How to Interpret the "Excess Debt" Problem: Debt Discipline or Debt Overhang?

First of all, let's take a look at how the amount of debt held by listed companies and their investment activities have changed since the bursting of the asset bubble. As can be seen in Table 1 (click here for the table), debt to asset ratios ((bank loans + outstanding bonds) / total assets in market value) for non-financial companies listed on the first section of the Tokyo Stock Exchange (TSE) remained relatively stable in the 1990s on average, but inter-company variance expanded rapidly, especially from 1997 onwards. The variance between the first and third quartiles in debt to asset ratios expanded from 15 percent in 1990 to 23 percent in 2000 for manufacturers, and from 21 percent to 34 percent for non-manufacturers during the same period. This shows that as deflation continues, some companies are accelerating debt reduction while others are watching their debt to asset ratios go up.

To be sure, high debt ratios do not necessarily translate into an "excessive" debt burden. Here, let's a take a look at the changes in the ratios of net interest-bearing debt to cash flow ((interest-bearing debts - cash equivalent) / cash flows, hereinafter referred to as debt to cash flow ratio), which has been increasingly focused on in recent years. In the non-manufacturing sector, the median in debt to cash flow ratios had already reached 5.8 and the third quartile 20.8, by as early as 1993, after which there was a conspicuous fall in asset prices; the median and threshold values subsequently rose and reached 8.6 and more than 23.9 respectively by 1999. Even for manufacturers, whose debt to cash flow ratios have been, and still are, substantially lower than those of non-manufacturers, the third quartile exceeded 8 in 2000. What is indicated by these figures is that more and more companies are coming under an "excessive" debt burden relative to their earnings.

As variance in debt ratios expands, investment ratios are falling. Investment ratios for both the manufacturing and non-manufacturing sectors used to be around 20 percent in 1990. The ratios, however, fell to less than half that level after the 1997 financial crisis. It appears that the presence of "excess" debt is becoming a major contributing factor to the ongoing slump in investment by making it difficult for companies to raise fresh funds, a situation that can be described as debt overhang.

However, it should be noted that the presence of debt can serve a disciplinary function to suppress over-investment in a mature economy. That is particularly true when there is a high chance that a great proportion of corporate earnings on business activities will remain unspent and become surplus funds, or free cash flow, due to a lack of profitable investment opportunities. Indeed, some empirical studies have corroborated such a disciplinary debt function by using a set of data on American companies in the 1980s. If debts held by Japanese companies are actually serving such a disciplinary function, the current high debt level would not have to be regarded as a problem of "excess" debt. Thus, in working out appropriate policy measures for the ongoing situation in Japan, it is important to examine empirically whether debts held by Japanese companies are causing under-investment or simply restricting over-investment.

In approaching this issue, we tried to construct a model based on a set of data from the 1990s taken from financial institutions listed on the TSE first section, incorporating debt to asset ratios and debt to cash flow ratios for a standard investment function that uses "Tobin's q" as an explanatory variable. The detailed analysis results of the study [PDF:212KB], which was presented at a RIETI academic conference on corporate governance in January 2003, is published on the RIETI website along with comments by University of Oxford Professor Colin Mayer [PDF:48KB]. But let me briefly introduce the essence of our analysis results here. The intuitive image of the basic ideas underlying the analysis is illustrated in Chart I (click here for the chart).

In a textbook world where no frictions such as asymmetric information exist, the level of corporate investment is determined by the availability of business opportunities (Tobin's q) and debt ratios play no role. The level of investment would be fixed at I* (the optimum level of investment under a given set of business opportunities) and uninfluenced by DA (debt to asset ratios), that is, the investment axis would run in parallel with the bottom axis. In the real world, however, the level of corporate investment, almost without exception, shows negative sensitivity to debt to asset ratios, and I (the level of investment) as measured by the vertical axis deviates from I* depending on DA. Such deviations can take place both in under- and over-investment situations, that is, regardless of whether the actual level of investment is below or above the optimum level. The first case is one in which higher debt to asset ratios restrict a company in raising fresh funds because of the high agency costs or bankruptcy costs involved, thus, the actual level of investment by the company falls below the optimum level. This case is illustrated in the area below the I* axis in Chart I and can be observed among companies with many investment opportunities. But this is not the only case in which the actual level of investment demonstrates negative sensitivity to the presence of debt. There is another case, which is often seen in companies with few investment opportunities. In this case, the actual level of investment tends to deviate upwardly from the optimum level when a company has ample free cash flow and the presence of debt functions as a restrictive factor that curbs over-investment, a situation illustrated in the area above the I* axis in Chart I. Thus, here, the presence of debt plays a role in providing appropriate discipline to corporate managers.

Does the situation above or below the I* axis most closely resemble the actual relationship between Japanese companies' investment activities and their level of debt? That is the question we hoped to answer by estimating this investment function. The findings of our study are as follow:

First of all, manufacturers and non-manufacturers differed with respect to how their investment activities were influenced by the presence of debt. From 1993 through 2000, manufacturers' investment activities consistently showed negative sensitivity to the amount of outstanding debt at the beginning of each business year. In contrast, non-manufacturers' investment activities during the same period surprisingly showed positive sensitivity to debt, though not at a significant level. Looking more closely at data broken down by time period and industry, we found that investment activities by those in the construction, real estate and retail industries demonstrated positive sensitivity - this time, at a significant level - to the amount of debt outstanding at the beginning of business years between 1993 and 1996, a period immediately after the bursting of the economic bubble. In the context of the aforementioned chart, the relationship between investment and debt illustrated by the dotted line above the I* axis was confirmed. This indicates that companies with high debt to asset ratios - despite having few growth opportunities - expanded investments, gambling to break the stalemate as they watched asset prices keep on falling, and that the banking sector supported such activities by implementing what is called "soft budgetary constraints." It is highly possible that these reckless investments between 1993 and 1996, which were accompanied by a moral hazard (asset substitutions), and a lack of appropriate monitoring by banks are responsible for the high debt to cash flow ratios that are seen among non-manufacturers today.

Secondly, manufacturers' investment activities, unlike those of non-manufacturers, consistently showed negative sensitivity to the amount of debt at the beginning of each business year throughout the 1990s and the sensitivity conspicuously rose after 1997 when the financial crisis took place in Japan. Our results show that a rise by two standard deviations in debt to cash flow ratios resulted in a 1.8 points decrease in investment ratios, averaged at 10.1 points, during the pre-financial crisis period between 1993 and 1996. In the post-financial crisis period between 1997 and 2000, however, the same rise in debt to cash flow ratios was translated into a 2.7 percentage point drop in investment ratios, which stood at 9.1 percent on average. This means that investments' sensitivity to debt increased 1.5 fold between before and after the financial crisis.

The question is how we interpret such a negative correlation between investment and the level of outstanding debt. Should it be taken to mean that the presence of debt restricts over-investment or causes under-investment? To find this out, we divided our sample companies into two sub-groups - mature companies with low expectations for growth and companies with high expectations of growth - as measured at the beginning of the sample period. Then, based on calculations for each sub-group, we attempted to discern whether the situation for Japanese manufacturers falls in the case of debt restricting over-investment (above the I* axis) or that of debt causing under-investment (below the I* axis).

We found that of the negative correlations between investment and debt seen in Japanese manufacturers during the 1990s, those observed prior to the financial crisis can be interpreted as being primarily debt restricting over-investment. However, neither of the two cases adequately explained the situation after the financial crisis. Rather, we concluded that both phenomena were occurring simultaneously. That is, the impact of debt on investment, which strengthened after the financial crisis in late 1997, has begun to cause under-investment among mature companies in addition to restricting their over-investment.

As such, the situation that has been highlighted as an "excess debt" situation in recent years has turned out to be very diverse in nature. On the other hand, however, it is also true that many problematic companies do hold excess debt, a great proportion of which is attributable to attempts to debt finance their investments at excessive risk. It is likely that such companies, which characteristically sustain high debt to cash flow ratios over a long period of time and should have been liquidated at a much earlier stage, have been preserved for one reason or another. Thus, the focus in devising policy measures to cope with the ongoing situation should be on designing institutional arrangements that appropriately facilitate the orderly exit of such nonviable companies.

At the same time, however, we must not forget that high debt ratios are serving a disciplinary function to curb over-investment by mature companies. What is required of companies falling in this group is the quick rebuilding of operations. And the presence of debt does prompt them to make such a move. Therefore, in cases of those companies, there is no necessity to implement an ex-post reduction of the debt burden, for instance by creditor banks forgiving part of loans. Providing an easy way out by alleviating their debt burden would induce a moral hazard.

The results of our analysis, however, also suggest that there exist other cases in which the presence of excess debt weighs heavily on companies and prevents the realization of investment opportunities and where the presence of debt facilitates can be described as a process of "creative destruction." This particularly holds true for the period after the financial crisis in late 1997. It is highly likely that this kind of restriction (debt overhang), which has been pointed out as a problem primarily of small and midsize companies, has begun to spread to listed companies following the 1997 financial crisis. This possibility may also deserve more attention. Indeed, it is these companies for which debt reduction measures - in the form of partial forgiveness of loans and the use of debt-for-equity swap scheme by creditor banks and/or financing by quasi-governmental financial institutions - should be taken.

The very fact that problems of such a diverse nature are part of the ongoing situation referred to simply as the "excess" debt problem shows the highly complex nature of problems facing Japanese companies ever since the financial crisis. Given such circumstances, it is extremely important that the soon-to-be-launched Industrial Revitalization Corp. properly discerns the role played by debt and selects appropriate policy measures for each case it handles.