During the "lost decade" of the Japanese economy, a substantial number of listed firms were regarded as "zombies" - troubled debtor firms with little or no chance of recovery that continued operating thanks to life support from their banks. Many economists argued that flourishing zombie firms were one of the major roadblocks to a quick turnaround of the Japanese economy. However, only a few of those zombie firms have actually gone bankrupt or been forced to delist. Moreover, a majority of the surviving zombies have in recent years achieved significantly improved business performance. In this article, I will summarize the major findings from a joint study with Professor Shinichi Fukuda of the University of Tokyo, on the factors behind this remarkable recovery of zombie firms.
Pros and cons of lender-borrower relationship and problem of zombie firms
The lender-borrower relationship between banks and their corporate customers has been the core of the financial intermediation system in postwar Japan. One of the major advantages of the system is the prevention of inefficient liquidation through the renegotiation of loan terms (e.g., interest concessions and additional loans) in a flexible manner when borrower firms are running into, but stand a good chance of overcoming, temporary financial difficulty. At the same time, such flexibility underpinned by the lender-borrower relationship has a negative side known as the "soft budget-constraint problem." Specifically, managers of borrower firms may have an incentive to relax their efforts to independently restore business in anticipation of a bailout by the creditor bank, while the bank may have the incentive to extend additional support to defer the liquidation of an otherwise insolvent firm even though it knows the firm has little or no chance of recovery. In this situation, the bank's expanding lending to inefficient borrower firms that should exit the market is often referred to as "forbearance lending," and the inefficient firm kept alive by its creditor bank as a zombie firm.
From the late-1990s through the early-2000s, as the Japanese economy faced an acute financial crisis, a view emerged that forbearance lending and zombie firms might be distorting the allocation of productive resources and imposing negative externalities on sound companies, thus hindering economic recovery. Many of the preceding empirical studies on this hypothesis generally supported the same idea though they somewhat differed in their theoretical frameworks and definitions of the zombie firm. Particularly, Caballero, Hoshi, and Kashyap (2006; hereinafter referred to as C-H-K) examined listed firms, identified specific zombie firms, and showed the presence of negative externalities in a more straightforward manner than many other studies.
However, if zombie firms truly are inefficient and have little or no chance of recovery, it is difficult to explain why a majority of these firms are still operating and, in recent years, significantly improving their business performance. Arguments at the time of the lost decade typically labeled "inefficient firms" as those that had a net capital deficiency. Any additional bank loans to such firms tended to be automatically regarded as forbearance lending. C-H-K also cites "interest concessions" as a condition defining a zombie firm. However, even when a company falls into net capital deficiency or is unable to pay contractual interest on its debt obligations, it is socially desirable to let the company continue to operate as a going concern, provided that it is expected to eventually yield earnings in excess of its liquidation value. It is extremely difficult, at least for outsiders, to accurately evaluate the future potential of a company using such forward-looking criteria (note 1). But if this cannot be done, it is impossible to clearly determine the pros or cons of the lender-borrower relationship.
Identifying zombies
Several years of financial data now exist since the problem of zombie firms was first recognized. With no special ability for evaluating the future potential of firms, yet by tracking the outcome of firms once regarded as zombies, their historical data can be analyzed on an ex post basis and forward-looking viewpoints can be back-tested to re-evaluate past judgments.
Based on this understanding of the problem, Professor Fukuda and I examined the recovery of zombie firms and the underlying factors, using non-aggregated data on a total of 2,228 firms listed either on the first or second section of the Tokyo Stock Exchange, belonging to five major industries (note 2). C-H-K analyzed data through 2002 and calculated a "hypothetical lower bound for interest payments" (the minimum amount of interest a borrower should be paying) based on the outstanding amount of interest-bearing debt and using the long- and short-term prime rates, wherein firms paying less interest than that hypothetical amount are regarded as receiving financial assistance (interest concessions) and thus defined as zombies. However, when these same criteria are applied to more recent data, the percentage of zombie firms continues to rise, exceeding 30% in and after 2002, despite the economic upturn. Several blue chip companies are also found to have been incorrectly labeled as zombies.
For our analysis, we developed original criteria by substantially improving on that in C-H-K. First, we introduced "profitability criteria" which define a company as a zombie if the amount of 1) operating income (loss) plus interest and dividend income or 2) earnings before interest and taxes (EBIT) falls below the hypothetical lower bound for interest payments. This solves the problem of misidentification of a blue chip company as a zombie in C-H-K.
Next, we incorporated forbearance lending, in addition to interest concessions, as an additional factor in the criteria for identifying financial assistance. That is, when a possible zombie firm in terms of profitability criteria is inferred to have received a "new loan" based on the comparison of the current-term-ending amount of outstanding debt with that of the preceding term, it is considered to have received financial assistance. The percentage of zombie firms defined by profitability criteria and refined financial assistance criteria peaked at 25% in 2001 and then rapidly declined from 2003 to below 10%. These changes are consistent with the observed changes in the ratios of nonperforming loans and loans classified as doubtful based on banking statistics.
Underlying factors behind recovery of zombie firms
Then, using these zombie indexes, we estimated a multinomial logit model covering the 10-year period of 1995-2004, in which the current term status of a company judged a zombie in the preceding term - whether the company remains a zombie, becomes a non-zombie, or is delisted - is the qualitative explained variable, to analyze underlying factors behind the recovery of zombie firms. Explanatory variables were selected by focusing on four possible contributing factors for recovery: improvement in the external macroeconomic environment, restructuring efforts by individual firms, corporate governance structure, and support from financial institutions.
The resulting estimates of sales growth and time dummy variables show that improvement in the macroeconomic environment, particularly the increase in overseas demand amid robust global economic growth from 2002, significantly increased the probability of recovery. At the same time, however, reductions in employee headcounts and fixed assets had a positive impact on the probability of recovery, as did the higher the ratio of shares held by financial institutions. Support from financial institutions was also found to increase the probability of recovery when it includes significant debt relief. These findings may indicate that not only favorable external factors but also steps taken by concerned parties - such as management efforts by the troubled company, financial institutions' encouragement of these efforts as a major shareholder of the company, and government economic policy to facilitate rapid disposal of bad debt - contributed to the recovery of zombie firms.
However, the impact of corporate restructuring or support from financial institutions has not always been positive. Wage cuts were not found to be effective, and protracted reductions of employees and fixed assets have not necessarily brought positive results. The relationship between the posted amount of an extraordinary profit or loss and the recovery of a zombie firm suggests that the implementation of downsizing by disposing of bad assets to eliminate future losses was effective in facilitating recovery, whereas downsizing by means of selling good assets tended to delay recovery. Debt relief offered in bits and pieces to a zombie firm shows a tendency to hinder the chance of recovery, as does prolonged zombie status. In other words, the recovery of zombie firms required clearly motivated reform to improve future profitability, instead of postponing the problem.
Right or wrong of hard-landing approach
"Appropriate restructuring + tailwind" can be seen as the simplified formula behind the recovery of zombie firms identified in this study. In order for the lender-borrower relationship to demonstrate its benefits, simply waiting for a tailwind is not enough; restructuring must be directed toward more efficient resource allocation, specifically by adequately compensating for what is needed and proactively eliminating what is not.
At the same time, given the successful recovery of many of the firms seen as zombies, whether it is right or wrong to take a hard-landing (liquidation) approach to dealing with the problem of bad loans is a question that must be carefully examined. In seeking an answer, further studies must be conducted on zombie firms.
