADT and margin financing balance series (falsification lines: 20 consecutive days below RMB 1.8 trillion / margin financing below RMB 2.4 trillion); 10-year government bond yield (2.1% first threshold / 2.3% second threshold); H-share price (below 6.2 revaluation, above 9.8 safety margin exhausted); FY2026 full-year payout ratio (below 25% thesis fails / ≥33% upgrade option exercised); quarterly credit impairments (above RMB 0.5 billion); any official moves on Huijin-system integration (two-way pulse); sustainability of international business segment revenue (annual report to verify pipeline and margins); whether the 2026-09-10 results briefing provides quantified "15th Five-Year Plan" targets.
This report is based on public materials (the company's 2026 interim report, FY2023-2025 annual reports, HKEX filings, data from the Securities Association of China and exchanges, public news and sell-side rating records) and this report's independent estimates. Sensitivities and falsification conditions for key assumptions are itemized in the main text; all valuations use the 2026-09-02 H-share closing price of HKD 7.82 and an exchange rate of 1 HKD = 0.85683 RMB as the baseline.