THE EFFECT OF BANKS'' HEALTH ON INVESTMENT IN JAPAN

The Effect of Banks'' Health On Investment In Japan During The Banking Crisis Of 1997-1998.

The Effect of Banks'' Health on Investment In Japan During The Banking Crisis Of 1997-1998.

Table of Contents

Chapter 33

Empirical Methodology3

Empirical Model7

Chapter 410

Empirical Results Descriptive Statistics10

Chapter 514

Basic Results14

Chapter 616

Robustness Checks16

Chapter 724

Discussion24

References27

Data Appendix32

Appendix 235

Diffusion Indeces35

The Effect of Banks'' Health on Investment In Japan During The Banking Crisis Of 1997-1998.

Chapter 3

Empirical Methodology

We study the effects of bank health on the level of the productivity of manufacturing firms during the 1990s. Specifically, we examine the manufacturing firms listed on any of the three major stock exchanges in Japan, Tokyo, Osaka, and Nagoya for FY1997— FY1998 and additionally 1999. There were more than 1300 listed manufacturing firms during the sample period. We use the TFP as a measure for the productivity of firms. Following the approach of (Gibson. 1995) we construct a hypothetical representative firm in the entire manufacturing industry in each year. The TFP level of each firm in each fiscal year is measured relative to that of a hypothetical firm in FY1990. The TFP level of a firm i in year t relative to that of a hypothetical firm in year 0 (the base year) is calculated as in the following equation:

Where ln Yit is the logarithm of the output of firm i in year t, Sk,it is the cost share of the input k of firm i in year t, and lnXk,it is the logarithm of the input k of firm i in year t. ln Yt, Sk,t, and lnXk,t denote the averages of ln Yit, Sk,it, and lnXk,it, respectively, across all the manufacturing firms in year t. We assume a hypothetical manufacturing firm with ln Yt, Sk,t, and lnXk,t in year t. Thus, in equation (1), the first two terms describe the difference between firm i and a hypothetical firm in year t while the last two terms chain together the hypothetical firms back to the base year. The output of a firm is the total sales. The inputs of a firm are intermediate input, labor, and capital.

The changes in working hours and capital utilization can affect the estimation of the productivity of firms. Unfortunately, we cannot obtain the firm-level data on working hours and capital utilization. We partially control for the effects of these factors by using the sector-level data. We focus on the relationship between firms and their main banks because many researchers argue that main banks play a special role in the case of Japanese firms (Aoki and Patrick, 1994). However, identifying a main bank for each firm is not an easy task, because there is no rigid definition of a main bank. Following Gibson (1997) and Hori et al. (2004), we identify a main bank for each firm by using Kaisha Shikiho (Japan Company Handbook), published by Toyo Keizai. The book reports the list of banks that each firm deals with. The first bank in the list is identified as the ...