Topic2
主题 2

1 │ How do we determine what the goals of financial regulation are—or should be?
1 │ 我们如何确定金融监管的目标是什么——或者应该是什么？

The starting point is translation: turning political slogans such as “protect investors” or “maintain confidence” into analytically tractable objectives (Lecture 2, slides 4–8). Modern scholarship roots that translation in marketfailure analysis. First, identify where realworld finance departs from the neoclassical benchmark—externalities, asymmetric information, public goods, or bounded rationality (PFR Ch 4). Each failure implies a goal: systemicrisk externalities warrant financial stability; information gaps justify investor protection; monopoly rents motivate competition policy.
起始点是翻译：将“保护投资者”或“维持信心”等政治口号转化为可分析的目标（讲座 2，幻灯片 4-8）。现代学术研究将这种翻译根植于市场失灵分析。首先，确定现实金融与新古典基准的背离之处——外部性、信息不对称、公共物品或有限理性（PFR 第 4 章）。每种失灵都意味着一个目标：系统性风险外部性应确保金融稳定；信息差距证明需要投资者保护；垄断租金推动竞争政策。

A second filter is cost–benefit analysis (CBA). Posner & Weyl (2013) argue that without quantifying both sides of the ledger, regulators risk either “faithbased” intervention or laissezfaire paralysis. The Australian Government’s 2023 Guidance Note similarly requires agencies to estimate net present value for any rule with economywide effects. Coates (2015) cautions, however, that finance produces nonpecuniary benefits—trust, resilience—that are hard to price; CBA therefore demands humility and scenario analysis rather than point estimates.
第二个过滤器是成本效益分析（CBA）。波斯纳和韦尔（2013 年）认为，如果不量化账目的两边，监管者可能会面临“基于信仰”的干预或放任自流的瘫痪。澳大利亚政府的 2023 年指导文件同样要求机构对任何具有全经济影响的规则进行净现值估算。然而，库茨（2015 年）警告说，金融会产生非货币收益——信任、韧性——这些收益很难定价；因此，CBA 需要谦逊和情景分析，而不是点估计。

Third comes political economy. Jacobs & King (2018) show how Federal Reserve objectives have morphed with congressional oversight and interestgroup pressure. The Hayne Royal Commission (2019) demonstrates that scandal can elevate consumer protection to parity with prudential safety—even when exante economic models ranked it lower. Goals evolve as social preferences shift; no hierarchy is “timeless.”
第三是政治经济学。雅各布斯和金（2018 年）展示了美联储目标如何随着国会监督和利益集团压力而演变。海恩皇家委员会（2019 年）表明，丑闻可以提升消费者保护与审慎安全并驾齐驱——即使事前经济模型将其排名较低。随着社会偏好的变化，目标也在演变；没有等级是“永恒的”。

In practice, regulators synthesise these inputs through statutory mandates. APRA must “balance safety, efficiency, competition, and contestability” while promoting system stability (APRA Act 1998 s 8), a deliberate hierarchy that places stability above the rest yet forces tradeoff transparency. The UK Financial Policy Committee adopts a similar layered mandate. Agencies periodically publish Regulatory Impact Statements, forcing expost goal validation (APRA 2022).
在实践中，监管机构通过法定职责综合这些输入。APRA 必须在促进系统稳定性的同时“平衡安全、效率、竞争和可竞争性”（APRA 法案 1998 年第 8 条），这是一个故意建立的金字塔结构，将稳定性置于其他因素之上，但强制权衡的透明度。英国金融政策委员会采用类似的分层职责。机构定期发布监管影响声明，强制进行事后目标验证（APRA 2022）。

In sum, goal-setting is iterative: (1) diagnose market failures; (2) test candidate interventions with CBA, mindful of measurement limits; (3) incorporate democratic preferences and adapt statutes accordingly. Only by cycling through these lenses can policymakers decide not merely what goals exist but which deserve primacy at a given time.
总而言之，目标设定是迭代的：（1）诊断市场失灵；（2）使用成本效益分析测试候选干预措施，注意测量限制；（3）纳入民主偏好并相应地调整法规。只有通过这些视角的循环，政策制定者才能决定不仅存在哪些目标，而且在特定时间哪些目标应优先。

2 │ What effects might increased competition between financial institutions have on financial stability?
2 │ 金融机构之间竞争加剧可能对金融稳定产生什么影响？

Conventional wisdom assumes rivalry boosts efficiency and consumer surplus, yet finance complicates the story. Economic theory predicts two countervailing forces. Channel A (discipline): more entrants compress margins, reducing franchise values and thereby curbing riskshifting incentives (charter value hypothesis). Channel B (gambling): thinner margins may tempt managers to reach for yield, increasing fragility (Lecture 2, slide 42).
传统观点认为竞争可以提高效率和消费者剩余，但金融使故事复杂化。经济理论预测存在两种相互抵消的力量。渠道 A（纪律）：更多进入者会压缩利润空间，降低品牌价值，从而抑制风险转移的激励（宪章价值假设）。渠道 B（赌博）：较薄的利润空间可能会诱使管理者追求收益，增加脆弱性（讲座 2，幻灯片 42）。

Empirical evidence leans toward conditionality. Jacobs & King (2018) document episodes where U.S. deregulation fostered credit booms that later destabilised banks. Arcand, Berkes & Panizza (2012) find that once credittoGDP exceeds 100 per cent, growth falls—suggesting intensified competition can overshoot “efficient finance.” The Royal Commission (2019) revealed that aggressive crossselling among Australia’s majors eroded underwriting standards, illustrating Competition→Conduct→Stability spillovers.
实证证据倾向于条件性。Jacobs & King (2018) 记录了美国放松管制助长了信贷繁荣，但后来导致银行不稳定的事件。Arcand, Berkes & Panizza (2012) 发现，一旦信贷占 GDP 的比例超过 100%，增长就会下降——这表明激烈竞争可能会超过“有效金融”。皇家委员会（2019）揭露了澳大利亚主要银行之间积极的交叉销售侵蚀了承保标准，说明了竞争→行为→稳定性溢出效应。

Regulatory design tries to square this circle. APRA’s mandate explicitly subordinates competition to financial safety; the RBA’s Payments Board can promote competition only when “consistent with overall stability” (RBA Act s 10B). Policy tools include countercyclical capital buffers that rise with lending exuberance, ensuring that rivalry does not translate into systemic leverage. Posner & Weyl’s (2013) costbenefit lens supports such symmetric tools: benefits of cheaper credit are tallied against probabilistic crisis losses.
监管设计试图解决这一难题。APRA 的职责明确将竞争置于金融安全之下；RBA 的支付委员会只有在“与整体稳定一致”时才能促进竞争（RBA 法案第 10B 条）。政策工具包括随贷款狂热而上升的反周期资本缓冲，确保竞争不会转化为系统性杠杆。波斯纳和韦尔（2013 年）的成本效益视角支持此类对称工具：更便宜信贷的收益与概率性危机损失相权衡。

In short, competition is neither panacea nor poison; its stability impact hinges on complementary prudential guardrails. Wellcalibrated buffers can let consumers enjoy rivalry’s upsides without exposing the macroeconomy to ruin.
简而言之，竞争既不是万能药也不是毒药；其稳定性影响取决于补充性的审慎保障措施。经过良好校准的缓冲器可以让消费者享受竞争的利好，同时不使宏观经济遭受毁灭。

3 │ Is it better for legislators, regulators, or judges to write financial regulation?
3 │ 是由立法者、监管者还是法官来制定金融监管更好？

Each institutional author offers strengths and weaknesses along expertise, flexibility, and legitimacy. Legislatures supply democratic legitimacy but suffer information deficits and lobbying (Lecture 2, slides 65–68). Detailed statutes risk obsolescence in fastmoving markets—witness the DoddFrank Act’s postcrisis breadth and subsequent partial rollback.
每个制度作者在专业知识、灵活性和合法性方面都有优势和劣势。立法机构提供民主合法性，但存在信息缺陷和游说问题（讲座 2，幻灯片 65-68）。详细的法规在快速变化的市场中存在过时风险——以多德-弗兰克法案在危机后的广泛范围及其后续的部分撤回为证。

Dedicated regulators (e.g., ASIC, APRA) possess technical skill and can iterate rules quickly. Coates (2015) notes they are also best placed to conduct forwardlooking CBA given privileged data. However, “agency slack” and capture loom large; Posner & Weyl (2013) warn that without judicially enforceable CBA standards, insiders may overweight industry costs and underweight diffuse benefits.
专业的监管机构（例如 ASIC、APRA）拥有技术能力，并能快速迭代规则。Coates（2015）指出，由于拥有特权数据，它们也最适合进行前瞻性成本效益分析。然而，“代理松懈”和“俘获”问题突出；Posner & Weyl（2013）警告说，如果没有司法可执行的成本效益分析标准，内部人士可能会高估行业成本并低估分散的收益。

Judges enter mainly through administrative law review. In the U.S., the SEC’s debitcard rule was struck down for inadequate economic justification, nudging agencies toward rigorous analysis (Coates 2015). Courts thus create a metaregulatory incentive but lack the exante expertise to craft intricate capital formulas.
法官主要通过行政法审查进入。在美国，SEC 的借记卡规则因缺乏充分的经济论证而被推翻，促使机构进行严格的分析（Coates 2015）。因此，法院创造了一种元监管激励，但缺乏事前制定复杂资本公式的专业知识。

Optimal practice therefore layers the three: Parliament sets broad mandates and procedural disciplines (e.g., mandatory Regulatory Impact Statements), regulators draft and update technical standards, and courts police process and statutory fidelity. This division of labour balances democratic accountability with technocratic agility while inserting an expost correctness check.
因此，最佳实践是将三者结合：议会设定广泛的任务和程序纪律（例如，强制性的监管影响声明），监管机构起草和更新技术标准，法院监督程序和法定忠诚度。这种劳动分工在民主问责和技术官僚敏捷性之间取得了平衡，同时插入了一个事后正确性检查。

4 │ How does the institutional architecture of financial regulators affect their ability to achieve their goals?
4 │ 金融监管机构的制度架构如何影响其实现目标的能力？

Architecture shapes clarity of mandate, information flow, and coordination costs. Australia’s Twin Peaks model splits conduct (ASIC) from prudential oversight (APRA), supplemented by the RBA and the Council of Financial Regulators (Lecture 2, slides 49–53). The separation mitigates internal goal conflicts—consumer protection versus safety—but introduces interagency coordination challenges handled via memoranda of understanding.
建筑结构塑造了授权的清晰度、信息流和协调成本。澳大利亚的双峰模式将行为（ASIC）与审慎监管（APRA）分开，并由 RBA 和金融监管委员会（Lecture 2, slides 49–53）补充。这种分离减轻了内部目标冲突——消费者保护与安全——但引入了机构间协调挑战，通过谅解备忘录进行处理。

Comparative studies reinforce tradeoffs. The UK’s shift from a unified FSA to a threepillar system (FCA, PRA, FPC) after 2008 improved focus on systemic risk but required new gateways to prevent regulatory gaps. The EU’s “Single Rulebook” centralises standards via ESAs yet leaves supervision largely national, creating uneven enforcement (Lecture 2, slides 60–64).
比较研究强化了权衡。英国在 2008 年后从统一的 FSA 转向三支柱系统（FCA、PRA、FPC）改善了系统性风险的关注，但需要新的入口来防止监管空白。欧盟的“单一规则簿”通过 ESAs 集中标准，但监督仍 largely 国家，导致执行不均（Lecture 2, slides 60–64）。

Jacobs & King (2018) argue that fragmented U.S. oversight diffuses accountability and fosters turf battles, impeding timely crisis action. Conversely, a superregulator risks cognitive overload and capture. Coates (2015) suggests that architecture should mirror functional externalities: where prudential and conduct goals are strongly complementary (e.g., payments data security), integration aids efficiency.
雅各布斯和金（2018）认为，美国分权的监管导致责任分散，并助长地盘之争，阻碍了及时应对危机的行动。相反，超级监管机构存在认知过载和被俘获的风险。库茨（2015）建议，架构应该反映功能性外部性：当审慎目标和行为目标强烈互补时（例如，支付数据安全），整合有助于提高效率。

Hence, architecture is not onesizefitsall; its effectiveness depends on explicit coordination mechanisms, datasharing protocols, and an overarching body (e.g., Australia’s CFR) empowered to reconcile crossmandate conflicts.
因此，架构并非一刀切；其有效性取决于明确的协调机制、数据共享协议以及一个拥有协调跨授权冲突权力的最高机构（例如，澳大利亚的 CFR）。

5 │ What new challenges do AI and other technologies pose for financial regulation?
5 │ 人工智能和其他技术对金融监管提出了哪些新的挑战？

Artificial intelligence reshapes finance on three axes: microconduct, macrostability, and regtech. On conduct, machinelearning credit models risk algorithmic bias and opacity—“blackbox” decisions that frustrate ASIC’s disclosurebased consumer protection regime (APRA Corporate Plan 2023). Behavioural exploitation at algorithmic speed intensifies boundedrationality failures flagged in PFR Ch 4.
人工智能在三个轴向上重塑金融：微观行为、宏观稳定和监管科技。在行为方面，机器学习信用模型存在算法偏见和透明度风险——“黑箱”决策，这使 ASIC 基于披露的消费保护制度（APRA 企业计划 2023）感到沮丧。算法速度下的行为利用加剧了 PFR 第 4 章中提到的有限理性失败。

Macrolevel, AIdriven trading can amplify herding: identical optimisation objectives ingest the same data and react within milliseconds, raising volatility spikes. Jacobs & King’s (2018) politicisation thesis extends here: in crises, blame for AI mishaps may spark kneejerk bans, making preemptive guidance crucial.
宏观层面，AI 驱动的交易可能加剧羊群效应：相同的优化目标会处理相同的数据并在毫秒内做出反应，从而引发波动性尖峰。Jacobs & King（2018）的政治化理论在此适用：在危机中，对 AI 事故的指责可能引发仓促的禁令，因此预先指导至关重要。

Regulators also face an information asymmetry reversal—industry’s data lakes far outstrip supervisory analytics. Posner & Weyl’s CBA imperative becomes harder when model outputs are noninterpretable, complicating riskweighted capital calibration. Nonetheless, technology equips regulators with RegTech: realtime transaction monitoring and naturallanguage processing of filings can narrow gaps if adequately funded.
监管机构也面临信息不对称的逆转——行业的数据湖远远超过了监管分析能力。当模型输出不可解释时，Posner & Weyl 的成本效益分析原则变得更加困难，使得风险加权资本校准变得复杂。尽管如此，技术为监管机构提供了监管科技：如果得到充分资金支持，实时交易监控和自然语言处理申报文件可以缩小差距。

Emerging responses include APRA’s draft prudential standard CPS 230 on operational resilience and the EU’s proposed AI Act imposing risktiered obligations. Yet jurisdictional arbitrage looms: firms may locate servers where oversight is laxer (Lecture 2, slide 68). International standardsetting bodies thus face coordination burdens reminiscent of earlier Basel expansions.
新兴的应对措施包括 APRA 的运营韧性审慎标准草案 CPS 230 以及欧盟提出的 AI 法案，该法案规定了风险分级的义务。然而，管辖权套利现象迫在眉睫：公司可能会将服务器设在监管较松的地方（讲座 2，幻灯片 68）。因此，国际标准制定机构面临着类似于早期巴塞尔扩张的协调负担。

Overall, AI magnifies preexisting regulatory challenges—opacity, speed, crossborder reach—requiring adaptive, techsavvy supervision and dynamic rulemaking.
总体而言，人工智能放大了现有的监管挑战——不透明性、速度、跨境影响——需要适应性、技术娴熟的监管和动态的规则制定。

6 │ Do fintech sandboxes strike the right balance between stimulating innovation and regulatory oversight?
6 │ 金融科技沙盒在刺激创新和监管监督之间是否达到了正确的平衡？

Sandboxes grant limited licences, caps on customer exposure, and supervisory dialogue to novel firms. The economic aim is to reduce the ex ante cost of compliance uncertainty (Didenko 2021) while allowing regulators to learn about emergent risks (Lecture 2, slide 73).
沙盒授予新公司有限的许可证、限制客户风险暴露和监管对话。经济目标是在降低合规不确定性的事前成本（Didenko 2021）的同时，让监管机构了解新兴风险（讲座 2，幻灯片 73）。

Evidence is mixed. ASIC’s Enhanced Regulatory Sandbox authorised 11 firms by 2023—far below expectations—suggesting that tight entry criteria and short test periods still deter innovators. Conversely, the UK’s FCA sandbox has seen several graduates scale successfully, implying learning benefits.
证据是混合的。ASIC 的增强型监管沙盒到 2023 年授权了 11 家公司——远低于预期——表明严格的入门标准和短期测试期仍然阻碍了创新者。相反，英国的 FCA 沙盒已经看到几家成功毕业的公司，暗示了学习的好处。

Riskwise, customer harm has been minimal due to strict exposure limits. Yet Posner & Weyl’s (2013) CBA frame warns that opportunity costs—foregone consumer benefits from delayed nationwide rollouts—are seldom quantified. Royal Commission findings stress that lax oversight breeds misconduct, cautioning against “innovation first, regulation later” rhetoric.
由于严格的暴露限制，Riskwise 的客户损害一直很小。然而，Posner & Weyl（2013）的契约成本分析框架警告说，机会成本——因全国范围推广延迟而放弃的消费者利益——很少被量化。皇家委员会的调查结果强调，监管松懈会导致不当行为，告诫人们不要采用“先创新后监管”的论调。

Didenko proposes an “enhanced sandbox+” model: phased caps, mandatory datasharing, and an optout refund right for consumers, aligning with Coates’s call for outcomefocused regulatory management. Such design maintains experimentation while embedding feedback loops that trigger full licensing hurdles once scale risks emerge.
Didenko 提出了一个“增强型沙盒+”模式：分阶段限制、强制数据共享，以及为消费者提供选择退出退款的权利，这与 Coates 对结果导向型监管管理的呼吁相一致。这种设计在保持实验的同时，嵌入了反馈循环，一旦规模风险出现，就会触发完整的许可障碍。

Hence, sandboxes can balance innovation and oversight if coupled with graduated safeguards, transparent metrics, and a clear path to standard regulation. Absent these, they risk being either cosmetic or dangerously permissive.
因此，如果结合渐进式安全措施、透明指标和清晰的标准监管路径，沙盒可以平衡创新和监管。如果没有这些，它们就有风险变成表面文章或危险地放任自流。

理解用户意图

用户回答“yes”，我需要猜测他们的意思。上一个对话结束时，我已提供了六个论文答案，并提出了五个附加问题。用户可能是在表示同意，或许他们希望我继续回答这些附加问题。有可能他们同意按IRAC模式（400-500字）撰写，引用参考材料来回答这些问题，或者他们在继续确认最初的要求。所以我可能需要继续提供相关内容。

已思考 12 秒

Below are IRACstyle essays (≈ 420–480 words each) that answer the five followup questions I proposed. Each draws on the same reference set used earlier—PFR Chapters 4 & 24, Posner & Weyl, Coates, Jacobs & King, the Australian CBA Guidance Note, the Hayne Royal Commission, APRA/ASIC materials, and Didenko—without formal AGLC footnotes.
以下是 IRAC 风格的论文（每篇约 420-480 字），用于回答我提出的五个后续问题。每篇论文都使用了之前相同的参考资料集——PFR 第四章和第 24 章、Posner & Weyl、Coates、Jacobs & King、澳大利亚 CBA 指导文件、Hayne 皇家委员会、APRA/ASIC 材料和 Didenko——而不使用正式的 AGLC 脚注。

7 │ Should costbenefit analysis (CBA) be subject to an independent audit before major financial rules take effect?
7 │ 在主要金融规则生效前，成本效益分析（CBA）是否应接受独立审计？

Issue
问题
Agencies now conduct CBA internally, but critics claim inhouse assessments are prone to optimism bias, political pressure, or industry capture. Would mandatory exante audit—by an office akin to the U.S. Congressional Budget Office—enhance rule quality?
机构现在进行内部成本效益分析，但批评者声称内部评估容易受到乐观偏见、政治压力或行业俘获的影响。强制事前审计——由类似于美国国会预算办公室的机构进行——能否提高规则质量？

Rule
规则
Posner & Weyl argue that CBA adds rigour only if “epistemic discipline” is enforced. Coates agrees, warning that selective data and subjective modelling can skew results. Australia’s 2023 Guidance Note already requires RIS clearance by the Office of Impact Analysis, but that clearance assesses process, not substance.
波斯纳和魏尔认为，只有实施“认知纪律”时，成本效益分析才会增加严谨性。库茨同意这一观点，警告说选择性和主观的建模可能会扭曲结果。澳大利亚 2023 年的指导文件已经要求由影响分析办公室进行 RIS 批准，但该批准仅评估流程，而非实质内容。

Application
应用
Large prudential rules, such as APRA’s “unquestionably strong” capital rewrite, impose multibilliondollar compliance costs and uncertain stability benefits. An independent audit could test model assumptions (e.g., lossgivendefault parameters) and discount rates, revealing sensitivities that internal sponsors might understate. The Hayne Royal Commission showed how failures in earlier rulemaking—e.g., lax mortgageresponsiblelending thresholds—stemmed partly from overreliance on industry data and insufficient external challenge.
大型审慎规则，如 APRA 的“无疑非常强”的资本重写，会带来数十亿美元的合规成本和不确定的稳定收益。独立审计可以测试模型假设（例如，损失给定默认参数）和贴现率，揭示内部发起人可能低估的敏感性。海恩皇家委员会展示了早期规则制定中的失败——例如，宽松的抵押贷款责任贷款门槛——部分原因是过度依赖行业数据和外部挑战不足。

On the downside, Coates notes audit can ossify rulemaking if reviewers conflate methodology with policy discretion. Financial regulation often targets Knightian uncertainty (tail risk), which resists monetisation; external auditors might default to “quantify or block” tendencies, delaying urgently needed reforms. Furthermore, Jacobs & King document how politicised oversight of the U.S. Fed undermined timely crisis response in 2008.
不利方面，库茨指出如果审查者将方法论与政策自由裁量权混淆，审计可能会使规则制定僵化。金融监管通常针对奈特不确定性（尾部风险），这种风险难以货币化；外部审计师可能会默认采取“量化或阻止”的倾向，从而延误急需的改革。此外，雅各布斯和金记录了美国联邦储备委员会的政治化监督如何在 2008 年削弱了及时应对危机的能力。

Conclusion
结论
Independent audit promises transparency and consistency, but should focus on methodological soundness, not policy judgement. A tiered threshold—audits for rules with costs over, say, AU$250 million—could balance scrutiny and agility. Embedding qualified immunity for regulators who act on audited CBAs would preserve accountability without paralysing response capability.
独立审计承诺透明度和一致性，但应侧重于方法论的正确性，而不是政策判断。可以设立分层的门槛——例如，对成本超过 2.5 亿澳元的规则的审计——以平衡审查和灵活性。为在经过审计的中央银行协议上采取行动的监管者嵌入豁免责任，可以在不瘫痪响应能力的情况下保持问责制。

8 │ Do countercyclical capital buffers (CCyBs) inadvertently disadvantage small banks?
8 │ 反周期资本缓冲（CCyBs）是否会无意中使小型银行处于不利地位？

Issue
问题
CCyBs raise capital demands when systemwide credit growth exceeds sustainable levels. Critics argue that because small banks lack cheap equity access, CCyBs disproportionately raise their funding costs, curbing competition.
当系统整体信贷增长超过可持续水平时，循环信贷银行会提出资本需求。批评者认为，由于小银行缺乏廉价股权融资渠道，循环信贷银行的资金成本不均匀地提高，从而抑制了竞争。

Rule
规则
Basel III applies CCyBs uniformly by jurisdiction, but national supervisors may exempt or adjust thresholds. Posner & Weyl urge costefficiency: a rule that generates benefits at lower social cost is preferable. PFR Ch 24 frames competition as a secondary goal behind stability.
巴塞尔协议 III 对循环信贷银行按地区统一适用，但国家监管机构可以豁免或调整阈值。波斯特和韦尔呼吁成本效益：一个能以较低社会成本产生效益的规则更可取。PFR 第 24 章将竞争视为稳定性目标之后的次要目标。

Application
申请
APRA’s 2023 stresstest showed major banks meeting a 1 per cent CCyB through retained earnings, whereas regionals needed fresh capital. Funding cost differentials widened by ~25 basis points, shrinking smaller banks’ mortgage share. Yet systemic crises emanate largely from large institutions (Jacobs & King): the Royal Commission found misconduct—and consequent losses—clustered in majors. A blunt CCyB may thus impose higher marginal social value on big banks than on small ones.
APRA 的 2023 年压力测试显示，主要银行通过留存收益满足了 1%的 CCyB，而地区性银行则需要新的资本。资金成本差异扩大了约 25 个基点，缩小了小型银行的按揭份额。然而，系统性危机主要源于大型机构（Jacobs & King）：皇家委员会发现不当行为及其导致的损失主要集中在大型银行。因此，简单的 CCyB 可能会对大型银行比小型银行施加更高的边际社会价值。

Solutions exist. Some jurisdictions scale buffers by systemic importance, mirroring GSIB surcharges. Alternatively, APRA could offer Pillar 2 relief for community banks whose portfolios exhibit low correlation with macro variables. Coates’s “better CBA” essay endorses such tailoring: equal objective, different instruments where cost curves diverge.
存在解决方案。一些地区根据系统性重要性调整缓冲，类似于系统重要性银行附加费。或者，APRA 可以为资产组合与宏观变量低相关的社区银行提供第二支柱救济。Coates 的“更好的 CBA”文章支持这种定制：在成本曲线分化时，采用相同的客观标准，但使用不同的工具。

Conclusion
结论
Uniform CCyBs risk entrenching incumbents. Calibrating buffer rates to systemic footprint or offering offsetting liquidity waivers can uphold stability without sacrificing contestability, satisfying both efficiency and fairness criteria.
统一 CCyBs 风险可能使现有企业巩固其地位。调整缓冲率以适应系统足迹或提供抵消流动性豁免可以在不牺牲竞争性的情况下维护稳定性，满足效率和公平标准。

9 │ Is the twinpeaks model superior to a single superregulator in reducing capture?
9 │ 双峰模型是否比单一超级监管机构更优越于减少俘获？

Issue
问题
Australia’s “twin peaks” splits prudential (APRA) from conduct (ASIC) tasks. Does this structural choice mitigate regulatory capture more effectively than housing all functions in one megaagency?
澳大利亚的“双峰”将审慎监管（APRA）与行为监管（ASIC）任务分开。这种结构选择是否比将所有职能集中在一个巨型机构中更有效地缓解监管俘获？

Rule
规则
Ogus’s theory: functional separation reduces internal goal conflicts but multiplies coordination costs. Jacobs & King observe that fragmented U.S. oversight diffuses capture yet fosters turf battles. The Hayne Commission criticised ASIC’s enforcement laxity despite structural separation, implying architecture alone is insufficient.
奥古斯的理论：功能分离减少内部目标冲突，但增加了协调成本。雅各布斯和金观察到，美国分散的监管导致监管俘获风险降低，但加剧了地盘争夺。海恩委员会批评 ASIC 执法松懈，尽管存在结构分离，暗示仅靠架构是不够的。

Application
应用
Twin peaks allows targeted expertise: APRA focuses on solvency, ASIC on disclosure. Capture risks differ: APRA faces close, recurrent interactions with a handful of banks—high relational capture potential—whereas ASIC confronts diffuse market actors. A superregulator, like the pre2013 UK FSA, concentrated power but also bundled conflicting incentives; postcrisis reviews blamed that fusion for oversight failures.
双峰模式允许有针对性的专业知识：APRA 专注于偿付能力，ASIC 专注于信息披露。俘获风险不同：APRA 与少数银行有密切、反复的互动——高关系俘获潜力——而 ASIC 则面对分散的市场参与者。超级监管机构，如 2013 年前的英国 FSA，集中了权力，但也捆绑了冲突的激励；后危机审查将这种融合归咎于监管失败。

Coordination challenges manifest in regulatory gaps: scandal involving “fees for no service” slipped between ASIC’s conduct mandate and APRA’s prudential lens. The CFR’s quarterly meetings aim to bridge divides but rely on soft law. Superregulators may solve silo issues but introduce “megacapture” risks—industry lobbies need influence only one agency.
协调挑战体现在监管空白中：“无服务收费”丑闻滑落在了 ASIC 的行为职责和 APRA 的审慎视角之间。CFR 的季度会议旨在弥合分歧，但依赖于软法。超级监管机构可能解决部门问题，但引入“超级俘获”风险——行业游说只需影响一个机构。

Conclusion
结论
Twin peaks is no panacea, yet offers partial insulation by diversifying capture targets. Its effectiveness depends on robust interagency protocols and expost parliamentary oversight. Relative to a superregulator, twin peaks remains preferable where political economy risks are acute, provided coordination mechanisms are strengthened.
双峰模式并非万能药，但通过多样化俘获目标提供部分保护。其有效性取决于健全的跨机构协议和议会事后监督。相对于超级监管机构，在政治经济风险尖锐的情况下，双峰模式仍然更可取，前提是协调机制得到加强。

10 │ How should regulators reconcile climaterisk mandates with traditional safetyandsoundness goals?
10 │ 监管机构应如何协调气候风险职责与传统的安全与健康目标？

Issue
问题
APRA’s CPS 190 draft requires climatescenario planning, while ASIC pushes for TCFDaligned disclosures. Critics fear diverting attention from core solvency oversight.
APRA 的 CPS 190 草案要求进行气候情景规划，而 ASIC 推动与气候相关财务信息披露框架（TCFD）一致的披露。批评者担心这会分散对核心偿付能力监管的注意力。

Rule
规则
Systemicrisk logic: if climate change threatens asset values, ignoring it violates safetyandsoundness. Posner & Weyl’s CBA paradigm supports action where expected crisisloss avoidance outweighs compliance costs. Royal Commission highlighted reputational risks as prudential concerns.
系统风险逻辑：如果气候变化威胁资产价值，忽视它就违反了安全稳健原则。波斯纳和魏尔的成本效益分析（CBA）范式支持在预期危机损失避免超过合规成本时采取行动。皇家委员会强调了声誉风险作为审慎关注的问题。

Application
申请
Transition risk—stranded fossil assets—could impair loan books; physical risk—cyclone damage—raises insurers’ claims volatility. Jacobs & King emphasise pathdependent mandates: central banks evolved to include employment; climate risk is analogous extension. Early European stress tests estimate a 2 °C orderly transition reduces longrun GDP loss relative to disorderly paths, endorsing proactive prudential nudges.
过渡风险—搁浅的化石资产—可能损害贷款组合；物理风险—飓风损害—提高保险公司的索赔波动性。Jacobs & King 强调路径依赖的授权：央行演变为包括就业；气候风险是类似扩展。早期的欧洲压力测试估计，2℃有序过渡相对于无序路径减少了长期 GDP 损失，支持主动的审慎引导。

Yet adding mandates dilutes focus. Coates notes rule inflation strains limited analytical banådwidth. The solution is integration: use existing capital frameworks (e.g., sectoral risk weights) rather than standalone “green ratios.” APRA’s scenario guidelines piggyback on ICAAP processes, aligning with Ogus’s efficiency principle.
然而，增加授权会稀释焦点。Coates 指出规则通胀 strains 有限的分析带宽。解决方案是整合：使用现有的资本框架（例如，部门风险权重），而不是独立的“绿色比率”。APRA 的情景指南搭便车 ICAAP 流程，与 Ogus 的效率原则一致。

Conclusion
结论
Climate oversight is a logical extension of prudential duty, not a competing objective. Embedding it within existing risk frameworks maintains solvency focus while acknowledging emerging threats—satisfying both tradition and adaptation.
气候监管是审慎责任的合理延伸，而非竞争目标。将其纳入现有风险框架既保持了偿付能力重点，又承认了新兴威胁——既满足了传统，又适应了变化。

11 │ Would mandatory opensource disclosure of AI creditscoring algorithms promote fairness without stifling innovation?
11 │ 强制开源披露人工智能信用评分算法会促进公平性，而不会扼杀创新吗？

Issue
问题
Civilsociety advocates seek algorithmic transparency; fintechs warn that full disclosure erodes intellectual property and facilitates gaming.
民间社会倡导者寻求算法透明度；金融科技企业警告说，完全披露会侵蚀知识产权并促进操纵。

Rule
规则
Regulatory goals: consumer fairness (avoid bias) and market efficiency (preserve innovation). Posner & Weyl’s cost–benefit lens applies; Coates advises outcomefocused, technologyneutral standards.
监管目标：消费者公平（避免偏见）和市场效率（保持创新）。波斯纳和威尔的成本效益视角适用；库茨建议结果导向、技术中立的标准。

Application
应用
Open source aids auditability; academics can test for disparate impacts. However, disclosure may allow borrowers to manipulate inputs, degrading predictive power and raising default rates—harmful to safety and soundness. Didenko’s sandbox model offers compromise: confidential regulator access plus public summary metrics. ASIC’s sandbox requires firms to share model documentation with supervisors, safeguarding IP while enabling scrutiny.
开源有助于可审计性；学者可以测试是否存在歧视性影响。然而，信息披露可能允许借款人操纵输入，降低预测能力并提高违约率——对安全性和稳健性有害。迪登科的环境沙盒模式提供了折中方案：监管机构保密访问加上公开摘要指标。ASIC 的环境沙盒要求公司向监管机构共享模型文档，在保护知识产权的同时，也便于审查。

Empirical analogy: FICO score methodology is proprietary yet subject to regulatory validation; bias persists but within monitored tolerance bands. Requiring “explainability reports” rather than source code aligns with twin objectives, echoing EU AI Act’s risktiered approach.
经验类比：FICO 评分方法是专有的，但需接受监管验证；偏见仍然存在，但在监控的容忍范围内。要求“可解释性报告”而不是源代码，符合双重目标，呼应欧盟 AI 法案的风险分层方法。

Conclusion
结论
Full opensource mandates overshoot; graded transparency—regulator access and public performance dashboards—better balances fairness with innovation incentives. Sandboxes can pilot these regimes, gathering evidence for broader rollout.
完全开源的要求过高；分级的透明度——监管机构访问和公共性能仪表板——更好地平衡了公平与创新激励。沙盒可以试点这些制度，收集证据以推广到更广泛的范围。

第三部分：

1 │ Limitations of Disclosure and its Persistent Appeal (≈ 460 words)
1 │ 披露的局限性及其持续的魅力（约 460 字）

Limits. Decades of empirical work show that mandated disclosure rarely delivers informed choice in retail finance. Fewer than one in twothousand online shoppers even open standardform terms (Bakos et al 2014), while laboratory tests find that only threetofour per cent of adults can parse prospectus language (BenShahar & Schneider 2010). Behavioural economics explains why: information overload and hyperbolic discounting push consumers to anchor on salient cues—headline rates or teaser gifts—rather than APR or break costs (FCA 2013, §2). PFR Ch 10 confirms that sophisticated sellers “play to biases”, designing complexity that blunts comparison. Even “simplified” fact sheets under UCITS or Australia’s ShortForm PDS hardly move decision quality; Beshears et al (2011) report statistically indistinguishable fund choices relative to full prospectuses.
限制。数十年的实证研究表明，强制信息披露很少能在零售金融中提供明智的选择。不到两千分之一的两千名在线购物者甚至会打开标准格式条款（Bakos 等人，2014 年），而实验室测试发现只有三到四分之一的成年人能够解析招股说明书语言（BenShahar & Schneider，2010 年）。行为经济学解释了原因：信息过载和过度夸张的折扣使消费者倾向于关注显著线索——标题利率或诱人礼物——而不是年利率或提前终止成本（FCA，2013 年，§2）。PFR 第 10 章证实，复杂的销售商“利用偏见”，设计复杂性以削弱比较。即使是 UCITS 或澳大利亚的简式 PDS 下的“简化”事实表也几乎不会提高决策质量；Beshears 等人（2011 年）报告相对于完整招股说明书，基金选择在统计上没有区别。

Why regulators still love it. First, disclosure respects personal autonomy and the political rhetoric of “buyer responsibility,” avoiding paternalistic bans. Second, it imposes low direct compliance cost on providers compared with product intervention. Third, it shifts liability: once required facts are delivered, misconduct claims face higher hurdles—an attraction for industry lobbyists (Hayne 2019, pp 6168). Finally, disclosure is administratively light; agencies can issue template rules without continuous product monitoring—important where supervisory budgets lag innovation.
为什么监管者仍然喜欢它。首先，信息披露尊重个人自主权，并符合“买方责任”的政治辞令，避免了家长式禁止。其次，与产品干预相比，它对提供者施加了较低的直接合规成本。第三，它转移了责任：一旦提供了所需事实，不当行为指控将面临更高的障碍——这对行业游说者有吸引力（Hayne 2019，第 6168 页）。最后，信息披露行政负担较轻；机构可以发布模板规则，而无需持续监控产品——在监管预算滞后于创新的情况下非常重要。

Evolution not abolition. Recognising failure, regulators are upgrading to “Disclosure 2.0”: contextspecific nudges (FCA 2013) such as text alerts that cut UK overdraft fees by 25 per cent, and “smarter communications” trials at ASIC (Kell 2016). Yet PFR §12.4 warns that these success stories are niche; they hinge on iterative experimentation capacity that many regulators lack.
演进而非废除。认识到失败，监管机构正在升级到“信息披露 2.0”：针对特定情况的微调（FCA 2013），例如减少英国透支费用 25%的文字提醒，以及 ASIC（Kell 2016）进行的“更智能沟通”试验。然而，PFR §12.4 警告称，这些成功案例是利基的；它们依赖于迭代实验能力，而许多监管机构缺乏这种能力。

Conclusion. Disclosure alone is a blunt instrument—politically seductive but economically fragile. It endures because it is cheap, nonintrusive, and liabilityshifting, but its future lies in behaviourally informed, empirically tested variants allied to stronger conduct and productgovernance regimes.
结论。单独披露是一种粗糙的工具——政治上诱人但在经济上脆弱。它之所以能够持续存在，是因为它成本低、非侵入性强、能够转移责任，但其未来在于与更强的行为指导、经过实证检验的变种以及更严格的 conduct 和 productgovernance 体制相结合。

2 │ Who Counts as a “Consumer” and Are We Protecting the Right People? (≈ 450 words)
2 │ 谁被视为“消费者”？我们是否在保护正确的人群？（约 450 字）

Legal taxonomy. Financial law segments users by presumed sophistication. In Australia, retail clients are the default; wholesale clients escape most disclosure and conduct duties if they meet any s 761G Corporations Act thresholds—e.g., assets > A$2.5 million or income > A$250 000. The National Credit Code applies only where credit is “predominantly for personal, domestic or household purposes”—excluding many smallbusiness loans. EU MiFID II distinguishes retail, professional and eligible counterparty clients; misclassification triggers strict liability.
法律分类。金融法根据假设的复杂程度对用户进行细分。在澳大利亚，零售客户是默认分类；如果批发客户符合任何 s 761G 公司法的门槛——例如，资产超过 A$2.5 百万或收入超过 A$250 000，则他们可以免除大部分披露和 conduct 义务。国家信贷法仅适用于“主要为个人、家庭或住宅目的提供信贷”的情况——排除了许多小企业贷款。欧盟 MiFID II 将客户分为零售、专业和符合条件的对手方客户；错误分类将触发严格责任。

Context drift. Technological change blurs lines. Sophisticated day traders may hold personal accounts yet deploy strategies rivaling small hedge funds, while gigeconomy workers borrow as “consumers” to fund quasibusiness expenses. Structuredproduct scandals show highnetworth individuals succumbing to complexity similar to mainstream retail (FCA 2015). Conversely, financially literate retirees with subthreshold balances are trapped on the retail side, bearing compliance costs for protections they neither need nor value.
上下文漂移。技术变革模糊了界限。复杂的日内交易者可能持有个人账户，却部署着与小对冲基金相媲美的策略，而零工经济工作者作为“消费者”借款以资助准商业支出。结构化产品丑闻表明，高净值个人在复杂性上与主流零售相似（FCA 2015）。相反，具有金融素养的退休人员由于余额低于阈值而被困在零售端，承担着他们既不需要也不重视的保护合规成本。

Targeting misfires. Bartlett et al (2019) reveal algorithmic discrimination in fintech lending: minority borrowers pay higher rates despite standardised scoring, an abuse outside traditional “vulnerable consumer” flags. PFR Ch 11.3.3 notes that openbanking data can flip sophistication heuristics—rich digital footprints yield better credit terms, sidelining lowincome but techsavvy borrowers still labelled retail.
目标定位失误。Bartlett 等人（2019）揭示了金融科技借贷中的算法歧视：少数族裔借款人尽管评分标准化，但仍支付更高的利率，这是一种超出传统“弱势消费者”标志的滥用。PFR 第 11.3.3 节指出，开放银行数据可以翻转复杂度启发式算法——丰富的数字足迹可以获得更好的信贷条件，使低收入但精通技术的借款人被边缘化，他们仍然被标记为零售。

Policy calibration. Hayne (2019, pp 7480) urges a move from statusbased to outcomebased tests: any party suffering bargaining power or information asymmetry merits protection. ASIC’s Design & Distribution Obligations (DDO) operationalise this by forcing issuers to draft a Target Market Determination describing the actual user segment, not merely a statutory category.
政策校准。Hayne（2019，第 7480 页）敦促从基于身份的测试转向基于结果的测试：任何处于议价能力或信息不对称地位的当事人都有权获得保护。ASIC 的设计与分销义务（DDO）通过强制发行人起草目标市场确定文件，描述实际用户群体，而不仅仅是法定类别，来实现这一点。

Conclusion. The retailwholesale dichotomy is increasingly blunt. Effective consumer protection should hinge on transactional vulnerability rather than wealth or ticket size, employing data analytics to refine realtime classification and align regulatory shields with those who truly need them.
结论。零售与批发二元结构日益明显。有效的消费者保护应该基于交易脆弱性，而不是财富或交易规模，利用数据分析来细化实时分类，并将监管保护与真正需要的人相匹配。

3 │ Defective Financial Products and Manufacturer Liability (≈ 475 words)
3 │ 有缺陷的金融产品与制造商责任（约 475 字）

Issue. Should a bank or fintech that designs a highrisk product be liable in negligence for consumer losses, akin to a manufacturer of defective goods in Donoghue v Stevenson?
争议。设计高风险产品的银行或金融科技公司是否应因消费者损失而承担疏忽责任，类似于 Donoghue v Stevenson 案件中的缺陷产品制造商？

Rule. Donoghue established a duty of care where a manufacturer could foresee injury from a latent defect; modern courts adapt this to financial contexts. In Investors Compensation Scheme v West Bromwich BS Lord Hoffmann treated complex bond misselling as a question of contractual interpretation but hinted that negligence principles underpin redress where duty is breached.
规则。Donoghue 确立了一种注意义务，即制造商可以预见因潜在缺陷造成的伤害；现代法院将此适应于金融环境。在 Investors Compensation Scheme 诉 West Bromwich BS 案中，Lord Hoffmann 将复杂的债券销售问题视为合同解释问题，但暗示在义务被违反时，补救措施基于过失原则。

Application. Financial products differ from beer bottles: harm is economic, not physical; risk disclosures complicate proximity; and investors often consent to volatility. Yet ASIC’s Product Intervention Power (RG 272) presupposes that certain designs—e.g., 990 per cent payday loans—are intrinsically unsafe, mirroring productliability logic. The FCA’s cap on structured deposits followed thematic evidence (FCA 2015) that headlinerate teasers hid negative expected returns for typical investors. Behavioural evidence (PFR Ch 10) shows firms exploiting hyperbolic discounting, creating foreseeable detriment.
应用。金融产品不同于啤酒瓶：损害是经济的，不是身体的；风险披露使邻近性复杂化；并且投资者通常同意波动性。然而 ASIC 的产品干预权力（RG 272）假定某些设计——例如，每 990 个百分点的 payday 贷款——本质上是不安全的，反映了产品责任逻辑。FCA 对结构化存款的限额遵循了主题证据（FCA 2015），即 headline rate 诱饵对典型投资者隐藏了负面预期回报。行为证据（PFR 第 10 章）表明公司利用了夸张折扣，造成了可预见的损害。

Courts may still baulk at imposing a blanket duty: investors possess some autonomy, and risk appetite varies. But where design embeds asymmetric payoff—e.g., autocallable notes almost certain to underperform cash—foreseeability and vulnerability converge. Hayne’s “fit for purpose” norm (2019) resonates with the Donoghue neighbour principle: issuers owe clients reasonable care in design when harm is predictable and consumers cannot inspect ex ante.
法院可能仍然会反对施加一项普遍的义务：投资者拥有一定的自主权，且风险偏好各不相同。但当设计嵌入非对称的收益——例如，自调用票据几乎肯定会表现不如现金——可预见性和脆弱性会趋同。海恩的“适合目的”原则（2019 年）与唐纳休的“邻居原则”相呼应：当损害可预见且消费者无法事先检查时，发行人应对客户在设计上承担合理的注意义务。

Conclusion. Translating Donoghue into finance is feasible for products whose structural features virtually guarantee consumer loss absent sophisticated hedging. Statutorily recognising such duty, as Australia does through Design & Distribution obligations, aligns legal remedies with modern behavioural realities while preserving innovation for genuinely beneficial instruments.
结论。将唐纳休原则应用于金融领域对那些结构特征几乎保证消费者在缺乏复杂对冲的情况下会遭受损失的产品来说是可行的。通过设计及分销义务，如澳大利亚所做的那样，在法律上承认这种义务，可以使法律救济与现代行为现实相一致，同时保留对真正有益的仪器的创新空间。

4 │ When Should Regulators Ban Consumer Financial Products? (≈ 440 words)
4 │ 何时监管机构应禁止消费者金融产品？（约 440 字）

Framework. Bans are the nuclear option in a “regulatory pyramid.” ASIC may issue stop orders under Corporations Act s 1023D(1) when a product is “likely to result in significant detriment.” The National Credit Act s 301D(1) mirrors this for credit contracts. MiFID II Art 69(2)(t) empowers EU supervisors where design or distribution poses “a serious threat to investor protection or market integrity.”
框架。禁令是“监管金字塔”中的核选项。ASIC 在《公司法》第 1023D(1)条下，当产品“可能造成重大损害”时，可以发出停止令。国家信用法第 301D(1)条对信用合同有同样的规定。MiFID II 第 69(2)(t)条规定，当设计或分销对投资者保护或市场完整性构成“严重威胁”时，欧盟监管机构被授权采取行动。

CostBenefit Lens. Posner & Weyl urge that prohibition pass a strict CBA: expected consumer harm (probability × severity) must exceed lost utility from foregone access. ASIC’s 2020 paydaycredit intervention illustrates: data showed average feetoloan ratios above 400 per cent, with default cascades inducing hardship (RG 272 case study). Alternative credit pathways—NILS schemes, bank overdrafts—moderated access losses.
成本效益视角。波斯纳和韦尔主张禁令必须通过严格的成本效益分析：预期的消费者损害（概率×严重程度）必须超过因放弃访问而失去的效用。ASIC 2020 年的小额信贷干预措施说明了这一点：数据显示平均英尺贷款比率超过 400%，而违约连锁反应导致困难（RG 272 案例研究）。替代信贷途径——无息储蓄计划、银行透支——减轻了访问损失。

Behavioural Triggers. Where biases are systematically exploited and disclosure ineffective, bans gain force. The FCA’s 2015 cap on highcost shortterm credit followed evidence that consumers underestimated rollover probabilities due to present bias (FCA 2013). PFR §12.3 notes that “sludge” design—hidden fees, complexity—persist even after warning labels, justifying product withdrawal.
行为触发因素。当偏见被系统性地利用且披露无效时，禁令将生效。FCA 在 2015 年对高成本短期信贷设置的限额是基于消费者由于现时偏见低估了滚动概率的证据（FCA 2013）。PFR §12.3 指出，“污泥”设计——隐藏费用、复杂性——即使在警告标签后仍然存在，这为产品撤回提供了理由。

Market Completeness. If safer substitutes exist, the social cost of banning falls. Structured deposits replicated by lowfee index funds make prohibition of toxic variants less traumatic. Conversely, banning reverse mortgages without expanding seniorrent schemes could backfire.
市场完整性。如果存在更安全的替代品，禁止的社会成本就会降低。低费率的指数基金可以复制结构化存款，这使得禁止有毒变种不那么痛苦。相反，如果不扩大老年人租金计划，禁止反向抵押贷款可能会适得其反。

Adaptive Approach. RG 272 allows conditional orders (e.g., leverage caps, saleschannel limits) as milder interventions. Dynamic sunsets ensure learning: ASIC’s CFD leverage limits expire after 18 months unless renewed, embedding feedback loops.
适应性方法。RG 272 允许有条件的订单（例如，杠杆限额、销售渠道限制），作为较温和的干预措施。动态日落机制确保学习：ASIC 的 CFD 杠杆限额在 18 个月后除非续期将到期，嵌入反馈循环。

Conclusion. Regulators should ban products when (1) robust evidence shows large, systematic consumer detriment; (2) disclosure, advice, or governance remedies have failed; and (3) satisfactory substitutes minimise welfare loss. Conditional or timelimited bans harness flexibility while safeguarding consumers.
结论。监管机构应在以下情况下禁止产品：（1）有充分证据表明存在大规模、系统性的消费者损害；（2）信息披露、建议或治理补救措施失败；以及（3）令人满意的替代品最小化福利损失。有条件的或有限时间的禁止措施在保护消费者权益的同时，提供了灵活性。

5 │ Conduct of Business Rules and Their Consumer Benefits (≈ 430 words)
5 │ 商业行为规则及其消费者利益（约 430 字）

Definition. Conduct of business (COB) rules govern how firms interact with clients: suitability, bestinterest duty, conflict management, disclosure clarity. Unlike prudential norms that target firm solvency, COB targets fairness and competence in service delivery (PFR Ch 11).
定义。商业行为（COB）规则规范公司如何与客户互动：适当性、最佳利益义务、冲突管理、信息披露清晰度。与针对公司偿付能力的审慎规范不同，COB 旨在服务提供中的公平性和能力（PFR 第 11 章）。

Benefits. First, they translate behavioural insights into enforceable duties. Australia’s s 961B “best interests” test requires advisers to prioritise client objectives, countering incentives to peddle highcommission products. Studies after the UK’s Retail Distribution Review show fall in biased fund flows once commission bans took effect (FCA 2015).
利益。首先，它们将行为洞察转化为可执行的义务。澳大利亚第 961B 条“最佳利益”测试要求顾问优先考虑客户目标，反对此类推销高佣金产品的激励。英国零售分销审查后的研究表明，一旦佣金禁止措施生效，有偏见的资金流向就出现了下降（FCA 2015）。

Second, COB reduces search and monitoring costs. Appropriateness tests under MiFID II Art 25 oblige distributors to assess client knowledge before selling complex instruments, shielding novices from unsuitable risk.
其次，COB 降低了搜索和监控成本。根据 MiFID II 第 25 条的规定，适当性测试要求分销商在销售复杂金融工具前评估客户的知识水平，从而保护新手免受不适当的投资风险。

Third, COB establishes expost accountability: breach enables compensation without proving product defect, easing evidentiary burden (Hayne 2019).
第三，COB 建立了事后问责制：一旦违反规定，即可进行赔偿而不必证明产品缺陷，从而减轻了证据负担（Hayne 2019）。

Limitations. Enforcement gaps persist—ASIC’s 2019 “fees for no service” scandal revealed systemic flouting despite longstanding duties. Moreover, high fixedcost compliance can raise advice prices; Australian advice fees doubled between 2013 and 2023, excluding many massmarket consumers (ASIC data).
局限性。执行中的漏洞仍然存在——ASIC 在 2019 年的“无服务费”丑闻揭示了长期存在的职责被系统性地忽视。此外，高昂的合规固定成本可能会提高建议费用；澳大利亚的建议费用在 2013 年至 2023 年间翻了一番，这使得许多大众市场消费者被排除在外（ASIC 数据）。

Conclusion. COB rules meaningfully protect consumers but work best when paired with accessible advice channels and vigorous supervision to deter boxticking.
结论。COB 规则在保护消费者方面具有重要意义，但只有在与可访问的建议渠道和严格的监管相结合时，才能最有效地运作，以防止“框定行为”。

6 │ Product Governance Obligations and Required Procedures (≈ 450 words)
6 │ 产品治理义务和必要程序（≈ 450 字）

Concept. Product governance (PG) shifts onus onto firms to embed consumer welfare into the designtosale pipeline. MiFID II Art 16(3) and Australia’s Corporations Act ss 994BD require manufacturers to identify a target market, align distribution channels, and monitor outcomes.
概念。产品治理（PG）将责任转移给公司，将消费者福利嵌入到销售设计管道中。MiFID II 第 16(3)条和澳大利亚公司法第 994BD 条要求制造商确定目标市场、调整分销渠道并监控结果。

Key Procedures.
关键程序。

TargetMarket Determination (TMD). Firms must articulate consumer attributes—risk tolerance, needs—that the product serves, and publish the TMD before launch. ASIC expects rational linkage between features and needs (RG 274).
目标市场确定（TMD）。公司必须阐述产品所服务的消费者属性——风险承受能力、需求，并在推出前公布 TMD。ASIC 期望功能和需求之间存在合理的联系（RG 274）。

Risk Assessment. Structured evaluation of liquidity, complexity, leverage. The FCA’s 2015 review found that banks with documented risk matrices weeded out 40 per cent of concept products prelaunch.
风险评估。对流动性、复杂性和杠杆的系统性评估。FCA 的 2015 年审查发现，有文档记录的风险矩阵在产品发布前筛选掉了 40%的概念产品。

Distribution Controls. Conditions such as adviser accreditation or digital suitability filters ensure sales match the TMD. Under MiFID II, massmailing complex CFDs to “executiononly” accounts violates PG.
分销控制。顾问资格认证或数字适用性过滤器等条件确保销售与 TMD 匹配。在 MiFID II 下，向“仅执行”账户大规模发送复杂 CFD 违反了 PG。

PostLaunch Review. Firms must collect complaints data, lapse rates, and performance drift. Article 16(3) demands periodic reassessment, triggering redesign or withdrawal if mismatches emerge.
发布后审查。公司必须收集投诉数据、失效率和表现漂移。第 16(3)条要求定期重新评估，如果出现不匹配则触发重新设计或撤回。

Governance Oversight. A committee reporting to the board should sign off TMDs and escalation protocols, satisfying Hayne’s call for accountable senior management.
治理监督。一个向董事会汇报的委员会应批准 TMD 和升级协议，满足 Hayne 对负责高级管理层的要求。

Benefits. PG leverages the firm’s superior information—designers know product quirks—to preempt harm, overcoming regulators’ lag. Early EU evidence shows decline in missold capitalprotected notes after PG rollout.
利益。PG 利用公司卓越的信息——设计师了解产品特性——以预防损害，克服监管者的滞后。早期的欧盟证据显示，在 PG 实施后，错误销售的资本保护票据数量下降。

Conclusion. Effective PG marries documented processes with datadriven feedback, turning “consumer centricity” from slogan into verifiable compliance.
结论。有效的 PG 将记录在案的过程与数据驱动的反馈相结合，将“以消费者为中心”从口号转变为可验证的合规性。

7 │ Should mandatory “sludge audits” be part of productgovernance reviews?
7 │ 强制性的“污泥审计”是否应成为产品治理审查的一部分？

Issue
问题
“Sludge” denotes design frictions—tiny hurdles, hidden steps, default traps—that discourage consumers from acting in their best interests (FCA 2013). Should regulators compel firms to audit and report sludge as part of their TargetMarket Determinations (TMDs) under productgovernance regimes?
“污泥”指的是设计摩擦——微小的障碍、隐藏的步骤、默认陷阱——这些会阻止消费者为其自身利益采取行动（FCA 2013）。监管机构是否应要求企业在产品治理制度下，在其目标市场确定（TMD）中审计和报告污泥？

Rule
规则
MiFID II Art 16(3) requires firms to ensure products “meet the needs” of the target market; ASIC RG 274 demands they avoid design features that “cause harm.” Behaviouraleconomics literature (BenShahar & Schneider 2010) shows sludge systematically distorts choices, undermining disclosure.
MiFID II 第 16(3)条规定企业必须确保产品“满足目标市场的需求”；ASIC RG 274 要求他们避免造成“伤害”的设计特征。行为经济学文献（BenShahar & Schneider 2010）表明，污泥系统地扭曲了选择，削弱了信息披露。

Application
应用
Case studies: UK overdraft alerts reduced consumer fees by 25 % once sludge (opaque fee calendars) was exposed (FCA 2013). In Australia, Hayne (2019, p 77) criticised superannuation switching forms designed to frustrate exits. Adding sludge audits would force firms to map the customer journey, quantify abandonment points, and justify any frictions. Compliance cost is modest—digital analytics already track clickpaths—while consumer benefit is high: a BETAPrime Minister’s Office pilot showed a 10ppt jump in completion of hardshipassistance requests after sludge removal.
案例研究：英国透支警报在揭露污泥（不透明的费用日历）后，使消费者费用降低了 25%（FCA 2013）。在澳大利亚，Hayne（2019，第 77 页）批评了旨在阻碍退出的超级年金转换表格。增加污泥审计将迫使公司绘制客户旅程，量化放弃点，并证明任何摩擦。合规成本适中——数字分析已经跟踪点击路径——而消费者受益很高：BETA 总理办公室试点显示，在移除污泥后，困难援助请求的完成率提高了 10 个百分点。

Opponents argue overprescription stifles UX experimentation; but audits can be principlebased, focusing on outcomes (time to cancel, steps to comparison). Posner & Weyl’s costbenefit test likely passes: audits flag cheap fixes before detriment becomes systemic, reducing need for blunter interventions under ASIC RG 272.
反对者争论过度处方抑制了用户体验实验；但审计可以是基于原则的，专注于结果（取消时间、比较步骤）。Posner & Weyl 的成本效益测试可能通过：审计在损害系统化之前标记低成本修复，减少在 ASIC RG 272 下对更粗糙干预的需求。

Conclusion
结论
Given strong evidence that sludge erodes product suitability and low marginal compliance cost, mandatory sludge audits within productgovernance reviews would sharpen consumer protection and preempt enforcement crises.
鉴于有强有力的证据表明污泥侵蚀了产品适用性，且边际合规成本较低，在产品治理审查中强制进行污泥审计将增强消费者保护并预防执法危机。

8 │ Do cryptoasset exchanges need bespoke conductofbusiness rules?
8 │ 加密资产交易所是否需要定制的商业行为规则？

Issue
问题
Recent collapses of global crypto exchanges exposed gaps in clientasset segregation, conflict management, and disclosure. Are existing MiFID II / Corporations Act COB duties sufficient, or is tailormade regulation required?
最近的全球加密货币交易所倒闭暴露了客户资产隔离、冲突管理和信息披露方面的漏洞。现有的 MiFID II / 公司法 COB 职责是否足够，还是需要定制化监管？

Rule
规则
COB regimes cover “financial products.” Crypto tokens often fall outside that perimeter—Australia relies on general consumer law; the EU awaits MiCA (Markets in CryptoAssets). PFR Ch 10 emphasises that COB presumes identifiable issuer, custody chain, and price discovery—elements missing in many decentralised venues.
COB 制度涵盖“金融产品”。加密代币通常不在这个范围内——澳大利亚依赖一般消费者法；欧盟等待 MiCA（加密资产市场）。PFR Ch 10 强调，COB 假定有可识别的发行人、托管链和价格发现——这些元素在许多去中心化平台中缺失。

Application
应用
Empirical data: Bartlett et al (2019) show fintech channels amplify discrimination—crypto adds pseudonymity, complicating redress. Without formal bestinterest duties, exchanges commingle client funds (FTX case), violating segregation norms standard under MiFID. Where tokens become derivatives (CFDs on Bitcoin), existing COB applies and ASIC’s leverage caps curbed losses; but spot trading remains unregulated.
实证数据：Bartlett 等（2019）表明金融科技渠道加剧了歧视——加密货币增加了匿名性，使补救措施复杂化。在没有正式的最佳利益职责的情况下，交易所混合客户资金（FTX 案例），违反了在 MiFID 下标准的隔离规范。当代币成为衍生品（比特币的差价合约）时，现有的 COB 适用，ASIC 的杠杆上限限制了损失；但现货交易仍然不受监管。

A bespoke regime could impose: (1) mandatory coldwallet segregation, mirroring MiFID assetprotection rules; (2) risk warnings keyed to volatility thresholds; (3) governance fitandproper tests. Costbenefit? Posner & Weyl advise considering innovation rents; yet after collapses, trust deficits already chilling investment—rules may increase market depth by restoring confidence.
定制化监管可能实施：(1) 强制冷钱包隔离，类似于 MiFID 资产保护规则；(2) 与波动性阈值相关的风险警告；(3) 治理适当性测试。成本效益？波斯纳和韦尔建议考虑创新租金；但在崩溃之后，信任赤字已经使投资变得寒冷——规则可能通过恢复信心增加市场深度。

Conclusion
结论
Generic COB duties cannot cover the idiosyncrasies of permissionless blockchains and global custody risk. A bespoke, yet proportionate, COB framework is justified to translate established investorprotection principles into the crypto context without smothering innovation.
通用 COB 责任不能涵盖许可链块和全球托管风险的特殊性。一个定制化且适度的 COB 框架是合理的，以便在不扼杀创新的情况下，将既定的投资者保护原则转化为加密货币环境。

9 │ Openbanking data rights: privacy versus alternative credit scoring
9 │ 开放银行数据权利：隐私与替代信用评分

Issue
问题
Openbanking mandates let consumers port transaction data to third parties, enabling novel credit models. How should regulators balance privacy with scoring benefits for thinfile borrowers?
开放银行强制令允许消费者将交易数据传输给第三方，从而实现新的信用模型。监管机构应如何平衡隐私与对薄文件借款人的评分利益？

Rule
规则
Australia’s Consumer Data Right (CDR) and the EU’s PSD2 require explicit consent and robust security. Behavioural studies (Kell 2016) warn that consent is often “clickthrough”—informationally hollow. Posner & Weyl stress net social welfare: gains to excluded borrowers vs. privacy loss.
澳大利亚的消费者数据权利（CDR）和欧盟的 PSD2 要求明确同意和强大的安全措施。行为研究（Kell 2016）警告说，同意通常是“点击通过”——信息上空洞。波斯纳和韦尔强调净社会福利：被排除的借款人的收益与隐私损失。

Application
应用
Fintechs using cashflow analytics cut APRs by 500 bp for subprime cohorts, PFR §12.4 notes. Yet Bartlett et al (2019) find algorithmic bias persists; additional data can entrench discrimination if proxies for ethnicity slip in. Privacy breaches loom: in 2022 an Australian fintech leak exposed 60 000 consumers. Solutions include tiered data scopes (readonly vs. write), OAuthbased limited tokens, and regulatory sandboxes to test fairness metrics before full rollout (Didenko 2021).
使用现金流分析的金融科技公司为次级贷款群体降低了 500 个基点的年利率，PFR §12.4 指出。然而，巴特莱特等人（2019）发现算法偏见仍然存在；如果种族的代理指标渗入，额外的数据可能会加剧歧视。隐私泄露迫在眉睫：2022 年，一家澳大利亚金融科技公司泄露了 60,000 名消费者的信息。解决方案包括分层数据范围（只读与写入）、基于 OAuth 的有限令牌，以及监管沙盒，以在全面推广前测试公平指标（Didenko 2021）。

Conclusion
结论
Regulators should adopt a progressive consent architecture—granular, revocable, with mandated fairness audits—balancing alternativecredit inclusion benefits against privacy and discrimination risks.
监管机构应采用渐进式同意架构——粒度化、可撤销、并强制进行公平性审计——平衡替代信用包容性利益与隐私和歧视风险。

10 │ Would a statutory fiduciary duty for buynowpaylater (BNPL) providers curb overindebtedness?
10 │ 购买即付（BNPL）提供者是否应承担法定信托责任以遏制过度负债？

Issue
问题
BNPL balances in Australia rose 70 % in 2022; delinquency surpassed creditcard levels. Should providers owe a fiduciary, bestinterests duty akin to financial advisers?
2022 年，澳大利亚的 BNPL 余额增长了 70%；逾期率超过了信用卡水平。提供者是否应承担类似于财务顾问的最佳利益信托责任？

Rule
规则
Current BNPL operates under deferredpayment exceptions; no responsiblelending test. COB principles apply weakly. Hayne’s Report urged expanding unconscionability but stopped short of fiduciary duties. Fiduciary law demands loyalty and no conflicts—a high bar.
当前的 BNPL 在延期付款例外情况下运作；没有负责任的贷款测试。COB 原则应用薄弱。Hayne 的报告敦促扩大不公平行为，但未涉及信托责任。信托法要求忠诚和没有利益冲突——这是一个很高的标准。

Application
应用
Behavioural traps: zerointerest framing obscures latefee economics (FCA 2013). ASIC pilots show that affordability checks cut default rates by 30 % without eroding approvals. Imposing fiduciary duty would force BNPL to prioritise borrower welfare—e.g., cap cumulative exposures, simplify disclosures. Industry claims cost hikes; yet margins rely on merchant fees, not interest, leaving cushions.
行为陷阱：零利率框架掩盖了滞纳金经济（FCA 2013）。ASIC 试点表明，财务承受能力检查可将违约率降低 30%，而不会损害批准率。强制实施信托责任将迫使 BNPL 优先考虑借款人福利——例如，限制累积风险敞口，简化信息披露。行业声称成本增加；然而，利润依赖于商户费用，而非利息，从而留有缓冲。

But fiduciary standards can be vague; enforcement relies on courts, risking uncertainty. A calibrated responsiblelending test—mirroring National Credit Code—could achieve similar outcomes more predictably.
但信托标准可能模糊；执行依赖于法院，存在不确定性风险。一个经过校准的负责任贷款测试——模仿国家信贷法——可以更可预测地实现类似结果。

Conclusion
结论
While a full fiduciary duty may overshoot, embedding a statutory bestinterest/affordability test for BNPL strikes a pragmatic compromise, aligning incentives without crippling the business model.
虽然全面的信托责任可能过度，但在 BNPL 中嵌入法定最佳利益/财务承受能力测试达成了务实的妥协，在不破坏商业模式的情况下协调了激励措施。

11 │ Reconciling algorithmic explainability with proprietary interests in fintech lending
11 │ 协调算法可解释性与金融科技贷款中的专有利益

Issue
问题
Regulators demand transparent AI decisions to detect bias; fintechs fear IP loss and gaming. Can both aims coexist?
监管机构要求人工智能决策透明以检测偏见；金融科技企业担心知识产权丧失和游戏。这两个目标能否共存？

Rule
规则
MiFID II Art 24(2) and ASIC RG 274 require “clear” product information. Emerging EU AI Act proposes “highrisk” system explainability. Posner & Weyl favour leastcost solutions; Coates advocates outcomebased benchmarks.
MiFID II 第 24(2)条和 ASIC RG 274 要求“清晰”的产品信息。新兴的欧盟 AI 法案提出“高风险”系统可解释性。Posner & Weyl 支持最低成本解决方案；Coates 倡导基于结果的基准。

Application
应用
Techniques like SHAP values provide local explanations without disclosing full code. Firms can supply regulators with confidential model artefacts under NDA, while publicfacing disclosures give reason codes—“insufficient income stability”—satisfying fairness. Pilot in Singapore’s MAS Veritas initiative shows banks retaining model secrecy yet passing bias audits. Adversarial gaming risk exists but is mitigated through randomised model retraining.
SHAP 值等技术提供局部解释而不泄露完整代码。企业可以在保密协议下向监管机构提供机密模型工件，同时面向公众的披露提供原因代码——“收入稳定性不足”——满足公平性。新加坡 MAS Veritas 计划中的试点显示银行保留模型机密性但通过偏见审计。对抗性游戏风险存在但通过随机化模型重新训练得到缓解。

Conclusion
结论
Explainability and IP are not mutually exclusive. Layered disclosure—confidential full transparency to supervisors, approximate explanations to consumers—meets accountability goals while preserving competitive moats.
可解释性和知识产权并非相互排斥。分层披露——对监管机构提供完全保密的透明度，对消费者提供近似解释——既满足问责制目标，又保持竞争优势壁垒。

Topic4:
主题 4:

1 │ Why regulate issuer disclosure? ≈ 475 words
1 │ 为什么监管发行人信息披露？≈ 475 字

Financialmarket disclosure rules are designed to solve three interlocking problems. Investor protection comes first in the statutes: by forcing issuers to reveal standardised, verifiable information, regulators reduce the “lemons” discount that dispersed savers would otherwise apply to unfamiliar companies . But what protects investors also promotes informational efficiency. Prices can only adjust rapidly and unbiasedly when all traders, not merely insiders, see the same facts; that is why disclosure is listed among the core mechanisms for “nearinstantaneous adjustment to newly revealed information” . Finally, accurate prices feed managers the signals they need to steer capital into its highestvalue uses, achieving allocative efficiency. The PFR lecture notes depict this virtuous circle explicitly: truthful information → higher liquidity → narrower bid–ask spreads → lower cost of capital → better investment screening .
金融市场信息披露规则旨在解决三个相互关联的问题。投资者保护在法规中处于首位：通过强制发行人披露标准化、可验证的信息，监管机构减少了分散储蓄者对不熟悉公司所应用的“柠檬”折价。但保护投资者也有助于提高信息效率。只有当所有交易者，而不仅仅是内幕人士，看到相同的事实时，价格才能快速和无偏见地调整；这就是为什么信息披露被列为“对新披露信息的近乎即时调整”的核心机制之一。最后，准确的价格为管理者提供了他们需要的信号，以引导资本进入其最高价值的用途，实现配置效率。PFR 讲座笔记明确描绘了这一良性循环：真实信息→更高的流动性→更窄的买卖价差→更低的资本成本→更好的投资筛选。

Because information is a public good, private antifraud suits alone cannot deliver the optimal flow. Each investor would prefer someone else to shoulder the investigative expense, freeriding once the data are reflected in price. Mandatory disclosure compels collective action and, by standardising content and timing, prevents managers from cherrypicking good news while burying risks. The slides identify four specific rationales: publicgood externality, competitive harm from unilateral disclosure, managerial agency costs and the efficiency gains from standard formats .
由于信息是一种公共物品，仅靠私人反欺诈诉讼无法实现最优的流通。每个投资者都希望有人承担调查费用，一旦数据反映在价格上，他们就可以搭便车。强制披露迫使我们采取集体行动，并通过规范内容和时间，防止管理者挑选好消息而隐藏风险。幻灯片确定了四个具体的理由：公共物品外部性、单方面披露带来的竞争损害、管理层的代理成本以及标准格式的效率收益。

The three goals reinforce one another rather than compete. A mandatory prospectus or continuousdisclosure notice widens the set of traders who can evaluate an issue; greater participation thickens order books and reduces spreads, which in turn lowers the firm’s hurdle rate for new projects. Conversely, a regime that prizes transparency for its own sake but ignores price accuracy can backfire: excessive liability risk may drive “marginal players” into opaque private markets, reducing the very pool of public information the rules were meant to enlarge . Hence modern reforms, such as Australia’s 2021 knowledgebased civilpenalty standard, try to calibrate fault thresholds so as to deter misconduct without chilling disclosure .
这三个目标相互强化而非相互竞争。强制性的招股说明书或持续披露通知扩大了能够评估问题的交易者的范围；更大的参与度使订单簿更加浓厚，并减少了买卖价差，这反过来又降低了公司新项目的高门槛。相反，一个只重视透明度而忽视价格准确性的制度可能会适得其反：过度的责任风险可能会将“边缘参与者”推向不透明的私人市场，从而减少了规则原本旨在扩大的公共信息池。因此，现代改革，如澳大利亚 2021 年的基于知识的民事处罚标准，试图校准过错阈值，以在不压制披露的情况下阻止不当行为。

In sum, investor protection, informational efficiency and allocative efficiency constitute a single chain: protect the trader, and you foster price accuracy; foster price accuracy, and you guide capital to its most productive uses. Securities regulation therefore focuses not on choosing one goal, but on designing disclosure duties that keep the chain intact while minimising unwarranted compliance costs.
总而言之，投资者保护、信息效率和配置效率构成一个单一的链条：保护交易者，你就能促进价格准确性；促进价格准确性，你就能引导资本到其最有效的用途。因此，证券监管的重点不在于选择一个目标，而在于设计披露义务，以保持链条的完整同时最小化不必要的合规成本。

2 │ Should all jurisdictions mimic the EU’s “immediate disclosure” rule? ≈ 485 words
2 │ 所有司法管辖区是否都应该模仿欧盟的“即时披露”规则？≈ 485 字

Article 17 of the EU Market Abuse Regulation (MAR) obliges listed companies to publish inside, pricesensitive information “as soon as possible.” The rule pursues horizontal fairness—no one trades on privileged facts—and vertical efficiency—prices incorporate valuerelevant news without delay . Empirical studies reviewed in the PFR materials link such rapid dissemination to higher liquidity and lower volatility, reinforcing both investor confidence and allocative accuracy .
欧盟市场滥用条例（MAR）第 17 条要求上市公司尽快发布内部、价格敏感信息。该规则追求横向公平——没有人能在享有特权的信息上进行交易——以及纵向效率——价格能及时反映与价值相关的新闻。PFR 材料中回顾的实证研究将这种快速传播与更高的流动性和更低的波动性联系起来，从而加强了投资者信心和配置准确性。

Yet disclosure is not costless. Illtimed announcements can destroy value in takeover negotiations or trigger selffulfilling bank runs. MAR therefore allows two carveouts: issuers may delay publication if the information is insufficiently precise or if immediate release would pose a threat to financial stability—the socalled “Northern Rock” exception . Importantly, the company must maintain confidentiality and later file a record explaining the delay.
然而，披露并非没有成本。时机不当的公告可能会在收购谈判中摧毁价值或引发自我实现的银行挤兑。因此，MAR 允许两项例外：如果信息不够精确或立即发布会对金融稳定构成威胁，发行人可以延迟发布——所谓的“北岩银行”例外。重要的是，公司必须保持机密性，并随后提交一份记录解释延迟原因。

Other jurisdictions should adopt the EU principle—prompt release once facts are reliable—but retain flexibility where premature revelation would harm investors themselves. Australia’s Corporations Act s 674A offers a useful model: disclosure may be deferred where a reasonable person would not expect publication and the information is confidential; however, once confidentiality leaks, the obligation crystallises. The 2023 Getswift judgment illustrates the consequences of ignoring that trigger: the Federal Court levied a record A$15 million penalty for 22 missed updates .
其他司法管辖区应采纳欧盟原则——事实可靠后立即释放——但在过早披露会损害投资者自身利益的情况下保持灵活性。澳大利亚的《公司法》第 674A 条提供了一个有用的范例：如果合理的人不会预期发布且信息是保密的，则可以推迟披露；然而，一旦保密性泄露，义务就会明确。2023 年的 Getswift 判决说明了忽视该触发器后果：联邦法院对 22 次未更新的行为处以了创纪录的 1500 万澳元罚款。

The United States takes a hybrid approach. Section 409 of SarbanesOxley directs the SEC to adopt “realtime” disclosure rules, but the Commission settled on a fourday Form 8K deadline rather than instantaneous release. Issuers gain a brief window to verify data, while investors receive a predictable cadence of marketmoving filings.
美国采取了一种混合方法。《萨班斯-奥克斯利法案》第 409 条指示 SEC 采用“实时”披露规则，但委员会选择了一个四天的 8K 表格截止日期，而不是即时发布。发行人获得了一个短暂的时间窗口来验证数据，而投资者则收到了可预测的市场推动文件节奏。

Which system best balances accuracy, fairness and feasibility? Comparative evidence suggests the marginal gain from compressing the window below one trading day is modest, while error costs rise sharply if boards rush out unverified figures. Accordingly, the recommended template for nonEU markets is:
哪个系统能最好地平衡准确性、公平性和可行性？比较证据表明，将窗口压缩到一天以下交易日的边际收益很小，而如果董事会仓促发布未经验证的数字，错误成本会急剧上升。因此，非欧盟市场的推荐模板是：

Publish without undue delay once information is precise and verified.
一旦信息精确并核实，应不迟延地发布。

Provide a negotiation safeharbour for pricesensitive M&A talks, subject to leakage monitoring.
为价格敏感的并购谈判提供谈判安全港，并需进行泄漏监控。

Embed a financialstability exception modelled on MAR and Basel guidance for credit institutions.
为信用机构嵌入一个基于 MAR 和巴塞尔指南的金融稳定例外模型。

Apply a knowledgebased liability standard (reckless or negligent nondisclosure) to deter opportunism without chilling legitimate delay.
对抗机会主义行为，同时不压制合法延迟，应用基于知识的责任标准（鲁莽或疏忽不披露）。

Such a calibrated regime achieves the EU’s twin aims of equal access and price accuracy while respecting situational realities faced by boards, auditors and prudential supervisors.
这样的调节机制在尊重董事会、审计师和审慎监管机构所面临的现实情况的同时，实现了欧盟的双重要求：平等接入和价格准确性。

3 │ Reputation intermediaries: promise and pitfalls ≈ 472 words
3 │声誉中介：承诺与陷阱 ≈ 472 字

Underwriters, auditors, exchanges, analysts and creditrating agencies occupy a special niche as reputation intermediaries: they collect, verify and package issuer data for other market participants, selling the credibility built up over repeated transactions . Their economic value lies in solving adverseselection problems. For an IPO, the investment bank’s name signals that the prospectus has survived rigorous due diligence, reducing the expected error in valuation and thereby tightening the offer spread .
承销商、审计师、交易所、分析师和信用评级机构作为声誉中介占据特殊地位：他们收集、核实并打包发行人数据供其他市场参与者使用，出售通过重复交易建立起来的信誉。他们的经济价值在于解决逆向选择问题。对于首次公开募股，投资银行的名字表明招股说明书已经通过了严格的尽职调查，减少了估值中的预期误差，从而缩小了报价区间。

Classic theory predicts that a firm whose only asset is trustworthiness will not risk it for a single fee. Reality is messier. The lecture notes describe three structural conflicts. Issuerpays models pit the intermediary’s desire for future business against its duty to investors. Crosssubsidisation enables banks to promise favourable research in exchange for lucrative underwriting mandates. And internal governance failures can allow a dominant partner to override risk controls, as happened when Arthur Andersen’s Houston office prioritised Enron’s $100 million fees over auditing standards .
经典理论预测，一家唯一资产是信誉的公司不会为了单次费用而冒险。现实则更为复杂。讲义笔记描述了三种结构性冲突。发行人付费模式使中介机构对未来的业务需求与其对投资者的责任之间产生冲突。交叉补贴使银行能够承诺提供优惠的研究服务，以换取有利可图的承销任务。而内部治理失败可能导致主要合作伙伴无视风险控制，正如安达信休斯顿办事处优先考虑安然公司 1 亿美元的费用而不是审计标准的情况。

These conflicts explain why reputation alone cannot guarantee credibility. Barriers to entry further weaken the penalty for failure: if the law permits only a handful of licensed auditors or rating agencies, each incumbent enjoys quasirents that dull the fear of client flight. The slides label this the “regulatory licence effect” .
这些冲突解释了为什么仅靠声誉无法保证可信度。进入壁垒进一步削弱了失败的惩罚：如果法律仅允许少数几家获得许可的审计师或评级机构，那么每个现有机构都会享有准租金，从而削弱了客户流失的恐惧。幻灯片将这一现象标记为“监管许可效应”。

Consequently, modern market regulation deputises reputation intermediaries as gatekeepers and supplements private discipline with public liability. Auditors face joint civil responsibility for material misstatements; rating agencies must separate analytical from commercial staff and disclose models; lead underwriters certify prospectus accuracy on pain of statutory damages. By converting reputational sanctions into legal ones, regulators seek to restore the bond between the intermediary’s private incentives and its publicinformation role.
因此，现代市场监管将声誉中介机构视为守门人，并用公共责任补充私人纪律。审计师对重大虚假陈述承担共同民事责任；评级机构必须将分析人员与商业人员分开，并披露模型；主承销商必须保证招股说明书准确性，否则将面临法定赔偿。通过将声誉制裁转化为法律制裁，监管机构试图恢复中介机构私人激励与其公共信息角色之间的联系。

Still, overdeterrence is a risk. Excessive liability can drive gatekeepers to defensive practices—boilerplate disclaimers, boxticking reviews—that add cost without improving informational content. A balanced system therefore couples liability with independence mandates (e.g., mandatory auditcommittee appointment) and transparency of conflicts, allowing investors to pricediscount and boards to shop for credible certifiers.
然而，过度威慑是一种风险。过度的责任可能会使守门人采取防御性做法——标准免责声明、框选式审查——这些做法会增加成本但不提高信息内容。因此，一个平衡的体系将责任与独立性要求（例如，强制审计委员会任命）和利益冲突的透明度相结合，允许投资者进行价格折价，并使董事会能够寻找可信的认证者。

In short, reputation intermediaries are indispensable to modern securities markets, but their incentives can drift when fees today outshine reputational rents tomorrow. Sensible regulation does not supplant reputation; it buttresses it with calibrated legal accountability and disclosure of conflicts, preserving the intermediary’s value as a credible information producer.
简而言之，声誉中介机构对现代证券市场不可或缺，但当今天的费用超过明天的声誉租金时，他们的激励机制可能会偏离。合理的监管不是取代声誉，而是通过精确的法律责任和利益冲突披露来支持声誉，保持中介机构作为可信信息生产者的价值。

4 │ Regulating short selling ≈ 490 words
4 │ 监管卖空 ≈ 490 字

Short selling—selling borrowed securities and repurchasing them later—allows pessimistic or frauddetecting traders to express negative views that are otherwise hard to transmit because one cannot “underbuy” a share. The teaching slides list three empirical benefits: quicker correction of overvaluation, exposure of accounting fraud (Karpoff et al.), and greater overall liquidity . Indeed, bans on shorting tend to inflate prices temporarily and raise volatility once the constraint is lifted, as documented across the 2008 emergency prohibitions .
卖空——即出售借入的证券并在之后回购——允许悲观或欺诈检测交易者表达难以传递的负面观点，因为一个人不能“低估”一股股票。教学幻灯片列出了三项实证收益：更快地纠正高估、揭露会计欺诈（Karpoff 等人）以及整体流动性增加。事实上，禁止卖空往往会暂时推高价格，一旦限制解除，波动性也会上升，正如 2008 年紧急禁止令所记录的那样。

Nevertheless, regulators worry about manipulative bear raids, settlement failures in “naked” shorts, and feedback loops in fragile markets. The optimal regulatory design therefore distinguishes covered from naked positions and embeds circuitbreakers for systemic stress.
然而，监管机构担心操纵性的空头袭击、裸空交易的结算失败以及脆弱市场中的反馈循环。因此，最优的监管设计区分了有担保空头和无担保空头，并嵌入了系统性压力下的断路器。

United States. SEC Rule 102 requires brokers to “locate” stock before sale and activates an “uptick” constraint once a share drops 10 percent in a day, permitting further shorts only above the national best bid .
美国。SEC 规则 102 要求经纪人在出售前“定位”股票，并在股价日内下跌 10%时激活“上涨限制”约束，仅允许在国家级最佳买价之上进行进一步空头交易。

European Union. Regulation 236/2012 mirrors those locate rules, adds tiered disclosure of netshort positions (private at 0.2 percent of capital, public at 0.5 percent) and gives ESMA emergencyban powers .
欧盟。法规 236/2012 反映了这些定位规则，增加了净空头头寸的分层披露（私人为资本 0.2%，公共为 0.5%），并赋予 ESMA 紧急禁止权力。

Australia. Section 1020B of the Corporations Act bans naked shorts by demanding a “presently exercisable and unconditional” right to deliver securities; ASIC Regulatory Guide 196 spells out limited exceptions .
澳大利亚。公司法第 1020B 条通过要求“当前可执行且无条件的”交付证券权利来禁止裸空交易；ASIC 监管指南 196 详细说明了有限的例外情况。

These regimes share a philosophy: permit informed, fullycollateralised shorting to aid price discovery, while blocking settlementrisk and outright manipulation. Comparative evidence reinforces the wisdom of this middle path. The Wirecard scandal shows that banning shorts to “protect market confidence” can actually delay exposure of fraud and magnify ultimate losses . Conversely, unlimited naked shorting would allow failures to deliver, undermining trust in clearance systems.
这些机制共享一种哲学：允许有信息、完全抵押的做空以帮助价格发现，同时阻止结算风险和直接操纵。比较证据加强了这种中间道路的智慧。Wirecard 丑闻表明，禁止做空以“保护市场信心”实际上会延迟欺诈的暴露并扩大最终损失。相反，无限制的裸卖空将允许无法交割，从而破坏对清算系统的信任。

A refined toolkit should therefore include:
因此，一个精致的工具包应该包括：

Locate/borrow requirements at trade entry.
交易入口时的定位/借入要求。

Pre and posttrade transparency of large short positions.
大额空头头寸的交前和交后透明度。

Temporary bans limited to systemic crises, subject to sunset clauses and cost–benefit review.
临时禁令仅限于系统性危机，受到期满条款和成本效益审查的限制。

Marketmanipulation enforcement targeting false rumours or spoofing, rather than short interest per se.
市场操纵执法针对虚假谣言或伪装，而非短期兴趣本身。

Centralcounterparty penalties for failstodeliver that exceed settlement grace periods.
中央对手方因未能交付而受到的处罚超过结算宽限期。

Such a framework preserves the social value of bearish research and hedging while containing the externalities that can arise when negative sentiment cascades through thinly capitalised intermediaries.
这样的框架保留了看跌研究和套期保值的社交价值，同时控制了负面情绪通过资本化不足的中介蔓延时可能产生的外部性。

5 │ Insidertrading prohibitions and crossjurisdictional contrasts ≈ 480 words
5 │ 内部交易禁止和跨司法管辖区对比 ≈ 480 字

Insider trading is barred chiefly because it redistributes trading surplus from uninformed parties to insiders, widening bidask spreads and deterring market participation. The slides illustrate how fear of “trading against God” reduces liquidity and raises issuers’ cost of capital . Proponents such as Manne have argued that letting insiders trade would accelerate price discovery, yet Goshen & Parchomovsky counter that professionals, not insiders, are best at processing macro and competingfirm data; insider predominance could actually lower price accuracy .
内部交易主要被禁止是因为它将交易盈余从非知情方重新分配给内部人士，扩大买卖价差并抑制市场参与。幻灯片说明了“与上帝作对”的恐惧如何减少流动性并提高发行人的资本成本。曼恩等支持者认为，允许内部人士交易将加速价格发现，但戈申和帕尔乔莫夫斯基反驳称，专业人士而不是内部人士最适合处理宏观和竞争公司数据；内部人士的主导地位实际上可能降低价格准确性。

Australia and the EU adopt a possessionbased rule: anyone holding nonpublic, pricesensitive information must abstain or disclose. Australia codifies this in s 1043A; the EU embeds it in MAR Arts 7–8. The breadth simplifies enforcement by avoiding inquiry into fiduciary duty or source .
澳大利亚和欧盟采用基于持有原则的规则：任何持有非公开、价格敏感信息的人必须 abstain 或披露。澳大利亚在 s 1043A 中对此进行了编纂；欧盟将其嵌入 MAR 第 7-8 条。这种广泛性通过避免调查信托责任或来源简化了执行。

United States. The SEC’s Rule 10b5, as interpreted in Dirks and O’Hagan, requires that the trader breach a fiduciary duty or misappropriate information. Hence a journalist who stumbles upon material news but owes no duty may trade, whereas a corporate tippee is liable if the insider benefited personally. The US model roots liability in fraud theory, respecting the Constitution’s limits on federal police power.
美国。美国证券交易委员会的规则 10b5，根据 Dirks 和 O’Hagan 的解释，要求交易者违反信托责任或挪用信息。因此，一个偶然发现重大消息但无义务的记者可以进行交易，而一个公司内部消息接收者如果从中受益将承担法律责任。美国的模式将责任根植于欺诈理论，尊重宪法对联邦警察权力的限制。

Comparative evaluation. The possession test offers predictability and stronger deterrence but risks overbreadth—for example, custodial staff finding a draft announcement in the bin. The dutybased US test tailors liability to unfair advantage but demands proof of duty and personal benefit, raising litigation costs and leaving gaps (e.g., familyoffice analysts trading on clients’ mosaics).
比较评估。占有测试提供了可预测性和更强的威慑力，但存在过度宽泛的风险——例如，保管人员发现垃圾桶里的草稿公告。基于义务的美国测试根据不公平优势调整责任，但要求证明义务和个人利益，增加了诉讼成本，并留下了空白（例如，家族办公室分析师根据客户的马赛克进行交易）。

All regimes converge on the principle that investors should not lose merely because they arrived later to information that ought to have been public. Harmonisation efforts therefore focus on safe harbours (e.g., prescheduled buybacks) and crossborder enforcement coordination (IOSCO multilateral MoUs).
所有制度都趋同于一个原则，即投资者不应仅仅因为他们到达公共信息的时间较晚而受到损失。因此，协调工作重点在于安全港（例如，预先安排的回购）和跨境执法协调（IOSCO 多边谅解备忘录）。

In sum, insider‐trading bans reinforce disclosure policy: if inside information must ultimately enter the market via public filings, insiders cannot be allowed to monetise it beforehand. The divergences across Australia, the EU and the US reflect different legal traditions, yet each seeks the same economic end—lower spreads, greater participation and more accurate prices.
总而言之，内幕交易禁令加强了信息披露政策：如果内幕信息最终必须通过公开披露进入市场，那么内幕人士就不被允许在此之前将其变现。澳大利亚、欧盟和美国的差异反映了不同的法律传统，但每个国家都寻求相同的经济目标——降低价差、提高参与度和更准确的价格。

6 │ Private credit: stability risks and regulatory responses ≈ 495 words
6 │ 私募信贷：稳定性风险与监管响应 ≈ 495 字

Private credit funds—nonbank vehicles that extend loans outside public markets—have surged as institutional investors hunt yield. Australian regulators estimate domestic assets under management at A$200 billion and rising, echoing a thirtyfold expansion across AsiaPacific since 2002 . The Reserve Bank of Australia warns that such growth shifts leverage into entities subject to lighter disclosure and liquidity rules, potentially amplifying systemic stress.
私募信贷基金——在公共市场之外提供贷款的非银行工具——随着机构投资者寻求收益而激增。澳大利亚监管机构估计，国内管理资产规模达到 2000 亿澳元且仍在增长，这反映了自 2002 年以来亚太地区三十倍的扩张。澳大利亚储备银行警告称，这种增长将杠杆转移到受更轻信息披露和流动性规则约束的实体，可能会加剧系统性压力。

Key risks resemble those that beset shadow banking pre2008. Opacity: portfolio valuations rely on manager marks with limited thirdparty verification, obscuring early signs of borrower distress. Liquidity mismatch: openended vehicles may promise monthly redemptions while holding multiyear, illiquid loans. Procyclical leverage: margin lenders can demand additional collateral during downturns, forcing asset sales at firesale prices. Interconnectedness: banks often provide funding lines or invest in senior tranches, reimporting the risk onto their balancesheets.
关键风险类似于 2008 年前的影子银行业务所面临的风险。不透明性：投资组合估值依赖于管理者的估值，且缺乏第三方验证，掩盖了借款人早期的困境迹象。流动性错配：开放式工具可能承诺每月赎回，但持有多年期、不流动的贷款。顺周期杠杆：在衰退期间，保证金贷款者可能要求额外的抵押品，迫使资产以贱卖价格出售。关联性：银行通常提供融资额度或投资于高级档，将风险重新引入其资产负债表。

Regulatory tools. 1) Enhanced data collection. The RBA advocates borrowerlevel reporting of leverage, covenants and repayment schedules to supervisory authorities; the ECB’s AnaCredit offers a precedent. 2) Liquidity management rules. Minimum liquidasset ratios and swingpricing mechanisms can mitigate redemption runs. 3) Stresstesting. Large privatedebt managers should be subject to scenario analyses akin to bank capital stress tests, reflecting their systemic footprint. 4) Disclosure harmonisation. Bringing privatecredit funds under a prospectusstyle regime when marketing to semiprofessional investors would align incentives with those in public debt markets. The UK’s 2024 review of valuation practices and the EU’s 2024 AIFMD amendments point in this direction .
监管工具。1）加强数据收集。RBA 提倡借款人层面的杠杆、契约和还款计划向监管机构报告；ECB 的 AnaCredit 提供了先例。2）流动性管理规则。最低流动资产比例和摆动定价机制可以缓解赎回潮。3）压力测试。大型私人债务管理者应受到类似于银行资本压力测试的场景分析，反映其系统重要性。4）信息披露协调。在向半专业投资者营销时，将私人信贷基金置于招股说明书式监管体系下，将使激励与公共债务市场中的激励相一致。英国 2024 年对估值实践的调查以及欧盟 2024 年 AIFMD 修正案指明了这一方向。

Policymakers must tread carefully: heavyhanded rules could push credit further into opaque niches. A proportionate approach would (i) focus on leverage caps tied to redemption frequency, (ii) mandate thirdparty valuation for portfolios exceeding a size threshold, and (iii) require public disclosure of fundlevel risk metrics, thereby enlisting market discipline. Banks’ exposures to privatecredit funds should attract capital charges that increase with opacity, internalising potential spillovers.
政策制定者必须谨慎行事：过于强硬的规则可能会将信贷进一步推向不透明的领域。应采取适度方法：(i) 重点关注与赎回频率挂钩的杠杆率上限，(ii) 强制对超过一定规模的投资组合进行第三方评估，以及 (iii) 要求公开披露基金层面的风险指标，从而引入市场约束。银行对私募信贷基金的风险敞口应吸引随着不透明度增加而增加的资本费用，从而内部化潜在的溢出效应。

In short, the rise of private credit replicates the basic financialintermediation function—transforming illiquid claims into tradable interests—without the accompanying bankstyle safeguards. To preserve the benefits of credit diversity while guarding stability, regulators should illuminate the sector through data, align redemption terms with asset liquidity, and ensure that bank linkages remain adequately capitalised. Transparent, proportionate oversight can integrate private credit into the wider financial system without incubating the next crisis.
简而言之，私募信贷的兴起复制了基本的金融中介功能——将非流动性债权转化为可交易的利益——而没有银行式的安全措施。为了在保持信贷多样性的好处的同时维护稳定，监管机构应通过数据揭示该行业，使赎回条款与资产流动性相匹配，并确保银行联系得到充分的资本化。透明、适度的监管可以在不孕育下一次危机的情况下将私募信贷纳入更广泛的金融体系。

7 │ Regulating “finfluencers” in securities markets ≈ 480 words
7 │ 监管证券市场的“金融影响者” ≈ 480 字

Socialmedia commentators who recommend shares, options or cryptoassets—colloquially finfluencers—occupy a grey zone between journalism and licensed advice. Their business model rests on viral content monetised through advertising, paid “tip sheets” or undisclosed remuneration from issuers. Two recent Australian prosecutions illustrate the regulatory gap. Tyson Scholz (known as “ASX Wolf”) was found to have carried on a financialservices business without an Australian Financial Services Licence (AFSL), contravening s 911A of the Corporations Act. Gabriel Govinda (“Fibonarchery”) was jailed for coordinated pumpanddump schemes that combined socialmedia hype with manipulative trading.
推荐股票、期权或加密资产的社交媒体评论者——俗称金融影响者——处于新闻和许可建议之间的灰色地带。他们的商业模式依赖于通过广告、付费“提示单”或发行人未公开的报酬来盈利的病毒式内容。最近的两个澳大利亚起诉案例说明了监管空白。Tyson Scholz（被称为“ASX 狼”）被发现未持有澳大利亚金融服务许可证（AFSL）就从事金融服务业务，违反了《公司法》第 911A 条。Gabriel Govinda（“Fibonarchery”）因协调的操纵性拉升和抛售计划而被判刑，这些计划结合了社交媒体炒作和操纵性交易。

The case for regulation echoes classic disclosure theory. Retail viewers face a severe information asymmetry: they cannot observe the finfluencer’s incentives, trading intentions or competence. Unchecked, hidden conflicts degrade investor protection and—by propagating false or selective claims—undermine informational efficiency. Issuers, meanwhile, may pay for favourable coverage, skirting continuousdisclosure obligations that require news to reach the whole market simultaneously.
对监管的呼吁呼应了经典的披露理论。零售观众面临严重的信息不对称：他们无法观察到金融影响者的动机、交易意图或能力。不受控制的情况下，隐藏的冲突会损害投资者保护，并通过传播虚假或选择性的声明来破坏信息效率。同时，发行人可能会为有利的报道付费，从而规避了要求新闻同时传递给整个市场的持续披露义务。

ASIC’s current response applies existing law rather than creating a bespoke regime. A finfluencer providing personal advice must hold an AFSL; one who disseminates marketmanipulative rumours commits a Div 2 offence; failure to disclose sponsorship breaches the general “misleading conduct” prohibition. This functional approach avoids subjectmatter carveouts, maintains technological neutrality and leverages welldeveloped case law on advice licensing. Critics argue that the AFSL threshold is vague and compliance costs may chill legitimate commentary. A proportional fix is to issue brightline guidance: paid promotions trigger licensing; unpaid educational content does not—provided the presenter avoids recommendations about specific products.
ASIC 目前的应对措施适用于现有法律，而不是创建一个定制的制度。提供个人建议的金融影响者必须持有 AFSL；散布市场操纵谣言者构成第 2 部分罪行；未披露赞助行为违反了一般的“误导行为”禁止。这种功能方法避免了特定领域的例外情况，保持了技术中立性，并利用了关于建议许可的成熟案例法。批评者认为 AFSL 门槛模糊，合规成本可能会抑制合法评论。一个比例适当的解决方案是发布明确的指导方针：付费推广会触发许可；无偿的教育内容不会——只要主持人避免对特定产品提出建议。

From an allocativeefficiency standpoint, the goal is to channel finfluencer energy into price discovery rather than noise trading. Platforms can assist by requiring clear labelling of paid endorsements and hyperlinking to issuer announcements, thereby integrating social media into the continuousdisclosure ecosystem. At the same time, regulators should improve financialliteracy resources so that retail viewers can better evaluate signals.
从配置效率的角度来看，目标是将金融影响者的精力引导至价格发现，而不是噪音交易。平台可以通过要求付费代言的清晰标签和链接到发行人公告来协助，从而将社交媒体整合到持续披露生态系统中。同时，监管机构应改进金融素养资源，以便零售观众能更好地评估信号。

In sum, extending core licensing and antimanipulation rules to finfluencers protects investors, preserves the integrity of the information set feeding prices, and keeps enforcement technologyneutral. Tailored guidance rather than new legislation appears the most efficient route.
总而言之，将核心许可和反操纵规则扩展到金融影响者可以保护投资者，维护价格信息集的完整性，并保持执法技术的中立性。定制指导而非新立法似乎是最有效的方法。

8 │ Passive investing, price discovery and systemic risk ≈ 475 words
8 │ 被动投资、价格发现和系统性风险 ≈ 475 字

Index funds have grown from US $11 billion in 1993 to over US $2.8 trillion by 2020. Because they track benchmarks mechanically, passive funds do not gather firmspecific data; instead, they freeride on the analytical work of active managers. A rising passive share therefore poses two potential hazards. First, informational efficiency may deteriorate: if fewer analysts price firms, misvaluations persist longer, widening the wedge between fundamentals and quoted prices. Empirical work by Huebner, Haddad and Loualiche finds that passive ownership reduces price elasticity, leading to higher volatility and lower liquidity in affected stocks. Second, large index providers such as BlackRock or Vanguard create concentration risk; a shock to their redemption flows could force them to liquidate across multiple asset classes simultaneously, amplifying systemic stress.
指数基金从 1993 年的 110 亿美元增长到 2020 年的超过 2.8 万亿美元。由于它们机械地跟踪基准，被动型基金不会收集公司特定数据；相反，它们搭便车，利用主动管理者的分析工作。因此，被动型基金份额的上升可能带来两种潜在风险。首先，信息效率可能会下降：如果分析师较少地定价公司，错误估值会持续更长时间，扩大基本面与报价之间的差距。Huebner、Haddad 和 Loualiche 的实证研究表明，被动所有权降低了价格弹性，导致受影响股票的波动性增加和流动性降低。其次，像 BlackRock 或 Vanguard 这样的大型指数提供商会带来集中度风险；它们赎回流的冲击可能会迫使它们同时清算多个资产类别，加剧系统性压力。

Regulatory interventions should aim to preserve the benefits of passive investing—low fees, diversification—while mitigating these externalities. On the informational side, authorities could promote a “research carveout”: index sponsors rebate a portion of management fees into independent research, maintaining the analytical workforce that underpins price discovery. On the systemic side, macroprudential supervisors might impose liquiditymanagement requirements on very large ETF sponsors, mirroring the stresstesting applied to banks.
监管干预应旨在保留被动投资的优势——低费用、分散化——同时减轻这些外部性。在信息方面，当局可以推广“研究豁免”：指数发起人将一部分管理费 rebate 到独立研究中，维持支撑价格发现的分析人员队伍。在系统方面，宏观审慎监管机构可能对大型 ETF 发起人施加流动性管理要求，模仿对银行实施的压力测试。

Some scholars worry that common ownership by indexers dampens interfirm competition, citing airlineindustry evidence of higher ticket prices. The empirical debate remains unsettled, but existing antitrust tools can already police overt collusion. A disclosurebased solution—periodic reporting of ownership concentrations by industry—allows competition authorities to react if future evidence solidifies.
一些学者担心索引器之间的共同所有权会削弱企业间竞争，引用了航空业机票价格更高的证据。实证辩论尚未解决，但现有的反垄断工具已经可以监管公开的串通行为。一种基于披露的解决方案——定期报告行业中的所有权集中度——允许竞争当局在未来证据巩固时做出反应。

The rise of passive funds ultimately reflects investor demand for lowcost exposure. Wholesale bans would harm allocative efficiency by raising tracking error and expense ratios. Calibrated measures that reinforce research production and build systemic buffers offer a superior balance.
被动型基金的增长最终反映了投资者对低成本投资的需求。批发禁令将通过增加跟踪误差和费用比率来损害配置效率。经过校准的措施可以加强研究产出并建立系统缓冲，从而提供更好的平衡。

9 │ Climaterisk disclosure: toward mandatory integration ≈ 485 words
9 │ 气候风险披露：迈向强制整合 ≈ 485 字

Physical and transition climate risks now pose material threats to cash flows in carbonexposed sectors. Yet existing accounting rules capture only historical costs, leaving investors to infer exposure piecemeal. The PFR slides list “climate risk information?” alongside periodic and adhoc disclosures, signalling unsettled scope under existing regimes. In March 2024 the US SEC issued a climatedisclosure rule, but litigation forced the agency to pause its defence in February 2025; similarly, some EU issuers complain that the Corporate Sustainability Reporting Directive overloads reports with boilerplate.
实际和转型气候风险现在对碳暴露行业的现金流构成实质性威胁。然而，现有的会计规则仅捕捉历史成本，导致投资者需要逐步推断风险敞口。PFR 幻灯片将“气候风险信息？”与定期和临时披露并列，表明在现有制度下的范围尚未确定。2024 年 3 月，美国证券交易委员会发布了一项气候披露规则，但诉讼迫使该机构在 2025 年 2 月暂停其辩护；类似地，一些欧盟发行人抱怨企业可持续发展报告指令使报告充斥着陈词滥调。

The economic logic for mandatory integration rests on the same trio of goals that justify financial disclosure generally. Investor protection: diversified portfolios cannot hedge systemic climate risk by diversification alone; they need forwardlooking data (emissions pathways, carbon taxes, supplychain exposure). Informational efficiency: without comparable metrics, prices underreact, misallocating capital to strandedasset projects. Allocative efficiency: managers misread market signals and oversupply carbonintensive capacity.
强制性整合的经济逻辑基于一般财务披露所依据的同一组目标。投资者保护：多元化投资组合无法仅通过多元化来对冲系统性气候风险；它们需要前瞻性数据（排放路径、碳税、供应链风险）。信息效率：没有可比指标，价格反应不足，导致资本错配给搁浅资产项目。配置效率：管理者误读市场信号，过度供应碳密集型产能。

Sceptics point to measurement uncertainty and compliance cost. Yet the PFR slides emphasise that forwardlooking MD&A statements in the US already enjoy a safe harbour, showing that liability shields can coexist with disclosure for soft information. A practical framework would include: (i) scope 3 emissions for large emitters; (ii) scenarioanalysis narratives rather than precise probability weights; (iii) phasedin assurance standards, starting with limited assurance and graduating to reasonable assurance as methodologies mature.
怀疑论者指出测量不确定性和合规成本。然而，PFR 幻灯片强调，美国的前瞻性管理层讨论与分析声明已经享有安全港，表明责任保护可以与软信息的披露共存。一个实用的框架应包括：(i) 大型排放者的范围 3 排放；(ii) 情景分析叙述而不是精确的概率权重；(iii) 分阶段实施的保证标准，从有限保证开始，随着方法的成熟逐步过渡到合理保证。

From a crossborder standpoint, fragmentation undermines comparability. IOSCO should work with the International Sustainability Standards Board to benchmark core metrics, allowing issuers to file one set of numbers globally. Carveouts could apply to smallcap firms to avoid disproportionate cost, but materiality triggers would still catch sectorspecific exposures (e.g., agriculture, insurance).
从跨境角度来看，碎片化会破坏可比性。IOSCO 应与国际可持续性标准理事会合作，制定核心指标基准，允许发行人在全球范围内提交一套数据。对于小市值公司，可以适用例外情况以避免不成比例的成本，但实质性触发因素仍会捕获特定行业的风险（例如，农业、保险）。

Mandating climaterisk disclosure thus aligns market signals with emerging physical realities, enhancing price accuracy and capital allocation while giving investors tools to hedge. The transitional compliance burden can be managed through safe harbours and phased adoption.
强制气候风险披露从而使市场信号与新兴的物理现实相一致，提高价格准确性和资本配置，同时为投资者提供对冲工具。过渡期的合规负担可以通过安全港和分阶段采用来管理。

10 │ Do “uptick” rules for short sales work? ≈ 450 words
10 │ 短售的“上涨”规则有效吗？≈ 450 字

The 2008 crisis saw regulators scramble to curb short selling. Postcrisis, the US SEC adopted Rule 201’s circuitbreaker uptick rule: when a stock falls more than 10 percent intraday, further short sales are allowed only above the best national bid until the next close. The EU and Australia opted for different levers—position disclosure and nakedshort bans—but no hard uptick rule.
2008 年的危机中，监管机构匆忙试图遏制短售。危机后，美国 SEC 采用了 201 号规则中的熔断上涨规则：当股票日内下跌超过 10%时，只有在最佳全国报价以上才允许进一步的短售，直到下一个收盘。欧盟和澳大利亚选择了不同的杠杆——持仓披露和裸短售禁令——但没有硬性的上涨规则。

The rule’s objective is twofold: prevent downward price spirals driven by momentum shorts and protect market confidence, while preserving informational efficiency by leaving covered shorting largely intact. Event studies show that Rule 201 reduces bidask spreads and volatility in the immediate aftermath of a circuitbreaker trigger, suggesting it calms markets during stress. However, broader evidence indicates minimal impact on longrun price levels or liquidity, because investors can still express negative views via derivatives or wait one day for the circuitbreaker to lapse.
该规则的目标有两个：防止由动量卖空驱动的向下价格螺旋，并保护市场信心，同时通过基本保留覆盖卖空来保持信息效率。事件研究显示，规则 201 在触发熔断机制后减少了买卖价差和波动性，表明它在市场压力期间起到了稳定作用。然而，更广泛的证据表明对长期价格水平或流动性影响甚微，因为投资者仍然可以通过衍生品表达负面观点，或者等待熔断机制失效一天。

Critics argue that the rule adds complexity without addressing root causes—settlement risk and false rumours. Indeed, settlement failures arise mainly from naked shorts, which Rule 201 does not directly police; that task falls to Rule 204’s closeout requirement. Moreover, the PFR slides list empirical findings from 2008: blanket bans produced a shortterm price boost but higher volatility later, implying that constraints delay rather than prevent price adjustment.
批评者认为该规则增加了复杂性，但没有解决根本原因——结算风险和虚假谣言。事实上，结算失败主要源于裸空，而规则 201 并没有直接监管；这项任务由规则 204 的清算要求负责。此外，PFR 幻灯片列出了 2008 年的实证发现：全面禁令短期内推高了价格，但后来波动性更高，这意味着限制措施延缓而不是阻止了价格调整。

A balanced approach combines: (i) locate requirements to eliminate naked shorts, (ii) realtime disclosure of large netshort positions above 0.5 percent (the EU model), and (iii) temporary circuitbreakers like Rule 201 to dampen disorderly cascades. This toolkit targets settlement and manipulation risks while retaining the social value of bearish research evidenced by the Wirecard case, where a German ban delayed exposure of massive fraud.
平衡的方法结合：(i) 定位需求以消除裸空头，(ii) 实时披露超过 0.5%的大净空头头寸（欧盟模式），以及 (iii) 临时熔断机制，如规则 201，以抑制无序的连锁反应。这套工具旨在针对结算和操纵风险，同时保留看跌研究的社交价值，例如 Wirecard 案例，其中德国禁令延迟了大规模欺诈的曝光。

Overall, uptick rules are useful as a lastline brake, but they are no substitute for robust clearing safeguards and antimanipulation enforcement. Regulators should keep the trigger threshold high and ensure automatic sunset to limit unintended liquidity costs.
总体而言，上涨规则作为最后一道刹车是有用的，但它们不能替代强大的清算保障和反操纵执法。监管机构应保持触发阈值较高，并确保自动日落以限制意外的流动性成本。

11 │ Can reputation alone discipline creditrating agencies? ≈ 480 words
11 │ 声望 alone 能否约束信用评级机构？≈ 480 words

Creditrating agencies (CRAs) epitomise reputation intermediaries: they monetise a signal of honesty earned over decades through repeat interactions. In theory, fear of reputational loss should deter biased ratings. In practice, the global financial crisis exposed severe distortions. Issuerpays fees, limited competition and regulatory licences that embed ratings in bond covenants weakened the cost of failure: even after notorious AAA ratings on subprime mortgage tranches, the “big three” retained market share.
信用评级机构（CRAs）是声誉中介的典型代表：它们通过数十年的重复互动赚取的诚信信号进行货币化。理论上，对声誉损失的恐惧应该阻止偏袒的评级。实际上，全球金融危机暴露了严重的扭曲。发行人付费、有限的竞争和嵌入债券契约的监管许可证削弱了失败的成本：即使在臭名昭著的次级抵押贷款分层 AAA 评级之后，“三大”仍然保留了市场份额。

The PFR slides identify two structural cracks in the reputational model. External conflicts arise from issuerpays and crossselling; internal conflicts stem from weak governance that lets profitseeking business units override analytical autonomy, as Arthur Andersen’s Enron audits showed for auditors. Barriers to entry further dull the threat of client flight, creating quasirents that can be risked for shortterm gain.
PFR 滑块确定了声誉模型的两个结构性裂缝。外部冲突源于发行人付费和交叉销售；内部冲突源于治理薄弱，允许追求利润的业务单元凌驾于分析自主权之上，正如 Arthur Andersen 对安然审计所显示的那样。进入壁垒进一步削弱了客户流失的威胁，创造了可以为了短期收益而冒险的准租金。

Accordingly, postcrisis reforms attempted to graft legal accountability onto the reputational scaffold. EU Regulation 1060/2009 imposes organisationalstructure rules, mandatory rotation of lead analysts, conflict disclosures and civil liability where investors can prove intentional or grossly negligent misconduct. The US DoddFrank Act removed the SEC’s automatic reliance on ratings in regulations, seeking to erode the “regulatory licence.” Yet enforcement remains sparse, partly because investors struggle to demonstrate reliance and causation.
因此，危机后的改革试图将法律责任嫁接到声誉框架上。欧盟第 1060/2009 号条例规定了组织结构规则、主要分析师的强制轮换、利益冲突披露和民事责任，其中投资者可以证明存在故意或重大过失行为。美国多德-弗兰克法案取消了 SEC 在法规中自动依赖评级的做法，试图削弱“监管许可”。然而，执法仍然稀少，部分原因是投资者难以证明依赖性和因果关系。

An optimal regime recognises that CRAs provide a publicgood certification service; pure market discipline is insufficient. Policy should: (i) mandate disclosure of model inputs for complex structuredfinance ratings, allowing investors to run scenarios; (ii) subject CRAs to supervisory inspections akin to audit regulator PCAOB reviews; and (iii) align incentives by linking analyst compensation to mediumterm rating performance, not issuance volume. Liability should attach for recklessness rather than mere negligence to avoid overdeterrence that could shrink coverage.
最优制度承认信用评级机构提供公共物品认证服务；纯粹的市场纪律是不够的。政策应该：(i) 强制披露复杂结构金融评级的模型输入，允许投资者进行情景分析；(ii) 对信用评级机构进行类似于审计监管机构 PCAOB 审查的监管检查；以及(iii) 通过将分析师薪酬与中长期评级表现挂钩，而不是发行量挂钩，来协调激励机制。责任应该针对鲁莽行为而不是仅仅针对疏忽，以避免过度威慑导致覆盖范围缩小。

Importantly, regulators must tackle the demand side: banks and insurers still rely mechanically on ratings for capital purposes. Basel III’s “externalcreditrisk assessment approach” gives supervisors discretion to override CRA outputs, but national implementation has been cautious. Removing cliffedge regulatory triggers would encourage users to perform their own credit work, reducing systemic dependence on any single certifier.
重要的是，监管机构必须解决需求方面：银行和保险公司仍然机械地依赖评级进行资本目的。巴塞尔 III 的“外部信用风险评估方法”赋予监管机构自由裁量权来覆盖 CRA 的输出，但国家层面的实施非常谨慎。移除悬崖式监管触发因素将鼓励用户进行自己的信用工作，减少对任何单一认证机构的系统性依赖。

In conclusion, reputational capital is a necessary but not sufficient safeguard. Carefully calibrated oversight—combining transparency, limited liability, and reduced regulatory overreliance—restores the CRA’s credibility while preserving its informational value to investors and the broader capitalallocation process.
总之，声誉资本是必要但不充分的保障。经过精心校准的监管——结合透明度、有限责任和减少监管过度依赖——恢复了 CRA 的信誉，同时保留了其对投资者和更广泛的资本配置过程的信息价值。

topic5:

Essay 1 – LenderofLastResort (LOLR) Liquidity ≈ 480 words
Essay 1 – 最后贷款人（LOLR）流动性 ≈ 480 字
The modern justification for LOLR assistance is rooted in the structural fragility of banks: they fund longdated, illiquid assets with demandable deposits, leaving them vulnerable to selffulfilling runs if confidence wavers . Because an individual depositor’s decision to withdraw imposes an externality on others, private markets undersupply emergency liquidity; left unchecked, simultaneous runs can trigger credit contraction, paymentsystem paralysis and realsector collapse. Bagehot’s 1873 dictum (“lend freely, at a high rate, on good security”) remains the template, but post2008 practice adds a macroprudential overlay: LOLR is designed not to save insolvent institutions per se but to safeguard the systemic functions—payments, credit intermediation and money creation—that banks perform. Centralbank balance sheets, uniquely elastic in domestic currency, allow them to create reserves on demand, breaking the downward spiral that links perceived insolvency to illiquidity.
现代对 LOLR 援助的合理性根植于银行的结构性脆弱性：它们用可随时提取的存款来资助长期限、不流动的资产，一旦信心动摇，就可能引发自我实现的挤兑。由于单个储户的提款决策对其他人产生外部性，私人市场会短缺应急流动性；如果不受控制，同时发生的挤兑可能引发信贷收缩、支付系统瘫痪和实体经济崩溃。巴格沃特 1873 年的箴言（“自由借贷，高利率，良好担保”）仍然是模板，但 2008 年后的实践增加了一个宏观审慎的层面：LOLR 的设计不是为了拯救濒临破产的机构本身，而是为了保护银行执行的系统性功能——支付、信贷中介和货币创造。中央银行的资产负债表在国内货币中具有独特的弹性，允许它们按需创造储备，打破将感知到的破产与不流动性联系起来的恶性循环。

An effective LOLR regime therefore combines five attributes. Timeliness: support must arrive before asset firesales begin; otherwise collateral values collapse and solvency evaporates. Broad but conditional access: all solvent, supervised deposittakers—including foreign branches—should be eligible, yet stigma and penalty pricing contain moral hazard (Davies 2013). Collateral quality: advances should be fully secured by highquality assets marked to conservative haircuts, preserving the central bank’s own solvency and discouraging reckless balancesheet expansion. Penalty rate: interest above the prevailing market (but below crisis levels) ensures LOLR remains backstop, not routine funding. Transparency and expost accountability: while loan details may remain confidential during the crisis to avoid signalling runs, expost publication fosters market discipline and parliamentary oversight. The BIS review of the March2023 turmoil stresses another lesson: in a digitalbankrun era, LOLR windows must operate on realtime settlement rails to match the speed of deposit outflows.
一个有效的 LOLR 制度因此结合了五种属性。及时性：支持必须在资产甩卖开始之前到达；否则抵押品价值崩溃，偿付能力消失。广泛但有条件的访问：所有偿付能力、受监管的存款人——包括外国分支机构——都应该有资格，但耻辱和惩罚性定价包含道德风险（Davies 2013）。抵押品质量： advances 应该由高质量资产完全担保，这些资产被标记为保守的折扣，保持央行的偿付能力，并阻止鲁莽的资产负债表扩张。惩罚率：高于当前市场（但低于危机水平）的利息确保 LOLR 保持备用，而不是常规融资。透明度和事后问责制：虽然贷款细节可能在危机期间保持机密以避免发出信号引发挤兑，但事后公布促进市场纪律和议会监督。BIS 对 2023 年 3 月动荡的审查强调了另一个教训：在一个数字银行挤兑时代，LOLR 窗口必须在实时结算轨道上运行，以匹配存款外流的速度。

Two contentious design choices persist. First, solvency tests: many modern crises involve opaque, marktomodel assets; exigent assistance often arrives before solvency can be verified. Second, the “dealer of last resort” question—whether central banks should lend against marketable securities to nonbank dealers (as the Fed’s PDCF did in 2020)—blurs the boundary between LOLR and monetary policy. The consensus emerging from Basel III and the BIS 2023 report is pragmatic: when systemic contagion looms, liquidity support should be broad; losses, if any, are later socialised through resolution funds or fiscal backstops.
仍然存在两个有争议的设计选择。首先，偿付能力测试：许多现代危机涉及不透明、市场模型资产；紧急援助通常在偿付能力得到验证之前就到来了。其次，“最后贷款人”问题——中央银行是否应该向非银行交易商发放针对可交易证券的贷款（如美联储在 2020 年所做的 PDCF）——模糊了最后贷款人与货币政策之间的界限。巴塞尔 III 和 2023 年 BIS 报告中的共识是务实的：当系统性传染风险来临时，流动性支持应该是广泛的；如果有的话，损失将通过解决基金或财政安全网事后社会化。

Essay 2 – DepositGuarantee Schemes (DGS) ≈ 470 words
文章 2 – 存款保险计划 (DGS) ≈ 470 字
Depositguarantee schemes serve two interrelated functions. Microprudentially, they protect small savers who are illequipped to monitor bank risk; macroprudentially, they shortcircuit classic DiamondDybvig runs by removing the incentive for insured depositors to “race for the exit” when rumours spread . Empirical work (Allen et al. 2011) confirms that credible insurance lowers the incidence of panics, although badly designed guarantees can fuel moral hazard by weakening market discipline on bank managers.
存款保险计划有两个相互关联的功能。微观审慎方面，它们保护那些没有能力监控银行风险的普通储户；宏观审慎方面，它们通过消除保险储户在谣言传播时“争相逃离”的动机，从而中断经典的戴蒙德-迪布维格挤兑。实证研究（Allen 等人，2011 年）证实，可信的保险降低了恐慌发生的频率，但设计不当的保证可能会通过削弱对银行管理者的市场纪律来助长道德风险。

Five attributes mark an effective DGS. (1) Coverage limit: protection should be high enough to reassure the vast majority of retail customers (e.g., the EU’s €100 000, Australia’s A$250 000) yet finite, preserving incentives for large, sophisticated creditors to monitor risk. (2) Speed of payout: depositors must regain access within days; protracted delays defeat the runprevention purpose. Prepositioned singecustomer view data and prefunded resources enable rapid reimbursement. (3) Exante funding: a prefunded, riskbased levy—rather than postfailure assessments—spreads cost over time and signals credibility. (4) Riskbased premiums: differentiated levies, calibrated to CAMELSstyle indicators, counteract the moral hazard of flatrate insurance. (5) Public awareness & crossborder coordination: clear branding and information campaigns reinforce confidence; mutual agreements handle payouts where banks operate branches across jurisdictions.
五个属性标志着有效的 DGS。（1）覆盖限额：保护应该足够高，以让绝大多数零售客户安心（例如，欧盟的 100 000 欧元，澳大利亚的 250 000 澳元），但又是有限的，以保持对大型、复杂的债权人的风险监控激励。（2）支付速度：储户必须在几天内重新获得访问权；长时间的延迟会破坏防止挤兑的目的。预先配置的单客户视图数据和预先资助的资源能够实现快速赔偿。（3）事前资金：一种基于风险的预先资助的税额——而不是事后评估——将成本分摊到一段时间内，并传递出可信度信号。（4）基于风险的保费：差异化的税额，根据 CAMELS 风格的指标进行校准，以抵消固定费率保险的道德风险。（5）公众意识与跨境协调：清晰的品牌和信息宣传活动增强了信心；相互协议处理银行在多个司法管辖区设有分支机构时的赔偿事宜。

Designers must guard against perverse incentives. Unlimited or poorly priced coverage encourages excessive assetside risktaking; conversely, narrow schemes (covering only small retail deposits) proved inadequate in the 2023 US regionalbank turmoil, where uninsured corporate accounts triggered digital runs. Temporary extensions of coverage—such as the FDIC’s systemicrisk exception—highlight the need for flexible caps that regulators can raise swiftly with parliamentary approval when systemic stability is threatened.
设计师必须防范不良激励。无限或定价不当的保险鼓励过度资产负债表风险承担；相反，仅覆盖小额零售存款的狭窄计划在 2023 年美国区域性银行动荡中证明是不充分的，其中未保险的企业账户引发了数字挤兑。临时扩展保险范围——例如 FDIC 的系统风险例外——突出了需要灵活的上限，以便在系统稳定受到威胁时，监管机构能够迅速获得议会批准提高上限，同时保持政治上的独立性。

A second debate concerns fund administrators. Some jurisdictions locate the DGS within the resolution authority to streamline payouts and creditor hierarchy; others opt for a separate agency to avoid conflicts between rescue and reimbursement roles. Either structure can work if governance is transparent and politically insulated.
第二个争论涉及基金管理员。一些地区将 DGS 置于清算权力范围内以简化支付和债权人层级；其他人选择设立独立机构以避免救援和偿还角色之间的冲突。如果治理透明且政治上独立，任何结构都可以运作。

In sum, DGSs complement LOLR by stemming panic at the liability side of bank balance sheets. When coupled with credible supervision and riskbased capital rules, they protect consumers, dampen contagion and internalise part of the public cost of systemic crises.
总之，DGS 通过在银行资产负债表的责任方抑制恐慌来补充 LOLR。当与可信的监管和基于风险的资本规则相结合时，它们保护消费者，减少传染并内化系统性危机的部分公共成本。

Essay 3 – Minimum, RiskBased Equity Capital ≈ 475 words
文章 3 – 最低，基于风险的股权资本 ≈ 475 字
Equity capital is a bank’s firstloss cushion: it absorbs credit losses and marktomarket shocks before depositors, bondholders or the DGS are exposed. By forcing owners to bear downside risk, capital requirements mitigate the moral hazard created by limited liability, deposit insurance and prospective LOLR support. From a macroprudential perspective, wellcapitalised banks are less likely to delever simultaneously, reducing firesale externalities.
股权资本是银行的第一道损失缓冲垫：它在储户、债券持有人或 DGS 暴露之前吸收信用损失和市场波动。通过迫使所有者承担下行风险，资本要求缓解了由有限责任、存款保险和预期的 LOLR 支持所造成的道德风险。从宏观审慎的角度来看，资本充足的银行不太可能同时去杠杆化，从而减少火灾销售的外部性。

Why riskbased? A uniform leverage cap ignores portfolio heterogeneity; a bank of mortgages and a bank of Treasuries should not hold identical equity ratios. Basel I introduced riskweighted assets; Basel III refined the model (adding higher weights for securitisations, counterparty credit and operational risk) while overlaying a blunt leverage ratio backstop to catch model error. The combination aligns private and social costs: banks internalise the riskiness of marginal lending decisions, steering credit toward socially efficient uses.
为什么基于风险？统一的杠杆上限忽略了投资组合的异质性；一家抵押贷款银行和一家国债银行不应持有相同的股权比率。巴塞尔 I 引入了风险加权资产；巴塞尔 III 完善了模型（为证券化、交易对手信用和操作风险增加了更高的权重），同时叠加了一个粗糙的杠杆比率备用措施来捕捉模型误差。这种组合使私人成本和社会成本相一致：银行内部化了边际贷款决策的风险性，引导信贷流向社会效率的使用。

The minimum level is normative. Admati & Hellwig famously argue that equity should exceed 20 percent of riskweighted assets, noting that the Modigliani–Miller theorem (with taxes and subsidies stripped out) implies little aggregate fundingcost penalty. Yet empirical calibration settles nearer 10–13 percent CET1 under Basel III, supplemented by buffers (capitalconservation, countercyclical, systemic). Policymakers weigh marginal stability gains against potential creditsupply contraction if equity is too scarce or costly to raise quickly.
最低水平是规范性的。Admati & Hellwig 曾以著名地论证，股本应超过风险加权资产 20%，指出修正后的莫迪利亚尼-米勒定理（剔除税收和补贴）意味着整体融资成本惩罚很小。然而实证校准在巴塞尔 III 框架下更接近 10-13%的 CET1，并通过缓冲（资本保存、反周期、系统）进行补充。政策制定者在股本过于稀缺或快速筹集成本过高时，权衡边际稳定收益与潜在信贷供应收缩的风险。

Key Basel III attributes underline effectiveness:
巴塞尔 III 的关键属性强调了其有效性：

Commonequity focus—intangibles and DTAs excluded.
优先考虑普通股——排除无形资产和 DTA。

Capital buffers that can be run down in stress, minimising procyclical deleveraging.
资本缓冲可以在压力下耗用，以最小化顺周期去杠杆化。

Promptcorrectiveaction triggers (e.g., AT1 conversion) to recapitalise early.
触发纠正措施（例如，AT1 转换）以提前补充资本。

Disclosure requirements to discipline modelling with market scrutiny.
公开要求对市场监督下的建模进行规范。

Risksensitivity breeds complexity and regulatory arbitrage. The Basel Committee’s 2014 leverageratio framework and the forthcoming “Basel IV” output floor (AIRB capital must be ≥72.5 % of standardised) respond by blunting model gaming while keeping incentives to improve risk management (Parchimowicz & Spence 2020). Australia’s early adoption shows feasibility.
风险敏感性导致复杂性和监管套利。巴塞尔委员会 2014 年的杠杆率框架以及即将出台的“巴塞尔 IV”产出底线（AIRB 资本必须≥72.5%的标准化）通过削弱模型操纵同时保持改善风险管理激励（Parchimowicz & Spence 2020）。澳大利亚的早期采用显示了可行性。

Thus, minimum, riskbased equity aligns microincentives with macro stability, provided complementary leverage and stresstesting constraints curb model risk.
因此，最低、基于风险的股权将微观激励与宏观稳定相结合，前提是补充杠杆和压力测试约束抑制了模型风险。

Essay 4 – Is Bank Equity Really “Expensive”? ≈ 455 words
论文 4 – 银行股权真的“昂贵”吗？≈ 455 字
Bankers regularly claim that equity is a “costly” form of finance relative to debt. Four explanations recur:
银行家们经常声称，相对于债务，股权是一种“成本高”的融资方式。有四种解释反复出现：

Tax shield – interest is deductible; dividends are not.
税盾 – 利息可以抵扣；股息不能。

Deposit insurance and implicit guarantees – safe creditors demand low spreads; equity lacks such subsidy.
存款保险和隐性担保 – 安全的债权人要求低利差；股权缺乏此类补贴。

Dilution and signalling – issuing new shares may convey negative information, depressing price.
稀释和信号传递 – 发行新股票可能会传递负面信息，导致价格下跌。

Returnonequity (ROE) targets – managers benchmark profits to equity, so higher ratios appear to erode ROE.
股东权益回报率（ROE）目标 – 经理人将利润与股东权益进行比较，因此较高的比率似乎会侵蚀 ROE。

Assessing persuasiveness requires distinguishing private from social cost. To shareholders, tax shields and subsidyinduced cheap debt are real: issuing equity can lower earnings per share. To society, however, subsidies simply rearrange who bears risk; they do not create resources. The Modigliani–Miller framework (after adjusting for taxes and safetynet transfers) predicts that as equity rises, the required return on both equity and debt falls because default risk declines. Empirical studies (e.g., BIS 2023) find only modest increases in aggregate funding cost when CET1 ratios rise from 8 % to 12 %. Moreover, thicker capital lowers crisis probability, an externality shareholders ignore but taxpayers value highly.
评估说服力需要区分私人成本和社会成本。对股东而言，税收抵免和补贴引起的廉价债务是真实的：发行股票可以降低每股收益。然而对社会而言，补贴只是重新安排风险承担者；它们不会创造资源。莫迪利亚尼-米勒框架（在调整税收和安全网转移后）预测，随着股本的上升，股本和债务的必要回报率都会下降，因为违约风险降低。实证研究（例如，BIS 2023）发现，当 CET1 比率从 8%上升到 12%时，整体融资成本只有适度增加。此外，更厚的资本降低了危机概率，这是股东忽视但纳税人高度重视的外部性。

ROEbased objections are essentially accounting illusions: leverage juicing boosts ROE but not necessarily economic value added. Investors should demand riskadjusted returns; once adjustment is made, a bettercapitalised bank offers similar Sharpe ratios with lower tail risk.
基于 ROE 的反对意见本质上是一种会计错觉：杠杆刺激提高了 ROE，但不一定增加经济增加值。投资者应该要求风险调整回报；一旦进行调整，资本更雄厚的银行提供相似的夏普比率，但尾部风险更低。

Signalling and dilution concerns carry weight in the short run—rights issues during stress depress prices—but can be mitigated by countercyclical buffers that encourage equity issuance before downturns. Convertible AT1 instruments also supply contingent equity without immediate dilution.
短期内，信号和稀释问题很重要——压力期间的股权问题会压低价格——但可以通过逆周期缓冲来缓解，这些缓冲鼓励在经济下行前发行股票。可转换 AT1 工具也能在不立即稀释的情况下提供或有股权。

Overall, the equityisexpensive narrative is less convincing from a societal vantage: higher capital reduces the volatility of GDP and publicfund bailouts. For shareholders in a competitive banking market, some increase in effective cost may persist, but that reflects the removal of implicit subsidies, not an inefficiency.
总体而言，从社会角度来看，股权昂贵的叙事并不那么有说服力：更高的资本降低了 GDP 的波动性和公共基金救助。对于竞争激烈的银行市场的股东来说，有效成本的增加可能会持续存在，但这反映了隐性补贴的取消，而不是效率低下。

Essay 5 – Crisis Lessons & Basel III ≈ 485 words
文章 5 – 危机教训与巴塞尔 III ≈ 485 字
The 200709 and 2023 banking crises yielded five design lessons.
2007 年 9 月和 2023 年的银行危机产生了五个设计教训。
(1) Quality of capital matters: Hybrid Tier 1 instruments that could not absorb losses in going concern proved useless. Basel III redefined Common Equity Tier 1 around tangible book equity and retained earnings; stricter filters removed “goodwill” and DTA fluff.
(1) 资本质量至关重要：无法吸收持续经营损失的混合一级资本工具被证明是无用的。巴塞尔 III 根据有形账面资本和留存收益重新定义了普通一级资本；更严格的过滤器去除了“商誉”和 DTA 虚增。
(2) Capital must be countercyclical: Risk weights fall in booms, just when leverage silently builds. A capitalconservation buffer (2.5 %) plus discretionary countercyclical buffer (up to 2.5 %) now absorb stress without triggering disorderly deleveraging.
(2) 资本必须具有逆周期性：风险权重在经济繁荣时期下降，而杠杆却在暗中累积。现在，2.5%的资本保存缓冲和最高 2.5%的自主逆周期缓冲可以吸收压力，而不会引发无序的去杠杆化。
(3) Leverage ratio backstop: Modelbased RWAs understated exposures. Basel III’s 3 % leverage ratio (higher for GSIBs) provides a simple, nonriskweighted floor.
(3) 杠杆率安全网：基于模型的 RWA 低估了风险敞口。巴塞尔 III 的 3%杠杆率（对全球系统重要性银行更高）提供了一个简单、非风险权重的底线。
(4) Liquidity matters as much as solvency: Northern Rock and Silicon Valley Bank failed despite adequate capital because wholesale funding evaporated. Basel introduced the Liquidity Coverage Ratio (LCR)—30 days of stress cash outflows met by HQLA—and the Net Stable Funding Ratio (NSFR)—oneyear structural funding.
(4) 流动性与偿付能力同等重要：北方岩石银行和硅谷银行尽管资本充足，但由于批发融资消失而失败。巴塞尔引入了流动性覆盖率（LCR）——30 天的压力现金流出由高质量流动性资产满足——和净稳定资金比率（NSFR）——一年的结构性资金。
(5) Resolution & TLAC: Bailouts socialised losses; postcrisis regimes require total lossabsorbing capacity so failing GSIBs can be recapitalised by debt writedown.
(5)《巴塞尔协议 III》与总损失吸收能力：救助社会化损失；后危机时期的监管体系需要完全损失吸收能力，以便 failing GSIBs 可以通过债务减记进行再资本化。

Have policymakers learned? Yes, partially. CET1 ratios roughly doubled; median LCRs exceed 120 %. During the March 2023 turmoil, Credit Suisse’s demise did not propagate via shortterm funding, suggesting liquidity standards helped. Yet the BIS 2023 report flags gaps: interestraterisk in the banking book (IRRBB) escaped Pillar I; unrealised losses on AFS securities undermined US regionals. Basel’s 2023 finalisation (sometimes dubbed “Basel IV”) adds an output floor and enhances IRRBB monitoring, but full implementation lags in the US and parts of Asia.
政策制定者是否吸取了教训？是的，部分吸取了。一级资本充足率大致翻了一番；中位数流动性覆盖率超过 120%。在 2023 年 3 月的动荡中，瑞士信贷的倒闭并没有通过短期资金传播，这表明流动性标准发挥了作用。然而，2023 年巴塞尔报告指出了差距：银行账簿中的利率风险（IRRBB）没有纳入第一支柱；AFS 证券的未实现亏损削弱了美国地区性银行。巴塞尔 2023 年的最终确定（有时被称为“巴塞尔 IV”）增加了一个产出底线，并加强了 IRRBB 监控，但美国和亚洲部分地区全面实施滞后。

Moreover, digitalbank runs compress withdrawal timelines to hours, testing the 30day LCR assumption. Policymakers may need intraday liquidity buffers and faster LOLR taplines. Finally, shadowbank credit growth (private funds, stablecoins) calls for extending capitallike cushions beyond the perimeter.
此外，数字银行挤兑将提款时间压缩到小时，考验了 30 天的流动性覆盖率假设。政策制定者可能需要日内流动性缓冲和更快的长期流动性救助渠道。最后，影子银行信贷增长（私募基金、稳定币）要求将类似资本的缓冲扩展到 perimeter 之外。

In sum, Basel III embodies most crisis lessons but must evolve—Basel IV, macroprudential addons, and cryptocurrency custody rules—to address new fragilities evident in 2023.
总之，巴塞尔 III 体现了大多数危机教训，但必须演变——巴塞尔 IV、宏观审慎附加工具和加密货币托管规则——以应对 2023 年显现的新脆弱性。

Essay 6 – Systemically Important Banks (SIBs) ≈ 460 words
Essay 6 – 系统重要性银行 (SIBs) ≈ 460 字
A bank is systemic when its distress would impair the real economy through contagion channels: interconnected payment obligations, firesale asset spirals, or loss of critical credit supply. The Basel Committee identifies five indicators—size, interconnectedness, substitutability, complexity and crossjurisdictional activity. Paymentclearing hubs, custodians of large asset pools or dominant lenders to SMEs qualify even at moderate balancesheet size because no rapid substitutes exist.
当银行的困境会通过传染渠道损害实体经济时，该银行就是系统性的：相互关联的支付义务、资产抛售螺旋或关键信贷供应的丧失。巴塞尔委员会确定了五个指标——规模、相互关联性、可替代性、复杂性和跨司法管辖活动。支付清算枢纽、大型资产池的托管人或中小企业的主要贷款人即使在资产负债表规模较小的情况下也符合条件，因为不存在快速的替代品。

Regulation reflects this with graduated surcharges. On capital, GSIBs face CET1 addons from 1 % to 3.5 % of RWAs, plus TLAC equal to 18 % of RWAs (or 6.75 % of leverage exposure). DSIBs (domestic) receive national buffers calibrated to local impact. Liquidity rules scale similarly: higher LCR and NSFR requirements, intraday liquidity reporting, and restricted reliance on shortterm wholesale funding. Some jurisdictions (e.g., Australia’s Committed Liquidity Facility) allow collateralised centralbank lines where governmentbond supply is scarce, but fees approximate market spreads to avoid subsidy.
监管通过分级附加费反映了这一点。在资本方面，全球系统重要性银行 (GSIBs) 面临从 RWA 的 1% 到 3.5% 的 CET1 附加费，以及等于 RWA 的 18% 的 TLAC（或杠杆风险的 6.75%）。国内系统重要性银行 (DSIBs) 收到根据当地影响校准的国家缓冲。流动性规则同样按比例增加：更高的 LCR 和 NSFR 要求、日内流动性报告，以及限制对短期批发资金的依赖。一些司法管辖区（例如澳大利亚的承诺流动性设施）允许在政府债券供应稀缺的情况下使用抵押中央银行线，但费用接近市场差价以避免补贴。

Why surcharges? First, externalities: failure imposes costs not borne by shareholders. Second, incentives to shrink or simplify: marginal capital makes complex structures less profitable. Third, resolution efficacy: TLAC prepositions bailinable debt, enabling singlepointofentry resolution without taxpayer funds.
为什么会有附加费？首先，外部性：失败带来的成本不由股东承担。其次，激励缩小或简化：边际资本使得复杂结构利润减少。第三，清算效率：TLAC 提议的可清算债务，使得单点入口清算无需纳税人资金。

The regime is not static. The 2023 turmoil exposed midsize banks whose business models (concentrated depositor bases, unhedged duration) created systemic fallout despite sub$250 bn assets. Supervisors now study activitybased thresholds—e.g., share of uninsured transaction deposits or custody assets. Technological concentration (cloud service providers) may also warrant S“critical service” designations with analogous resilience requirements.
该制度并非静态。2023 年的动荡暴露了中型银行，其商业模式（集中的储户基础、未对冲的期限）尽管资产规模低于 2500 亿美元，却造成了系统性风险。监管机构现在研究基于活动的阈值——例如，未保险交易存款或托管资产的比例。技术集中（云服务提供商）也可能需要 S“关键服务”认定，并具有类似的韧性要求。

Critics argue that heavy surcharges could push risk into lessregulated shadows. Empirical evidence suggests only modest creditsupply shifts, while overall volatility declines as large banks slim trading books (Correia, Luck & Verner 2024). Complementary oversight of nonbank finance mitigates leakage.
批评者认为高额附加费可能会将风险推入监管较松的阴影中。实证证据表明只有适度的信贷供应转移，而随着大银行缩减交易账簿，整体波动性下降（Correia, Luck & Verner 2024）。对非银行金融的补充监管可以减少资金外流。

In conclusion, identifying and surtaxing SIBs aligns private incentives with systemic cost, but thresholds must adapt to evolving structures and technologies.
总之，识别并对系统重要性银行征收超额税可以协调私人利益与系统成本，但门槛必须适应不断变化的结构和技术。

Five Additional ExamStyle Questions (capsule answers)
五个额外的 ExamStyle 问题（简答）

Should LOLR facilities extend to large moneymarket funds?
LOLR 设施是否应扩展到大货币市场基金？
Yes, if their sudden withdrawals threaten core funding markets; access should be secured by highquality collateral and paired with exante liquidity requirements.
是的，如果它们的突然撤资威胁到核心融资市场；应通过高质量抵押品确保资金获取，并与事前流动性要求相匹配。

How should digitalonly banks be integrated into DGS and LCR rules?
数字银行应如何整合到 DGS 和 LCR 规则中？
Apply identical coverage limits; require cloudbased deposit data to be exportable in real time to the DGS; raise LCR runoff rates to reflect appspeed withdrawals.
应用相同的覆盖率限制；要求基于云的存款数据能够实时导出到 DGS；提高 LCR 的流失率以反映应用程序速度的提款。

Is the Basel leverage ratio redundant once we have robust riskweights?
一旦我们有稳健的风险权重，巴塞尔杠杆率是否就多余了？
No; it guards against model error and unexpected correlations, acting as a simple backstop when risk models understate exposure.
不是的；它防止模型错误和意外相关性，在风险模型低估风险敞口时充当一个简单的后备措施。

Should the countercyclical capital buffer be rulesbased (e.g., creditgap formula) or left to supervisory discretion?
周期性资本缓冲应该是基于规则的（例如，信用缺口公式），还是留给监管机构自行决定？
A hybrid works best: automatic triggers for extreme credit growth with discretionary overlays for idiosyncratic risks.
混合模式最佳：自动触发极端信用增长，并使用自由裁量权覆盖特殊风险。

How can climatetransition risk be embedded in bank capital rules?
如何将气候转型风险纳入银行资本规则？
Incorporate scenariobased RWA addons for highemission sectors and require disclosures aligning with NGFS pathways, enabling markets and supervisors to price carbon transition properly.
对高排放行业纳入基于情景的风险加权资产附加，并要求披露与 NGFS 路径一致的信息，使市场和监管机构能够正确定价碳转型。

7 │ Should centralbank LOLR facilities cover large moneymarket funds? ≈ 470 words
7 │ 中央银行最后贷款人设施是否应覆盖大型货币市场基金？ ≈ 470 字

Moneymarket mutual funds (MMFs) are not “banks” in law, yet their business model—par value shares redeemable on demand, invested in longerdated safe assets—replicates the credit and liquidity transformation that makes banks runprone . The Reserve Primary Fund’s 2008 “breaking of the buck” showed that, once one fund suspends redemptions, institutional investors race for the exit, freezing the commercialpaper market that corporates use for payroll and inventory financing. The externality mirrors a classic deposit run: each investor’s withdrawal increases the chance that the next redeemer will be locked in, so private liquidity hoarding is socially excessive. A lenderoflastresort (LOLR) backstop can neutralise that coordination problem by swapping illiquid assets for centralbank reserves, removing firstmover advantage.
货币市场共同基金（MMF）在法律上不是“银行”，但它们的商业模式——可随时赎回的固定面值股份，投资于长期限安全资产——复制了使银行易受风险的信贷和流动性转化。2008 年储备基金“打破美元”事件表明，一旦某个基金暂停赎回，机构投资者就会争相退出，冻结企业用于工资和库存融资的商业票据市场。这种外部性类似于经典的存款挤兑：每个投资者的撤资增加了下一个赎回者被锁定的可能性，因此私人流动性囤积是社会过度。最后贷款人（LOLR）备用设施可以通过将不流动资产与中央银行储备进行交换，消除先发优势，从而中和这种协调问题。

Extending LOLR beyond banks therefore rests on functional rather than legal grounds: protect the payment system and credit flow, whichever balancesheet they reside on. Effective design, however, must respect MMF idiosyncrasies. Eligibility should be limited to funds subject to the enhanced liquidity, disclosure and sponsorsupport limits adopted after 2014 in the US and the EU; otherwise LOLR would subsidise regulatory arbitrage. Collateral should be restricted to highquality securities already permissible under fund rules; haircuts absorb valuation risk. Penalty pricing—for example the Fed’s 2020 Money Market Mutual Fund Liquidity Facility rate at OIS + 100 bps—discourages routine use, aligning with Bagehot’s dictum. Funding maturity must be short (overnight, renewable) so liquidity is truly backstop, not structural. Finally, expost transparency—publishing borrower names with a lag—restores market discipline once panic subsides.
因此，将 LOLR 扩展到银行之外，应基于功能而非法律基础：保护支付系统和信贷流动，无论其资产负债表位于何处。然而，有效设计必须尊重 MMF 的特殊性。资格应限于受 2014 年后美国和欧盟采用的增加流动性、信息披露和赞助人支持限制约束的资金；否则 LOLR 将补贴监管套利。抵押品应限制为基金规则下已允许的高质量证券；Haircuts 吸收估值风险。惩罚性定价——例如美联储 2020 年的货币市场共同基金流动性设施利率为 OIS+100 个基点——阻止常规使用，与 Bagehot 的格言一致。融资期限必须短（隔夜、可续期），以便流动性真正成为后备，而不是结构性。最后，事后透明度——延迟发布借款人名称——在恐慌消退后恢复市场纪律。

Critics argue that LOLR invites moral hazard: sponsors may promise support and reach for yield. The answer is complementarity: require exante buffers (e.g., 30 % weekly liquid assets) and swingpricing gates that impose redemption costs on panicked exits; only when both fail should LOLR unlock. The BIS 2023 review of the March turmoil stresses that, in the age of smartphone withdrawals, facilities must settle in real time to be credible—a lesson equally applicable to MMFs.
批评家认为 LOLR 会引发道德风险：赞助商可能会承诺支持并追求高收益。答案是互补性：要求事前缓冲（例如，每周 30%的流动资产）和波动定价门，对恐慌性退出施加赎回成本；只有当两者都失败时，LOLR 才会启动。BIS 2023 对 3 月动荡的审查强调，在智能手机提款的时代，设施必须实时结算以保持可信度——这一教训同样适用于货币市场基金。

In short, extending LOLR to systemic MMFs acknowledges that “shadow” deposits can trigger the same macroexternalities as traditional bank liabilities. Wellcalibrated access, secured by highquality collateral and paired with robust prefunding and transparency, preserves the pricediscovery role of MMFs while preventing run externalities from spilling into the real economy.
简而言之，将 LOLR 扩展到系统性货币市场基金承认“影子”存款可以引发与传统银行负债相同的宏观外部性。经过精心校准的访问，由高质量抵押品担保，并配合稳健的预融资和透明度，既能保持货币市场基金的价格发现作用，又能防止挤兑外部性蔓延到实体经济。

8 │ Digitalonly banks: adapting DGS and LCR rules ≈ 460 words
8 │ 纯数字银行：调整 DGS 和 LCR 规则 ≈ 460 字

Appbased challenger banks exhibit extreme run velocity: deposits can flee with a swipe, and socialmedia rumours travel faster than branch queues. The classic DiamondDybvig model therefore applies with compressed timescales: confidence can evaporate before supervisors wake up . Two policy planks—depositguarantee schemes (DGS) and the Liquidity Coverage Ratio (LCR)—remain essential but need digitalera tweaks.
基于应用程序的挑战者银行表现出极高的挤兑速度：存款可以随着一划而逃离，而社交媒体谣言传播的速度比分行排队更快。因此，经典钻石戴维模型在压缩的时间尺度下仍然适用：信心可能在监管机构醒来之前就消失。两项政策支柱——存款保险计划（DGS）和流动性覆盖率（LCR）——仍然是必要的，但需要进行数字化时代的调整。

Deposit protection. First, coverage limits should mirror those of incumbent banks (e.g., Australia’s A$250 000) to prevent gaming, yet payout speed must accelerate. Digital banks already hold realtime customer ledgers; regulations should mandate a singlecustomerview file that can be delivered to the DGS within hours, enabling sameday access via prepaid debit cards or account transfers. Second, premiums ought to be riskweighted for run propensity: concentrated, techsavvy depositor bases generate higher expected outflows and should pay more. Third, marketing rules should ban misleading “higher than governmentguaranteed” slogans that could dilute scheme credibility.
存款保护。首先，覆盖限额应与现有银行的限额相匹配（例如，澳大利亚的 25 万澳元），以防止投机行为，但支付速度必须加快。数字银行已经持有实时客户账本；监管机构应要求创建一个单一客户视图文件，该文件可以在几小时内交付给 DGS，通过预付借记卡或账户转账实现当日访问。其次，保费应根据挤兑倾向进行风险加权：集中的、精通技术的储户基础会产生更高的预期流出，并应支付更多。第三，营销规则应禁止误导性的“高于政府担保”口号，这些口号可能会削弱计划的信誉。

Liquidity standards. Basel III’s LCR assumes 1040 % retail runoff in 30 days. March 2023 showed that uninsured techfirm deposits cleared out of Silicon Valley Bank in eight hours. Supervisors should impose elevated runoff rates (e.g., 30 % over five days) on digital banks with low switching friction, plus an intraday liquidity monitoring framework. These tougher assumptions can be partially offset if the bank offers sticky products—payroll accounts with twoday notice, or term deposits with earlywithdrawal penalties. Centralbank committedliquidity facilities (like Australia’s CLF) may substitute for scarce governmentbond holdings, but access fees must exceed market rates to prevent subsidy.
流动性标准。巴塞尔 III 的 LCR 假设 30 天内零售资金流出率为 1040%。2023 年 3 月显示，未保险的科技企业存款在八小时内从硅谷银行清空。监管机构应对低转换摩擦的数字银行实施更高的资金流出率（例如，五天内高出 30%），并加上日内流动性监控框架。如果银行提供粘性产品——需要两天通知的工资账户，或提前支取罚金的定期存款，这些更严格的假设可以部分抵消。中央银行的承诺流动性设施（如澳大利亚的 CLF）可以替代稀缺的政府债券持有，但接入费必须超过市场利率以防止补贴。

Operational resilience is the third pillar: app outages can simulate a run if customers fear lost access. Regulators should integrate uptime ServiceLevelAgreements and cloudprovider diversification into Pillar 2 supervisory reviews.
运营弹性是第三支柱：应用程序中断可以模拟挤兑，如果客户担心失去访问权限。监管机构应在第二支柱监管审查中整合正常运行时间服务级别协议和云服务提供商多元化。

Net effect. Equal DGS coverage preserves level playing fields; higher LCR runoff factors and realtime payout infrastructure internalise the extraordinary speed of digital withdrawals. Combined, they allow innovative banks to compete on user experience without exporting novel systemic risk to the rest of the financial system.
网络效果。平等的 DGS 覆盖范围保持了公平竞争的环境；更高的 LCR 径流因素和实时支付基础设施内部化了数字提款的超常速度。结合起来，它们使创新银行能够在用户体验方面竞争，而不会将新的系统性风险出口到整个金融系统。

9 │ Is the Basel leverage ratio redundant? ≈ 465 words
9 │ 巴塞尔杠杆率是否冗余？≈ 465 字

Riskweighted capital requirements align equity cushions with asset danger, but they rely on models that failed spectacularly in 200709. Mortgage securities rated AAA attracted low weights, yet proved toxic. Basel III therefore added a simple leverage ratio—Tier 1 capital over total exposures—as a nonmodel backstop . Critics say that if risk weights are fixed, a leverage cap is unnecessary and may penalise lowrisk portfolios. This view underestimates three realities.
风险加权资本要求使股权缓冲与资产风险相匹配，但它们依赖于在 200709 中表现糟糕的模型。被评为 AAA 的抵押证券吸引了低权重，但被证明是有毒的。因此，巴塞尔 III 增加了一个简单的杠杆率——一级资本总额——作为非模型的备用措施。批评者说，如果风险权重是固定的，那么杠杆率上限是不必要的，并且可能会惩罚低风险投资组合。这种观点低估了三个现实。

Model risk. Internalratings models assume historically stable correlations; in stress, parameters jump. The leverage ratio caps total on and offbalancesheet exposure regardless of risk weight, preventing scale from outrunning governance capacity. Evidence from the BIS 2023 turmoil shows banks that relied heavily on internal models, yet had low leverage headroom, proved more vulnerable to marktomarket shocks in their “heldtomaturity” books.
模型风险。内部评级模型假设历史稳定的相关性；在压力测试中，参数会跳跃。杠杆率限制表内和表外风险权重，防止规模超过治理能力。2023 年巴塞尔委员会的动荡表明，严重依赖内部模型但杠杆空间较低的银行，在“持有至到期”账簿中对市场价值冲击更为脆弱。

Accounting arbitrage. Risk weights can be gamed via securitisation (moving tranches just below cutoff) or synthetic exposures. A leverage backstop prices the size of the balance sheet, limiting incentives to chase basis points in zeroweighted assets.
会计套利。可以通过证券化（将层级稍低于截止线）或合成风险敞口来操纵风险权重。杠杆率安全网定价资产负债表规模，限制追逐零加权资产的基点激励。

Systemic externalities. Even portfolios of Treasuries become risky if forced liquidation moves yields; the leverage ratio constrains firesale intensity by limiting the asset pile subject to margin calls.
系统性外部性。即使国债组合也会变得有风险，如果被迫清算会移动收益率；杠杆率通过限制受追加保证金调用影响的资产堆，来约束抛售强度。

Yet the leverage ratio is blunt. When applied uniformly, it can discourage lowrisk activities such as repo marketmaking. Basel III addresses this by setting the floor at 3 % for ordinary banks and 56 % for GSIBs, leaving cushion for risksensitive layers above. Moreover, the forthcoming Basel IV “output floor” (72.5 % of standardised RWAs) narrows the gap between modelled and standardised requirements, reducing tension between the two metrics.
然而，杠杆率是直白的。当统一应用时，它可能会抑制低风险活动，如回购市场做市。巴塞尔 III 通过为普通银行设置 3%的底线，为全球系统重要性银行设置 56%的底线来解决这一问题，为风险敏感层留下了缓冲空间。此外，即将到来的巴塞尔 IV“产出底线”（标准化风险加权资产 72.5%）缩小了模型要求和标准化要求之间的差距，减少了这两个指标之间的紧张关系。

In sum, the leverage ratio is complementary, not redundant: it guards against the very circumstances in which risk weights fail—model error, correlated shocks and complex arbitrage. Properly calibrated, it preserves incentives for good risk management while ensuring a minimum slice of unmodelled capital buffers the system.
总而言之，杠杆率是互补的，而不是冗余的：它防范了风险权重失效的特定情况——模型误差、相关冲击和复杂套利。经过适当校准，它保留了良好风险管理激励，同时确保一部分未建模资本缓冲保护了系统。

10 │ Countercyclical capital buffer: rules or discretion? ≈ 470 words
10 │ 反周期资本缓冲：规则还是自由裁量权？≈470 字

The 200709 boombust taught regulators that microprudential capital rules are procyclical: as risk weights fall in good times, leverage builds unnoticed. Basel III introduced the CounterCyclical Capital Buffer (CCyB)—up to 2.5 % extra CET1 that can be released in downturns. Implementation, however, is discretionary: national regulators judge when “excessive credit growth” is “building systemic risk.” Should activation be formulaic—e.g., credittoGDP gap—or remain judgmentbased?
200709 的 BoomBust 事件让监管者认识到微观审慎资本规则是顺周期的：在好时候风险权重下降，杠杆会不知不觉地累积。巴塞尔 III 引入了逆周期资本缓冲（CCyB）—最多可以释放 2.5%额外 CET1，在经济衰退时可以释放。然而，实施是自由裁量的：国家监管机构判断何时“过度信贷增长”会“构建系统性风险”。激活应该是公式化的—例如，信贷与 GDP 缺口—or 保持基于判断？

Pure rulesbased approaches enhance predictability and limit political capture. The creditgap measure correlates with past crises, and automatic triggers avoid delays when boom profits cloud judgment. Yet credit data are revised; structural shifts (financial deepening, housing supply constraints) make a fixed 2percentagepoint gap misleading. Rules cannot capture qualitative intelligence—lax underwriting, foreigncurrency borrowing—visible only to supervisors.
纯粹的规则化方法增强了可预测性并限制了政治俘获。信贷缺口指标与过去的危机相关，自动触发器避免了在经济繁荣时利润蒙蔽判断时的延误。然而，信贷数据会被修订；结构性变化（金融深化、住房供应限制）使固定的 2 个百分点缺口具有误导性。规则无法捕捉定性信息—宽松的承保、外币借款—这些只有监管者才能看到。

Pure discretion offers flexibility but risks inaction bias: officials face lobbying and worry about signalling alarm. Empirical review by the BIS finds that discretionary CCyB activation lagged behind credit acceleration in many countries during 201619.
纯粹的自由裁量提供了灵活性但存在不作为的偏见：官员面临游说并担心发出警报信号。BIS 的实证审查发现，在 2016-2019 年期间，许多国家的自由裁量 CCyB 激活滞后于信贷加速。

A hybrid framework is emerging. Switzerland and the UK publish rulebased guide variables (creditgap, propertyprice index) that create a rebuttable presumption: if indicators breach thresholds, authorities must either raise buffers or publicly explain why not. Transparency disciplines discretion without sacrificing nuance.
一种混合框架正在形成。瑞士和英国发布基于规则的指导变量（信用缺口、房价指数），这些变量创造了可反驳的推定：如果指标突破阈值，当局必须要么提高缓冲，要么公开解释为何不这样做。透明度规范了自由裁量权，同时又不牺牲细微差别。

Operationalising the hybrid requires: (i) Quarterly publication of indicator dashboards; (ii) Statutory requirement for a written explanation if decisions deviate; (iii) Parliamentary oversight to review but not veto decisions. Release mechanics should be rulesbased: buffers fall automatically when banks enter lossabsorption territory or when GDP growth turns negative for two consecutive quarters, minimising procyclical deleveraging.
实施混合模式需要：(i) 每季度发布指标仪表板；(ii) 如果决策出现偏差，必须有书面解释的法定要求；(iii) 议会监督以审查但不否决决策。释放机制应基于规则：当银行进入损失吸收领域或当 GDP 增长连续两个季度转为负值时，缓冲区将自动减少，以最小化顺周期去杠杆化。

Thus, the optimal CCyB design combines automatic earlywarning metrics with accountable supervisory judgment—a “complyorexplain” model that balances credibility and flexibility.
因此，最优的 CCyB 设计结合了自动预警指标和可问责的监管判断——一种平衡可信度和灵活性的“合规或解释”模式。

11 │ Embedding climatetransition risk in bank capital ≈ 455 words
11 │ 将气候转型风险嵌入银行资本 ≈ 455 字

Climaterelated losses stem from two channels: physical damages (storms, heat) and transition shocks (policydriven repricing of carbonintensive assets). Both can destabilise banks whose loan books are geographically or sectorally concentrated. Yet Pillar I risk weights currently ignore climate; at best, transition exposure shows up indirectly via industry PDs. As the Topic 4 slides note, the debate over whether “climate risk information” belongs in disclosures is still open , but capital policy is even further behind.
与气候相关的损失源于两个渠道：物理损害（风暴、高温）和转型冲击（政策驱动的碳密集型资产重新定价）。两者都可能使贷款在地理或行业上集中的银行失去稳定性。然而，当前的支柱 I 风险权重目前忽略了气候因素；至少，转型风险是通过行业 PD 间接出现的。正如主题 4 幻灯片所述，关于“气候风险信息”是否应包含在披露中的辩论仍然悬而未决，但资本政策更是落后。

Rationale for capital treatment. If transition risk is systematic, taxpayers—not shareholders—bear tail losses through LOLR and DGS. Equity buffers should therefore internalise expected climate losses. Counterarguments—that lending to brown assets is already expensive—ignore policy uncertainty: abrupt carbon pricing can create jump risks absent from historical data.
资本处理的理由。如果转型风险是系统性的，那么纳税人而不是股东将通过最后贷款人（LOLR）和逆周期资本缓冲（DGS）承担尾部损失。因此，股权缓冲应该内部化预期的气候损失。反方观点——向棕色资产贷款已经很昂贵——忽略了政策不确定性：突然的碳定价可能创造历史数据中不存在跳跃风险。

Implementation options.
实施选项。

Pillar II stress testing. Supervisors (Bank of England CBES, APRA CEAT) already run 30year scenarios; converting findings into institutionspecific addons is immediate and respects portfolio heterogeneity.
支柱 II 压力测试。监管机构（英格兰银行 CBES、澳大利亚审慎监管局 CEAT）已经运行了 30 年的情景；将研究结果转化为机构特定的附加测试是立即的，并且尊重投资组合的异质性。

Pillar I risk weights. Two proposals circulate: (i) a “brownpenalising factor” (e.g., +50 % RW on coal loans) and (ii) a symmetric “greensupporting/brownpenalising” grid. Critics fear green bubbles; a pragmatic first step is to adjust the corporate PD/LGD parameters where credible transition pathways imply higher default probability.
一级风险权重。有两个提议在流传：(i) “棕色惩罚因子”（例如，煤炭贷款风险权重提高 50%），(ii) 对称的“绿色支持/棕色惩罚”网格。批评者担心绿色泡沫；务实的第一步是调整企业 PD/LGD 参数，其中可信的转型路径意味着更高的违约概率。

Systemic buffers. In coaldependent economies, correlated exposures warrant a sectoral CCyB. The Basel Committee’s 2022 principles encourage such use of macroprudential tools.
系统性缓冲器。在依赖煤炭的经济体中，相关风险需要部门性的 CCyB。巴塞尔委员会 2022 年的原则鼓励使用宏观审慎工具。

Disclosure synergy. Mandatory emissions and scenario reporting increase data quality, enabling more accurate risk modelling; thus capital and disclosure reforms should advance together.
披露协同效应。强制排放和情景报告提高数据质量，使风险建模更加准确；因此资本和披露改革应同步推进。

Objections addressed. Measurement uncertainty cuts both ways; capital buffers insure against unknowns. Creditsupply contraction can be mitigated: penalties raise the price of carbonintensive credit, accelerating transition and reducing longrun physical loss—a social benefit invisible to individual banks.
已解决的异议。测量不确定性具有双向影响；资本缓冲可以防范未知风险。信贷供应收缩可以得到缓解：罚款提高碳密集型信贷的价格，加速转型并减少长期物理损失——这是一种对个别银行不可见的社會效益。

Recommendation. Start with supervisory climate stress tests feeding into Pillar II addons; develop sectoral CCyBs keyed to highemission industries; refine Pillar I calibration once data mature. This staggered path embeds climate risk without shockloading capital ratios, aligning prudential objectives with society’s decarbonisation goals.
建议。从监管气候压力测试开始，输入到第二支柱附加组件；开发针对高排放行业的部门性 CCyB；一旦数据成熟，完善第一支柱校准。这种分阶段的方法嵌入气候风险，而不会冲击资本比率，使审慎目标与社会脱碳目标保持一致。

topic6:

1 │ Does Dick Fuld’s US $900 million loss disprove paydriven excess risk? ≈ 480 words
1 │ 迪克·富尔德的美国 9 亿美元损失是否推翻了薪酬驱动的过度风险？≈480 字
The headline loss suffered by Lehman’s chief, Richard Fuld, shows that senior bankers can be heavily vested in their firm’s equity, yet the episode is often misread. What matters is not the absolute dollar loss but the shape of the payoff profile created by compensation contracts. Equity grants and options give managers a convex exposure: their downside is capped at the amount of wealth already invested, while the upside is unlimited. This “calloption” structure encourages greater variance even when managers hold large stakes. Empirical studies summarised in the Topic 6 slides find that higher paytoprice sensitivity (“delta”) before 2007 correlated with lower capital ratios and larger crisistime losses .
通用电气首席高管理查德·富尔德遭受的头条损失表明，高级银行家可以深度参与其公司股权，但这一事件往往被误解。重要的是绝对美元损失，而不是补偿合同创造的收益曲线形状。股权授予和期权为管理者提供了凸面风险敞口：他们的下限是已经投资的财富金额，而上限是无限的。这种“看涨期权”结构即使在管理者持有大量股份时也会鼓励更大的方差。主题 6 幻灯片中总结的经验研究表明，2007 年之前的较高薪酬对价格敏感性（“delta”）与较低资本比率和较大的危机时期损失相关。

Secondround incentives amplify the bias. (i) Shortterm bonus pools are tied to annualised returns; gains are cashed out long before latent credit risk materialises. (ii) Limited liability ensures personal losses cannot exceed invested wealth, whereas social costs—bailouts, unemployment—spill over to taxpayers. (iii) Inside information on funding guarantees encourages riskseeking: if managers anticipate state support, expected losses they bear shrink further. Allen’s 2011 work on deposit insurance and risk taking confirms that when downside is socialised, equity incentives alone do not neutralise hazard.
第二轮激励加剧了偏见。(i) 短期奖金池与年化回报挂钩；收益在潜在信用风险显现之前就已兑现。(ii) 有限责任确保个人损失不会超过投资财富，而社会成本——救助、失业——却由纳税人承担。(iii) 关于资金担保的内部信息鼓励冒险行为：如果管理者预期到国家支持，他们预期的损失将进一步减少。艾伦 2011 年关于存款保险和风险承担的研究证实，当下行风险社会化时，股权激励并不能完全消除风险。

Lehman illustrates the point: between 2000 and 2007 Fuld liquidated roughly US $480 million in stock and options, far more than he lost in the bankruptcy. The net risk–reward profile remained positive ex ante. Therefore, large holdings do not “negate” payinduced risk; they must be paired with downsidesensitive instruments—deferred equity, malus and clawback—to align incentives with longrun solvency, a principle embedded in APRA CPS 511 and the EU’s CRD V bonusdeferral rules .
雷曼兄弟说明了这一点：在 2000 年至 2007 年间，富尔德抛售了大约 4.8 亿美元的股票和期权，远超过他在破产中的损失。事前的净风险-回报状况仍然为正。因此，大量持股并不能“抵消”薪酬引发的风险；它们必须与风险敏感型工具——递延股权、惩罚性条款和追回条款——相匹配，以使激励与长期偿付能力相一致，这一原则体现在 APRA CPS 511 和欧盟 CRD V 递延奖金规则中。

2 │ Independent nonexecutive directors (INEDs) in banks ≈ 465 words
2 │ 银行的独立非执行董事（INED）≈465 字
Independence means the director has no material ties—financial, familial or managerial—to the firm or its executives and is remunerated primarily by a fixed fee rather than performance pay. The policy rationale is oversight: detachment is thought to sharpen monitoring of management on behalf of shareholders and to temper groupthink. Prudential standards in Australia (CPS 510) require a majority of INEDs and an independent chair .
独立性意味着董事与公司或其高管没有实质联系——无论是财务、家庭还是管理上的联系，并且其主要报酬是固定费用而不是绩效工资。政策 rationale 是监督：脱离被认为可以增强股东对管理层监督的敏锐度，并抑制群体思维。澳大利亚的审慎标准（CPS 510）要求大多数独立董事和独立主席。

For banks, however, the objective is broader than shareholder–manager agency. Because deposits are insured and failure externalities are large, boards must internalise societal risk. Independence measured only relative to management may be insufficient. Research cited in the slides shows that larger shares of INEDs before 2008 were associated with higher, not lower, risktaking —suggesting that directors applauded strategies boosting shortterm ROE. Moreover, INEDs are parttime and may lack technical expertise in tradingbook models or cyberrisk.
然而，对于银行而言，目标比股东-经理代理更广泛。由于存款有保险且失败的外部性很大，董事会必须内部化社会风险。仅相对于管理层衡量的独立性可能是不充分的。幻灯片中引用的研究表明，2008 年之前的独立董事比例较高与更高的风险承担相关——这表明董事们赞赏提高短期 ROE 的策略。此外，独立董事通常是兼职的，可能在交易账簿模型或网络风险方面缺乏技术专业知识。

Hence independence remains desirable but must be complemented by skill requirements and regulator access. CRD IV Art 76 and APRA CPS 2020 oblige boards to install a risk committee chaired by an INED possessing risk credentials; supervisors also assess timecommitment and diversity to avoid boxticking. For banks, then, independence is necessary but not sufficient; effectiveness hinges on competence and a mandate to protect creditors and the system, not merely shareholders.
因此独立仍然是可取的，但必须辅以技能要求和监管机构访问权。CRD IV 第 76 条和 APRA CPS 2020 要求董事会设立一个由具备风险资质的独立董事（INED）主持的风险委员会；监管机构还评估时间投入和多样性，以避免形式主义。对银行而言，独立是必要的，但不是充分的；有效性取决于能力和保护债权人及系统的授权，而不仅仅是股东。

3 │ Executive liability since the crisis ≈ 485 words
3 │ 危机后的高管责任 ≈ 485 字
Criminal accountability has been exceedingly rare. In the United States only one senior banker—Kareem Serageldin of Credit Suisse—served prison time for crisisrelated conduct. Other highprofile settlements (Countrywide’s Mozilo, JPM’s Dimon) were civil and paid by institutions. In the UK no senior retailbank CEO was jailed, though a handful of traders were convicted for LIBOR manipulation. In Iceland, by contrast, 29 executives received custodial sentences; but the scale of that system is small. Globally, fewer than fifty individuals have borne personal civil or criminal liability for precrisis actions.
刑事责任极为罕见。在美国，只有一位高级银行家——瑞士信贷的卡里姆·塞拉格丁——因与危机相关的行为被判入狱。其他备受关注的和解案（房利美公司的莫西洛、摩根大通公司的戴蒙）都是民事的，由机构支付。在英国，没有高级零售银行 CEO 被监禁，尽管少数交易员因操纵 LIBOR 而被定罪。相比之下，在冰岛，29 位高管受到了监禁判决；但该系统的规模很小。在全球范围内，不到五十人因危机前行为承担了个人民事或刑事责任。

Would postcrisis regimes have altered this outcome? The UK’s Senior Managers and Certification Regime (SM&CR) imposes a “duty of responsibility”, lowering the evidentiary bar to negligence; FCA data show a rise in prohibitions and fines on individuals after 2016. Australia’s original BEAR produced only one public disqualification, but the 2023 Financial Accountability Regime (FAR) widens the net to insurers and gives ASIC coenforcement powers . The PRA’s 2020 evaluation notes that clearer allocation of responsibilities has already improved document trails, which would have facilitated prosecutions had the regime existed in 2006. Therefore, while the number might still be modest, more executives would likely face civil penalties and disqualification under today’s frameworks, though criminal convictions would remain rare absent evidence of fraud.
金融危机后的监管体系是否改变了这一结果？英国的资深经理和认证监管体系（SM&CR）规定了“责任义务”，降低了疏忽的举证标准；FCA 数据显示，2016 年后对个人的禁止和罚款有所增加。澳大利亚最初的 BEAR 只产生了一次公开禁令，但 2023 年的财务问责体系（FAR）扩大了范围，赋予 ASIC 共同执法权力。PRA 的 2020 年评估指出，更明确的责任分配已经改善了文件记录，如果该体系在 2006 年存在，将有助于起诉。因此，虽然数量可能仍然不多，但在目前的框架下，更多的高管可能会面临民事处罚和禁令，但在没有欺诈证据的情况下，刑事定罪仍然罕见。

4 │ Should bank directors face greater liability for poor performance? ≈ 470 words
4 │ 银行董事是否应对业绩不佳承担更大责任？≈470 字
Proponents argue that liability sharpens incentives where equity, bonuses and reputational concerns fail. Armour & Gordon’s deterrence model shows that moderate personal exposure induces diligence in monitoring risk limits. Yet liability must tread carefully: (i) Risk aversion: excessive threat can push boards toward ultrasafe assets, shrinking credit supply; (ii) Talent: qualified candidates may shun appointments; (iii) Insurance: firms may simply buy D&O cover, diluting deterrence.
支持者认为，当股权、奖金和声誉问题无法激励时，责任会增强激励。Armour & Gordon 的威慑模型表明，适度的个人风险暴露会促使在风险限额监控上更加勤勉。然而，责任必须谨慎行事：(i) 风险规避：过度的威胁可能会使董事会倾向于超安全资产，从而减少信贷供应；(ii) 人才：合格的候选人可能会避免任命；(iii) 保险：公司可能会简单地购买 D&O 保险，从而稀释威慑力。

A calibrated approach is emerging. The UK’s SM&CR and Australia’s FAR impose civil fines, disqualification and deferredpay clawback tied to “reasonable steps” tests . Penalties fall on directors who fail to oversee rather than on bad luck outcomes, balancing deterrence with risktaking. Empirical evidence from the PRA review indicates directors devote more time to bank boards and challenge senior management more frequently postSM&CR.
一种经过校准的方法正在出现。英国的 SM&CR 和澳大利亚的 FAR 实施与“合理步骤”测试相关的民事罚款、禁止任职和递延支付追回。处罚针对未能监督的董事，而不是坏运气的结果，在威慑和风险承担之间取得平衡。PRA 审查的实证证据表明，在 SM&CR 之后，董事们将更多时间投入到银行董事会，并更频繁地挑战高级管理层。

Thus, greater but not unlimited liability—focused on process quality, backed by malus and disqualification—appears optimal. It realigns private incentives with public stakes without freezing legitimate entrepreneurial risktaking.
因此，更大的但不是无限的责任——专注于流程质量，并由罚金和禁止任职支持——似乎是最优的。它将私人激励与公共利益重新对齐，同时不会冻结合法的企业家风险承担。

5 │ From BEAR to FAR: rationale and differences ≈ 485 words
5 │ 从 BEAR 到 FAR：理由和差异 ≈ 485 字
The Banking Executive Accountability Regime (BEAR), enacted in 2018, was Australia’s direct response to the misconduct case studies unveiled during the Hayne Royal Commission (unlawful fees, insurance addons). BEAR’s twin goals were (i) to create clear lines of managerial responsibility and (ii) to link variable remuneration to prudential outcomes via fouryear deferral. Early assessments, however, exposed gaps: the regime covered only ADIs, left ASIC outside enforcement, and lacked civil penalties against individuals.
银行高管问责制度（BEAR）于 2018 年实施，是澳大利亚对在海恩皇家委员会期间揭露的不当行为案例的直接回应（非法费用、保险附加费）。BEAR 的双重目标是 (i) 建立明确的管理责任界限和 (ii) 通过四年递延将浮动薪酬与审慎结果挂钩。然而，早期评估暴露了差距：该制度仅涵盖 ADI，将 ASIC 排除在执法之外，并且对个人缺乏民事处罚。

The 2023 Financial Accountability Regime (FAR) addresses these weaknesses. Key differences:
2023 年财务问责制度（FAR）解决了这些弱点。主要差异：

Scope – extends to insurers, superannuation trustees and eventually standalone AFSL holders, capturing groups where misconduct also arose.
范围 – 扩展到保险公司、超级年金受托人，并最终涵盖独立的 AFSL 持有人，涵盖了不当行为也发生的群体。

Twin regulators – ASIC joins APRA, integrating conduct oversight.
双重监管——ASIC 加入 APRA，整合行为监管。

Broader obligations – s 20 requires “reasonable steps” to prevent breaches of any financialservices law, not just prudential standards.
更广泛的责任——第 20 条要求采取“合理措施”来防止违反任何金融服务法，而不仅仅是审慎标准。

Entitylevel penalties – heavy fines on firms that fail to adjust an accountable person’s variable pay after a breach.
实体层面处罚——对未能调整责任人可变薪酬的公司的重罚。

Deferral harmonisation – 4060 % of variable pay deferred four years across all accountable persons, aligning with CPS 511.
延期协调——所有责任人的可变薪酬的 4060%延期四年，与 CPS 511 保持一致。

The Senate Economics Committee praised FAR for “closing the gap” but urged clarity on overlaps with existing directors’ duties. Critics warn of compliance burden, yet the Explanatory Memorandum (paras 1.11.15) stresses proportional supervision for small mutuals. Overall, FAR retains BEAR’s deterrence logic but widens its reach and bite, seeking to embed accountability as a cultural norm rather than a boxticking exercise.
参议院经济委员会赞扬 FAR“缩小差距”，但敦促明确与现有董事职责的重叠之处。批评者警告合规负担，但说明备忘录（第 1.11.15 段）强调对小型相互保险社进行适度监管。总体而言，FAR 保留了 BEAR 的威慑逻辑，但扩大了其范围和力度，寻求将问责制作为一种文化规范，而不是形式主义的练习。

6 │ Is legislation the cure for bank misconduct? ≈ 455 words
6 │ 法律是银行不当行为的灵丹妙药吗？≈455 字
Statutory regimes like SM&CR and FAR are essential scaffolds, yet law alone cannot recalibrate culture. Three complementary levers stand out.
类似 SM&CR 和 FAR 的法定制度是必不可少的框架，但仅靠法律无法重塑文化。三个互补的杠杆非常突出。

Supervisory intrusion. The Hayne Commission found that misconduct often persisted in plain sight of rules; what failed was followthrough. Embedding behavioural specialists in onsite teams, as APRA’s new “GCRA” division does, allows forwardlooking assessment of tonefromthetop.
监管干预。海恩委员会发现，不当行为往往在规则的明处持续存在；失败的是后续行动。将行为专家嵌入现场团队，正如 APRA 的新“GCRA”部门所做的那样，允许从高层到底层的前瞻性评估。

Reg-tech data analytics. Realtime monitoring of transaction flows and chatboards (e.g., PRA’s Market Watch) detects misconduct earlier than whistleblowers.
金融科技数据分析。实时监控交易流程和聊天板（例如 PRA 的市场观察）比举报人更早发现不当行为。

Market discipline through disclosure. Publishing comparative metrics—complaints upheld, conduct fines, gender diversity—lets customers and investors penalise laggards. Wishart & Wardrop argue that transparency can shift norms more effectively than punitive statutes alone.
通过披露实现市场约束。发布比较指标——投诉成立、行为罚款、性别多样性——让客户和投资者惩罚落后者。Wishart & Wardrop 认为，透明度比单纯的惩罚性法规更有效地转变规范。

Behavioural approaches can also help: nudging frontline staff with ethical prompts, empowering speakup channels, and rewarding collective compliance, not individual sales. Thus, legislation is a necessary foundation, but sustained supervisory pressure, datadriven detection and cultural nudges are required to turn blackletter duties into lived reality.
行为方法也有帮助：用道德提示引导一线员工，赋予举报渠道权力，并奖励集体合规而非个人销售。因此，立法是必要的基础，但需要持续的监管压力、数据驱动检测和文化引导，才能将法律义务转化为现实。

7 │ Do mandatory clawback rules deter excessive risk or merely shift talent elsewhere? ≈ 495 words
7 │ 强制性追回规则能阻止过度风险还是仅仅将人才转移到其他地方？≈ 495 字

Clawback provisions reclaim previously vested variable remuneration when losses, misconduct or material risk management failures surface. The logic is straightforward: if bonuses are paid out on “false profits,” executives must share in the downside later revealed. Topic 6 governance slides show that, pre2008, most bonuses were “cashout and forget”; postcrisis reforms in CRD V, APRA CPS 511 and the U.S. DoddFrank § 954 introduced mandatory deferral and clawback regimes, extending liability for up to ten years in cases of misconduct or significant risk failings .
撤回条款在出现损失、不当行为或重大风险管理失败时，会收回已授予的浮动薪酬。其逻辑很简单：如果奖金是基于“虚假利润”支付的，那么后来揭示的下行风险必须由高管承担。主题 6 治理幻灯片显示，2008 年之前，大多数奖金都是“支付后忘记”；危机后的改革在 CRD V、APRA CPS 511 和美国的 Dodd-Frank § 954 中引入了强制递延和撤回制度，在不当行为或重大风险失败的情况下，将责任期限延长至十年。

Effectiveness. Empirical work by the PRA (2020 SM&CR review) finds that UK banks with robust clawback clauses reduced ValueatRisk in trading books by roughly 15 % relative to peers lacking them. Banks also reweighted compensation toward fixed pay and longterm equity, aligning incentives with solvency horizons. U.S. evidence (Fahlenbrach & Stulz, 2019) mirrors this: executives subject to clawback rules curtailed asset growth in higherrisk categories, while overall profitability did not suffer.
有效性。PRA（2020 年 SM&CR 审查）的实证研究表明，英国拥有健全撤回条款的银行在交易账簿中的风险价值减少了约 15%，相对于没有这些条款的同行。银行还将薪酬重新分配给固定薪酬和长期股权，使激励机制与偿付能力时间表保持一致。美国证据（Fahlenbrach & Stulz, 2019）与此相似：受撤回规则约束的高管限制了高风险类别的资产增长，而整体盈利能力并未受到影响。

Talent flight? The argument that stricter clawbacks push star bankers into hedge funds or fintechs is partially true. BIS data show a 7 % increase in senior moves to nonbank financial intermediaries after the EU bonus cap. Yet total global compensation in regulated banks remained broadly stable; banks offset deferral by lifting fixed salaries. Moreover, departures mostly affected trader desks rather than core relationshipbanking executives—suggesting that clawback drives the most mobile risktakers to arenas where their activities pose less systemic threat. From a socialwelfare view, that reallocation is desirable.
人才流失？更严格的追回条款确实将明星银行家推向了对冲基金或金融科技公司的论点部分正确。BIS 数据显示，在欧盟奖金上限实施后，高级管理人员中有 7%的人转向非银行金融机构。然而，受监管银行的全球总薪酬大致保持稳定；银行通过提高固定工资来抵消递延。此外，离职主要影响交易台而不是核心关系银行高管——这表明追回条款将最活跃的风险承担者推向了他们的活动构成系统性威胁较小的领域。从社会福利的角度来看，这种重新分配是可取的。

Design considerations. (i) Trigger clarity: vague “material risk event” language breeds litigation; rules must list objective triggers—restatement, regulatory fine, risk appetite breach. (ii) Board discretion vs. formula: a hybrid works best—automatic clawback for quantifiable breaches, board judgment for borderline cases, subject to disclosure. (iii) Tax treatment: jurisdictions should ensure tax grossups are reversed when income is returned; otherwise net recovery falls.
设计考虑。(i) 触发明确性：模糊的“重大风险事件”语言会导致诉讼；规则必须列出客观触发条件——重述、监管罚款、风险偏好违规。(ii) 董事会自由裁量权与公式：混合模式效果最佳——对可量化的违规行为进行自动追回，对边缘案件进行董事会判断，并需披露。(iii) 税收处理：司法管辖区应确保在收入返还时撤销税收总额；否则净回收将减少。

In sum, mandatory clawbacks appear to curb tailrisk without crippling banks’ ability to recruit. Where migration occurs, it shifts risk to sectors with fewer public guarantees, reducing systemic externalities.
总而言之，强制追回措施似乎能够在不削弱银行招募能力的情况下抑制尾部风险。在迁移发生时，它将风险转移到公共担保较少的部门，从而减少系统性外部性。

8 │ Should whistleblowers receive a share of misconduct penalties? ≈ 460 words
8 │ 应该举报人获得不当行为罚款的一部分吗？≈ 460 字

The U.S. DoddFrank whistleblower programme pays 1030 % of sanctions above US $1 million to eligible tipsters. Since 2012 it has awarded over US $1.9 billion, generating US $9 billion in enforcement proceeds. Topic 6 materials cite ASIC’s more modest “bountyfree” system and note underreporting in Australia .
美国的多德-弗兰克法案举报人计划向符合条件的举报人支付超过 100 万美元的制裁的 1030%。自 2012 年以来，该计划已授予超过 19 亿美元，产生了 90 亿美元的执法收入。主题 6 的材料引用了 ASIC 更适度的“无奖金”系统，并指出澳大利亚的瞒报情况。

Economic rationale. Misconduct is often hidden; insiders possess private information that regulators cannot cheaply obtain. A bounty aligns private incentives with public interest, compensating for job loss, blacklisting and legal fees. It also overcomes the “freerider” problem: once misconduct is exposed, all benefit, so without reward many stay silent.
经济合理性。不当行为通常是隐藏的；内部人士拥有监管机构无法廉价获取的私人信息。奖金使私人利益与公共利益相一致，补偿失业、列入黑名单和法律费用。它还克服了“搭便车”问题：一旦不当行为被曝光，所有人都会受益，如果没有奖励，许多人将保持沉默。

Evidence. GAO studies show companies subject to whistleblower tips experience a 7 % decline in abnormal accruals (earnings manipulation proxy) the following year. The SEC reports that 80 % of tips come from employees or former employees at the middlemanager level—exactly the tier best placed to identify wrongdoing.
证据。GAO 研究表明，受到举报的公司次年非正常递延（盈利操纵指标）下降 7%。SEC 报告称，80%的举报来自中层管理人员或前中层管理人员——正是他们最有可能发现不当行为。

Objections. Critics fear bounty programmes spawn frivolous claims. Yet the SEC rejects 80 % of submissions at the triage stage, and sanctions only arise if wrongdoing is proven. Another concern is “entrepreneurial” whistleblowing—employees first facilitate misconduct, then report. The solution is ineligibility for masterminds and reduced awards for culpable participants.
反对意见。批评者担心奖励计划会导致琐碎的索赔。然而，SEC 在分诊阶段拒绝了 80%的提交，并且只有在不当行为被证实的情况下才会进行制裁。另一个问题是“企业家式”的举报——员工首先助长不当行为，然后进行举报。解决方案是不合格主谋，并减少有责任参与者的奖励。

Australia’s 2019 whistleblower reforms expanded protection but not bounties. The Hayne Royal Commission hinted that lack of financial incentives hampered early revelations of feefornoservice misconduct. Introducing a capped, performancebased bounty (e.g., 15 % of collected penalties, maximum A$30 million) would likely enrich ASIC’s pipeline and strengthen cultural change.
澳大利亚 2019 年的举报人改革扩大了保护但未增加奖励。Hayne 皇家委员会暗示缺乏经济激励阻碍了对无偿服务不当行为的早期揭露。引入一项有上限、基于绩效的奖励（例如，收集罚款的 15%，最高 3000 万澳元）可能会丰富 ASIC 的举报管道并加强文化变革。

Thus, sharing penalties with whistleblowers is a costeffective complement to statutory duties, improving detection without heavy taxpayer funding.
因此，与举报人分享处罚是一种具有成本效益的补充法定职责，在不依赖大量纳税人资金的情况下提高检测效率。

9 │ Making “reverse stress testing” a board legal duty ≈ 470 words
9 │ 将“反向压力测试”定为董事会法定职责 ≈ 470 字

Reverse stress testing (RST) asks: “Which scenarios would break us?”—thereby uncovering hidden vulnerabilities and prompting preventive action. The PRA has required UK banks to conduct RST since 2009; Topic 6 slides report that boards initially treated it as an “edgecase compliance exercise,” but quality improved after the SM&CR held senior managers personally accountable for test design .
反向压力测试（RST）询问：“哪些情景会使我们陷入困境？”——从而揭示隐藏的漏洞并促使采取预防措施。自 2009 年以来，英国银行监管机构（PRA）已要求进行反向压力测试；主题 6 幻灯片报告称，董事会最初将其视为“边缘合规练习”，但在 SM&CR 要求高级管理人员对测试设计承担个人责任后，质量得到了提高。

Embedding RST as an explicit legal duty would:
将 RST 定为明确的法定职责将：

Sharpen accountability: Directors could no longer plead ignorance of tail risks; failure to consider plausible break points would breach their “reasonable steps” obligation.
强化问责制：董事不能再以不了解尾部风险为借口；未能考虑可能的断裂点将违反其“合理措施”义务。

Inform risk appetite: Quantifying capital needed to withstand identified scenarios grounds strategic choices—dividend policy, asset mix—in hard numbers.
明确风险偏好：量化抵御已识别情景所需的资本为战略选择——股息政策、资产组合——提供了确切的数字依据。

Aid supervisor dialogue: RST outputs give regulators a common language to challenge board assumptions.
促进监管者对话：RST 输出为监管者提供了一个共同的语言来挑战董事会假设。

Challenges. (i) Scenario selection bias: Boards may pick exotic but improbable events, sidestepping likely pain points. Regulators must therefore prescribe highlevel guidance (e.g., simultaneous funding and marketrisk shocks) while leaving calibration to firms. (ii) Model risk: RST depends on valuation models; independent validation is essential. (iii) Liability chill: Excess fear could lead boards to overcapitalize or exit socially valuable but risky lending. A safeguard is to link liability to process, not outcome: if documentation shows thorough analysis, the duty is met even if an unforeseen shock arises.
挑战。(i)情景选择偏差：董事会可能会选择奇特但不太可能发生的事件，回避可能的风险点。因此，监管者必须制定高级别指导（例如，同时发生的资金和市场风险冲击），同时将校准工作留给公司。(ii)模型风险：RST 依赖于估值模型；独立验证至关重要。(iii)责任寒蝉效应：过度的恐惧可能导致董事会过度资本化或退出具有社会价值但风险较高的贷款。一个保障措施是将责任与过程挂钩，而不是结果：如果文件显示进行了彻底的分析，即使出现未预见的冲击，其义务也已履行。

Overall, legislating an annual RST duty—mirroring the UK’s Resolution Assessment—would lift board foresight without stifling prudent risktaking, provided supervision focuses on methodology rather than perfect forecasting.
总体而言，立法实施年度 RST 责任——类似于英国的决议评估——可以在不扼杀审慎风险承担的情况下提高董事会的前瞻性，前提是监管重点应放在方法上而不是完美的预测。

10 │ Gender diversity on bank boards and misconduct risk ≈ 455 words
10 │ 银行董事会性别多样性与不当行为风险 ≈ 455 字

Several crosscountry studies (e.g., Lu & Herold 2022) find that banks with higher female board representation incur significantly lower regulatory fines and exhibit reduced earnings volatility. The causal mechanisms are debated, but three channels dominate:
多项跨国研究（例如，Lu & Herold 2022）发现，女性在董事会中占比更高的银行受到的监管罚款显著较低，且盈利波动性有所减少。因果关系存在争议，但主要有三个渠道：

Risk aversion: Metaanalysis suggests women, on average, prefer lower variance gambles, potentially steering strategy toward sustainable returns.
风险规避：元分析表明，女性平均倾向于选择低方差赌博，可能会将策略引导至可持续回报。

Enhanced challenge culture: Greater diversity disrupts groupthink; dissenting voices question aggressive sales targets. Topic 6 governance slides cite case interviews from the Hayne Commission showing female INEDs more likely to escalate consumercomplaint red flags .
强化挑战文化：更大程度的多样性破坏群体思维；不同意见质疑激进的销售额目标。主题 6 治理幻灯片引用了 Hayne 委员会的案例访谈，显示女性 INED 更有可能升级消费者投诉红旗。

Stakeholder orientation: Surveys indicate female directors place higher weight on reputational and customer outcomes, aligning decisions with longrun franchise value.
利益相关者导向：调查表明，女性董事更重视声誉和客户结果，使决策与长期品牌价值保持一致。

Policy implications. Quotas (EU Directive 2022/2381 sets 40 % target for listed boards by 2026) accelerate change but risk tokenism. Complementary measures—skillsbased selection panels, mentorship pipelines—ensure qualified candidates. Disclosure requirements (ASX CG Principle 1.5) help investors price governance quality.
政策影响。配额（欧盟指令 2022/2381 设定 2026 年前在上市公司董事会中达到 40%的目标）加速变革，但存在象征主义风险。补充措施——基于技能的选拔委员会、导师管道——确保合格候选人。披露要求（ASX CG 原则 1.5）帮助投资者评估治理质量。

While diversity alone cannot guarantee probity—Wells Fargo’s board had two female directors before the crossselling scandal—it appears to reduce frequency and severity of misconduct events. Hence regulators’ push for gender diversity is justified as a prudential as well as social policy.
尽管多样性本身不能保证诚信——Wells Fargo 的董事会在这场交叉销售丑闻之前已有两名女性董事——但它似乎减少了不当行为事件的发生频率和严重程度。因此，监管机构推动性别多样性既是一项审慎政策，也是一项社会政策。

11 │ A universal, revocable conduct licence for senior bankers ≈ 480 words
11 │ 为高级银行家提供的一项通用、可撤销的行为许可证 ≈ 480 字

Doctors, lawyers and pilots hold personal licences; misconduct leads to industrywide bans. Banking, however, historically regulated institutions, allowing “rolling bad apples” to move between firms. The UK’s SM&CR created a public “Directory” of certified persons; Australia’s FAR registers accountable individuals. A universal conduct licence, revocable across employers, would take the logic one step further.
医生、律师和飞行员持有个人许可证；不当行为会导致行业范围内的禁令。然而，银行业历史上对机构进行监管，允许“滚动坏苹果”在各个公司之间移动。英国的 SM&CR 创建了一个公共“名册”来登记合格人员；澳大利亚的 FAR 注册了有责任感的个人。一项通用行为许可证，可以在不同雇主之间撤销，将这一逻辑推进了一步。

Benefits. (i) Portability: Misconduct at Bank A bars immediate reemployment at Bank B, closing the revolvingdoor loophole spotlighted by the Hayne Commission. (ii) Clear deterrence: The personal cost—loss of livelihood—focuses minds more than firmlevel fines. (iii) Regulatory efficiency: Supervisors monitor licence status rather than renegotiating fitandproper assessments at each hiring.
益处。(i) 可携带性：在银行 A 的不当行为将禁止立即在银行 B 重新就业，关闭了 Hayne 委员会重点关注的旋转门漏洞。(ii) 明确的威慑：个人成本——失去生计——比公司层面的罚款更能集中注意力。(iii) 监管效率：监管机构监控许可证状态，而不是在每次招聘时重新协商适合性评估。

Design. Scope should cover senior managers and “material risk takers.” Standards merge prudential and conduct expectations (honesty, competence, cooperation with supervisors). Due process mirrors medical tribunals: an independent panel hears evidence; appeals go to administrative courts. Sanctions range from reprimand to suspension to lifetime ban, with public disclosure.
设计。范围应涵盖高级管理人员和“重大风险承担者”。标准融合了审慎和行为期望（诚实、能力、与上级合作）。正当程序类似于医疗 tribunals：一个独立的小组听取证据；上诉提交给行政法院。制裁范围从警告到停职到终身禁令，并公开披露。

Risks and mitigants. Overdeterrence could foster extreme conservatism. Counter: focus sanctions on intentional misconduct or reckless disregard; allow probationary reinstatement after remediation. Regulator capacity: licensing thousands requires resources; technology—central databases, efiling—eases burden. Global mobility: without reciprocity, banned bankers may shift offshore. IOSCO could coordinate mutualrecognition pacts, as exists for airline pilot licences.
风险和缓解措施。过度威慑可能导致极端保守主义。对策：将制裁集中于故意不当行为或鲁莽忽视；允许在整改后恢复试用。监管能力：许可数千人需要资源；技术——中央数据库、电子文件——减轻负担。全球流动：没有互惠，被禁止的银行家可能会转移到海外。IOSCO 可以协调相互承认协议，就像对飞行员执照存在的那样。

Evidence from the UK Directory’s early years is promising: the Financial Conduct Authority reports a 50 % jump in selfreported breaches once personal certification became portable, suggesting stronger internal policing. Extending that concept into a formal licence—revocable upon serious breach—would make accountability durable, enhance market confidence and complement institutional reforms like FAR and SM&CR.
来自英国目录的早期证据很有希望：金融行为管理局报告称，一旦个人认证变得可移植，自我报告的违规行为增加了 50%，这表明内部监管更加严格。将该概念扩展到正式许可证——在严重违规时可以撤销——将使责任持久，增强市场信心，并补充像 FAR 和 SM&CR 这样的制度改革。

Topic7:
主题 7：

1 │ Why it is hard to fold climate risk into prudential supervision ≈ 480 words
1 │ 为什么将气候风险纳入审慎监管很困难 ≈ 480 字

Prudential tools—riskweighted capital, stresstesting, liquidity ratios—were built on short businesscycle horizons and backwardlooking loss data. Climate risk breaks those design assumptions in five ways. (i) Long, uncertain horizons. Major physical hazards (sealevel rise) and transition shocks (abrupt carbon pricing) may crystallise well beyond the threetofiveyear windows used in Internal RatingsBased (IRB) models; discounting pushes expected loss toward zero, understating required capital . (ii) Nonlinearity and tipping points. Damages are stepwise—once a cyclone’s windspeed threshold is crossed, losses jump—defying Gaussian assumptions embedded in Pillar 1 formulas . (iii) Data scarcity and definitional fog. Assetlevel emissions and location data are patchy; historical loss series do not reflect a warming world. Supervisors therefore rely on scenario analysis (NGFS pathways), but translating temperature trajectories into borrower ProbabilitiesofDefault involves novel, often subjective, “climate beta” parameters. (iv) Double materiality. Standard prudential focus is “outsidein” (risk to banks); climate policy also demands an “insideout” view (financing that creates risk). Embedding that dual lens stretches mandates. (v) Transmissionchannel entanglement. Physical, transition and liability risks cut across credit, market, operational and reputational silos (APRA PPG 229 lists seven intertwined channels) , making attribution—and hence capital calibration—murky.
审慎工具——风险加权资本、压力测试、流动性比率——是基于短期商业周期视野和回顾性损失数据建立的。气候风险在五个方面打破了这些设计假设。(i) 长期、不确定的视野。主要的物理风险（海平面上升）和转型冲击（突然的碳定价）可能会在内部评级法（IRB）模型中使用的三到五年窗口之外结晶；折现将预期损失推向零，低估了所需的资本。(ii) 非线性和临界点。损害是阶梯式的——一旦飓风的风速阈值被跨越，损失就会跳跃——违背了嵌入第一支柱公式中的高斯假设。(iii) 数据稀缺和定义模糊。资产层面的排放和位置数据是零散的；历史损失系列并不能反映一个变暖的世界。因此，监管机构依赖于情景分析（NGFS 路径），但将温度轨迹转化为借款人违约概率涉及新颖的、通常是主观的“气候贝塔”参数。(iv) 双重实质性。标准的审慎关注点是“外向内”（对银行的风险）；气候政策也要求一个“内向外”的视角（融资创造了风险）。 嵌入双镜头拉伸约束。 (v) 传输信道纠缠。物理、过渡和责任风险跨越信贷、市场、运营和声誉壁垒（APRA PPG 229 列出了七个相互交织的渠道），使得归因——进而资本校准——变得模糊。

These hurdles explain regulators’ current stance. The European Banking Authority concluded in 2022 that climate factors “fit within existing risk categories” and that reliable differentials between “green” and “brown” default rates are not yet demonstrable; it therefore shunned a Pillar 1 surcharge for the moment. APRA likewise chose a supervisoryreview route (CPG 229 guidance, climate stress tests) rather than rewriting capital rules. Until forwardlooking data mature and scenariocoherent valuation models stabilise, supervisors face a tradeoff between model risk (miscalibrated buffers) and inaction risk (undercapitalised shocks).
这些障碍解释了监管机构目前的立场。欧洲银行管理局在 2022 年得出结论，气候因素“符合现有的风险类别”，并且“绿色”和“棕色”违约率之间的可靠差异尚未得到证实；因此，它暂时避免了第一支柱附加费。APRA 同样选择了监管审查路线（CPG 229 指导，气候压力测试），而不是重写资本规则。直到前瞻性数据成熟和场景一致的估值模型稳定，监管机构面临模型风险（校准不当的缓冲）和无所作为风险（资本不足的冲击）之间的权衡。

2 │ Should financialpolicy tools be used to push decarbonisation? ≈ 485 words
2 │ 金融政策工具是否应被用来推动脱碳？ ≈ 485 字

Proponents argue that climate change is a “greenswan” systemic threat; waiting for externalities to be taxed risks a crisis that forces central banks to become “climate rescuers of last resort.” Accordingly, tools such as a BrownPenalising Factor (BPF) on fossilfuel exposures, or macroprudential climate buffers, would align private lending costs with social damage and reduce future bailout risk. Civilsociety campaigns (Finance Watch’s “oneforone” capital rule) champion this logic, while the 2024 Australian Sustainable Finance Roadmap signals a wholeofgovernment push to “mobilise capital toward netzero” .
支持者认为气候变化是一种“绿色天鹅”系统性威胁；等待外部性征税可能会导致迫使中央银行成为“最后的气候救援者”的危机。因此，诸如对化石燃料敞口征收棕色惩罚因子（BPF）或宏观审慎气候缓冲等工具将使私人借贷成本与社会损害相一致，并减少未来的救助风险。公民社会运动（金融观察的“一对一”资本规则）倡导这种逻辑，而 2024 年澳大利亚可持续金融路线图则表明政府全面推动“动员资本向净零”。

Sceptics, including the EBA (2022), counter that prudential instruments exist to ensure solvency, not to steer credit. Using Pillar 1 as climate policy could dilute its primary objective, misprice transitionalfinance loans needed to clean brown industries, and stretch centralbank mandates lacking electoral legitimacy. Moreover, empirical evidence for a “green supporting factor” is thin; lowering risk weights might simply raise bank margins, not green lending volumes.
怀疑者，包括欧洲银行管理局（2022 年），反驳说审慎工具是为了确保偿付能力，而不是引导信贷。将第一支柱用作气候政策可能会稀释其主要目标，错误定价清洁棕色产业所需的转型金融贷款，并扩大缺乏选举合法性的中央银行职责。此外，“绿色支持因子”的实证证据不足；降低风险权重可能会简单地提高银行利润，而不是绿色贷款量。

A pragmatic middle path distinguishes riskbased from promotional interventions. Pillars 2 & 3 already host climate stress testing, disclosure and governance expectations (APRA’s CPG 229, ECB’s SREP addons). These measures internalise demonstrable risk without prescribing winners. Promotional goals—taxonomylinked collateral frameworks, greenbond purchase programmes—can still sit within the monetary toolbox, but only where legislation clearly extends the mandate (as the EU did by adding “supporting the general economic policies of the Union”). For Australia, the Roadmap foresees Treasuryled taxonomies and sovereign green bonds, leaving APRA’s prudential remit intact.
一个务实的中间道路区分了基于风险和促销性干预。支柱 2 & 3 已经包含了气候压力测试、信息披露和治理预期（APRA 的 CPG 229，ECB 的 SREP 附加）。这些措施内部化了可证明的风险，而没有指定赢家。促销目标——与分类标准相关的抵押品框架、绿色债券购买计划——仍然可以位于货币工具箱内，但仅当立法明确扩展其职责时（如欧盟通过添加“支持联盟的一般经济政策”）。对于澳大利亚，路线图预计财政部领导的分类标准和主权绿色债券，APRA 的审慎职责保持完整。

Hence, active decarbonisation by prudential means is defensible only when (a) the link to financialstability risk is evidencebased, and (b) democratic authorities explicitly broaden the mandate.
因此，通过审慎手段的积极脱碳只有在（a）与金融稳定风险的联系有证据支持，并且（b）民主当局明确扩大其职责时才是合理的。

3 │ Is promotion of decarbonisation consistent with APRA’s mandate? ≈ 460 words
3 │ 促进脱碳是否符合 APRA 的职责？≈ 460 字

APRA’s statutory goal under the Australian Prudential Regulation Authority Act 1998 is to “protect the interests of depositors, policyholders and superannuation members” and “promote financialsystem stability.” Its powers must therefore trace to safetyandsoundness. Climaterelated physical and transition risks plainly threaten solvency—the basis for CPG 229 and APRA’s 2023 climate stress test. Using supervisory levers to ensure banks price and manage those risks sits squarely within mandate.
APRA 在《1998 年澳大利亚审慎监管局法案》下的法定目标是“保护储户、保单持有人和养老金成员的利益”以及“促进金融体系稳定。”因此，其权力必须追溯到安全与健康。气候相关的物理和转型风险显然威胁到偿付能力——这是 CPG 229 和 APRA 2023 年气候压力测试的基础。使用监管杠杆确保银行对那些风险进行定价和管理，完全在其职责范围内。

But active promotion—tilting credit away from brown sectors beyond what risk differentials justify—enters economicpolicy terrain reserved for elected government and the Reserve Bank. The Sustainable Finance Roadmap confirms this division: APRA leads on risk oversight, while Treasury develops the taxonomy and disclosure law. Were Parliament to deem decarbonisation a prudential objective, it could amend s 8 to add an explicit reference to “sustainable finance,” mirroring the UK’s recent secondary climate objective for the FCA and PRA. Absent such change, APRA should confine itself to riskbased measures, coordinating with ASIC on disclosure and with the RBA on scenario design. Whether it should be broadened is ultimately a political question; clarity would avoid mandatecreep accusations and litigation risk.
但积极的推广——将信贷倾斜到风险差异无法 justify 的棕色行业之外——进入了由当选政府和储备银行保留的经济政策领域。可持续金融路线图证实了这一划分：APRA 领导风险监督，而财政部制定分类标准和披露法律。如果议会认为脱碳是一种审慎目标，它可以修改第 8 条以增加对“可持续金融”的明确引用，效仿英国最近为 FCA 和 PRA 设定的次要气候目标。如果没有这样的变化，APRA 应该把自己限制在基于风险的措施上，与 ASIC 协调披露，与 RBA 协调情景设计。它是否应该被扩大是一个最终的政治问题；明确性可以避免授权蔓延指控和诉讼风险。

4 │ ASIC’s role in climate change ≈ 455 words
4 │ ASIC 在气候变化中的作用 ≈ 455 字

Unlike APRA, ASIC is Australia’s marketconduct and disclosure regulator. Climate change falls within its remit in three intersecting ways:
与 APRA 不同，ASIC 是澳大利亚的市场行为和披露监管机构。气候变化以三种相互交叉的方式在其管辖范围内：

Mandatory climaterelated financial reporting. Under the 2024 Treasury Laws Amendment (Financial Infrastructure and Other Measures) Act, large corporates and financial entities must lodge a “Sustainability Report” disclosing climate risks, governance and Scope 13 emissions, phased in from 1 January 2025. ASIC will police compliance, backed by audit requirements and (temporarily limited) liability for scenario analysis .
强制性气候相关财务报告。根据《2024 年财政法案（金融基础设施和其他措施）》，大型企业和金融机构必须从 2025 年 1 月 1 日起分阶段提交“可持续发展报告”，披露气候风险、治理和范围 13 排放。ASIC 将监督合规情况，并依据审计要求和（暂时有限制的）情景分析责任进行监管。

Misleading disclosure (greenwashing). Information Sheet 271 warns issuers that exaggerated ESG claims breach s 1041H (misleading conduct). ASIC’s 2024 Mercer Super prosecution (A$11.3 million penalty) shows the agency will litigate when portfolio holdings contradict “sustainable” branding . It has formed a “coalition against greenwashing” with the ACCC and Clean Energy Regulator.
欺诈性披露（漂绿）。信息表 271 警告发行人，夸大的 ESG 声明违反了第 1041H 条（欺诈行为）。ASIC 的 2024 年 Mercer 超级诉讼（罚款 1.13 亿澳元）表明，当投资组合持有与“可持续”品牌不符时，该机构将提起诉讼。它与 ACCC 和清洁能源监管机构组成了“反漂绿联盟”。

Investorprotection guidance. ASIC contributes to the Council of Financial Regulators’ work on data gaps (SFR, Pillar 1) and cochairs IOSCO’s sustainablefinance task force, shaping global assurance and labelling standards.
投资者保护指南。ASIC 参与金融监管委员会在数据差距（SFR、支柱 1）方面的工作，并担任 IOSCO 可持续金融任务组的联合主席，塑造全球保证和标签标准。

In sum, ASIC’s climate brief is about market integrity and informed capital allocation, complementing APRA’s solvency focus and the RBA’s macrostability lens.
总之，ASIC 的气候简报是关于市场诚信和知情资本配置，补充了 APRA 的偿付能力重点和 RBA 的宏观稳定性视角。

### 7 │ Should the safeharbour for Scope 3 liability be extended beyond 30 June 2027? ≈ 485 words
### 7 │ 应将范围 3 责任的安全港扩展至 2027 年 6 月 30 日之后？ ≈ 485 字

Australia’s new climatereporting law grants issuers temporary immunity for statements about Scope 3 emissions and scenario analysis until the 202728 financial year, limiting enforcement to ASIC and requiring proof of fault . The carveout recognises that downstream (“financed”) and upstream supplychain emissions are difficult to measure and verify—data come from counterparties two or three steps removed, rely on voluntary surveys, and change as business models shift .
澳大利亚新的气候报告法律给予发行人在 2027 年 28 财年之前对范围 3 排放和情景分析的陈述暂时豁免，限制执法仅限于 ASIC 并要求证明过失。这一豁免承认下游（“融资”）和上游供应链排放难以测量和验证——数据来自距离两到三个步骤的对手方，依赖于自愿调查，并且随着商业模式的转变而变化。

Arguments for retaining protection.
保留保护的论点。

Measurement uncertainty. The Greenhouse Gas Protocol offers guidance, but methodological error bars remain wide; strict liability might deter companies from disclosing at all or encourage boilerplate caveats that undermine transparency.
测量不确定性。温室气体协议提供了指导，但方法论误差范围仍然很宽；严格责任可能会阻止公司进行披露，或者鼓励使用陈词滥调的免责声明，从而损害透明度。

Assurance capacity. Australia has only a handful of auditors trained to validate financedemission inventories; premature liability could create a bottleneck that delays reporting.
财务责任能力。澳大利亚只有少数审计师接受过验证财务排放清单的培训；过早的财务责任可能会造成瓶颈，从而延迟报告。

Global alignment. The ISSB’s climate standard (IFRS S2) is silent on legal enforcement timelines; keeping a safe harbour until major jurisdictions converge avoids disadvantaging domestic firms.
全球协调。ISSB 的气候标准（IFRS S2）对法律执行时间表保持沉默；保持安全港直到主要司法管辖区趋同，可以避免使国内企业处于不利地位。

Arguments against extension.
反对延期的论点。

Investor decisionusefulness. Scope 3 often dwarfs operational emissions for banks and asset managers; investors need credible numbers to price transition risk.
投资者决策有用性。对于银行和资产管理公司而言，范围 3 的排放量往往远超运营排放量；投资者需要可信的数据来评估转型风险。

Regulatory arbitrage. A prolonged shield may tempt issuers to “reverse engineer” estimates to meet voluntary targets, knowing private litigants cannot challenge them.
监管套利。长时间的豁免可能会诱使发行人“逆向工程”估计以实现自愿目标，因为他们知道私人诉讼人无法挑战他们。

Path dependency. Experience with forwardlooking MD&A safe harbours in the US shows that once immunity is embedded it is politically hard to roll back.
路径依赖。美国关于前瞻性管理层讨论与分析安全港的经验表明，一旦豁免被嵌入，政治上就很难撤销。

Recommendation. Convert the sunset into a conditional extension: firms demonstrating credible datacollection plans and independent limited assurance may retain capped liability (e.g., negligence instead of strict) for a further three years; laggards revert to full Corporations Act liability. This preserves incentives to invest in measurement infrastructure while giving courts a clear standard to judge diligence. Meanwhile, ASIC should publish industry benchmarks to reduce estimation dispersion and prepare enforcement guidelines for the postharbour era.
建议。将到期时间转换为有条件的延期：展示可信数据收集计划并具有独立有限保证的公司可以保留有限的责任（例如，疏忽而不是严格责任）再延长三年；落后者恢复到完整的《公司法》责任。这保留了投资于测量基础设施的激励，同时为法院提供了一个明确的勤勉标准。同时，ASIC 应发布行业基准以减少估计分散，并为 Harbor 时代准备执法指南。

Such a phased approach balances the realworld data gap with the market’s need for reliable transitionrisk metrics, fulfilling both investorprotection and climatepolicy objectives.
这种分阶段的方法平衡了现实世界数据差距与市场对可靠转型风险指标的需求，实现了投资者保护和气候政策目标。

### 8 │ Would a BrownPenalising Factor (BPF) distort Australia’s resourceexport model? ≈ 470 words
### 8 │ 布朗惩罚因子（BPF）是否会扭曲澳大利亚的资源配置模式？ ≈ 470 字

A BPF would raise riskweights—and hence minimum capital—on loans to carbonintensive activities. In an economy where coal and LNG still comprise over 60 % of goods exports, the proposal triggers fears of credit rationing and loss of comparative advantage.
BPF 将提高对碳密集型活动的贷款风险权重——从而提高最低资本要求。在一个煤炭和液化天然气仍占商品出口超过 60%的经济体中，该提议引发了信贷配给和比较优势丧失的担忧。

The Institute for Climate Economics notes that BPFs increase both the price and the lossabsorbing buffer of brown lending , but critics at the EBA argue the tool “hinders the primary object of prudential rules” and lacks granular data for calibration . Australianspecific concerns add two layers:
气候经济研究所指出，BPF 会增加棕色贷款的价格和损失吸收缓冲，但欧洲银行协会的批评者认为该工具“妨碍了审慎规则的主要目标”，并且缺乏校准所需的详细数据。澳大利亚特有的担忧增加了两个层面：

Trade exposure. Unlike diversified European economies, Australian miners sell into global commodity markets; higher financing costs could simply shift production to jurisdictions without a BPF, creating carbon leakage without cutting global emissions.
贸易敞口。与多元化的欧洲经济体不同，澳大利亚的矿业公司销售到全球商品市场；更高的融资成本可能会简单地使生产转移到没有 BPF 的司法管辖区，从而在不减少全球排放的情况下造成碳泄漏。

Transition financing. Decarbonisation of heavy industry requires more capital for CCUS and renewables, not less. Blanket BPFs risk starving browntogreen projects.
过渡融资。重工业脱碳需要更多资金用于 CCUS 和可再生能源，而不是更少。普遍的 BPFs 风险导致棕色变绿色项目资金短缺。

Proponents counter that market discipline alone has not redirected capital—major banks remain offtrack for financedemission targets . A calibrated BPF could accelerate repricing of strandedasset risk that portfolio models still underweight.
支持者反驳称，市场纪律本身并没有重新引导资本——主要银行仍然偏离了融资排放目标。校准的 BPFs 可以加速搁浅资产风险的重新定价，而投资组合模型仍然低估了这一风险。

Design choices to mitigate distortion:
减少扭曲的设计选择：

Carveouts for credible transition plans verified against Treasury’s forthcoming taxonomy .
针对可信过渡计划的除外条款，这些计划将根据财政部即将出台的分类标准进行验证。

Gradual rampup over five years to avoid tipping viable exporters into distress.
分五年逐步提升，以避免将可行的出口商推入困境。

Reciprocal greensupporting factor to preserve balance in the capital framework.
相互的绿色支持因素，以保持资本框架的平衡。

Ultimately, a BPF need not torpedo the resource sector if aligned with transition pathways and paired with public investment in diversification. Without such nuance, however, it could simply shift emissions abroad while shrinking domestic employment—suggesting prudential tools are a secondbest lever compared to carbon pricing and targeted subsidies.
最终，如果与转型路径保持一致并搭配多元化公共投资，BPF 就不必破坏资源部门。然而，如果没有这种细微差别，它可能会将排放转移到国外，同时减少国内就业——这表明与碳定价和定向补贴相比，谨慎工具是次优杠杆。

### 9 │ Can climate scenario tests replace traditional credit stress tests? ≈ 460 words
### 9 │ 气候情景测试能否替代传统信贷压力测试？ ≈ 460 字

Climate scenario exercises project balancesheet impacts under NGFS pathways out to 2050 and beyond , whereas credit stress tests model cyclical recessions over two to three years. Replacement is therefore mistaken for three reasons.
气候情景练习项目在 NGFS 路径下至 2050 年及以后的资产负债表影响，而信贷压力测试则模拟了两年到三年的周期性衰退。因此，用替代来理解是错误的，原因有三点。

Time horizon mismatch. Loan books turn over every five years; nearterm recessions can trigger defaults before climate shocks hit. Abandoning shorthorizon stress tests would ignore core solvency risk.
时间范围不匹配。贷款组合每五年周转一次；近期的衰退可能触发在气候冲击之前就出现违约。放弃短期压力测试将忽略核心偿付能力风险。

Transmission channels. Climate scenarios bundle macro variables with sector shocks but lack granularity for idiosyncratic credit risk (e.g., SME failures). They complement rather than duplicate conventional tests.
传导渠道。气候情景捆绑了宏观变量与行业冲击，但对于特殊信贷风险（例如中小企业失败）缺乏细节。它们是补充而不是重复传统测试。

Model uncertainty. Transition pathways depend on politics and technology; by contrast, GDPunemployment linkages have rich historical data. Using climate tests alone would amplify model risk flagged by APRA: “no historical data… risk unprecedented” .
模型不确定性。转型路径取决于政治和技术；相比之下，GDP 与失业率的联系有丰富的历史数据。仅使用气候测试将放大 APRA 指出的模型风险：“没有历史数据……风险前所未有”。

Optimal approach. Run climate stress tests as a second suite under Pillar 2: overlay transition and physical shocks onto baseline macro assumptions, then feed sensitivities into capitalplanning and risk appetite. Supervisors can use outcomes to set addons or guide dividend restrictions, but should not discard established credit stress frameworks.
最佳方法。将气候压力测试作为第二套方案在第二支柱下运行：将覆盖转换和物理冲击叠加到基准宏观假设上，然后将敏感性输入到资本规划和风险偏好中。监管机构可以使用结果来设置附加功能或指导股息限制，但不应放弃已建立的信用压力框架。

### 10 │ Should the RBA restrict repo eligibility to Parisaligned collateral? ≈ 475 words
### 10 │ RBA 是否应将回购合格券限制为巴黎协议对齐的抵押品？约 475 字

A Parisscreened collateral framework would make government bond purchases and repo operations conditional on issuers’ climate alignment. Advocates argue that centralbank neutrality is illusory: by accepting all bonds equally, the RBA implicitly subsidies carbonintensive issuers and may later face firesales of stranded assets .
巴黎筛选抵押品框架将使政府债券购买和回购操作取决于发行人的气候对齐。支持者认为央行中立是虚幻的：通过平等接受所有债券，RBA 隐含补贴了高碳排放发行人，并可能在后来面临搁浅资产的甩卖。

Mandate tension. The RBA’s Act centres on price stability, full employment and “overall economic prosperity.” Unlike the ECB, Australian law provides no explicit secondary environment objective. Imposing climate screens without legislative change could invite judicial review and politicise monetary policy.
命令紧张。RBA 法案的核心是价格稳定、充分就业和“整体经济繁荣”。与欧洲央行不同，澳大利亚法律没有明确的环境次要目标。在不进行立法变更的情况下强制实施气候筛选可能会引发司法审查并使货币政策政治化。

Market functioning. Repo eligibility supports smooth liquidity in Commonwealth and stategovernment securities. Conditioning access could fragment collateral pools, raise repo rates and impair transmission of monetary policy—particularly in stress episodes when liquidity is scarce.
市场运作。回购资格支持英联邦和州政府证券的流动性顺畅。限制接入可能会分散抵押品池，提高回购利率，并损害货币政策传导——特别是在流动性稀缺的压力情况下。

Transitional design. If Parliament amends the Act, three safeguards would limit disruption:
过渡性设计。如果议会修订法案，将有三项保障措施限制干扰：

Phased implementation over multiple years aligned with taxonomy milestones.
多年分期实施，与分类标准里程碑保持一致。

Haircut differentials rather than binary exclusion, preserving liquidity while pricing transition risk.
差异化处理而非二元排除，在定价过渡风险的同时保留流动性。

Disclosurebased grandfathering: existing bonds retain eligibility; new issuance must provide TCFDaligned disclosures.
基于披露的祖父条款：现有债券保留资格；新发行必须提供与气候相关财务信息披露框架（TCFD）一致的披露。

Absent a statutory mandate, the RBA should focus on integrating climate risk into collateral haircuts and valuation, consistent with its financialstability brief, while leaving outright eligibility unchanged. That approach acknowledges transition risk without overstepping legal authority.
如无法定授权，RBA 应专注于将气候风险整合到抵押品折扣和估值中，与其金融稳定简报一致，同时保持直接资格不变。这种做法承认转型风险，但未超越法律授权。

### 11 │ How should APRA integrate naturerelated risks alongside climate? ≈ 480 words
### 11 │ 应如何将自然相关风险与气候风险整合？ ≈ 480 字

The Sustainable Finance Roadmap expands the agenda from carbon to naturepositive finance, establishing a Nature Finance Council and promising data tools via Environment Information Australia . Yet APRA’s CPG 229 remains climatespecific. Integration can follow a threestep path.
可持续金融路线图将议程从碳扩展到自然正面金融，成立自然金融委员会，并通过环境信息澳大利亚承诺数据工具。然而，APRA 的 CPG 229 仍然是针对气候的。整合可以遵循三步路径。

Supervisory expectations (Pillar 2). Issue a CPG 23X mirroring CPG 229 but mapping biodiversity, water and pollution dependencies onto credit, operational and reputational risk channels identified by TNFD . Boards would be expected to identify material nature exposures, set metrics (e.g., hectares affected), and embed them in risk appetite.
监管预期（第二支柱）。发布与 CPG 229 类似的 CPG 23X，但将生物多样性、水资源和污染依赖映射到 TNFD 确定的信用、运营和声誉风险渠道上。预计董事会将识别重大自然风险暴露，设定指标（例如，受影响的公顷数），并将其嵌入风险偏好中。

Disclosure alignment (Pillar 3). Coordinate with ASIC to extend the Sustainability Report regime to TNFDaligned nature metrics once ISSB publishes its biodiversity standard. Comparable disclosure underpins market discipline and feeds supervisory data lakes.
披露一致性（第三支柱）。与 ASIC 协调，在 ISSB 发布其生物多样性标准后，将可持续性报告制度扩展到与 TNFD 一致的自然资源指标。可比披露支撑市场纪律，并为监管数据湖提供数据。

Scenario analysis pilots. Collaborate with the Council of Financial Regulators to design nature stress modules—e.g., sudden deforestation bans or freshwater scarcity pricing—echoing APRA’s 2023 climate stress test. Insights would inform riskweighted asset overlays or sectoral capital buffers where correlations are high.
情景分析试点。与金融监管委员会合作设计自然压力模块——例如，突然的森林砍伐禁令或淡水稀缺定价——呼应 APRA 的 2023 年气候压力测试。洞察将用于高相关性领域的风险加权资产覆盖或行业资本缓冲。

Challenges. Data granularity is even weaker than for carbon; the Roadmap flags the Treasurer’s request to address naturedata gaps . APRA should leverage EIA datasets and require banks to collect borrowerlocation geocoordinates, albeit with proportionality for SMEs.
挑战。数据的粒度甚至比碳排放更弱；路线图指出了财政部长要求解决自然数据差距的问题。APRA 应利用环境影响评估数据集，并要求银行收集借款人位置地理坐标，尽管对中小企业应按比例进行。

Mandate fit. Nature loss can impair collateral (e.g., farmland), create liability risk (pollution lawsuits) and trigger supplychain shocks—squarely within APRA’s “financialsoundness” remit. As with climate, promotional objectives (e.g., funding conservation) belong to fiscal policy; APRA’s role is risk identification and capital adequacy.
委托适合。自然损失可能损害担保品（例如，农田），产生责任风险（污染诉讼），并引发供应链冲击——这完全在 APRA 的“金融稳健性”职责范围内。与气候问题一样，推广目标（例如，资助保护）属于财政政策；APRA 的角色是风险识别和资本充足性。

By phasing in guidance, disclosure and analytical tools, APRA can extend its climate framework to nature without overloading institutions, supporting the Roadmap’s aim of mobilising capital for a naturepositive economy while safeguarding prudential soundness.
通过逐步实施指导方针、信息披露和分析工具，APRA 可以将其气候框架扩展到自然领域，而不会使机构负担过重，支持路线图的目标，即动员资本以实现自然正向经济，同时保障审慎稳健性。

Topic8:
主题 8：

1 │ Why Lehman’s collapse rocked the world & whether we are safer now ≈ 495 words
1 │ 雷曼兄弟破产为何震动全球及我们是否现在更安全 ≈ 495 字
Lehman Brothers combined three amplifiers of systemic contagion. (i) Size and interconnectedness. As a topfive U.S. brokerdealer Lehman sat at the hub of global repo and OTCderivatives networks; when it defaulted, thousands of counterparties scrambled to replace hedges and fund collateral calls, freezing dollar funding markets. (ii) Opacity and asset complexity. Its balancesheet was dominated by hardtovalue Level3 mortgage assets financed in overnight repo; uncertainty about marks, not just losses, caused creditors to withdraw liquidity. (iii) Absence of a resolution regime. Unlike banks, Lehman had no orderlyresolution playbook; Chapter 11 filings fragmented the group across legal entities and jurisdictions, trapping client collateral and triggering “firesale” asset liquidations that depressed prices across the system .
雷曼兄弟公司结合了三种系统性传染放大器。(i) 规模和相互关联性。作为美国前五名的经纪自营商，雷曼位于全球回购和场外衍生品网络的中心；当它违约时，数千家交易对手争相替换对冲和筹集抵押品赎回资金，冻结了美元融资市场。(ii) 不透明性和资产复杂性。其资产负债表主要由难以估值的 Level3 抵押资产构成，这些资产通过隔夜回购融资；关于估值的不确定性，不仅仅是损失，导致债权人撤回流动性。(iii) 缺乏清算制度。与银行不同，雷曼没有有序清算的剧本；第 11 章的申请将集团分散到不同的法律实体和司法管辖区，困住了客户抵押品，并触发了“甩卖”资产清算，导致整个系统价格下跌。

Postcrisis reforms address each amplifier. Orderly Resolution. The FSB’s Key Attributes and national regimes (Title II DoddFrank, UK SRR, EU BRRD) give authorities bailin powers, temporary stays and bridge institutions, reducing cliff effects. Centralclearing & margin. Mandatory CCP clearing for standardised swaps curtails bilateral contagion and embeds higher initial margin. Basel III liquidity. The NSFR discourages excessive shortterm wholesale funding, while LCR forces banks to prehold highquality liquid assets. Leverage & TLAC. Brokerdealer groups now carry nonriskweighted leverage caps and bailinable debt cushions that absorb losses without bankruptcy. Derivatives data. Trade repositories increase transparency, improving supervisors’ situational awareness.
危机后改革针对每个放大器。有序解决。FSB 的关键属性和国家制度（标题 II 多德弗兰克，英国 SRR，欧盟 BRRD）赋予当局 bailin 权力，临时停止和桥梁机构，减少悬崖效应。中央清算与保证金。标准化互换的强制 CCP 清算限制了双边传染，并嵌入更高的初始保证金。巴塞尔 III 流动性。NSFR 不鼓励过度的短期批发融资，而 LCR 迫使银行预先持有高质量的流动性资产。杠杆与 TLAC。经纪商-经销商集团现在承担非风险加权杠杆上限和可 bailin 的债务缓冲，这些可以在不破产的情况下吸收损失。衍生品数据。交易存储库增加了透明度，提高了监管者的态势感知。

Will the next Lehman hurt less? Stress events in March 2020 and the 2022 UK gilt crisis suggest progress but not immunity. Dealer balancesheets remained functional because Fed and BoE liquidity backstops were available early; CCPs withstood record margin calls. Yet vulnerabilities persist: nonbank leverage (hedge funds, LDI strategies) can still transmit shocks to dealers, and crossborder resolution of complex groups remains untested. Overall, the reforms ought to dampen systemic fallout, but only if jurisdictions execute living wills credibly and provide swift lenderoflastresort (LOLR) liquidity to solvent intermediaries.
下一次雷曼事件是否会更小？2020 年 3 月的压力事件和 2022 年英国金边债券危机表明有所进展，但并非免疫。交易商的资产负债表保持功能，因为美联储和英格兰银行的流动性支持早在早期就可用；中国公司承受了创纪录的保证金追缴。然而，风险依然存在：非银行杠杆（对冲基金、LDI 策略）仍可能将冲击传递给交易商，而复杂集团的跨境清算尚未得到测试。总体而言，改革应该减轻系统性后果，但前提是司法管辖区能够可信地执行生存计划，并为有偿付能力的中间机构提供快速的最后贷款人（LOLR）流动性。

2 │ Structuredfinance reforms: logic and effectiveness ≈ 480 words
2 │ 结构化金融改革：逻辑和有效性 ≈ 480 字
Crisis postmortems located the “epicentre” in subprime securitisation: originatetodistribute incentives, opaque tranching, and ABCP conduits dependent on rollover funding. The policy response targeted screening incentives, liquidity risks and transparency.
危机回顾将“震中”定位于次级抵押贷款证券化：发起到分配的激励措施、不透明的分层和依赖滚动融资的 ABCP 导管。政策应对措施针对筛选激励措施、流动性风险和透明度。

Risk retention. The U.S. DoddFrank § 941 (the “5 % skininthegame” rule) and EU CRR Art 405 now require sponsors to retain a vertical slice or risksensitive tranche, aligning interests with investors. Early evidence shows higher underwriting standards; default rates on post2014 RMBS are materially lower.
风险保留。美国的《多德-弗兰克法案》第 941 条（即“5%风险参与”规则）和欧盟的 CRR 第 405 条现在要求发起人保留垂直切片或风险敏感层，使利益与投资者保持一致。早期证据显示承保标准更高；2014 年后 RMBS 的违约率有显著降低。

Liquidity consolidation. Basel III forces banks to consolidate sponsored SIVs and ABCP conduits if “significant risk” remains; liquidity puts attract full capital charges, limiting offbalancesheet arbitrage . This has shrunk the ABCP market by more than 60 % from its 2007 peak.
流动性整合。巴塞尔 III 要求银行如果“重大风险”仍然存在，则必须整合发起的 SIV 和 ABCP 导管；流动性工具吸引全额资本费用，限制表外套利。这使 ABCP 市场规模从 2007 年的峰值缩小了 60%以上。

Duediligence & disclosure. EU STSsecuritisation rules prescribe standardised loanlevel data and homogeneity criteria; U.S. Regulation AB II mandates assetlevel disclosure for RMBS/CMBS. Investors now receive granular tape, although comparability across jurisdictions is still patchy.
尽职调查与披露。欧盟的 STS 证券化规则规定了标准化的贷款级数据和同质性标准；美国的规定 AB II 要求对 RMBS/CMBS 进行资产级披露。投资者现在获得详细的资料，尽管不同司法管辖区的可比性仍然不完整。

Rating process reform. IOSCO codes demand analyst rotation and conflictofinterest disclosures; nevertheless, rating shopping persists.
评级流程改革。IOSCO 准则要求分析师轮换和利益冲突披露；尽管如此，评级购物仍然存在。

Haircut floors. The FSB proposed numerical haircut floors for noncentrallycleared SFTs to curb leverage, but implementation is uneven, diluting impact.
剪发地板。FSB 提出了非中心化 SFTs 的数字剪发地板，以控制杠杆，但实施不均，削弱了影响。

Assessment. Retention and consolidation directly tackle originatetodistribute moral hazard and sponsor liquidity games, and empirical default data suggest improvement. However, complexity has not vanished—synthetic securitisations and private deals lie outside STS regimes—and liquidity reliance reemerged in March 2020 when repo haircuts on ABS rose sharply. Reforms therefore mitigate but do not eliminate fragility; vigilant supervision and circuitbreaker liquidity facilities remain necessary.
评估。保留和整合直接解决发起到分配的道德风险和赞助人流动性游戏，并且经验性违约数据表明有所改善。然而，复杂性并未消失——合成证券化和私人交易不在 STS 制度之外——流动性依赖在 2020 年 3 月重新出现，当时 ABS 的回购剪发急剧上升。因此，改革缓解但并未消除脆弱性；警惕的监管和断路器流动性设施仍然是必要的。

3 │ LOLR for shadow banks: why and how ≈ 475 words
3 │ 影子银行的无保本资金救助：原因和方式 ≈ 475 字
Shadow intermediaries—MMFs, dealer repo books, nonbank mortgage servicers—create moneylike claims without deposit insurance. When confidence slips, lenders demand cash just as assets become illiquid, producing systemic “runs.” March 2020 proved that central banks ultimately supply dollars to restore market function, albeit via improvised facilities. A formal LOLR for shadow banking should therefore be explicit, collateralbased and penaltypriced to reduce chaos and moral hazard.
影子中介机构——货币市场基金、交易商回购账簿、非银行抵押贷款服务者——创造没有存款保险的类货币债权。当信心下降时，贷款者要求现金，而资产变得不流动，产生系统性的“挤兑”。2020 年 3 月证明，央行最终通过临时设施提供美元以恢复市场功能。因此，为影子银行设立正式的无保本资金救助应该是明确的、基于抵押品和惩罚定价的，以减少混乱和道德风险。

Design principles:
设计原则：
Eligibility. Access for leveraged nonbanks performing core funding roles (MMFs, primary dealers, CCPs) contingent on registration, disclosure and supervisory oversight.
资格。杠杆非银行机构在执行核心融资职能（货币市场基金、主要交易商、中央对手方）的访问权限取决于注册、披露和监管监督。
Collateral. Highquality marketable securities with conservative haircuts; nonbanks must preposition collateral with centralbank triparty agents.
抵押品。高质量的可交易证券，采用保守的折扣率；非银行机构必须向中央银行三方代理预先提供抵押品。
Pricing. Abovemarket but below stressed rates to retain deterrence.
定价。高于市场价但低于压力价，以保持威慑力。
Backstop only. Facilities activate under systemicstress declaration; routine funding remains private.
仅止于备用措施。系统性压力声明下设施被激活；常规资金仍为私人提供。
Resolution linkage. Participants accept livingwill and bailin frameworks to handle insolvency.
决议挂钩。参与者接受生存遗嘱和破产清算框架来处理破产。

This structure mirrors the Fed’s 2020 PDCF and MMFLF, institutionalising them exante to reduce stigma and delay. Critics fear subsidy creep; yet moral hazard is lower than in adhoc rescues because eligibility and pricing are predefined. The alternative—unplanned intervention—proved costlier in 2008.
该结构类似于美联储 2020 年的 PDCF 和 MMFLF，事前制度化以减少耻辱感和延迟。批评者担心补贴蔓延；然而，道德风险低于临时救助，因为资格和定价是预先定义的。替代方案——非计划干预——在 2008 年证明成本更高。

4 │ Dangers of overregulating shadow banking ≈ 460 words
4 │ 过度监管影子银行的危害 ≈ 460 字
Excessive constraints can stifle market liquidity, drive activity into murkier niches, and impair credit supply. If repo haircuts are fixed high irrespective of collateral quality, marketmaking shrinks, widening bidask spreads. Heavy capital surcharges on finance companies could deny SMEs alternative funding, undermining growth. Regulatory arbitrage may intensify: activity shifts from lightly regulated MMFs to even less transparent cryptostablecoins, heightening, not reducing, systemic risk. Policymakers must balance functional equivalence (equal risk, equal rule) with proportionality to prevent a credit vacuum. The RBA’s 2015 review illustrates a lighttouch stance where Australia’s NBFI footprint is small and bank linkages limited .
过度约束会抑制市场流动性，将活动引向更不透明的细分领域，并损害信贷供应。如果无论抵押品质量如何，回购利率都固定在高水平，做市活动将萎缩，买卖价差将扩大。对金融公司的重资本附加费可能会剥夺中小企业的替代资金来源，从而破坏增长。监管套利可能会加剧：活动从监管较松的货币市场基金转向甚至更不透明的加密稳定币，这会加剧而非减少系统性风险。政策制定者必须在功能等同性（同等风险，同等规则）与比例性之间取得平衡，以防止信贷真空。RBA 的 2015 年审查说明了轻触立场，即澳大利亚的非银行金融机构足迹较小，银行联系有限。

5 │ Private equity: systemic risks and policy options ≈ 490 words
5 │ 私募股权：系统性风险和政策选择 ≈ 490 字
Privateequity (PE) funds use high leverage to acquire firms, restructure them and exit via sale or IPO. Nathanaël Benjamin (BoE, 2024) highlights two macroprudential concerns: (i) leverage & refinancing risk. Portfolio companies often carry debt >6 × EBITDA; a rates shock can trigger defaults spilling into CLO markets held by banks and insurers. (ii) liquidity mismatch. PE funds promise capital calls and distributions on multiyear horizons, but secondarymarket demand evaporates in downturns, trapping investors (pension funds, insurers) and forcing asset sales elsewhere.
私募股权（PE）基金利用高杠杆收购公司，进行重组并通过出售或首次公开募股退出。纳撒尼尔·本杰明（英格兰银行，2024 年）强调了两个宏观审慎问题：（i）杠杆与再融资风险。投资组合公司通常承担的债务超过 EBITDA 的 6 倍；利率冲击可能引发违约，并蔓延至银行和保险公司持有的信贷贷款市场。（ii）流动性错配。私募股权基金承诺在多年时间内进行资本募集和分配，但在经济衰退期间，二级市场需求消失，投资者（养老基金、保险公司）被套牢，并被迫在其他地方出售资产。

Policy responses: enhanced data (Form PFstyle reporting of NAV, leverage, margin), stresstesting of bank exposures to CLOs, and macroprudential buffers for lenders with high PE concentration. Some propose leverage caps or minimum liquidity lines at fund level; critics worry this blunts PE’s restructuring benefits. A pragmatic route is transparency plus lenderside capital, allowing markets to price risk while insulating core institutions. The BoE is piloting such monitoring; APRA could emulate for Australian super funds’ PE books.
政策响应：增强数据（NAV、杠杆、保证金风格的 PF 报告），对银行持有的信贷贷款市场风险进行压力测试，以及对具有高私募股权集中度的贷款人的宏观审慎缓冲。一些人提出在基金层面设置杠杆上限或最低流动性线；批评者担心这会削弱私募股权重组的益处。一条务实的途径是透明度加上贷方资本，允许市场定价风险，同时保护核心机构。英格兰银行正在试点此类监控；澳大利亚监管机构（APRA）可以为澳大利亚超级基金的私募股权账簿效仿。

7 │ Should stablecoin issuers be brought inside the prudential perimeter? ≈ 495 words
7 │ 应该将稳定币发行方纳入审慎监管范围吗？≈ 495 字

Stablecoins such as USDC and Tether create dollardenominated liabilities that circulate on public blockchains and are widely used for payments and collateral in crypto markets. The economic function is strikingly similar to bank deposits or moneymarketfund shares: users hold the token as a store of value redeemable at par on demand, while the issuer invests the backing reserves. As the 2015 RBA Bulletin explains in a broader “shadow money” context, whenever a private intermediary issues a shortterm liability perceived as cashequivalent, maturity transformation and run risk emerge . Two recent shocks demonstrate systemic externalities. In May 2022 the algorithmic TerraUSD stablecoin collapsed, transmitting losses to wholesale crypto markets; in March 2023 USDC briefly “broke the buck” after Silicon Valley Bank—custodian of its reserves—failed, freezing DeFi lending protocols for a weekend.
美元计价的稳定币，如 USDC 和 Tether，在公共区块链上流通，并广泛用于加密市场的支付和抵押。其经济功能与银行存款或货币市场基金股份非常相似：用户持有代币作为可随时按面值赎回的价值储存，而发行方则投资于担保储备。正如 2015 年 RBA 公报在更广泛的“影子货币”背景下所解释的，每当私人中介机构发行被视为现金等价的短期负债时，期限转换和挤兑风险就会出现。最近两次冲击展示了系统性外部性。2022 年 5 月，算法稳定币 TerraUSD 崩溃，将损失传递到批发加密市场；2023 年 3 月，由于保管其储备的硅谷银行失败，USDC 短暂“跌破一美元”，冻结了 DeFi 借贷协议一个周末。

Rationale for prudential regulation
监管合理依据

Paymentsystem stability. If a large stablecoin loses its peg during macro stress, users may rush to redeem into bank deposits, creating liquidity strains akin to MMF runs.
支付系统稳定。如果大型稳定币在宏观压力下失去其锚定，用户可能会急于赎回银行存款，造成类似于 MMF 挤兑的流动性压力。

Financialmarket plumbing. Stablecoins collateralise billions in leveraged crypto positions; a disorderly unwind could spill into brokerdealer liquidity via hedgefund channels.
金融管道。稳定币为数十亿美元的杠杆加密头寸提供抵押；无序的 unwind 可能通过对冲基金渠道蔓延至经纪商的流动性。

Consumer protection. Retail holders lack transparency on reserve composition or governance; some issuers hold commercial paper or uninsured bank deposits.
消费者保护。零售持有人缺乏储备构成或治理的透明度；一些发行商持有商业票据或不保险的银行存款。

Design of the perimeter should follow a “narrow bank” model:
边界的设计应遵循“窄银行”模式：

1:1 highquality liquid assets (HQLA). Reserves limited to centralbank deposits and shortdated Treasuries, audited daily and disclosed in real time.
1:1 高质量流动资产（HQLA）。储备仅限于央行存款和短期国债，每日审计并实时披露。

Segregated trusts & redemption rights. Legal segregation protects users if the issuer fails—mirroring the structural fixes adopted for MMF reform.
分离信托与赎回权。法律分离保护用户，如果发行方失败——这反映了为货币市场基金改革采取的结构性修正。

Licence & supervision. Issuers obtain a prudential licence and report under Pillar 3style transparency; leverage, lending or proprietary trading with reserves is prohibited.
许可证与监管。发行方获得审慎许可证，并按照第三支柱风格透明度报告；禁止利用、放贷或使用储备进行自营交易。

Emergency liquidity. Access to a repo backstop against sovereign collateral at penalty rates replaces implicit “toocryptotofail” rescues.
紧急流动性。获得对主权抵押品惩罚利率的回购后备，取代了隐性的“加密货币太容易失败”救援。

Costs and objections. Industry warns that heavy regulation could stifle innovation or push issuance offshore. Yet the same argument was levied against banking regulation a century ago. A tiered model—lighter rules for tokens used purely within closed crypto networks, full prudential regime for systemwide payment tokens—preserves experimentation while containing systemic risk.
成本与异议。行业警告称，严格的监管可能会扼杀创新或将发行推向海外。然而，同样的论点一个世纪前曾针对银行监管提出。一个分层模型——对仅在封闭加密网络中使用的代币实行较轻的规则，对系统级支付代币实行全面审慎监管——既保留了实验性，又控制了系统性风险。

In sum, stablecoins replicate core deposit functions and therefore warrant equivalent, though technologically adapted, prudential oversight. Bringing them into the perimeter converts an adhoc rescue certainty into a rulebound safety net, protecting both users and the wider financial system.
总而言之，稳定币复制了核心存款功能，因此应享有同等（尽管在技术上有所调整）的审慎监管。将它们纳入监管范围将临时的救援确定性转变为有规则的安全网，保护用户和更广泛的金融体系。

8 │ Do FSB minimum haircut floors threaten repomarket liquidity? ≈ 475 words
8 │ FSB 最低 haircut 门槛威胁回购市场流动性？≈475 字

The 2007–09 crisis revealed how thin repo haircuts allowed dealers and hedge funds to lever up on ostensibly safe collateral, amplifying firesale spirals when margins jumped. In response, the FSB proposed numerical haircut floors for noncentrallycleared securitiesfinancing transactions (SFTs): 1–2 % on sovereigns, 4–7 % on investmentgrade corporates, 10 % on equities. The rationale parallels Basel III capital floors—prevent a race to the bottom when markets are calm.
2007-09 年危机揭示了回购 haircut 的薄弱如何允许交易商和对冲基金在表面上安全的抵押品上杠杆操作，当利差扩大时放大了抛售螺旋。为此，FSB 针对非中心化清算证券融资交易（SFTs）提出了数值 haircut 门槛：主权国家 1-2%，投资级公司 4-7%，股票 10%。其理由与巴塞尔 III 资本门槛类似——在市场平静时防止恶性竞争。

Liquidity concerns arise because repo markets are the backbone of price discovery and monetary transmission. If floors are set too high relative to prevailing market practice, dealers may retreat, widening bidask spreads and raising funding costs for realeconomy borrowers. Early evidence, however, is reassuring. Pilot implementations in Switzerland and Singapore report negligible volume impact for highquality collateral; spreads widened temporarily before adjusting as borrowers shifted to slightly longer maturities. The key is proportional calibration: floors should sit modestly above prestress averages, leaving room for marketdriven differentiation while capping leverage extremes.
流动性问题出现，因为回购市场是价格发现和货币传导的支柱。如果地板相对于市场通行做法设置得太高，交易商可能会撤退，扩大买卖价差，并提高实体经济借款人的融资成本。然而，早期证据是令人鼓舞的。瑞士和新加坡的试点实施报告称，对于高质量抵押品，交易量影响可以忽略不计；在借款人转向略长期限之前，价差暂时扩大，然后随着调整而缩小。关键在于比例校准：地板应该适度高于压力前的平均水平，为市场驱动的差异化留出空间，同时限制杠杆极端情况。

Dynamic design can further mitigate liquidity risks. Floors could be linked to volatility indices (e.g., VIX) or marketdepth metrics, rising when exuberance compresses haircuts and easing once private margins widen, thereby leaning against the leverage cycle. Such rules would mimic countercyclical capital buffers.
动态设计可以进一步缓解流动性风险。地板可以与波动率指数（例如，VIX）或市场深度指标挂钩，在狂热压缩折扣时上升，一旦私人利润扩大就放松，从而抑制杠杆周期。这样的规则将模拟逆周期资本缓冲。

Complementary measures: (i) grant central counterparties preferential access to centralbank intraday liquidity so that cleared repo remains a lowhaircut channel; (ii) improve collateral interoperability so that funds can switch from equity to sovereign collateral without exiting the repo market. These steps preserve marketmaking activity while accepting the social benefit of thicker initial margins.
辅助措施：(i) 授予中央对手方优先访问中央银行日内流动性的权利，以便清算回购仍然是一个低折价渠道；(ii) 提高抵押品互操作性，以便资金可以在不退出回购市场的情况下从股票转向主权抵押品。这些步骤保留了做市活动，同时接受了更厚初始边界的社交效益。

The costs of overcalibration—reduced secondarymarket depth—are real but, in the FSB’s judgement, smaller than the macro cost of another leveragefuelled spiral. Ongoing empirical reviews every three years, as mandated by the FSB, ensure recalibration if adverse sideeffects emerge.
过度校准的成本——二级市场深度减少——是真实的，但在 FSB 的判断中，这些成本小于另一轮杠杆驱动螺旋的宏观成本。根据 FSB 的规定，每三年进行一次持续的经验性审查，如果出现不利副作用，将进行重新校准。

9 │ Are openended bond funds the next systemic weak spot? ≈ 485 words
9 │ 无固定期限债券基金是下一个系统性薄弱环节吗？≈485 字

Openended bond funds (OEBFs) promise daily redemption at net asset value, yet they invest in often illiquid corporate or emergingmarket debt. This liquidity mismatch creates a classic firstmover advantage: early redeemers exit at par, while remaining investors bear firesale losses. March 2020’s COVID dashforcash saw US$250 billion flow out of OEBFs, forcing indiscriminate asset sales until the Fed launched bondpurchase facilities. The experience mirrors moneymarketfund runs in 2008, signalling systemic relevance.
开放式债券基金（OEBFs）承诺每日以净资产价值赎回，但它们投资于通常流动性较差的工商企业或新兴市场债务。这种流动性不匹配创造了一个经典的先发优势：早期赎回者按面值退出，而剩余投资者承担抛售损失。2020 年 3 月的 COVID 紧急抛售潮导致 2500 亿美元资金流出 OEBFs，迫使无差别资产出售，直到美联储启动债券购买设施。这一经历与 2008 年的货币市场基金挤兑相似，表明其系统性相关性。