摘要伴随着利率市场化的深入推进，以及金融脱媒时代的到来，商业银行在管理负债结构方面更需要借助主动负债这以工具来实施管理。自2008年金融危机肇始，国际学界渐渐开始关注商业银行负债结构调整的经济影响。一些新近的理论推演和实证检验表明，银行主动负债管理可能会显著影响其风险承担、脆弱程度和经营稳健性等风险特征。当商业银行面临被动存款-信贷需求缺口时，往往依赖同业负债和金融债等主动负债工具来弥补，以继续扩张信贷规模和配置其他资产。一方面，主动负债可能增加银行风险暴露敞口 (Altunbas et al. , 2011)，降低银行经营稳健性 (Kohler, 2015)，推高银行风险承担程度；另一方面，负债提供方可能为银行提供严格外在约束和间接监督，继而降低其经营激进程度 (Huang & Ratnovski, 2011)，具体引致怎样的风险特征有待甄别，而其产生的确切的现实结果即为银行资产的质量特征。相较于银行实务的日新月异，目前主动负债管理的理论推演和实证检验偏少偏新，基本都是欧美银行业的经验概略和2008年金融危机启示录，而来自国内银行业的理论构架和经验证据均严重匮乏。相较于银行实务的日新月异，目前主动负债管理的理论推演和实证检验偏少偏新，基本都是欧美银行业的经验概略和2008年金融危机启示录，而来自国内银行业的理论构架和经验证据均严重匮乏。本研究旨在回答如下问题：银行的主动负债管理如何影响其资产质量？我们将先兆性的银行风险承担与现实性的银行资产质量区分开来，提供了新的分析维度，基于A股商业银行2008-2017年的面板数据，构造固定效应模型和中介效应检验模型，分析银行主动负债管理与其资产质量间的经济关联与渠道机制。首先，我们构建了主动负债比重与银行资产质量关系的基准模型，考察潜在经济关联及其非线性关系。其次，我们基于国有和非国有银行两组样本子集进行分组分析。再次，我们着眼银行风险承担，探讨前述经济关联中的渠道机制，构造了以风险承担为中介变量的层级回归分析模型组。研究结果表明，银行主动负债比重可以对其资产质量产生显著倒U型影响，也就是说相对于适当的主动负债比重，偏低或偏高的主动负债比重都将引致银行资产质量降低；与国有银行相比较的话，非国有银行依赖主动负债时将更容易降低其资产质量；中介效应分析显示，银行风险承担程度是主动负债管理对其资产质量影响中的重要作用渠道，且其中介部分在总效应中占比约为41.85%。本文的创新体现在：其一，区别风险承担与资产质量，识别主动负债管理经济影响内在逻辑。既有文献在银行风险或稳健性的衡量中，存在破产风险指数Z-score、不良贷款率等常见指标，其中Z-score是经验性和先兆性的，径直探讨银行的破产风险；而不良贷款率等指标具备现实性和精确性，是银行经营中积蓄、蔓延和呈现的现实结果，直接衡量银行资产质量。两类指标在实证检验中往往混淆使用，不做更进一步区分。本文将先兆性的银行风险承担与现实性的银行资产质量区分开来，提供了新的分析维度，并识别了风险承担在主动负债管理经济效应中的中介作用。其二，裨补主动负债管理研究，提供中国银行业主动负债管理经验证据。本研究从银行资产质量的现实结果出发，为主动负债管理的经济效应和渠道机制提供了新的经验证据。本研究从银行资产质量的现实结果出发，为主动负债管理的经济效应和渠道机制提供了新的经验证据。最后为促进银行合理有效地开展主动负债管理提出政策建议。关键词：主动负债管理；资产质量；风险承担；中介效应ABSTRACTUnder the current trend of interest rate liberalization and financial disintermediation, active debt is becoming an increasingly important debt structure management tool for commercial banks. Compared with the rapid development of banking practice, the theoretical deduction and empirical tests of active debt management are fewer and new, which are basically the experience of European and American banking industry and the apocalypse of the 2008 financial crisis, and the theoretical framework and experience from domestic banking industry are relatively scarce. When commercial banks face a passive deposit-credit demand gap, they often rely on active debt instruments such as interbank debt and financial bonds to make up for the expansion of credit scale and allocation of other assets. On the one hand, active debt may increase exposure to banks risks (Altunbas et al., 2011), reduce bank operations robustness (Kohler, 2015), and increase bank risk exposure; on the other hand, debt providers may provide banks with Strict external constraints and indirect supervision, and then reduce its operational aggressiveness (Huang & Ratnovski, 2011), it still needs to be identified of specific risk characteristics , and the exact actual results produced by them are the quality characteristics of bank assets.The purpose of this study is to answer the following question: How does the banks active debt management affect its asset quality? We distinguish the precursory bank risk exposure from the actual bank asset quality, and provide a new analytical dimension. Based on the panel data of a-share commercial banks from 2008 to 2017, we construct a fixed effect model and a mediation effect test model. The economic linkage and channel mechanism between the banks active debt management and its asset quality.The main work of this study is as follows. First, we constructed a benchmark model of the relationship between the proportion of active debt and the quality of bank assets, examining potential economic relationships and their nonlinear relationships. Second, we conduct group analysis based on a subset of the two groups of state-owned and non-state-owned banks. Thirdly, we focus on bank risk-taking, explore the channel mechanism in the aforementioned economic linkages, and construct a hierarchical regression analysis model group with risk-bearing as a mediator.The research results show that the proportion of active debt of banks can have a significant inverse U-shaped effect on their asset quality, that is to say, the proportion of active debt that are low or high relative to the proportion of appropriate active debt will lead to the deterioration of bank asset quality; Compared with banks, non-state-owned banks will be more likely to reduce their asset quality when they rely on active debt. The mediation effect analysis shows that bank risk-taking is an important channel for active debt management to affect its asset quality, and its mediation is The total effect is about 41.85%.Based on the realistic results of bank asset quality, this study provides new empirical evidence for the economic effects and channel mechanisms of active debt management, and complements the empirical test and theoretical discussion of relevant subject areas. Finally, we outline the conclusions of this paper and propose policy recommendations for promoting the rational and effective implementation of active debt management by banks. First, the bank level should change management co