Overlapping Indebtedness from the Viewpoint of Economics: Adequate restraint needed for new loans

UCHIDA Hirofumi
Professor, Kobe University

UESUGI Iichiro
Faculty Fellow, RIETI

In the aftermath of the Great East Japan Earthquake, there remain a host of problems that must be tackled as we proceed with the reconstruction of disaster-affected areas. Amongst all, the problem of overlapping indebtedness is a critical issue. Taking out new loans for post-disaster rehabilitation and reconstruction would create an overlapping indebtedness to those local companies and households that had already been in such situations before the quake hit, and the burden of repayment of both pre- and post-quake debts could cause serious disruptions to their businesses and lives. In this regard, the government has drawn up a set of measures and taken necessary legislative and budgetary actions to work out existing (old) debts and mitigate the burden of new ones.

Together with Arito Ono of Mizuho Research Institute, Kaoru Hosono of Gakushuin University, and Daisuke Miyakawa of the Development Bank of Japan, we examined the problem of overlapping indebtedness from the viewpoint of economics by specifically focusing on firms rather than on households. In what follows, we would like to introduce some of our findings.

♦  ♦  ♦

From the viewpoint of economics, we should first point out that the so-called "overlapping indebtness problem" actually consists of three different problems.

The first problem concerns the efficient use of funds. In terms of resource distribution, overlapping indebtedness is a problem in which a disaster-affected company is unable to obtain funds necessary for making a new start or a bank rejects a request for a loan that should have been made available. This is called a type I error. Here, the "loan that should have been made available" refers to a loan that has a reasonable prospect of repayment, that is, a case in which the net present value of the anticipated cash flow from business activity to be carried out with the loan is positive.

In terms of economics, old debts--those that had existed prior to the earthquake--are a sunk cost or an expense that has already been incurred. Thus, the presence of such prior debts must be excluded as factors for banks' decisions on new loans. Failure to provide loans to promising businesses simply because of the presence of old debts would impede progress in the rehabilitation and restoration not only of the affected companies but also of the affected areas.

The second problem is one in which, contrary to a type I error, a bank provides a loan that should not have been made available, a case in which the net present value of the anticipated cash flow is negative. This is called a type II error. Loans classified as such include: those to high-risk, poorly performing companies that are seeking funds in an attempt to make a miraculous recovery; those taken out with ill intentions; and those to borrowers who would not be diligent in their repayment efforts. As experienced with the special treatment of the so-called shinsai tegata or "quake-affected bills" following the 1923 Great Kanto Earthquake, post-disaster support measures designed to ease temporary cash flow problems may end up becoming a life-prolonging measure for otherwise failing companies that were ridden with problems before the March 11 quake. If gone too far, measures intended to prevent type I errors could cause an increase in bad loans and lead to serious problems in the future.

The third problem is who should bear the costs incurred on the disposal of old debts held by affected borrowers and how. This is not a question regarding the allocation of resources but the distribution of sunk costs, i.e., the distribution of negative income. The involvement of many stakeholders--debtors, creditors, and the government (taxpayers)--makes this problem more difficult to solve.

These three problems comprising the overlapping indebtedness problem are caused by different sets of reasons and therefore must be addressed differently. To begin with, type I errors occur in the following four kinds of situations.

First, they occur where a bank cannot see the prospect of being able to collect on a new loan to a prospective borrower because a portion of the borrower's future income has to be appropriated as repayment for the old debt, a problem referred to as debt overhanging. In this case, pardoning or mitigating the old debt would enable the borrower to take out a new loan. To achieve that end, the following schemes are expected to be effective: court-administered debt restructuring in the form of a reorganization-type bankruptcy, out-of-court debt restructuring in accordance with the Guidelines for the Out-of-Court Workouts or other relevant guidelines, the provision of new loans on a non-recourse basis, and the subordination of old debt claims through debt-to-equity swap or other similar means. A debt purchasing organization, to be established in Iwate and each of the affected prefectures, is counted on to facilitate these schemes.

Second, where banks have written off loans to certain affected borrowers and reclassified them into a higher risk category, they might find it impossible, as a matter of risk management, to provide new loans. In order to prevent the occurrence of such a situation, it is necessary to enable banks to make new loan decisions independently from the disposal of old debts.

Third, there may be situations in which the continuation of a business with a new loan is expected to have positive externalities (spillovers) on the local economy and its business partners. For instance, where the affected company is a supply chain hub, it would be socially desirable to provide a new loan to enable the continuation of business. However, we cannot expect private-sector banks to reflect such spillover effects in their loan decisions. Indeed, the overlapping indebtness problem is not the sole source of externalities, and it is the government--not banks--that should be addressing matters associated with externalities. In this regard, the government should take a broad point of view and provide a comprehensive response that includes fiscal and other measures.

Fourth, banks' ability to lend might have been undermined by the earthquake. For local banks, which are the main providers of funds to local businesses, it is difficult to diversify geographically the credit risk exposures. Thus, the risk-taking capacity of those operating in the affected areas has declined as their non-performing debt ratios increased after the quake. However, mega banks are not prepared to take over immediately the roles of those local banks as providers of relationship banking services, having no accumulation of borrower data.

In this regard, it is appreciative that the government paved the way for disaster-affected financial institutions to replenish their capital bases by asking for a public fund injection under the Act on Special Measures for Strengthening Financial Functions. A sudden increase in nonperforming loans (NPLs) could induce moral hazard on the part of banks, prompting them either to postpone writing off NPLs or resort to the so-called "forbearance lending" tactics, i.e., continuous rollover of loans to borrowers with little prospect of being able to repay. From the viewpoint of avoiding this potential problem, capital replenishment by means of public fund injections is an important scheme.

♦  ♦  ♦

As to the second problem, type II errors may occur in the form of moral hazard on the part of both lenders and borrowers as a result of government support measures that are too readily available, be it interest rate subsidies, loan guarantees, or else. In order to prevent this, it is desirable to have both lenders and borrowers bear a degree of risk. Since the purpose of government support measures is to accommodate post-disaster reconstruction, it is only natural to set out clearly an exit strategy for ending such measures.

Regarding the third and last question of what to do with the costs associated with old debts, there exists no equitable cost sharing scheme to which all of the stakeholders would agree. Furthermore, in working out a solution to this question, we must consider a range of issues, namely, the burden bearing capacity of each party concerned, the level of living standards to be guaranteed for debtors, the advisability of public remedy to redress damage to private property, and the equitability of such measures relative to those made available at the time of other disasters. Some of these issues, by their nature, should be dealt with, not by means of financial support which is designed to address the overlapping indebtedness problem, but by fiscal and/or other measures. This will require a political decision, and it must be made from a broad point of view, through a due democratic process, and in a timely manner.

Thus, in summary, government measures must be designed to achieve two goals--preventing type I errors and promoting the disposal of old debts. At the same time, however, it is necessary to avoid making assistance too readily available so as not to invite the risk of type II errors.

Now, specifically in the case of the Great East Japan Earthquake, how should we consider the risk of type I and II errors? Type I errors are more likely to occur when debt-ridden but promising companies--those that are saddled with large debts but capable of generating new, profitable business after the quake--are present in a large number. Conversely, if there are many debt-ridden and poorly performing companies, a scheme for facilitating new loans tends to invite the risk of type II errors.

Table: Capital-to-assets ratio and operating income-to-sales ratio of disaster-affected companies, as compared to the average ratios for all Japanese companies (%)Table: Capital-to-assets ratio and operating income-to-sales ratio of disaster-affected companies, as compared to the average ratios for all Japanese companies (%)

The table compares the average performance of companies affected by three major earthquakes--the Great East Japan, Great Hanshin-Awaji, and Niigata-ken Chuetsu-oki earthquakes--prior to each disaster with the corresponding average performance of all Japanese companies, measured in terms of the ratio of shareholders' equity to assets and the ratio of operating income to sales. Companies affected by the Great East Japan Earthquake compare more poorly to the all-companies average, relative to the average ratio of those affected by the previous two earthquakes. With their average shareholders' capital-to-assets ratio significantly lower than the all-companies average (meaning a higher leverage ratio), it is highly likely that there exist serious overlapping indebtedness problems among these affected companies in the Tohoku region. Meanwhile, the underperformance in the operating income-to-sales ratio indicates the presence of a large number of unprofitable businesses.

These conclusions remain unchanged when we examine the situation more closely. Therefore, measures for facilitating new loans must be carefully designed because the thoughtless expansion of such measures would invite type II errors. Needless to say, this suggestion is based on our analysis from the viewpoint of resource distribution, and it is not our intention to deny the need to implement support measures from the viewpoint of income redistribution. In doing so, however, the government should consider a broad range of measures including non-financial support.

♦  ♦  ♦

How can we evaluate the government measures implemented to date? The first and second supplementary budgets for the current fiscal year include generous allocations for measures designed to facilitate new loans to quake-hit firms such as the expansion of the government's emergency loan guarantee scheme and the availing of special loans from governmental financial institutions. It is thus fair to say that sufficient care has been taken to prevent type I errors.

It is warranted that sufficient consideration should be given to those affected by the disaster. At the same time, however, we are concerned that not enough consideration has been given to the risk of type II errors, as the scope of companies eligible to support measures is extremely broad and the government's loan guarantee scheme covers 100% of new loans. From the viewpoint of the need to solve the question of how the cost of disposing of old debts should be borne, and in order to avoid the risk of type I errors, it is imperative to work out quickly measures to alleviate the old debt burden. However, government efforts in this regard are slow to make progress.

Needless to say, the most effective solution to the old debt problem is to restore the restoration and reconstruction of the disaster-affected areas. Increasing promising private-sector businesses by promoting infrastructure development and deregulation in the special reconstruction zone will be effective in alleviating the problem of overlapping indebtedness. These points should be taken into careful consideration in formulating and implementing government measures, including the third supplementary budget plans.

>> Original text in Japanese

* Translated by RIETI.

October 10, 2011 Nihon Keizai Shimbun

December 7, 2011

Article(s) by this author