Positive Analyses of Information System Investment in Money and Banking Industry

1997年，我们分析了27家日本银行对我们邮寄问卷的回答所获得的1995年的横截面数据。以信息系统开发成本和净利润为因变量，以存款和人均信息系统开发成本为自变量，估计了两条显著达95%水平的回归曲线。1998年和1999年，东京证券交易所只有50%的日本银行在向财务省提交的财务报告中披露了其软件资产。在所有银行的数据中，它们的软件资产与员工数量的正相关性最大。2000年，我们对33家银行1995-1998年的邮寄问卷和访谈的回答进行了面板数据分析。如果将计算机软件和硬件归结为计算机资产，得出的结论是1日元的计算机相关资产与10日元的市场价值的关系最好，这一结果与日本对计算机革命的传统看法不一致。首先，计算机软硬件投资对各家银行市值的积极影响并不像他们所说的那么大。第二，与美国宏观经济学家的发现相反，计算机软硬件投资对资本生产率的影响并不是负的，也不像他们所认为的那么小；第三，美国微观经济学家高估了计算机软硬件投资对市场价值的影响。如果他们采取和我们一样的做法，日本和美国之间的差异可以忽略不计。

In 1997 we analyzed the 1995 cross section data taken by the answers of 27 Japanese banks to our mailing questionnaires. Employing the developing costs of information system and net profit as dependent variables and taking deposit and the developing costs of information system par capita as independent variables we estimated two regression curves which were significant at 95% level.In 1998 and 1999 we found that only 50% of Japanese banks in Tokyo stock exchange disclosed their software assets on their financial reports to the ministry of finance. We concluded that the positive correlation between their software assets and the number of employee took the maximum one among the data of all banks.In 2000 we took panel data analyses to the answers of 33 banks to our mailing questionnaires and interviews during 1995-1998. We concluded that one yen of computer related assets had the best relation to ten yen of market value of each bank if we summarized computer software and hardware as computer assets.Above results are not consistent with the conventional opinion to the computer revolution in Japan. First the positive effect of computer software and hardware investment to the market value of each bank are not so large as they said. The estimated effect took the same level of manufacturing industries in the United States.Second, on the contrary to the fact found by macroeconomists in the United States, the effect of computer software and hardware investment to the capital productivity are not negative and not so small as they have insisted.Third the effect of computer software and hardware investment to the market value is overestimated by the microeconomists in the United States. If they took the same approach as we did, the difference between Japan and the United States are negligible small.