Machine learning for financial transaction classification across companies using character-level word embeddings of text fields

An important initial step in accounting is mapping financial transfers to the corresponding accounts. We devised machine-learning-based systems that automate this process. They use word embeddings with character-level features to process transaction texts. When considering 473 companies independently, our approach achieved an average top-1 accuracy of 80.50%, outperforming baselines that exclude the transaction texts or rely on a lexical bag-of-words text representation. We extended the approach to generalizes across companies and even across different corporate sectors. After standardization of the account structures and careful feature engineering, a single classifier trained on 44 companies from 28 sectors achieved a test accuracy of more than 80%. When trained on 43 companies and tested on the remaining one, the system achieved an average performance of 64.62%. This rate increased to nearly 70% when considering only the largest sector.     

Operational risk quantified with spectral risk measures: a refined closed-form approximation

The quantification of operational risk has become an important issue as a result of the new capital charges required by the Basel Capital Accord (Basel II) to cover the potential losses of this type of risk. In this paper, we investigate second-order approximation of operational risk quantified with spectral risk measures (OpSRMs) within the theory of second-order regular variation (2RV) and second-order subexponentiality. The result shows that asymptotically two cases (the fast convergence case and the slow convergence) arise depending on the range of the second-order parameter. We also show that the second-order approximation under 2RV is asymptotically equivalent to the slow convergence case. A number of Monte Carlo simulations for a range of empirically relevant frequency and severity distributions are employed to illustrate the performance of our second-order results. The simulation results indicate that our second-order approximations tend to reduce the estimation errors to a great degree, especially for the fast convergence case, and are able to capture the sub-extremal behavior of OpSRMs better than the first-order approximation. Our asymptotic results have implications for the regulation of financial institutions, and may provide further insights into the measurement and management of operational risk.     

严监管背景下农信社合规风险管理体系建设

2018年,银行业监管环境进一步趋严,不但强调对银行的监管,更提出了对监管人员职责的监管,这种严格的外部监管环境对农信社的合规风险管理带来了极大的挑战。农信社合规风险管理体系的建设和完善已势不容缓。论文提出了农信社合规风险管理体系建设要从战略定位、文化渗透、防线搭建、专业专职化、考核和追究六个维度入手,推进合规风险管理长效机制的建设。     

The mechanism of credit risk contagion among internet P2P lending platforms based on a SEIR model with time-lag

The credit risk contagion of Internet peer-to-peer (P2P) lending platforms is an important part of Internet financial risk management and supervision. This study analyzes the contagion path of credit risk in Internet P2P lending. Based on complex network theory and the theory of infectious disease dynamics, the characteristics of Internet P2P lending development are combined to construct a SEIR model of credit risk transmission among Internet P2P lending platforms with time lag, and the robustness of the model is analyzed and proven. The influence of platform correlations, the susceptible immune rate, the platform elimination rate, contagion latency, the saturation coefficient, and the susceptibility input rate on credit risk contagion behavior among Internet P2P lending platforms is analyzed, using the equilibrium point and threshold value. The impact of each variable is analyzed by simulation. Corresponding countermeasures and suggestions are proposed to prevent and control credit risk contagion among these platforms.     

时间压力对网络冲动性购买的影响研究：交易效用和感知风险的调节

以往关于时间压力对冲动性购买影响的研究结论具有两面性，即促进或抑制。文章引入交易效用和感知风险两个调节变量，通过两个实验探讨了时间压力对网络冲动性购买倾向影响的边界条件。首先，当交易效用较高时，时间压力的增加会导致其网络冲动性购买倾向的提高；而当交易效用较低时，时间压力的增加会导致其网络冲动性购买倾向的降低（实验一）。其次，当人们感知风险较低时，时间压力的增加会导致其网络冲动性购买倾向的提高；而当人们感知风险较高时，时间压力的增加会导致其网络冲动性购买倾向的降低（实验二）。研究结论可以帮助电商企业进一步了解消费者的冲动性购买行为，充分利用时间压力的影响，制定更灵活有效的促销策略。     

Unemployment insurance benefits and income smoothing

Labor unemployment insurance reduces unemployment concerns. We argue that these benefits moderate incentives to smooth earnings to reduce employees’ concerns about unemployment risk. Using exogenous variations in unemployment insurance benefits, we find evidence consistent with this argument. We also find that the link between unemployment insurance benefits and income smoothing is stronger when there is higher unemployment risk and when the firm is likely to employ more low-wage workers, who find unemployment insurance benefits especially useful. Our paper contributes to the literature by showing that public policy decisions such as unemployment insurance have significant, albeit probably unintended, externalities on corporate financial reporting.     

A hybrid information approach to predict corporate credit risk

This study proposes a hybrid information approach to predict corporate credit risk. In contrast to the previous literature that debates which credit risk model is the best, we pool information from a diverse set of structural and reduced‐form models to produce a model combination based on credit risk prediction. Compared with each single model, the pooled strategies yield consistently lower average risk prediction errors over time. We also find that while the reduced‐form models contribute more in the pooled strategies for speculative‐grade names and longer maturities, the structural models have higher weights for shorter maturities and investment grade names.     

新发展格局视野中的消费券发放问题与重构研究

2020年面对新冠肺炎疫情冲击,各地纷纷发放消费券刺激消费。本轮发放消费券的省市和规模远超以往。考察当前各地消费券的实践,其存在三方面的问题:短期刺激消费与构建发展新格局缺乏衔接、地方各自为政与顶层设计阙如、财政补贴与财政约束的张力。针对这些问题,本文从新发展格局视野提出将当前消费券重构为财政和金融融合的消费金融券,使之成为构建新发展格局中扩大内需的长期政策和重要的宏观调控工具。消费金融券的制度设计要与社会结构新特征和新趋势相契合。针对低收入群体的消费金融券,应以财政补贴为主、以重大生活项目消费的利息补贴与信用担保为辅。针对中等收入群体的消费金融券,应以利息补贴为主,通过消费金融的方式扩大消费。加大消费金融券的全国顶层设计,明确中央和地方财政支持产业的范围和力度。