Development of Economical Evaluation Method on Information System Investment during Monetary Crisis

本研究首先认识到，信息技术资产对企业价值贡献的计量方法适用于行业信息系统投资的评估，而不适用于企业特定信息系统投资的评估。首先，我们选取了由实际投入和产出组合距生产前沿的距离定义的技术效率指数，对1993-99年间各银行多种资产与这种效率之间的相互作用进行了横截面分析。我们的统计计算机软件是经济计量软件公司的LIMDEP 8.0版。要达到这个效率，最困难的条件是如何定义银行的生产。我们尝试了五种定义来衡量它。当我们使用存款和净贷款总额的定义时，统计上有意义的案例最多：贷款减去不良贷款。研究还表明，计算机软件资产对这一效率始终具有稳定的正向影响。将该评价方法应用于个别银行的时间序列分析。其次，我们认识到个别银行的组织结构和人力资本是信息系统运行的决定性因素。我们开发了一种对这些因素的统计评估方法。我们以分层抽样的方式，向日本邮政各白化局邮寄了600份IT战略问卷。最终返回的答案数量为326个。对这些答案的交叉列表分析得出了我们的第三个发现，即在职培训和管理分权对于提高IT投资的技术效率非常重要。最后，介绍了日本四大银行集团的平衡计分卡，以实现IT投资评估的多元化。财务数据与包括IT指数在内的其他数据是相互影响的。这是我们的第四个发现。

This research started from the recognition that the measurement of information technology assets contribution to the firms' value in the stock market is proper to the evaluation of industries' information system investment but not to one of firms' specific information system. At first, we selected technical efficiency index defined by the distance of the actual inputs and output combination from the production frontier in order to take cross section analyses of the interaction between individual banks' several types of assets and this efficiency during the time of 1993-99. Our statistical computer software was LIMDEP ver.8.0, Econometric Software Inc. the most difficult condition to this efficiency was how to define the production of bank. We tried five types of definition to measure it. The largest number of the statistically significant cases was founded when we used the definition of total value of deposit and net loans: loans deducted by nonperforming loans. It was also cleared that the computer software asset always had stable positive effect to this efficiency. This evaluation method will be applied to time series analysis of individual bank.Second, we realized that organization constitution and human capital of individual bank were decisive factors to information system working. We developed a statistical evaluation method of these factors. We mailed 600 IT strategy questionnaires to the blanch bureaus of Japan Postal Service by stratified sampling. The final number of returned answers was 326. The analyses of cross tabulation for these answers derived our third finding that on the job training and management decentralization were important to improve the technical efficiency concerning IT investment. Finally, balanced scorecards of four largest Japanese banking groups were introduced in order to diversify IT investment evaluation. Financial data were reciprocal movement to the other data including IT index. This was our forth finding.