I 摘要 近年来，随着我国居民用汇需求的持续增加，商业银行个人外汇管理也产生了 一些新的问题，特别是互联网时代下个人外汇分拆行为呈现出手段更隐蔽、办理时 区更分散、事后更难追查等新特点。个人外汇分拆行为不仅影响国家外汇管理稳定， 更会影响商业银行的监管表现、内部风险管理及业务平稳发展等。因此，顺应个人 外汇业务发展趋势和监管要求，提高商业银行反分拆管理水平至关重要。 本文在介绍研究背景和明确研究问题的基础上，首先梳理了个人外汇管理特 别是商业银行个人外汇业务中分拆行为的监管要求与政策背景。其次，以 B 银行 为典型案例，分析了结售汇、收付汇和现钞存取三种个人外汇业务中分拆行为的特 征与经济后果。再者，利用问卷调查，从事前、事中和事后三个层面，剖析了 B 银 行反分拆管理的现状与问题。最后，提出基于大数据技术的反分拆管理框架，并针 对三种业务中的分拆行为，给出相应的反分拆风险控制策略要点，并从业务风险、 运营效率和合规管理三个角度进行效果评价。 研究结论表明：基于大数据技术的反分拆管理模式，不仅能够实时识别和有效 拦截分拆行为，还可大幅降低识别差错率，具体效果体现为三个方面。一是实施大 数据反分拆管理的一年内，个人外汇业务中分拆笔数下降比例为 98.28%，涉及人 数也下降了 94.29%。二是分拆行为识别的问题差错率下降了 50%，事后核查频次 由原来的每季度一次提升到每天一次。三是银行外部合规管理效果及员工业务办 理积极性明显提升，监管年度考核评级从 B 级提升到 A 级，个人外汇业务办理网 点覆盖率从 53%上升到 79%，大幅提高了 B 银行个人外汇业务的市场占比。总体 而言，实施大数据反分拆管理模式，不仅使得 B 银行实现了提高分拆风险防范能 力和减轻柜面人工识别压力的双重目的，还有效地提升了银行公信力与合规管理 水平，并对商业银行个人外汇风险管理实践具有重要的借鉴价值。 关键词：个人外汇业务，大数据，分拆行为，反分拆管理ABSTRACT II ABSTRACT In recent years, with the continuous increase of residents’ demand for foreign exchange,some new problems have arisen onthe personal foreign exchangemanagement in commercial banks. In particular, the individual foreign exchange splitting behavior presents more new characteristics in the era of internet, such as more concealing approaches, more scattering time zones and more difficult to trace after the event. Individual foreign exchange splitting not only affects the stability of national foreign exchange management, but also affects the regulatory performance, internal risk management and the stable business development of commercial banks. Therefore, it’s crucial for commercial banks to enhance the capability of anti-splitting management to fulfill both the individual foreign exchange development requirement and regulatory requirement. Onthebasisofintroducingtheresearchbackgroundand definingtheresearchissues, this paper firstly sorts out the regulation requirement and policies background of individual foreign exchange management, especially the anti-splitting management of individual foreign exchange business in commercial banks. Secondly, taking Bank B as a typical example, this paper analyzes the splitting behavior, behavior characteristics and economic consequences of individual foreign exchange transactions including foreign exchange settlement and sale, collection and payment and cash deposit and withdrawal. Finally, this paper proposes an anti-splitting management framework based on big data technology, proposes corresponding anti-splitting risks control points for three types of individual foreign exchange splitting behaviors, and evaluates the effect from three perspectives of business risk, operational efficiency and compliance management. The research results show that the anti-splitting management model based on big data technology not only has the effect of real-time identification and inception of splittingbehavior,but alsocansignificantlyreduce theidentificationerrorrateof splitting transaction behavior, which is embodied in three aspects. First, within one year after the big data management model was launched, the number of splitting transactions of Bank B decreased by 98.28%, and the number of people involved also dropped by 94.29%. Second, the splitting identification error rate has been reduced by 50%. The frequency of post-verification checks increased from once a quarter to once a day. Third, theABSTRACT III effectiveness of the bank's external compliance management and the enthusiasm of staffs for business handling have been significantly improved. The annual regulatory assessment rating has been upgraded from Grade B to Grade A, and the coverage rate of individual foreign exchange outlets has increased from 53% to 79%, significantly increasing the market share of individual foreign exchange business of Bank B. In summary, the implementation of Bank B’s anti-splittingmanagement model with big data not only achieves the dual purposes of enhancing the risk prevention of splitting and reducing the pressure of manual identification at the counter, but also has the effect of enhancing the bank's credibility and compliance management, which has important reference value for the individual foreign exchange risk management of commercial banks. Key words：Individual foreign exchange business, Big data, splitting behavior, Anti- splitting management目 录 IV 目 录 第一章 绪论....................... 1 1.1 研究背景及意义... 1 1.2 研究内容与结构安排.......................... 2 1.3 研究方法............... 4 1.4 主要创新点........... 4 第二章 个人外汇分拆行为与反分拆管理的制度背景分析......... 5 2.1 相关概念................ 5 2.1.1 个人外汇业务............................. 5 2.1.2 分拆行为..... 5 2.2 我国个人外汇管理及监管制度变迁.. 6 2.2.1 个人外汇管理的相关政策......... 6 2.2.2 个人外汇监管制度的变迁........ 7 2.2.3 商业银行个人外汇管理面临的挑战........................ 8 2.3 个人外汇业务分拆行为的动机与后果.............................. 9 2.3.1 分拆行为的动机分析................ 9 2.3.2 分拆行为的后果分析................ 9 2.4 个人外汇业务反分拆管理.................. 9 2.4.1 反分拆的常用手段.................... 9 2.4.2 外汇反分拆管理的相关研究.. 10 2.5 本章小结............. 13 第三章 B 银行个人外汇的分拆行为分析 .... 14 3.1 B 银行简介 .......... 14 3.2 个人外汇业务中的分拆行为概况.... 14 3.2.1 结售汇业务.............................. 15 3.2.2 收付汇业务.............................. 15 3.2.3 现钞存取业务.......................... 16 3.3 个人外汇分拆行为的特征分析........ 17 3.3.1 资金流向... 17 3.3.2 资金用途... 18 3.3.3 办理渠道... 21目 录 V 3.4 个人外汇分拆的经济后果................ 22 3.4.1 影响国家的外汇管理.............. 23 3.4.2 影响银行的监管表现.............. 23 3.4.3 影响 B 银行的业务风险管理 . 24 3.4.4 影响个人用汇及触犯法律...... 25 3.5 本章小结............. 27 第四章 B 银行个人外汇反分拆管理的现状分析 ....................... 28 4.1 B 银行个人外汇业务的治理框架 ..... 28 4.1.1 部门设置... 28 4.1.2 内控制度... 29 4.2 B 银行个人外汇反分拆的管控流程 . 31 4.2.1 事前环节：识别客户身份及业务办理动机......... 31 4.2.2 事中环节：审核业务办理要素及资料真实性..... 31 4.2.3 事后环节：重点筛查大额分拆业务..................... 32 4.3 B 银行现有反分拆管理存在的问题 . 32 4.3.1 问卷调查与数据收集.............. 32 4.3.2 调查结果分析.......................... 33 4.3.3 事前环节：线上业务资金用途真实性识别存在难度........................ 34 4.3.4 事中环节：无法实时拦截具有分拆行为的业务. 34 4.3.5 事后环节：数据提取难度大，客户配合度低..... 35 4.4 本章小结............. 36 第五章 B 银行基于大数据的个人外汇反分拆管理及效果评价 .............................. 37 5.1 基于大数据的反分拆管理模式探讨 37 5.1.1 可行性分析.............................. 37 5.1.2 基本原理... 38 5.1.3 反分拆策略设计...................... 39 5.1.4 保障措施... 42 5.2 反分拆风险管理的效果分析............ 43 5.2.1 分拆行为的识别...................... 43 5.2.2 问题差错率.............................. 44 5.2.3 事后核查效率.......................... 44 5.3 内部运营的效率分析........................ 45 5.3.1 资源投入... 45目 录 VI 5.3.2 精细化管理程度...................... 46 5.4 外部合规管理的效果分析................ 46 5.4.1 外部公信力.............................. 46 5.4.2 监管处罚频率.......................... 46 5.4.3 经济犯罪事例.......................... 47 5.5 本章小结............. 47