Efficiency analysis and prediction in the financial services industry

金融服务业效率分析与预测

• 批准号: 261426-2006
• 负责人: Yang, Zijiang
• 金额: $ 1.17万
• 依托单位: York University
• 依托单位国家: 加拿大
• 项目类别: Discovery Grants Program - Individual
• 财政年份: 2008
• 资助国家: 加拿大
• 起止时间: 2008-01-01 至 2009-12-31
• 项目状态: 已结题
• 来源: https://www.nserc-crsng.gc.ca/ase-oro/Details-Detailles_eng.asp?id=395013
• 关键词: Efficiency; analysis; prediction; financial; services

The financial services industry plays a vital role in the economic development and growth of a country. Its economic significance as well as the highly competitive market structure motivates research evaluating the industry's performance and monitoring its financial condition. Data Envelopment Analysis (DEA) is a very appropriate approach when considering the complex nature of the companies in the financial services industry since it directly incorporates multiple inputs and outputs, which means that the results will be explicitly sensitive to the complexity and mix of inputs and outputs. In addition, DEA can identify best practices and inefficient units by comparing their actual operating results without the need for any pre-defined production function.  Thus, DEA has won wide acceptance in the financial services industry. However, there are some limitations that have to be considered. Firstly, DEA can not deal with missing data. Secondly, the DEA frontier is very sensitive to the presence of the outliers and statistical noise and the measured efficiency scores can be contaminated by the observations on the frontier due to data errors or statistical noise. This motivates the further investigation of optimizing the frontier. Thirdly, DEA can hardly be used to predict the performance of other Decision Making Units (DMUs). This research will propose to integrate two methods, DEA and Neural Networks (NNs) to solve the above three issues. Furthermore, the proposed research will integrate fuzzy logic to the DEA formulation to address the cross-system comparison with multiple quantitative and/or qualitative environmental variables. As a result, the proposed research provides an innovative and unique approach to address key issues in DEA literature. It will open up new possibilities for performance analysis in the financial services industry. It is also the target of this research to gain further insight of the financial services industry as a whole in terms of its sustainable growth rate and potential.

金融服务业在一个国家的经济发展和增长中起着至关重要的作用。它的经济意义以及高度竞争的市场结构促使研究评估行业的业绩和监测其财务状况。数据包络分析（DEA）是一个非常合适的方法，当考虑到金融服务行业公司的复杂性时，因为它直接包含多个输入和输出，这意味着结果将对输入和输出的复杂性和组合非常敏感。此外，DEA可以通过比较实际运营结果来识别最佳实践和低效单位，而无需预先定义任何生产函数，因此DEA在金融服务业中获得了广泛的认可。但是，必须考虑一些限制。首先，DEA无法处理缺失数据。其次，DEA前沿对异常值和统计噪声的存在非常敏感，由于数据误差或统计噪声，测量的效率得分可能会被前沿上的观测值污染。这激发了对优化边界的进一步研究。第三，DEA方法很难用来预测其他决策单元的绩效。本研究将提出整合DEA与神经网路两种方法来解决以上三个问题。此外，拟议的研究将模糊逻辑集成到DEA的制定，以解决跨系统的比较与多个定量和/或定性的环境变量。因此，拟议的研究提供了一个创新和独特的方法来解决DEA文献中的关键问题。它将为金融服务业的绩效分析开辟新的可能性。本研究的目的也是进一步了解整个金融服务业的可持续增长率和潜力。