GF Securities Additional Discussion Report: SSC Discussion on Balance-Sheet Risk, Support and Financial Policy

• Report date: 2026-07-17
• Issuer / Theme: GF Securities Co., Ltd. — SSC discussion on balance-sheet expansion, support, controls and financial policy
• Report type: additional_discussion
• Discussion scope: External SSC Q&A conducted on 2026-07-16; this report retains the questions, answers, follow-up challenges and monitoring implications.
• Reference context: GF Securities Issuer Summary dated 2026-05-21, Working Note dated 2026-06-12, and the 2026-07-16 SSC discussion log.

1. Purpose and Treatment

This is a supplementary record of an external SSC discussion. It is not a new verified finding, a replacement for the existing issuer summary, or an investment recommendation. Numerical sensitivities, warning levels and proposed portfolio responses in the discussion are analytical illustrations unless explicitly described below as context already confirmed in the existing project materials. They should not be read as disclosed management limits, regulatory intervention thresholds or rating-agency triggers.

The existing issuer summary already establishes the core context: GF Securities is a market-based Chinese securities company with improved 2025 and Q1 2026 earnings and supportive reported parent regulatory ratios, but with material growth in trading assets, repo and other market-sensitive funding. The SSC Q&A tested how that position could become credit-negative, how far parent support can be extended to GFHK and offshore instruments, and what management behaviour would preserve or consume the reported buffers.

2. Discussion Takeaway

The central read-through was not that one identified loss, ratio or compliance event currently undermines GF Securities' credit. Rather, the discussion treated the enlarged balance sheet as increasing the importance of a correlated stress transmission: trading-book valuation losses, repo haircuts, derivative margin calls, client activity, risk-capital charges and funding access could interact more quickly than an ordinary earnings slowdown.

The discussion also sharpened two distinctions that matter for bondholders. First, reported parent capital and liquidity metrics are meaningful supports, but their usability cannot be assessed from public disclosure without the composition of unencumbered liquidity, collateral, trading sensitivity and regulatory-risk calculations. Second, domestic regulatory interest in preserving market functioning, GFHK's strategic role and an identified parent guarantee are different forms of support. They do not establish identical protection or timely liquidity for every GFHK or offshore-SPV obligation.

The January 2026 capital raising was discussed as initially credit-supportive but strategically growth-oriented: the issue is whether it remains a durable buffer after overseas deployment, rather than whether the issuance itself was positive. A prolonged earnings downturn would become more significant if management continued capital-intensive offshore and trading expansion while retained earnings, liquidity buffers and regulatory headroom weakened.

3. Q&A Discussion Notes

3.1 What drove the balance-sheet expansion, and when would it become a liquidity problem?

Question intent. The first question asked whether the increase from RMB758.7bn of assets at end-2024 to RMB1.122tn at end-March 2026 was principally client-driven and reversible, or instead reflected proprietary positions, leveraged bonds, derivatives and margin financing that could create simultaneous valuation and funding pressure.

Answer points. The discussion identified the largest disclosed movement as financial assets held for trading, alongside growth in margin financing, client cash and settlement balances, repo liabilities and trading liabilities. It therefore argued that the expansion was not explained solely by client cash. It also noted that the available category data cannot distinguish outright proprietary risk from hedged market-making, client facilitation or structured-product hedging. Existing project context confirms that total assets, trading activity and repo expanded materially; the precise risk and liquidity composition remains unconfirmed.

Follow-up deepened the issue. The next question correctly challenged whether high reported ratios can be translated into a specific market shock. The response concluded that public disclosures do not provide the trading-book duration, spread sensitivity, equity delta, repo-collateral haircuts, unencumbered asset inventory, derivative margin terms or LCR cash-flow components needed to calculate a verified breaking point. Its reverse-stress examples treated the following as a correlated warning combination, not a forecast: a 100-150bp bond-spread widening, a 20% equity decline, a five-percentage-point repo-haircut increase, gross client outflows of about 10%, RMB8bn-RMB10bn of derivative margin demand and higher risk-capital reserves.

Credit implication. A moderate trading loss could remain an earnings event. The more relevant credit inflection would be a decline in usable proprietary liquidity at the same time as collateral demands rise and funding tenors shorten. The Q&A proposed that risk coverage moving below about 180%, LCR below about 150% or NSFR below about 130%, especially while repo and short-term funding stay elevated, should prompt closer review. These are discussion-derived portfolio markers, not disclosed GF Securities thresholds.

Doubts and unconfirmed matters. The key unknowns are the share of the trading book that is hedged or client-driven, the amount and quality of unencumbered proprietary liquidity excluding client funds, collateral encumbrance, derivative netting and stress-margin outcomes, and the management actions expected before a regulatory minimum is approached. Client balances should neither be treated as stable bank deposits nor automatically equated to a like-for-like drain on house liquidity.

3.2 What support can creditors assume, and when would GFHK cease to be fully aligned?

Question intent. The discussion asked how much practical support creditors could expect from regulators, local-government-linked shareholders and major shareholders in severe stress, and then asked for observable developments that would invalidate the assumed support benefit or GFHK alignment.

Answer points. The answer separated domestic operational continuity from contractual creditor protection. It described GF Securities as relatively non-government-controlled and treated regulatory liquidity facilitation or orderly domestic deleveraging as more plausible than an unconditional government rescue. It further noted, as reported in the discussion, that GFHK is wholly owned, strategically important and has received parent capital support, and that some identified offshore instruments have direct parent guarantees. These forms of evidence are not interchangeable: an enforceable parent guarantee offers stronger instrument-level protection than discretionary shareholder action, a keepwell-style arrangement or an expectation of regulatory coordination.

Follow-up deepened the issue. The support follow-up focused on timing and fungibility. It proposed that support assumptions could weaken before any missed payment if parent regulatory pressure caused domestic capital preservation to take priority, cross-border approvals became delayed, GFHK increasingly relied on unguaranteed structures, or offshore access diverged from parent domestic access. The Q&A proposed an amber review zone where parent buffers weakened, shareholder participation in capital raising was below expectations, transfers were delayed or adjusted offshore spreads diverged persistently; it treated removal of a support uplift, loss of GFHK's core status, formal transfer restrictions or failed parent support as decisive red flags.

Credit implication. GFHK should not be automatically passed through at the parent level merely because it is strategic. For each offshore obligation, bondholders need to distinguish parent willingness, parent capacity, regulatory permission and legal enforceability. The existing issuer summary's instruction to verify issuer, guarantor, guarantee scope, remittance arrangements and covenants remains directly applicable.

Doubts and unconfirmed matters. The exact regulatory-support mechanism, shareholder capacity and willingness, GFHK's immediately drawable parent facilities, stress-time cross-border approval process, the proportion of offshore debt covered by direct parent guarantees and a comparable parent-versus-GFHK market-spread measure were not established by the SSC discussion.

3.3 Which control failure would change the credit analysis?

Question intent. The Q&A considered whether a conduct or regulatory event could migrate from a remediation issue into a broader capital, liquidity and refinancing concern. The follow-up sought one observable escalation event rather than a broad watchlist.

Answer points. The response identified a group-wide failure in trading, derivatives, valuation, collateral or counterparty-risk control as the most damaging case, because it could challenge reported risk measures while creating losses, margin needs and counterparty withdrawals. It also discussed the historical Kangmei-related restriction as evidence that a transaction-level control failure can lead to direct business restrictions. The SSC log stated that the 2025 annual report described effective internal controls but also recorded several separate compliance or self-regulatory findings. This supports monitoring recurrence and severity; it does not demonstrate a current enterprise-wide failure.

Follow-up deepened the issue. The selected primary trigger was a regulator suspending or materially restricting a balance-sheet-intensive activity — for example derivatives, proprietary trading, repo, margin financing or overseas institutional activity — because this would combine external validation of a control problem with an immediate operational and funding consequence. The Q&A regarded verified coordinated reductions by several material counterparties as functionally comparable, while noting that individual bilateral limit changes may be difficult to interpret.

Credit implication. A formal restriction could affect earnings capacity, the ability to hedge legacy positions, liquidity, counterparty confidence and GFHK's operating role before a reported regulatory ratio breaches its minimum. A reassessment would need to include both the parent and GFHK, the affected business, collateral and liquidity effects, and whether parent-to-offshore support remains operationally available.

Doubts and unconfirmed matters. The current CSRC classification, the completion and acceptance of remediation for recent findings, any operational-risk capital add-on, and counterparty-limit behaviour have not been independently established in this report.

3.4 Has the January 2026 capital raising strengthened buffers or started a releveraging cycle?

Question intent. The Q&A tested whether the January capital raising was being retained as a durable loss-absorption and liquidity buffer or being deployed to accelerate overseas trading, derivatives, margin financing and institutional expansion.

Answer points. The discussion stated that the combined placement and convertible-bond proceeds were allocated predominantly to international-business expansion, with a smaller portion for liquidity support and overseas-liability repayment. It also stated that Q1 2026 parent regulatory ratios improved while consolidated trading assets, repo liabilities and derivative assets expanded faster than attributable equity. The appropriate interpretation was therefore active growth with temporarily improved buffers, not evidence that rating discipline had already been abandoned. The convertible-bond component should not be treated as identical to immediately available common equity before conversion.

Follow-up deepened the issue. The discussion proposed that a shift to a releveraging cycle would require three persistent elements: GFHK deploying actual injected capital into balance-sheet-intensive exposure at a high multiple; two or more parent capital or liquidity buffers declining despite positive earnings; and management continuing expansion or distributions rather than restoring buffers. It suggested review markers around risk coverage below 200%, LCR below 175%, NSFR below 140% or capital leverage below 10%, when linked to offshore risk growth. These are analytical markers only.

Credit implication. The capital raise supports credit only while it improves loss absorption, unencumbered liquidity and funding tenor after deployment. The concern is not GFHK growth by itself, but whether that growth consumes parent and subsidiary buffers faster than earnings and stable funding rebuild them. Unguaranteed GFHK and SPV claims would be more exposed than direct parent or directly parent-guaranteed debt if the pattern emerges.

Doubts and unconfirmed matters. The actual timing and amount of capital transferred to GFHK, its deployed balance sheet, derivatives, repo funding, offshore liquidity and regulatory metrics, and whether repaid liabilities were replaced with short-term funding were not available in the discussion.

3.5 Can diversified earnings replenish capital through a prolonged weak cycle?

Question intent. The final Q&A pair examined a correlated China capital-markets downturn and the management response required during its first year.

Answer points. The discussion characterised wealth management and trading/institutional business as the principal 2025 profit engines, with investment management providing an additional but market-sensitive source and investment banking too small to offset broad weakness. It noted the issuer's history of remaining profitable through weaker years, but argued that diversification would cease to protect credit if weak turnover, trading performance, fee income, margin-finance credit costs and fixed expansion costs coincided for two to three years.

Follow-up deepened the issue. The response treated one weak year as manageable only if management promptly reduced capital consumption: staging or deferring GFHK deployment, slowing trading and derivatives growth, terming out or reducing repo and short-term funding, preserving liquidity, tightening margin standards, adjusting costs and reducing distributions if retained earnings weakened. The defining failure to act would be continued GFHK and group trading expansion while recurring earnings and retained capital generation declined, parent buffers fell and short-term funding or dividends were not reduced.

Credit implication. The key early signal is not a single fall in headline profit. It is a pattern in which retained earnings fail to cover capital consumption, market-dependent trading gains mask weak recurring earnings, margin costs rise and management continues to use buffer capacity. That could affect refinancing confidence and GFHK support capacity before a statutory ratio is breached.

Doubts and unconfirmed matters. Public information considered in the discussion does not isolate normalised earnings from fair-value effects, fixed versus variable costs, asset-management fee sensitivity, or a reliable through-the-cycle margin-financing loss rate. The numerical earnings and ratio levels in the Q&A remain proposed portfolio indicators, not issuer guidance.

4. Candidate Items For issuer_notes.md

The following are candidates for later consideration in issuer_notes.md; they are not updates to permanent issuer memory.

5. Monitoring / Next Check

The next issuer update should seek primary disclosure or instrument documentation rather than treat the SSC sensitivities as facts. Highest-value materials are: the 2026 interim and annual reports; asset and funding composition for trading, repo and derivatives; actual GFHK capital-injection tranches and use of proceeds; GFHK liquidity, maturity and secured-funding information; current rating-agency reports; current CSRC, exchange and SFC actions; and offering documents that establish the precise guarantor and payment mechanics for offshore obligations.

In later reporting, the analytical focus should remain on combinations: buffer trends together with funding tenor and collateral capacity; overseas deployment together with parent regulatory headroom; and control events together with counterparty or refinancing response. No single discussion threshold should replace a review of the underlying exposures and legal structure.

6. Unverified / Pending Items

The SSC discussion included web-checked claims and issuer disclosure references, but this additional discussion did not independently re-verify them. In particular, it does not verify: trading-book market sensitivities; proprietary versus client facilitation exposure; repo-collateral haircuts and encumbrance; derivative margin and netting; internal management floors for regulatory metrics; current rating-agency support assumptions; stress-time cross-border approvals; GFHK liquidity facilities and maturity profile; the current status of remediation and regulatory classification; or the legal protection of every offshore bond.

Accordingly, the reverse-stress scenarios, internal limit reductions and numerical warning levels reported in the SSC Q&A are hypotheses for future verification. They must not be represented as GF Securities commitments, regulator requirements, rating-agency triggers or confirmed forecasts.

7. Reference Context

• GF Securities Issuer Summary, dated 2026-05-21, for the existing parent-credit, market-risk, liquidity and offshore-structure context.
• GF Securities Working Note, dated 2026-06-12, for the current knowledge snapshot and issuer follow-up items.
• SSC external discussion log, dated 2026-07-16, for the questions, discussion claims, illustrative sensitivities and extracted follow-up candidates organised in this report.
• The SSC discussion referred to GF Securities' 2025 Annual Report, 2026 First Quarterly Report, a January 2026 capital-raising announcement and rating-agency materials. These references are retained as discussion context only; no new source verification was performed for this auxiliary report.