Are Japanese Companies Subject to Natural Selection or Unnatural Selection?

UESUGI Iichiro
Fellow, RIETI

Many people say that the economy's "entry-exit system" has failed to function properly for Japanese companies from the latter half of the 1990s until recently. Normally, companies that are supposed to be forced out of the market exit while those that are supposed to stay in the market continue to exist. But critics say the opposite is happening in Japan. If true, this means "unnatural selection," rather than natural selection, is taking place in the Japanese corporate sector. The perverse incentives of Japanese banks have been cited as one reason behind such unnatural selection. Hoping to avoid bringing to light the crisis of borrowing companies, banks have an incentive to continue to lend or to expand lending to such troubled companies despite the fact that it is unlikely they can successfully reconstruct their businesses. This creates an environment that keeps virtually bankrupt companies alive.

Unnatural selection of major companies versus natural selection of SMEs

Previous studies have gathered a substantial volume of evidence underlining the presence of unnatural selection. It therefore seems certain that such a phenomenon is actually taking place in the Japanese economy, partly due to banks' behavior. Previous studies have found the amount of borrowing tends to increase at poorly performing companies (i.e., those with high debt ratios, low productivity and so forth). Based on these findings, analysts insist that unnatural selection is taking place against the backdrop of banks' lending behavior.

Despite this, it is premature to conclude that such unnatural selection is occurring extensively in the Japanese economy. First of all, most of the analyses carried out to date have targeted listed companies, thus leaving out most small- and medium-size enterprises (SMEs), which account for 99 percent of Japanese companies and employ 60% to 70% of the working population of Japan. Meanwhile, from the viewpoint of banks, which are said to be promoting unnatural selection, the greatest concern is not the level of debt ratios or the productivity of borrowing companies, but their capability to continue to pay principal and interest; that is, whether these companies will stay in business or not. Of course, banks suffer huge losses if these borrowers actually go bankrupt but as long as they continue to make repayments, it is rational for banks to continue lending to these troubled companies. In this respect, however, very few attempts have been made to compare companies doomed to fail with those that survive because bankruptcy is rare among the major companies that have been the focus of previous studies.

A RIETI study group on corporate finance empirically examines the current state of corporate finance, in particular for SMEs. As part of this project, we have analyzed the process of selection with respect to SMEs. Specifically, based on panel data containing information from the financial statements of a total of 1.2 million SMEs for the period from 1997 to 2002, we examined the rate at which doomed-to-fail companies were borrowing from the debt market and how these companies' exit from the market affected the interest rates paid by surviving companies. This data set, including data on several thousand failed companies for each year, enables statistically reliable analysis. The failed companies referred to here include those with delinquent repayments for six months or longer, as well as those that have de facto or actually gone bankrupt.

Our analysis has found the following:

  • The interest rates applied to doomed-to-fail companies - those that actually went bankrupt in the following year - were significantly higher than those applied to surviving companies (see figures below).
  • Because of the exit of failed companies, which had been paying relatively high interest rates, the rates applied to surviving companies gradually decline over time.

Figures: Differences in interest rates applied between surviving companies and those doomed to fail in the following year
(The vertical axis shows the interest rate while the horizontal axis shows the age of companies.)

Figures: Differences in interest rates applied between surviving companies and those doomed to fail in the following year

The figures show the interest rates paid by each group of companies established in a given year gradually decline with the number of years in operation. (Companies are classified into cohorts by the year in which they were established, with a separate line for each cohort. The companies to the right of the horizontal axis are older.) Also, by comparing the two figures, we can see that the rates applied to companies doomed to fail in the next year are higher than those paid by those that survived. These results indicate corporate borrowers are gradually subjected to higher interest rates in accordance with the degree of credit risk they pose before they are eventually forced into bankruptcy. In this sense, we can assume natural selection is functioning properly. Admittedly, the question remains whether the interest rates banks apply to doomed-to-fail companies reasonably reflect the degree of credit risk the respective borrowers pose. For instance, with regard to companies that would fail with 100% certainty, banks should demand an infinitely high rate of interest or try to immediately withdraw outstanding loans because it is virtually meaningless to receive a premium of less than 1% in such a situation. Nevertheless, we can still say that banks are fostering natural selection among borrowing companies in the sense that lending banks, using information not necessarily included in the financial statements of borrowers, identify those close to bankruptcy and charge them higher rates of interest. Also, it should be noted that such natural selection was at work from the latter half of the 1990s through the first few years of this decade, a critical period during which Japanese banks supposedly had the greatest incentive to engage in forbearance lending to failing companies with little prospect of successful reconstruction, a practice that fosters unnatural selection.

Issues requiring further analysis to get a better picture of corporate selection

There remain some issues that require further analysis. First of all, not all SMEs are subjected to natural selection and we need to understand the reasons for this. For instance, in the real estate sector, which is one of the most conspicuous exceptions to natural selection, interest rates applied to exiting companies are significantly lower than those paid by surviving companies. Is this phenomenon attributable to the peculiar nature of the real estate industry, in which many asset management companies are set up for the purpose of facilitating smooth business succession, thereby creating unique factors influencing the exit of real estate companies - factors that are not observed in other sectors? Or is it because of factors on the side of lending banks, which often waive interest payments partially or completely on existing loans in addition to providing forbearance loans? The answer to these questions has yet to be found.

Second, apart from natural selection, there are other factors that have a profound influence on the relationship between lending interest rates and the age of borrowing companies. Lending interest rates usually fall with the age of the borrower under the influence of natural selection, but this tendency is not observed in areas such as Tokyo and Osaka. In these areas, interest rates do not fall with the age of borrowing companies, due to factors other than natural selection. It is assumed that competition among banks is one such factor, but we have yet to determine the specific cause.

Finally, there is the question of how to account for two seemingly contradictory phenomena, namely, the unnatural selection of major companies as shown in a number of earlier studies and the natural selection of SMEs seen in our current research. Has unnatural selection of major companies played a more important role in determining the overall performance of the Japanese economy by exerting a strong influence through such companies' close relationships with their subcontractors, or has the natural selection of SMEs, which are far greater in number, been more important? This question also needs to be answered.

In order to clarify all these points, we plan to carry out further research in the corporate finance study group. For a discussion of the efficiency of corporate finance in the broader sense, covering not only the issue of corporate selection but also the process of interest rate determination to reflect risk, please refer to the Policy Analysis Paper entitled "Nihon no Kigyo-Kinyu ha Hikoritsu ka: Chusho-Kigyo no Kinri ni Motozuku Kensho (Is Japanese Corporate Finance Ineffective? Evidence Based on Interest Rates Applied to SMEs)," which will soon be published on the RIETI website.

May 17, 2005

Sakai, K., Uesugi, I., and Watanabe, T., 2005, "Natural or Unnatural Selection? Evidence from Japanese Credit Market in 1997-2002," (Unfinished manuscript).

May 17, 2005