Print
E-MailFlorencio Lopez-de-Silanes is Professor of Finance and Associate Dean for International Affairs at SKEMA Business School, and a Research Associate at NBER. This post is based on a recent paper by Professor Lopez-de-Silanes; Erasmo Giambona, Michael Falcone Chair in Real Estate, and director of the James D. Kuhn Real Estate Center at Syracuse University Whitman School of Management; and Rafael Matta, Assistant Professor of Finance at University of Amsterdam Business School.
Empirical evidence suggests that asset pledgeability, debt complexity, and valuable control rights of dispersed debt influence the resolution of financial distress. Firms that borrow from multiple uncoordinated creditors and have more tangible assets often fail to renegotiate debt out of court and inefficiently file for bankruptcy. But the current theoretical literature is at odds with the evidence. Existing models predict inefficient bankruptcy filings to be negatively associated with both asset pledgeability and the number of debtholders. In our paper, we bridge the gap between existing models and evidence on distress resolution proposing a broader financial contracting model and testing its predictions using an exogenous variation in the court’s ability to price assets. We analyze the effect on bankruptcy filings and firm debt capacity after the 1999 U.S. Supreme Court ruling that reorganization plans in which equity holders keep an interest must be exposed to a “market test.”
Our model builds on the incomplete contracting framework of Aghion and Bolton (1992), Hart and Moore (1998, 1994), and Bolton and Scharfstein (1990), but introduces three innovations. First, creditors’ ownership rights over assets in place are limited because courts cannot fully verify them. In the spirit of Diamond (2004) and Bolton and Scharfstein (1996), this forces the borrower to pledge a portion of the verifiable assets to multiple creditors to make financing feasible. Second, the out-of-court value of the assets in distress states depends on the aggregate control exerted by creditors, which is costly and non-contractible. This creates a coordination problem among creditors as their incentive to exert control depends on the control exerted by the other creditors. Third, creditors are better informed than the borrower about their own alignment of interests over the actions to be taken; hence about the out-of-court value of the assets in distress. As a result, the borrower cannot induce efficient coordination on exerting control, often resulting in costly bankruptcy. To our knowledge, we are the first to incorporate all these three frictions in the analysis of distress resolution. Our model uniquely determines that bankruptcy increases with the verifiability of assets in place and creditors’ degree of misalignment of interests. These predictions seem to match the existing empirical evidence mentioned above.
In the second part of the paper, our empirical analysis provides more direct evidence on the model’s predictions. In 1999, the U.S. Supreme Court decision in Bank of America v. 203 North LaSalle established that reorganization plans in which equity holders keep an interest in the firm must be exposed to a “market test” allowing competing plans or bids for the equity interest. This court decision amounts to an exogenous increase in the verifiability of assets in place during the bankruptcy process where the unavailability of market-based information can lead to large valuation errors by the court (Gilson, Hotchkiss, and Rubak, 2000; and Pulvino, 1999). These errors are often detrimental to creditors since the 1978 U.S. Bankruptcy Code granted the debtor the exclusive right to propose a reorganization plan while retaining control of the firm once in Chapter 11. The situation before the 1999 ruling amounted to “stiffing the creditor” (Forbes, October 5th 1998)
Our identification strategy uses a difference-in-difference approach comparing Chapter 11 filings for low- and high-verifiability firms (our treated and control groups) before and after the 1999 U.S. Supreme Court ruling. To classify low- and high-verifiability firms, we rely on the ratio of industry-year sales of Property, Plant, & Equipment (sales of PP&E) to PP&E. According to our model, the Supreme Court’s “market test” should increase verifiability more in those situations where the bankruptcy judge has a harder time verifying the equitability of a restructuring plan using information from the sales of PP&E used in the industry of the distressed firm. Indeed, relative to high-verifiability firms, Chapter 11 filings of low-verifiability firms increased by 1.2 percentage points in the two years after the Supreme Court decision. This increase is large considering that Chapter 11 filings for the full sample period are only one percent on average.
We test for a series of additional predictions of the model in terms of the type of firms that are most likely to be affected by the Supreme Court’s ruling. The effect of an increase in verifiability on Chapter 11 is predicted to be stronger when firms are more likely to face financial difficulties and when creditors are more likely to face coordination problems. In line with the predictions of our model, we find that Chapter 11 filings increased by a sizable 2.9 percentage points for firms in financial alert, while there was close to no change in Chapter 11 filings for financially sound firms. When we test for differences in debt complexity of firms, our findings also corroborate the empirical prediction of the model showing that the bulk of the effect is among firms with higher debt complexity. We find that Chapter 11 filings increase between 2.9 and 6.7 percentage points when creditors are likely to face large coordination problems.
A second set of empirical findings shows the effect of the higher asset verifiability of the 1999 Supreme Court decision on firm debt capacity. An increase in debt capacity should lead to an increase in the borrower’s equity value to the extent that the additional credit is used to finance positive net present value projects. In line with the model’s predictions, we find significant positive abnormal returns for all affected firms in the days surrounding the 1999 Supreme Court ruling. Cumulative abnormal returns are stronger for low-verifiability firms reaching almost 1.7 percent in the five days surrounding the ruling. The analysis of the market response of firms with different risk levels of financial distress shows that significant positive returns are concentrated among firms of low-to-moderate risk of financial distress. Arguably, higher verifiability has little effect when the risk of bankruptcy is imminent or financial distress is remote. We complement the event study analysis with evidence showing that firms facing moderate to low risk of financial distress increased leverage between 3 and 6 percent in the two years following the Supreme Court decision. Meanwhile, leverage did not change for firms with high financial distress risk or for financially sound firms.
Our empirical findings are robust to various specifications and tests, including the parallel trend assumption, placebo events, the effect of the burst of the dotcom bubble, alternative verifiability measures and cutoffs, and propensity score matching. Overall, our paper shows that the improvement in creditor rights brought about by the 1999 Supreme Court decision more than doubled Chapter 11 filings and significantly increased debt capacity of low-verifiability firms. To our knowledge, we are one of the first to examine how asset verifiability in combination with valuable control rights of dispersed debt affect distress resolution outcomes. This approach provides new research opportunities in the analysis of financial distress and other areas of finance such as risk management, firm pay-out policy and M&A.
The complete paper is available for download here.
