A Theoretical and Empirical Study on Structural Change in Flow-of-Funds of the Japanese Economy in the 1990s

本文从银行行为、企业行为和银企关系三个角度考察了20世纪90年代日本经济资金流的结构性变化。我们总结了我们研究的主要发现。首先，我们研究了在2001年3月至2006年3月采取的数量宽松政策下，银行超额准备金的决定因素。利用银行面板数据集，从银行的预期利润最大化出发，估计了商业银行超额准备金的需求函数。我们发现，由于脆弱的金融体系和接近零利率的情况下，银行持有超额准备金。其次，我们调查了日本央行短观调查中报告的金融机构贷款态度的决定因素。估计是根据公司规模和行业分别进行的。我们特别感兴趣的是银行资产负债表恶化对银行贷款态度的影响。我们发现银行业不良贷款的增加导致金融机构贷款态度的严谨性。首先，应该指出的是，公司的外部融资溢价在很大程度上取决于贷方和借款人之间的信息不对称程度或代理成本。外部融资溢价本质上是不可观测的，但我们尝试使用贝叶斯统计工具对其进行估计。我们发现，在泡沫时期代理成本较低，泡沫破灭后代理成本迅速上升。我们还发现，企业投资与代理成本的变动密切相关。其次，我们考察了企业在调整投资计划时，企业投资会对什么样的信息做出反应。我们特别研究了投资修正是否对股票市场的评价敏感。我们发现企业投资的修正对销售和利润的修正比较敏感，而对股票市场事件的修正比较敏感。银行与公司的关系在日本，人们常常认为公司的信息集中在一家银行或其主要银行。虽然公司的信息集中在其主要银行，但公司在许多情况下与多家银行有关系。基于大型上市公司和小型非上市公司的微观数据，研究了企业与银行存在多重关系的原因。我们发现，主行及其关联公司的资产负债表状况会影响银行关系的数量。少

We investigated the structural change in flow of-funds of the Japanese economy during the 1990s from the following three viewpoints: bank behavior, firm behavior and bank-firm relationship. We summarize our main findings of our research.1. Bank BehaviorFirst, we investigated the determinants of banks' excess reserves under the quantity easing policy adopted in the period of March in 2001 to March in 2006. We estimated the demand function for excess reserves by commercial banks derived from expected profit maximization by banks, using bank panel data set. We found that banks held excess reserves due to fragile financial system and almost zero interest rate situations. Secondly, we investigated the determinants of lending attitude of financial institutions reported in the BOJ Tankan survey. Estimation was conducted separately by firm size and industry. In particular we were interested in the impact of deteriorating balance sheet of banks on lending attitude of banks. We found that mounti … More ng non-performing loans in the banking sector led to severity of lending attitude of financial institutions.2. Firm BehaviorFirst, it should be noted that external finance premium for the firm crucially hinges upon the degree of asymmetric information between lenders and borrowers or agency cost. The external finance premium is intrinsically unobservable, but we made an attempt to estimate it by using Bayesian statistical tools. We found that the agency cost was lowered in the bubble period and then it increased rapidly after the bubble collapse. We also found that business investment was closely related to the movement of agency cost. Second, we investigated what kind of information business investment responds to when the firm adjusts its investment plans. In particular we examined whether investment revision is sensitive to evaluation in stock market. We found that revision of business investment is sensitive to revision of sales and profits rather than stock market events.3. Bank-Firm RelationshipIn Japan it is often argued that information of the firm is concentrated on one bank or its main bank. Although information of the firm is concentrated on its main bank, the firm has in many cases multiple bank-firm relationships. We investigated why the firm has multiple relations with banks based on micro-data of large listed firms and small unlisted firms. We found that balance sheet conditions of the main bank as well as its affiliated firms affected the number of bank relations. Less