Rating: Cautiously Bullish | Target Price Range: RMB 10.2–11.3 (base-case fair value = 0.45–0.50x PB × 2026H1 book value per share of RMB 22.63) Current Price: RMB 9.82 (close on 2026-10-27) | Market Cap RMB 327.06 billion | PB 0.43x / PE (TTM) 6.38x / Dividend Yield 4.28% Margin of Safety: +4% (vs. lower bound of base-case fair range at RMB 10.2; probability-weighted expected return +7.6%) | Time Horizon: 6–12 months
At RMB 9.82, SPDB is priced on a "perpetual status quo" basis: reverse DDM analysis shows the current price implies a perpetual ROE of 6.6%–7.3%, essentially identical to the company's actual ROE (6.76% weighted in 2025) — after the retail risk exposure in H1 2026, the market has withdrawn the repair premium it once paid (52-week high of RMB 12.97; current price is down 24.3% from that peak). The current price embeds almost no optimism about NIM stabilization continuing, asset quality cleanup, or a higher payout ratio. On fundamentals, NIM has stabilized for two consecutive periods (1.42%→1.43%), the NPL ratio has declined for six consecutive years to 1.25%, and after 99.67% conversion of the RMB 50 billion convertible bond, the CET1 capital adequacy ratio rose to 9.05% — the repair trajectory is clear. However, retail NPLs rose 20bp to 1.67% in six months, special-mention loan migration doubled to 59.30%, restructured loans grew 28.7% in six months, and H1 2026 profit growth relied on unsustainable pillars such as expense cuts and impairment-offset maneuvers — these form the main drivers of the downside scenario. The probability-weighted fair-value center across three scenarios is roughly RMB 10.6, +7.6% vs. the current price — odds are positive but not extreme. Cautiously bullish: avoid one-sided heavy positioning until confirmation signals of valuation repair (NIM and retail NPL formation data in the Q3 report due 10-31) land. Key constraints: profit growth downshift in 2027 (growth falling back to 0–5%) and near-zero organic capital headroom.
The most important changes over the next 12 months: ① Redemption of the liability-cost repricing dividend — the deposit cost ratio fell to 1.35% in 2026H1 (down 32bp YoY); the concentrated maturity of high-rate time deposits in 2026Q4–2027H1, combined with the window for lower posted deposit rates, is the deterministic variable for whether NIM stabilization can continue; ② A showdown on the asset-quality narrative — the divergence of "corporate recovery, retail stress" will be verified in the window between the 2026 Q3 report (10-31) and the annual report (March 2027). Confirmation of an inflection in retail NPL formation would remove the core rationale for the 0.43x PB discount; confirmation of deterioration would reverse the provision-replenishment logic.
Primary share-price driver: Retail asset quality and NPL formation (determining the provisions→profit→dividend chain and PB discount convergence), followed by the durability of NIM stabilization.
Verifiable expectation gap: The current price implies "no deterioration but no repair either," whereas the actual direction is ROE rising for two consecutive years (5.21%→6.28%→6.76%) and NIM stable for two periods. That said, it must be honestly noted that the sell-side consensus target price median of RMB 12.8 (implying roughly +30%) is itself closer to the bull case — our base-case fair value (RMB 10.2–11.3) is more conservative than the street. The expectation gap lies in the contradiction between "optimistic street target prices and net downward earnings revisions (-2.9% over the past 90 days)"; direction will be adjudicated by Q3 data.
Validation catalysts: Q3 report on 10-31 (a two-way test event); the LPR quote window in December 2026; the 2027-03 annual report and profit distribution plan (whether the payout ratio stays above 30%). Falsification triggers: provision coverage falling below 185%, NIM below 1.38%, retail NPL ratio above 1.75%, or a decisive break below RMB 8.8 (any single trigger prompts re-rating).
Structurally, this must be decomposed: of the RMB 6.324 billion increase in net interest income, the +2bp YoY NIM improvement contributed roughly RMB 1.8 billion (~29%), while the remaining ~70% came from balance-sheet expansion (total assets +3.21% vs. prior year-end, ~+7.9% YoY) and lower liability costs; moreover, about half of the +2bp reflects the low base of 2025H1 (1.41%) — 2026H1 was actually flat QoQ vs. full-year 2025. Asset yields are still declining (-36bp); "stabilization" is entirely driven by the liability side alone. Looking ahead, the concentrated maturity/repricing of high-rate time deposits in 2026Q4–2027H1 and posted deposit rate cuts (from September 2026, small and mid-sized banks have resumed cuts of 15/35bp on 2/3-year terms) form the basis for continuation; but if the LPR is cut further (market institutions expect another 5–10bp within the year) while industry average deposit rates stabilize at the margin, the dividend's momentum will fade. Reverse test: if the single-quarter NIM in the Q3 report is <1.42% and the deposit cost ratio's decline narrows to within 10bp, or NIM falls below 1.38%, this thesis fails.
The improvement on the corporate side is real: the bank-wide NPL ratio has declined for six consecutive years, real estate balances have begun to shrink, and asset quality in the Yangtze River Delta core region is better than the bank-wide average. But signal density on the retail side is not encouraging: the jump in migration rates cannot be explained by the new risk-classification rules (reclassification of existing balances was completed on 2025-12-31), and under the Commercial Banks Financial Asset Risk Classification Measures, restructured loans belong to the "front queue of future NPLs"; this report estimates an annualized NPL formation rate of roughly 1.19%–1.45% (methodological differences), consistently pointing to accelerating formation. Provision coverage of 197.81% declined for the first time in four years, falling 2.91pp; in 2026H1, write-offs and disposals reached RMB 40.124 billion (+53.2% YoY), sustaining the decline in book NPLs through high-intensity cleanup. Stress test: if the NPL ratio rises another 20bp with the provision balance flat, coverage would slide toward the 170% range; coverage below 185% is a yellow light, below 175% a red light — at that point the "endgame of cleanup" narrative is falsified, and provisions shift from replenishing profit to consuming it.
Of the +4.08% growth in profit attributable to parent, a substantial part of the +9.98% pre-tax profit growth came from one-off expense cuts and a structural switch within impairment lines (the non-loan provision pool being drawn down), plus the fading drag reversal from a low tax-rate base. In 2027, all three pillars face normalization: room for expense cuts narrows after years of reductions, financial-investment impairment reversals cannot repeat, and the tax-exemption tailwind has faded. The key is the state of market pricing: 2026E consensus of +3.5% is essentially aligned with this report's forecast (RMB 51.5–53.5 billion, +3.0%~+7.0%) and largely priced in; but 2027E consensus implies +5.9%, above the upper bound of our judged 0–5% downshift range — the risk of forward downward revisions has not yet been absorbed by consensus. If the Q3 report confirms narrowing expense declines or accelerating impairments, a collective 2027E downgrade is possible. Reverse test: if 2026 full-year net profit growth is ≥4.5% and total credit impairment growth is <8%, the three-pillar concern is falsified.
The two-sided nature of the capital position must be stated in parallel: on one hand, after the RMB 50 billion convertible bond conversion concluded at 99.67% (the largest single A-share issuance), CET1 of 9.05% carries a nominal buffer of about 105bp over the 8.0% regulatory requirement, making equity refinancing unlikely within the next 12–18 months; on the other hand, at a 30% payout ratio, retained earnings can support annualized RWA growth of only about 4.5% at most, while 2026H1 annualized RWA growth has already reached 4.4% — organic capital headroom is near zero, and any medium-term payout increase is hard-constrained by capital. The real slope of shareholder returns is "stable": DPS RMB 0.41→0.42 (+2.4%), payout ratio flat around 30%, dividend yield 4.28% (a spread of about 2.67 percentage points over the 1.61% 10-year Treasury) is the substantive anchor for the valuation floor — but the company has no buyback program and has long given boilerplate responses to market-cap management demands. Reverse test: CET1 below 8.8%, or FY2026 payout ratio falling below 25%.
Valuation structure: the current price embeds almost no repair expectations — the implied perpetual ROE matching the status quo means that even if ROE rises above 7.5% or the payout ratio is raised, the market is paying nothing for it; conversely, if retail risk ferments, the bear-case floor (0.35–0.38x PB) sits roughly -12%~-16% from the current price. The base-case fair value of RMB 10.2–11.3 is anchored independently of the current price: Gordon-implied fair PB (assuming ROE 7% / payout 30% / Ke 9%–9.5% → 0.46–0.51) cross-checked against the peer PB-ROE cross-section (Everbright ROE 6.56%/PB 0.36 ↔ CITIC 9.39%/0.68; SPDB's asset quality is clearly better than Everbright's and worse than CITIC's). Caveat: steady-state value is sensitive to assumptions (Ke 8.5%–10% and g 3.5%–4.7% imply a range of RMB 7.2–11.6); "current price at the steady-state center" is a range judgment, not a precise equation; this report's base case is roughly 15% more conservative than the sell-side consensus target median of RMB 12.8. Reverse test: a decisive break below RMB 8.8 (near the bear-case center, implying the market is pricing in ROE decline) falsifies this thesis and shifts to a defensive assessment.
| Metric | 2023FY | 2024FY | 2025FY | 2026H1 (latest period) |
|---|---|---|---|---|
| Operating revenue (RMB bn) | 173.434 | 170.748 | 173.964 | 93.777 |
| Revenue YoY | -7.98% | -1.55% | +1.88% | +3.55% |
| Net profit attributable to parent (RMB bn) | 36.702 | 45.257 | 50.017 | 30.951 |
| Net profit YoY | -28.28% | +23.31% | +10.52% | +4.08% |
| Recurring net profit attributable to parent (RMB bn) | 33.882 | 44.207 | 50.144 | 30.971 |
| Net interest margin (NIM) | 1.52% | 1.42% | 1.42% | 1.43% |
| NPL ratio | 1.48% | 1.36% | 1.26% | 1.25% |
| Special-mention loan ratio | — | 2.34% | 2.09% | 1.97% |
| Overdue loan ratio | — | 2.06% | 1.89% | 1.84% |
| Provision coverage ratio | 173.51% | 186.96% | 200.72% | 197.81% |
| CET1 capital adequacy ratio | — | 8.92% | 8.99% | 9.05% |
| Weighted average ROE | 5.21% | 6.28% | 6.76% | 3.96% (non-annualized) |
| Cost-to-income ratio | 29.65% | 29.16% | 28.50% | 22.52% |
| Total assets (RMB tn) | — | 9.46 | 10.08 | 10.41 |
| Total loans (RMB tn) | — | 5.39 | 5.70 | 5.87 |
| Total deposits (RMB tn) | — | 5.15 | 5.58 | 5.86 |
| Book value per share (RMB) | 20.95 | 22.36 | 22.13 | 22.63 |
| Dividend per share (RMB) | — | 0.41 | 0.42 | — |
| Dividend yield (at current price) | — | 4.18% | 4.28% | — |
Note: As a bank, SPDB has no gross margin, free cash flow, or net debt/EBITDA concepts — structurally not applicable and thus omitted. Operating cash flow is dominated by net deposit/loan flows (2024FY -RMB 333.7 bn / 2025FY +RMB 375.8 bn) and does not reflect earnings quality, so it is omitted here. Dividend yields are computed on the 2026-10-27 close of RMB 9.82 and each year's DPS.
Reasons for metric changes (items with YoY ≥±20%): ① Write-offs and disposals +53.2% YoY in 2026H1 — per the company, "adhering to both controlling new NPLs and reducing legacy ones, continuously intensifying the disposal of existing non-performing assets," maintaining the NPL ratio's decline through high-intensity cleanup; ② Loan impairment provisions +29.9% coexisting with RMB 1.636 billion of financial-investment impairment reversals, total impairment up only +5.01% — the company did not specifically explain the line-item switch; discussed in this report's accounting-maneuvers section; ③ Income tax expense +56.97% — official explanation "due to increased pre-tax profit and reduced tax-exempt income" (tax-exempt income impact -RMB 4,208 million vs. -RMB 4,843 million prior year); ④ Fair-value gains/losses -927.66%, FX -185.29% — official explanations "valuation changes in trading financial instruments" and "market exchange-rate fluctuations," with trading-line gains internally offset (investment income +34.31%); ⑤ Minority interest income +156.69% — "increased profit of subsidiaries attributable to minority shareholders."
2026H1 revenue was RMB 93.777 billion (+3.55%), net profit attributable to parent RMB 30.951 billion (+4.08%), recurring profit RMB 30.971 billion (+3.29%; non-recurring items only -RMB 0.22 bn — a clean earnings-quality reading). Growth versus full-year 2025 (+1.88%/+10.52%) shows a "revenue accelerating, profit decelerating" combination, with Q2 alone (revenue ~+5.8%, profit ~+7.8%) better than Q1. Against a backdrop of 42 listed banks posting 2026H1 revenue growth of +7.42% and negative attributable profit growth for joint-stock banks overall, SPDB's profit growth outpaced the joint-stock bank peer average.
Structural interpretation: The main growth engine was net interest income (RMB 64.631 bn, +10.85%), driven by liability costs (deposit cost ratio 1.35%, -32bp) and balance-sheet expansion; the drag was other non-interest income (-14.30%): fair-value losses of -RMB 5.206 bn and FX losses of -RMB 1.328 bn together eroded about RMB 6.5 bn, offset by investment income of +RMB 6.77 bn (realized bond gains) — trading-related items account for ~14.5% of revenue with intense internal offsetting, so bond-market direction cannot be ignored as a driver of revenue growth. Fee income of RMB 11.924 bn (-1.92%) remained weak.
Deviation explanation: Profit growth (+4.08%) lagged pre-tax profit growth (+9.98%), with the gap in income tax (effective tax rate up 4.18pp); the 2.91pp decline in provision coverage to 197.81% coexisting with heavier loan provisioning reflects a "high write-off + high provision" cleanup combination, not provision release. Management gave no quantitative guidance; for comparison, after the interim report Guosen Securities cut its 2026E attributable profit to RMB 52.8 bn (vs. consensus ~RMB 52.4 bn); this report's forecast range of RMB 51.5–53.5 bn is aligned with that — the interim report itself did not beat expectations; the market's pricing dispute lies beyond 2027.
Business model in brief: A national joint-stock commercial bank (one of 12, in Group 2 of domestic systemically important banks), dominated by corporate banking (corporate loans account for 65.26%), concentrated in the Yangtze River Delta region (segment revenue share 25.29%, operating profit share 44.28%), with group-level operations spanning trust/wealth management/fund/financial leasing/investment banking licenses. Revenue is primarily net interest income (68.92% in 2026H1), fee income is weak (12.72%, consecutive negative growth), and other non-interest income is highly volatile (18.36%). The asset-light concept does not apply to banks; the key variables are the asset quality cycle and liability cost management.
"Cash content" test of earnings (bank basis replacing FCF): The credibility of a bank's reported profits depends on whether provisions are adequately made. SPD Bank's provisioning intensity (impairment losses/revenue) was 40.7%/37.9%/36.1% in 2024/2025/2026H1, significantly above the peer average—this is the lagged cost of its historical legacy (corporate risk from 2015–2020), and it means the income statement is still dominated by the disposal of existing risk. Estimated NPL formation rate (provisions minus net provision changes) was ~0.91% in 2024FY, ~0.89% in 2025FY, and ~1.19% annualized in 2026H1 (Guosen basis: 1.45%)—formation pressure is re-emerging in 2026, and the "scissors gap" versus the declining book NPL ratio is the core of the market discount. ROE recovered from 5.21% in 2023 to 6.76% in 2025, driven by lower impairment provisions (-5.15%) and cost reduction (cost-to-income ratio 29.65%→28.50%), not revenue expansion (revenue still -1.55% in 2024)—the slope of a recovery-driven ROE rebound must inevitably flatten.
Return on capital: ROE of 6.76% is still only half of China Merchants Bank's (13.42%), below CITIC (9.39%) and Industrial Bank (9.15%), in the same tier as Minsheng (6.05%) and Everbright (6.56%); however, SPD Bank's asset quality (NPL 1.25%/provision coverage 197.81%) is significantly better than the latter two (Minsheng 1.47%/142%, Everbright 1.44%/150%), delivering superior risk-adjusted returns at the same ROE tier.
Improved liability structure is the most solid part of this earnings recovery: cost of deposits 1.97% (2024)→1.59% (2025)→1.35% (2026H1), marginal easing of time-deposit migration (time deposit share 62.55%→61.87%, demand deposits recovering to 37.85%), deposits +5.05% in six months outpacing loans +2.88%—the liability side has shifted from "high-rate deposit solicitation" to "settlement deposits + low-cost financial bonds (RMB 30 billion at 1.56% coupon)", which is the micro foundation for NIM stabilization amid industry-wide margin compression.
Quality red flags (extracted from filing): ① Profit growth relies on line-item switching (loan provisions +29.9% vs. reversal of investment impairments); ② Retail NPLs rising across the board; ③ Trading gains account for ~14.5% of revenue with extreme volatility; fair value plus FX losses of RMB -6.534 billion in six months.
Consistency between words and deeds: pragmatic (4 of 5 commitments delivered). The 2024 annual report committed to "double decline in NPLs, higher provisions"→2025 delivered NPL balance -RMB 1.164 billion / NPL ratio 1.26%; "strive for better NIM performance"→2026H1 NIM 1.43% (+1bp) delivered; "payout ratio above 30%"→30.22% delivered; "revenue stabilization"→+1.88%/+3.55% expanding sequentially, delivered (modest magnitude); "overall asset quality remains stable"→book metrics delivered, but retail deterioration and the jump in migration rates were not separately attributed—partially delivered. Since Chairman Zhang Weizhong and President Xie Wei took office at end-2023, guidance has been conservative and the track record of delivery is good.
Shareholder-friendliness: neutral to friendly. Payout ratio above 30% for three consecutive years; the FY2025 dividend (DPS RMB 0.42, +2.4% YoY) was completed on 2026-07-16; but there is no buyback record, and repeated investor inquiries about "share price down 30%, market cap management" received only boilerplate responses (SSE e-interaction, Jun–Sep 2026), with no commitment to buybacks or share increases; the bank also did not join the industry's interim dividend wave in the 2026 interim report—returns are "stable" but lack incremental signals. Historical dilution: the RMB 50 billion convertible bond issued in 2019 took nearly 7 years to complete conversion by 2025 (+13.5% share count); the dilution from that round has been absorbed, but the process was lengthy.
Risk signals: ① Dense related-party credit—credit lines approved in Aug–Sep 2026 to Shanghai International Group (RMB 35 billion), China Mobile (RMB 40.5 billion), and Bailian Group (RMB 16 billion), all processed through risk committee + independent director + board procedures; properly disclosed but the scale warrants attention; ② Compliance fines at subsidiaries—SPD Wealth Management fined RMB 9.1222 million (failure to identify related parties / liquidity indicators not met), credit card center fined RMB 3.2 million (8 violations); retail-line internal controls still catching up; ③ Executive appointments—on 2026-10-16, Vice President Cui Bingwen was approved to concurrently serve as Chief Risk Officer/Chief Compliance Officer; the dual-chief risk setup strengthens governance; ④ 24 village banks converted into branches within the year, with changes to consolidation scope.
| Segment | Scale/Share (latest period) | YoY/Trend | One-line Business Logic |
|---|---|---|---|
| Corporate Banking | Corporate loans 65.26%; corporate net operating income RMB 68.579 billion (2025FY, parent-bank basis) | Corporate loans +8.23% in 2025FY | Five key tracks (technology/supply chain/inclusive finance/cross-border/treasury) are the main growth pole; NPL ratio 1.08%, the best |
| Retail Banking | Retail loans 32.41%; retail AUM RMB 4.66 trillion (+20.26%) | Retail balances -1.37% in six months; NPL 1.67% (+20bp) | AUM/wealth management is the second curve, but credit-side risk exposure is the biggest current flaw |
| Financial Markets & Treasury | Investment interest income 17.55% of total business income (2026H1); custody RMB 20.21 trillion | Investment interest income -0.91% | "Strong trading" strategy is highly dependent on the bond market; volatile returns |
| Yangtze River Delta Segment | Revenue share 25.29%, operating profit share 44.28% (2025FY) | Loans RMB 2.00 trillion (+RMB 163 billion) | Core base: leading regional scale in deposits/loans among joint-stock banks; asset quality better than the whole bank |
| Overseas & Subsidiaries | Revenue share 9.48%, operating profit share 11.13% | Overseas revenue +23.98% | Hong Kong/Singapore/London branches + SPD International; group license synergies |
| Corporate Real Estate (risk exposure basis) | Loans RMB 379.589 billion, 6.47% share, NPL ratio 3.32% | Balance -5.1% in six months; NPL ratio stabilizing at elevated level | Workout period: list-based management + group coordination; deleveraging has begun |
Profit engine: Corporate banking is the absolute mainstay (65%+ of loans, lowest NPL ratio of 1.08%), and the Yangtze River Delta segment contributes 44.28% of group operating profit—the group's profit center lies in the region, not the business line. The retail line accounts for one-third of revenue but is in a risk exposure period; its marginal changes (not corporate) will determine profit elasticity over the next four quarters. The biggest divergence in earnings quality structure: retail loan yields are higher than corporate, but rising current risk costs are eroding the spread; the mortgage interest subsidy policy (borne 90%+10% by fiscal authorities, contract rates unchanged) has limited impact on the asset side of mortgages.
Accounting red flags: ① Restructured loans +28.7% in six months to RMB 48.811 billion (0.83% of loans), far exceeding loan growth of 2.88%—under the new risk classification rules, these represent a "future NPL front-run queue," possibly involving real estate/LGFV extensions—moderately severe; ② Structural switching in impairment provisions—loan provisions +29.9% (RMB 34.590 billion) coexisting with reversal of financial investment impairments (-RMB 1.636 billion), flattening total impairment growth to +5.01%, showing traces of profit smoothing—low-to-moderately severe; ③ Large swings in effective tax rate (2025H1 9.79%→2026H1 13.97%, 2025FY 5.56%)—net profit is sensitive to tax-exempt asset allocation; the company has explained this—low severity.
Cross-period consistency: ① Loan migration rates—special mention 27.50%→29.31%→59.30%, substandard 85.99%→140.11%; the deterioration in forward-looking indicators coexists with book NPL stabilization, and the company has not explained it (this report believes it cannot be attributed to the new classification rules, as reclassification was completed at end-2025); ② Retail NPL ratio rebounded rapidly at 1.61%→1.47%→1.67%; management stated in the summary that "overall asset quality remains stable" but did not separately attribute the retail deterioration; ③ Provision coverage ratio 173.51%→186.96%→200.72%→197.81%, first decline after four consecutive years of increases, consistent with the stated "intensified disposal efforts"; ④ Corporate real estate NPL ratio 2.50%→3.36%→3.32%, balance beginning to decline in 2026, consistent with management's "resolve existing, strictly control new" statement; ⑤ NIM 1.77%→1.52%→1.42%→1.42%→1.43%, deep decline followed by stabilization, consistent with the "liability-cost-driven" explanation.
Conclusion: No fatal accounting-manipulation suspicion; provision and write-off bases are consistent. However, the absence of management explanation for two forward-looking indicators—migration rates and retail NPLs—is where information asymmetry is most concentrated; the disclosure quality of the Q3 report is itself an observation point.
Current market data (2026-10-27 close): Share price RMB 9.82, market cap RMB 327.06 billion; PB 0.4339x (1/3/5/10-year percentile 48%/46%/58%/29%, 10-year median 0.55); PE(TTM) 6.38x (5-year 65% percentile—earnings are in a recovery channel, so PE percentile is less meaningful than PB); dividend yield 4.28% (2026E ~4.7%), spread of 2.67 percentage points over the 10-year government bond yield of 1.61%, historically wide. Peer comparison:
| Company | PB | PE(TTM) | ROE (latest available) | 2026H1 Net Profit Growth | NPL Ratio / Provision Coverage |
|---|---|---|---|---|---|
| China Merchants Bank | 0.95 | 7.15 | 13.42% (H1 annualized) | +2.02% | 0.94%/385% |
| China CITIC Bank | 0.68 | 7.07 | 9.39% (2025 weighted) | +3.08% | 1.15%/203% |
| Industrial Bank | 0.48 | 5.36 | 9.15% (H1 annualized) | -4.66% | 1.08%/226% |
| SPD Bank | 0.43 | 6.38 | 6.76% (2025 weighted) | +4.08% | 1.25%/198% |
| China Everbright Bank | 0.36 | 5.53 | 6.56% (H1 annualized) | -24.01% | 1.44%/150% |
| Minsheng Bank | 0.27 | 5.40 | 6.05% (H1 annualized) | Slight decline | 1.47%/142% |
(PB/PE are the 2026-10-27 close cross-section, calculated in this report; ROE bases differ by company, as noted)
Market-implied expectations: Solving back from 0.43x PB and 4.28% dividend yield (Gordon: PB=(ROE−g)/(Ke−g)), the current price implies a perpetual ROE of 6.6%–7.3% (Ke 9%–10%) and an implied perpetual DPS growth of 4.5%–5.5%—overlapping with the actual ROE of 6.76% and sustainable growth of 4.7%. In one sentence: the current price demands that SPD Bank's "ROE stays forever around 6.7% and the payout ratio never rises," while in reality ROE has recovered for two consecutive years, NIM has stabilized, and the provision buffer is thicker than same-ROE-tier peers. The market is not pricing in "overstated book value" (otherwise PB would be below 0.30), nor paying for any recovery—a premium did exist over the past 12 months (52-week high of RMB 12.97 ≈ 0.57x PB) but was withdrawn after the 2026 retail exposure.
Three-layer value (EPV): ① Asset value (floor reference): extremely pessimistic liquidation basis of 0.30–0.40×BPS = RMB 6.8–9.1 (Minsheng at 0.27x is the going-concern reference); ② Zero-growth EPV (mechanical basis): under a zero-growth full-payout assumption with sustainable dividend of RMB 0.42/share and cost of equity of 9.5%, EPV/share is RMB 4.42, 45% of the current price, with the residual (growth option) at 55%—but this mechanical basis is distorted for banks: a bank retains ~72% of earnings and mechanically compounds ~4.7% in book value and dividends at ROE 6.5%; zero growth means zero improvement, not zero dividend growth; ③ Zero-improvement steady state (growth-inclusive DDM): discounting sustainable DPS of RMB 0.42 at g=4.7%, Ke 9%–10%, steady-state value is RMB 8.26–10.18, midpoint RMB 9.12, with the current price near the midpoint—the recovery option is worth only ~RMB 0.7/share (~7%); the market pays nearly zero premium for "ROE back to 8%, payout 35%." Overall, the current price is mainly supported by the zero-improvement steady-state layer; this conclusion is sensitive to Ke/g (steady-state range RMB 7.2–11.6) and is a range judgment, not a precise equation.
Three scenarios and odds (fair range based on PB × 2026H1 BPS of RMB 22.63):
| Scenario | Probability | Fair Range | vs. Current Price | Core Assumptions and Swing Factors |
|---|---|---|---|---|
| Bear | 30% | RMB 7.9–8.6 (midpoint 8.25) | -12%~-20% (midpoint -16.0%) | Retail formation sustained above 1.4%+, provisions sliding to the 150–170% tier eroding profits, NIM below 1.35%, net profit turning negative (reference: Industrial -4.7%/Everbright -24% samples); PB 0.35–0.38 (anchors: own 3-year PB floor of 0.32; Everbright 0.36/Minsheng 0.27 as going-concern references) |
| Base | 45% | RMB 10.2–11.3 (midpoint 10.75) | +3.9%~+15.1% (midpoint +9.5%) | NIM 1.42%±2bp, NPL ratio ≤1.30%, provisions ≥190%, net profit ~+4%, payout 30%; PB 0.45–0.50 (anchors: Gordon fair PB 0.46–0.51 + peer PB-ROE cross-section + quality premium positioning between Everbright/CITIC, independent of current price) |
| Bull | 25% | RMB 12.4–13.6 (midpoint 13.0) | +26%~+38% (midpoint +32.4%) | Retail formation inflection confirmed, ROE back to 8% by 2028, payout 35% (DPS≈RMB 0.61; yield still 4.7% at RMB 13, meeting insurer thresholds), insurance/passive allocation inflows; PB 0.55–0.60 (anchor: Gordon ROE8%/p35%→0.58–0.65, conservative side taken; CITIC's 0.68x this cycle as the demonstrated ceiling) |
Probability-weighted fair midpoint is ~RMB 10.6 (+7.6%); the current price sits at the lower edge of the three-scenario distribution—odds are mildly favorable but not deeply undervalued, with the bear case's 30% probability (higher than typical names, reflecting the real uncertainty of the retail headwind) constraining certainty. External cross-check: the cautious end of the street (Guosen, 2026-08-28) maintains Neutral with 2026–28E of RMB 52.8/55.7/59.0 billion, whose earnings forecasts are no lower than this report's base case; the optimistic end (Zheshang Securities, 2025-10-09, target price RMB 16.45 ≈ 0.71x PB) was formed before the retail exposure and is significantly incompatible with this report's PB-ROE framework—used only as an upper-bound reference. This report's base-case range is ~15% more conservative than the sell-side consensus target price median of RMB 12.8.
Own earnings forecast (earnings reconciliation anchor):
| Period | Revenue (RMB bn) | Net Profit Attributable to Parent (RMB bn) | Core Drivers |
|---|---|---|---|
| FY2026E | 176.5–180.0 (+1.5%~+3.5%) | 51.5–53.5 (+3.0%~+7.0%) | Corporate loans +5~6% supporting volume (retail continues to shrink amid industry-wide household loan contraction headwinds), NIM 1.41–1.44%, other non-interest income as the biggest variable, provisions fluctuating narrowly at 190–200% |
| FY2027E | 181.0–186.0 (+2%~+4%) | 54.0–56.5 (+3%~+7.6%) | Continued release of deposit repricing benefits, zero-growth cost assumption, impairment normalization; lower bound corresponds to renewed NIM pressure |
(Compared with consensus 2026E ~RMB 52.4bn / 2027E ~RMB 55.5bn, and Guosen's 52.8/55.7; the seasonal low base for banks' H2 profits is already factored in—2026E range midpoint implies H2 net profit of ~RMB 21.5bn, +6.3% YoY)
Conclusion: Reasonably cheap + cautiously bullish. Quality is that of "a joint-stock bank in mid-recovery"—absolute ROE is only half of CMB's, but its direction, NIM, and asset quality trends rank among the top of the five comparable joint-stock banks; pricing is "steady-state priced, recovery free"—dividend yield of 4.28% and the historical floor of 0.35–0.38x provide a double floor; the base-case fair value of RMB 10.2–11.3 implies +3.9%~+15.1% upside. To state quality and price separately: this is not a "cheap good bank," but a "fairly priced, recovery option free of charge, modestly favorable risk-reward" recovery play.
Industry Size: Total assets of China's commercial banking system stood at RMB 433.04 trillion at end-June 2026 (+7.5%), of which joint-stock commercial banks held RMB 80.4 trillion (16.1% share) and large commercial banks RMB 221.6 trillion (44.5% share) (NFRA, published August 2026). Commercial banks' full-year 2025 net profit was RMB 2.4 trillion (NFRA), while listed banks (58 A+H) posted combined 2025 revenue of RMB 5.94 trillion (EY basis). Over the next 2–3 years, the industry will enter a "new normal" of "low NIM + steady scale growth": volume growth centering at 6%–8%, NIM stabilizing at low levels around 1.4%, and low-single-digit profit growth. 2026H1 has already confirmed this: 42 A-share listed banks posted revenue +7.42% and net profit attributable to parent +2.96% (revenue recovering faster than profit, mainly due to high growth in bond-investment-related non-interest income), while commercial banks' net profit was RMB 1.2 trillion (-0.6%).
The cyclical position of net interest margin is the primary variable in banking sector pricing this cycle: official trajectory 1.69% (2023) → 1.52% (2024) → 1.42% (2025) → 1.40% (2026Q1) → 1.41% (2026Q2, up 1bp QoQ — the first quarterly rebound since 2022Q1). The rebound is driven by the liability side: high-rate time deposits originated in 2023–2024 are repricing at maturity en masse in 2026, a certain cost tailwind; on the asset side, loan yields are still declining (LPR unchanged for 17 consecutive months after the 10bp cut in May 2025 through October 2026, at 3.0% for 1-year and 3.5% for over-5-year), with the "volume up, price down" pattern intact (loan-to-deposit ratio 80.08%). Industry consensus (Fitch Bohua, etc.): stabilization but not reversal; retail loan risk exposure constrains the magnitude of profit improvement.
Industry Chain and Value Distribution: Upstream are liability-side funding providers (depositors/central bank/interbank market) — under the market-based deposit rate adjustment mechanism and interest rate self-discipline mechanism, banks' bargaining power over depositors is strengthening (ban on manual interest supplements plus multiple rounds of posted-rate cuts), which is the institutional foundation of this NIM stabilization; midstream is the bank's own role as funding intermediary and risk manager; downstream are asset-side borrowers — effective credit demand is insufficient (household deleveraging; nationwide household loans -RMB 1.03 trillion in the first three quarters of 2026), top-quality borrowers' bargaining power is rising, and LGFV debt resolution swaps high-rate assets with low-rate government bonds (approximately RMB 10 trillion of swaps, reducing interest costs by about 2.5 percentage points) — asset quality improvement comes at the cost of yield concession, which is the key to understanding why "NPLs falling but NIM also falling" coexist. SPDB's position: a nationwide funding intermediary focused on corporate credit; deposit growth (+5.05%) outpacing loan growth (+2.88%) places it on the beneficiary side of the liability cost tailwind.
Competitive Landscape: Banking is a tiered oligopoly — asset share structure of 44.5% for large banks vs. 16.1% for joint-stock banks, nationwide banking licenses frozen for over a decade, extremely high entry barriers. Within joint-stock banks: CMB (RMB 13.79 trillion), Industrial Bank (RMB 11.46 trillion), and SPDB (RMB 10.41 trillion, overtaking CITIC in 2026Q1) rank in the top three, CR3 ≈ 44%; profit concentration is even higher — of the RMB 271.4 billion in 2026H1 net profit across 10 listed joint-stock banks, CMB alone accounts for 28.2%. Competition takes a "sandwich" structure: large banks, leveraging funding cost advantages, are moving down into inclusive/retail/corporate banking (large banks' asset growth 8.5% vs. joint-stock banks' 6.2%, share gap widening ~2.3 percentage points per year), while leading city commercial banks (Ningbo/Jiangsu/Hangzhou with 2026H1 net profit growth of around +12%) squeeze the mid-tier through regional depth. With NIM near 1.4%, there is no room for price wars; competition is shifting to non-interest income and customer base management. Substitution threats come from direct financing (bond markets siphoning quality corporate clients) and wealth management/public funds (siphoning deposits). Bargaining power: strengthening versus depositors, weakening versus top-quality borrowers, weakest versus government-related clients.
Cycle and Regulation: The industry is in a dual-cycle overlap of "tail end of NIM decline + mid-stage asset quality improvement." Leading indicators: LPR and posted deposit rate adjustments, the pace of time deposit maturity repricing (peaking in 2026), retail NPL formation rates (industry-wide increases in 2026H1: mortgages +11bp, credit cards +27bp, consumer loans +22bp), and the progress of LGFV/real estate exposure swaps (whitelist loans exceeding RMB 7 trillion). The regulatory direction is overall neutral-to-positive: debt resolution + whitelist mitigate two major historical burdens (positive for asset quality and provision releases); new capital rules (effective 2024) + D-SIB list expansion to 21 banks (SPDB in the second group, +0.5% additional capital) raise capital requirements; normalization of interim dividends (21 banks totaling RMB 266.587 billion in 2026) reinforces banks' fixed-income-like attributes — SPDB has not yet participated.
Positioning within the Industry: A repairer among chasers. Third in asset scale among joint-stock banks (total assets RMB 10.41 trillion), fourth in revenue and net profit, 18th in the global Top 1000 banks (+1 versus the prior year); profitability ranks lower (ROE on par only with Minsheng/Everbright within its tier), but asset quality ranks higher (NPL 1.25% vs. Minsheng 1.47%/Everbright 1.44%; provision coverage 198% vs. 142%/150%) — "stable-to-slightly-declining scale share, rising profit and quality rankings" describes its trajectory over the past three years. Share trends: total asset growth of 3.21% (half-year) below the joint-stock bank average of 6.2% means slow dilution of scale share; but profit growth (+4.08%) above the joint-stock bank average (about -3.4%) means recovering profit share. Moat sources: Yangtze River Delta corporate client base (Shanghai headquarters; infrastructure/manufacturing/tech-innovation foundations), liability-side repair capability (2025 deposit cost rate -38bp, among the best), and the group's full licenses. Core weaknesses: retail and wealth management far behind CMB, low absolute ROE, and retail lending is shrinking (balances -1.37% over six months).
Note on industry positioning methodology: The company's own annual-report claims do not materially conflict with third-party data (asset scale ranks third among joint-stock banks under both; the 2026Q1 overtaking of CITIC is a third-party estimate by 21st Century Business Herald and others — the company does not disclose rankings itself).
Bank stock investing differs from most sectors: the valuation anchor is PB-ROE and the dividend yield–treasury yield spread, while the operational battleground is the dynamic balance among four corners — NIM, asset quality, capital, and dividends. SPDB's four-corner coordinates:
NIM: 1.43% (2026H1), flat with the industry average (1.41%), but the path of achievement is better than peers — liability cost rate -39bp YoY fully offsetting asset yield -36bp; Industrial Bank's NIM in the same period was -11bp versus end-2024, Minsheng +7bp, while SPDB at +1bp sits in the improving group. Future elasticity depends on deposit repricing (time deposit share still high at 61.87%, tailwind not exhausted) and the LPR path.
Asset Quality: Book-level three declines (NPL 1.25%/special mention 1.97%/overdue 1.84%) combined with rising formation (estimated 1.19%–1.45%) make SPDB unique among the five comparables: "best on-book, coexisting off-book pressure." The market's 0.43x PB (19%–59% premium to Minsheng/Everbright, 37% discount to CITIC) precisely prices this structure: a quality premium exists, but formation-side uncertainty suppresses re-rating.
Capital: CET1 at 9.05%, a nominal 105bp buffer; but at a 30% payout ratio, internal capital generation supports only ~4.5% annualized RWA growth vs. actual 4.4% — this zero-headroom state is the root of the mutual exclusivity between raising the payout ratio and expanding scale, and the constraint that gives the bull case (35% payout) only a 25% probability. D-SIB second-group status means capital requirements can only get tighter.
Dividends: A 4.28% dividend yield plus 2.67pp spread over treasuries provides a fixed-income-like floor of value at the current price; but DPS growth (+2.4%) is essentially in line with net profit growth (+4%), and with the payout ratio constant at 30%, yield improvement can only come from a stagnant share price or accelerated earnings — dividends are a floor, not a catalyst, unless the ratio is raised in the 2026 annual report.
Policy implications of the four-corner interplay: if regulators push interim dividend expansion to joint-stock banks, SPDB faces the trade-off of "payout↑ → capital headroom↓ → scale growth↓"; if retail risk clears faster than expected, provision releases could simultaneously support profit and dividends. The direction of rebalancing across the four corners over the next 12 months will determine where SPDB lands within the 0.43–0.60x PB range.
Rating: Cautiously bullish; confidence 0.55; time horizon 6–12 months. Logic chain: ① The current price of RMB 9.82 is near the no-improvement steady-state value (central case RMB 9.12), making the repair option nearly free while dividends and historical floors provide dual downside support; ② the repair direction of NIM stabilization + corporate risk clearing is data-supported, but accelerating retail NPL formation and the 2027 profit transition constitute substantive headwinds — odds are moderately positive rather than extreme; ③ catalysts are dense (Q3 report on Oct 31 → December LPR → 2026 annual report and dividend plan in March 2027), each a two-way test at every node. Strategically: build positions in stages at current levels, sized against the 30% probability of the bear case; add after confirmation signals (Q3 NIM ≥1.42% and retail formation declining); exit immediately on falsification signals (provision coverage <185% or NIM <1.38%).
Key Risks: ① Retail NPL formation sustained at 1.4%+ dragging provision coverage toward the 150–170% band, turning provision releases into provisions drag (bear case -16%); ② normalization of the three pillars of fees/impairments/tax rate in 2027, compounded by a second leg of NIM decline, turning net profit growth negative and driving collective downward revisions of sell-side 2027E; ③ with zero internal capital headroom, payout ratio and scale expansion are mutually exclusive — interim dividend expansion could become a capital burden; ④ rising bond yields turning other non-interest income from support into drag + a reversal of crowded trades in high-dividend stocks; ⑤ compliance rectification at wealth management/credit card subsidiaries constraining retail expansion.
Monitoring Checklist (Next 6–12 Months):
This report is based on public information (company periodic reports/announcements, regulatory data, sell-side research, and public news) and the author's independent estimates, with data as of the market close on 2026-10-27; figures marked "estimated," such as NPL formation rates, are not company-disclosed values. This report does not constitute investment advice.