Lawyer Networks and Corporate Bankruptcies
by | posted in: Bankruptcy Administration and Jurisdiction, Bankruptcy Roundtable Updates | 0

By Vidhan K. Goyal (Hong Kong University of Science and Technology), Joshua Madsen (Carlson School of Management, University of Minnesota), and Wei Wang (Smith School of Business, Queen’s University)

Vidhan Goyal
Joshua Madsen
Wei Wang

Does having a lawyer who has previously interacted with the judge matter for bankruptcy outcomes? While knowledge obtained through past interactions about the judge’s views and preferences could improve the efficiency of court process, lawyer familiarity with the judge could also result in a capture of economic rents, leading to delays due to the difficulties in measuring lawyer efforts. Furthermore, connected lawyers could also exploit their connections to obtain biased outcomes in favor of their clients.

We examine these questions in the context of corporate bankruptcies by assembling a comprehensive dataset that contains detailed biographical information, professional experiences, and past in-court interactions of 162 bankruptcy judges overseeing 650 large Chapter 11 cases from 1996–2013, and 2,426 unique lawyers from 775 law firms representing those cases as debtor’s counsel. Our results show that cases with a lead counsel lawyer connected to the judge spend 16–21% less time in bankruptcy, a 2.6–3.5-month reduction in bankruptcy duration, translating into aggregated savings of $3.2–4.5 billion in professional fees for our sample firms.

Our empirical strategy exploits a setting where lead counsel lawyers are selected by the firm before the bankruptcy is filed and thus the assignment of a judge, minimizing concerns that connected lawyers are endogenously hired. The results are robust to the inclusion of controls for case complexity, industry effects, lawyer’s expertise, law firm quality, and judges’ fixed characteristics. Our specifications therefore ensure that any effect from having a connected lawyer is not due to unobserved heterogeneity that is specific to courts, judges, or lawyers.

We further document that the most effective lead counsel connections arise through previous clerkships and in-court interactions with the judge assigned to the case. The effects concentrate in cases with smaller legal teams where connected lawyers presumably have more influence. Having a connected non-lead counsel lawyers’ or connected lawyer representing the unsecured creditors committee only weakly affects case duration.

Lastly, we investigate other bankruptcy outcomes, including the probability of emergence, the bankruptcy refiling rate, operating performance post emergence, the likelihood of a Chapter 7 conversion, and the likelihood of loss of exclusivity extension. We find no evidence that the faster restructurings come at a cost of higher refiling rates or poorer operating performance after emergence. More importantly, there is no evidence that connections lead to judge favoritism or pro-debtor biases.

How do connected lawyers accelerate the bankruptcy process? The most likely explanation is connected lawyers’ knowledge of a judge’s preferences. Judges are extremely busy, and must devote enormous effort to keep straight all the facts and legal nuance under consideration. Connected lawyers are plausibly more familiar with the assigned judge’s preferences and expectations as well as the cases, legal precedents, and statutes that the judge will rely on. They can exploit this knowledge to help the “light shine through.” Idiosyncrasies across judges and their preferences imply that lawyers’ experience with other judges may not be as useful as a connection to the assigned judge and that there is likely no one “magic bullet” used by all connected lawyers. That is, lawyers’ knowledge of judges’ preferences are largely non-transferrable. These findings have implications for the design of bankruptcy institutions, where institutions that lead to lawyers’ increased awareness of a judge’s preferences could produce efficiency gains.

The full article is available here.

Rent Extraction by Super-Priority Lenders
by | posted in: Bankruptcy Administration and Jurisdiction, Bankruptcy Roundtable Updates, Cramdown and Priority | 0

By B. Espen Eckbo (Tuck School of Business at Dartmouth), Kai Li (Sauder School of Business at University of British Columbia) and Wei Wang (Smith School of Business at Queen’s University)

After filing for Chapter 11 bankruptcy, it takes large firms on average 16 months to restructure debt obligations and emerge as a going concern. With little cash on hand at filing, many firms need an infusion of new debt capital in order to fund continued operations while in bankruptcy. The standard debt instrument for this purpose is a debtor-in-possession (DIP) loan. Clearly, for a lender to be willing to supply a DIP loan, the loan must be fully collateralized and grant the lender extensive control rights. With an unprecedented large sample of DIP loan packages over the period 2002-2014 – totaling $120 billion in constant 2017 dollars – we first show that DIP loan contracts are fully collateralized, highly restrictive instruments. Our main research question is whether the interest rate on DIP loans reflects the actual risk of the firm defaulting on its DIP loan obligation. Our evidence on actual loan defaults strongly indicates that DIP loans are nearly risk-free, with only a single economic default (without full recovery) going back to 1988 (a default rate of 0.13% or lower). Nevertheless, loan spreads (the interest rate in excess of the London Inter-bank Offered Rate or LIBOR) average 604 basis points (bps), which exceeds even the average spread of even high-risk (“junk”) bonds. While desperate borrowers are willing to pay supra-competitive DIP loan prices, the central question is why competition among lenders fails to bring down DIP-loan spreads.

To answer this question, we first show that prepetition lenders dominate the supply of DIP loans (more than 70% of the cases). This is hardly surprising since, under §364(d) of the Bankruptcy Code, granting collateral to the DIP lender requires “priming” the lien of prepetition lenders, the debtor must obtain their consent. Moreover, a prepetition lender may “roll up” portions of the existing debt into the DIP loan package, which lowers the risk of the prepetition loan as well. By blocking the debtor’s access to DIP loans from new lenders, prepetition DIP-loan providers are in a strong monopolistic bargaining position vis-à-vis the debtor – resulting in opportunities for rent extraction. However, when collateralizing the DIP loan does not require priming prepetition lenders, the debtor may turn to new lenders such as hedge funds (HF) or private equity funds (PE). In these cases, it is reasonable to expect competition among the prospective DIP-loan providers to lower spreads. However, we find the opposite: While there is no evidence that DIP loans provided by new lenders face a greater risk of default, loans spreads are significantly higher.

It is possible that, notwithstanding the strong contractual protection afforded by DIP loans, there may be unobservable heterogeneity in the risk of firms seeking DIP loans that only a skilled DIP-loan provider is able to detect ex ante. If so, a high loan spread may be viewed as a return to the loan provider’s unique screening ability. We investigate this possibility by comparing spreads and fees in DIP loans on leveraged loans (“junk” debt). Presumably, the much longer maturities of leveraged loans (on average five years), combined with their much lower control rights and degree of collateralization, renders leveraged loans more risky than DIP loans. In fact, using Moody’s rating information, the typical spread on a B-rated leveraged loan has an expected default rate that is much higher than what we estimate for our DIP loan sample. Therefore, we expect leveraged loans that are supplied by sophisticated financial institutions to have higher spreads. Instead, we find the opposite: DIP-loan spreads are 236 bps higher than leveraged loans matched on size, industry and year of issuance, 255 bps (152 bps) higher than leveraged loans by the same firm within three years (one year) of filing.

Last, but not least, we show that junior claimholders (unsecured creditor committees and suppliers) file objections to the DIP-loan terms in as much as over 60% of the cases in our sample. Moreover, spreads are 80+ bps higher when objections occur, suggesting that high spreads are a concern. However, reading case files, we do not find a single case where the court lowered the loan spread (or fee). Although both the spirit and the letter of §364 require the terms of DIP-loans to be “fair, reasonable and adequate”, courts appear not to act as a backstop for what our data strongly suggest is significant extraction of economic rents by DIP-loan providers.

The full article is available here.

Practice Makes Perfect: Judge Experience and Bankruptcy Outcomes
by | posted in: Bankruptcy Administration and Jurisdiction, Bankruptcy Reform | 0

By Benjamin Charles Iverson (Brigham Young University), Joshua Madsen (University of Minnesota, Twin Cities, Carlson School of Management), Wei Wang (Queen’s School of Business), and Qiping Xu (University of Notre Dame, Department of Finance).

Prior studies document the influence of bankruptcy judges’ discretion on restructuring outcomes, yet we know little about how judicial experience affects the bankruptcy process. We study how the accumulation of job-specific human capital influences judges’ efficiency in handling large corporate bankruptcy filings, using 1,310 Chapter 11 filings by large U.S. public firms overseen by 309 unique bankruptcy judges in 75 bankruptcy courts between 1980 and 2012.

Using random assignment of judges to cases for empirical identification, we show that cases assigned to a judge with twice as much time on the bench realize a 5.5% decrease in time spent in reorganization. This reduced time in court translates into savings of approximately $2 million in legal fees alone for a typical case in our sample. Judges’ time on the bench is associated with higher probability of emergence but not higher recidivism. The combined evidence suggests that more experienced judges are overall more efficient. We also find that it takes up to four years for a new judge to become efficient and that judges who see a higher volume of business filings and a greater diversity of cases by size and industry early in their tenure become efficient faster than those who don’t. We find little evidence that judges’ general experience and personal attributes consistently affect case outcomes.

Our analyses highlight a potential benefit of allowing firms to file in courts with more experienced judges. Restricting this flexibility (e.g., through the proposed Bankruptcy Venue Reform Act of 2017) may impose a cost on firms by forcing them to file in courts with less experienced judges.

The full article is available here.

The Roundtable has previously posted on potential Bankruptcy venue reforms, including a summary of the Bankruptcy Venue Reform Act of 2018 introduced by Senators John Cornyn, R-TX, and Elizabeth Warren, D-MA. For a critique of current venue rules—and a possible solution—see Prof. Lynn LoPucki, “Venue Reform Can Save Companies.” For a defense of the current system, see the Roundtable’s summary of the Wall Street Journal’s “Examiners” Panel on venue reform.

Selling Innovation in Bankruptcy
by | posted in: 363 Sale, Valuation | 0

By Song Ma (Yale School of Management), (Joy) Tianjiao Tong (Duke University, Fuqua School of Business), and Wei Wang (Queen’s School of Business).

The past decades have witnessed the emergence of patent sales in corporate bankruptcies. Yet we know little about the facts and rationales of these important economic transactions.

In this working paper, we assemble a comprehensive data set of US Chapter 11 filings, USPTO patent transaction documents, and court records on assets sales from the past three decades. We document three stylized facts on patent sales in bankruptcy. First, patent sales are pervasive — more than 40% of bankrupt firms sell at least one patent, and on average they sell 18% of their patent portfolios. Second, patent transactions occur immediately after bankruptcy filing — concentrating largely within the first two quarters after filing. Third, patents are frontloaded in general asset sales in bankruptcy — firms sell a disproportionately large quantity of patents in asset sales during the early period of reorganization.

Why do firms sell patents during bankruptcy? We design a set of empirical tests to study the economic decisions behind patent sales based on the two economic views on assets reallocation in bankruptcy, namely asset restructuring and financing through asset sales. Our results show that bankrupt firms reallocate patents that are more redeployable and trade in a more liquid market . We find no evidence that they sell underexploited or underperforming patents. This pattern of selling more liquid patents holds stronger in firms with financial distress, firms undergoing poor industry conditions, and firms lacking external financing. The combined evidence lends support to the view that firms sell innovation during bankruptcy for financing purposes rather than for asset restructuring. Additionally, we find that bankrupt firms try to retain the inventors of sold patents and continue to cite sold patents after their sale. The evidence overall suggests that a firm’s imminent financing needs interact with its intent to avoid bankruptcy costs in shaping a firm’s decision to sell patents in bankruptcy.

The full paper is available here.

 

The Roundtable will be off for the holidays. We’ll be back early after the New Year.