Asset Allocation in Bankruptcy

• Date: 01 February 2019
• Author: SHAI BERNSTEIN,BENJAMIN IVERSON,EMANUELE COLONNELLI
• Published date: 01 February 2019
• DOI: http://doi.org/10.1111/jofi.12740

THE JOURNAL OF FINANCE •VOL. LXXIV, NO. 1 •FEBRUARY 2019

Asset Allocation in Bankruptcy

SHAI BERNSTEIN, EMANUELE COLONNELLI, and BENJAMIN IVERSON∗

ABSTRACT

This paper investigates the consequences of liquidation and reorganization on the

allocation and subsequent utilization of assets in bankruptcy. Using the random as-

signment of judges to bankruptcy cases as a natural experiment that forces some

firms into liquidation, we find that the long-run utilization of assets of liquidated

firms is lower relative to assets of reorganized firms. These effects are concentrated

in thin markets with few potential users and in areas with low access to finance.

These findings suggest that when search frictions are large, liquidation can lead to

inefficient allocation of assets in bankruptcy.

DECLINING INDUSTRIES,INSOLVENCY,AND DISTRESSED firms are unavoidable con-

sequences of an evolving economy. The ability of an economy to subsequently

direct assets to better uses has important implications for productivity and

the speed of recovery following adverse economic shocks (Eisfeldt and Rampini

(2006), Hsieh and Klenow (2009), Bartelsman, Haltiwanger, and Scarpetta

(2013)). Since economies rely on courts to resolve insolvency, bankruptcy in-

stitutions play an important role in allocating the assets of distressed firms.

Broadly speaking, bankruptcy is resolved through two approaches: liquida-

tion and reorganization (Hart (2000), Str¨

omberg (2000), Djankov et al. (2008)).

While liquidation involves winding down the firm and putting all firm assets

back on the market, reorganization aims at rehabilitating the company when-

ever possible.

Despite the importance of the bankruptcy system, empirical evidence on key

questions is scarce. For instance, how does the bankruptcy regime affect asset

∗Shai Bernstein is with Stanford School of Business and NBER. Emanuele Colonnelli is with

Booth Business School, University of Chicago. Benjamin Iverson is with Marriott School of Busi-

ness, Brigham Young University. We thank Manuel Adelino; Ken Ayotte; Nick Bloom; Dave Don-

aldson; Darrell Duffie; Xavier Giroud; Dirk Jenter; Pete Klenow; Charlie Nathanson; Adriano

Rampini; Antoinette Schoar; Per Str¨

omberg; Francisco Urz´

ua; and seminar participants at the

Finance UC Conference, FOM Conference, Harvard Business School, HKUST, Hong Kong Univer-

sity,IDC Herzliya, Maastricht University, MIT Junior Finance Conference, Ohio State University,

Tilburg University, University of Illinois at Urbana-Champaign, Washington University in St.

Louis; brownbag participants at Stanford and Northwestern Universities for helpful comments

and suggestions. Wenhao Li provided superb research assistance. Any opinions and conclusions

expressed herein are those of the authors and do not necessarily represent the views of the U.S.

Census Bureau. All results have been reviewed to ensure that no confidential information is dis-

closed. The authors do not have any potential conflicts of interest to disclose, as identified in the

Journal of Finance Disclosure Policy.

DOI: 10.1111/jofi.12740

5

6The Journal of Finance R

allocation and utilization? Are assets in liquidation utilized similarly to assets

in reorganization? And, if not, what frictions lead to different effects under the

two resolution approaches?

Theoretically, in frictionless markets, both bankruptcy approaches should

lead to similar outcomes, as both regimes should allocate assets to their best

use. This null hypothesis may not hold, however, in the presence of frictions.

For example, conflicts of interests among claimholders, information asymmetry,

and coordination costs in reorganization may lead to inefficient continuation

and in turn to inefficient asset allocation (Baird (1986), Gertner and Scharfstein

(1991), Aghion, Hart, and Moore (1992), Ivashina, Iverson, and Smith (2015)).

In liquidation, assets may not be reallocated to best uses if they are specific

to the firm and markets are thin with few potential users (Williamson (1988),

Gavazza (2011)). Misallocation may be further exacerbated if potential users

of the assets are financially constrained (Shleifer and Vishny (1992)).

To address these questions, one must tackle two important issues. First, lit-

tle information is available with respect to how assets are reallocated between

firms and how assets are subsequently utilized, particularly in bankruptcy,

when plants are shut down and firms are dissolved. Second, distressed firms

that go through liquidation may be fundamentally different from firms that

are reorganized. Any comparison between two insolvent firms that experience

different bankruptcy regimes may therefore be biased due to unobserved differ-

ences in firm prospects and other characteristics. This is a common limitation

to papers that explore the implications of different bankruptcy codes.

In this paper, we focus on the U.S. bankruptcy system and compare the

effects of liquidation (under Chapter 7 of the bankruptcy code) with those of

reorganization (under Chapter 11 of the bankruptcy code) on asset allocation

and utilization. To do so, we focus on the real estate assets of bankrupt firms

and construct a novel data set that tracks the allocation and utilization of these

assets over time. Real estate assets represent a significant portion of firms’ total

capital.1Moreover, these assets are likely to be highly specific, as the optimal

user varies significantly with building features and location characteristics. For

example, an industrial warehouse is unlikely to be suitable for a retail store,

and a restaurant is unlikely to be replaced with a hotel. Furthermore, location

benefits in terms of access to customers and suppliers, local labor markets, and

knowledge spillovers vary across firms (Ellison, Glaeser, and Kerr (2010)).

We combine the U.S. Census Bureau’s Longitudinal Business Database

(LBD) with bankruptcy filings from LexisNexis Law to obtain a data set with

rich information on 129,000 establishments belonging to 28,000 bankrupt firms

that employ close to 4.7 million workers at the time of bankruptcy.The compre-

hensive nature of these data allow us to examine the population of bankrupt

firms in the United States, including small and private businesses. An im-

portant methodological contribution of this work is the creation of geographic

1Based on Flow of Funds tables from the Federal Reserve, nonresidential structures (value

of buildings, excluding the value of the land) accounted for $8.2 trillion of real assets, while

nonresidential equipment comprised only $4 trillion, at the end of 2014.

Asset Allocation in Bankruptcy 7

linkages that track occupier identities and economic activity at real estate as-

sets over time. This allows us to capture the allocation and utilization of assets

when plants shut down and the real estate is vacant or when it is used for a

different purpose from the original plant.2

To explore long-run (i.e., five-year) allocation and utilization of these assets,

we rely on several measures. We first examine the length of time a location

continues to be operated by the bankrupt firm and, if it does not continue,

whether it is occupied by a new firm or falls vacant. We further study the

average number of employees at a given location over time. While the former

measure captures whether economic activity takes place in a given asset, the

latter also captures the intensity of such economic activity.

Tracking assets in bankruptcy reveals several interesting stylized facts.

First, both liquidation and reorganization lead to substantial asset realloca-

tion. Second, when an asset is redeployed to a different user, it is typically to

a local firm, and to a firm in the same industry, consistent with a significant

degree of asset specificity and search costs, as highlighted by Williamson (1988)

and Ramey and Shapiro (2001). Third, we find that industry conditions, espe-

cially local economic activity, are important determinants of asset reallocation

and utilization, consistent with the importance of market liquidity and eco-

nomic conditions for asset redeployment, as discussed by Shleifer and Vishny

(1992) and Gavazza (2011).

In the main analysis, we address potential endogeneity of the bankruptcy

regime by employing an instrumental variables strategy that exploits the fact

that U.S. bankruptcy courts use a blind rotation system to assign cases to

judges, effectively randomizing filers to judges within each court division. While

there are uniform criteria by which a judge may convert a Chapter 11 case to

Chapter 7, there is significant variation in the interpretation of these criteria

across judges.

Our empirical strategy compares bankrupt firms that are reorganized under

Chapter 11 to firms that file for Chapter 11 but are converted to Chapter 7

liquidation due to the assignment of the judge. In effect, otherwise identical

filers are randomly placed in either reorganization or liquidation by the random

judge assignment, which allows us to compare asset outcomes across the two

regimes. Our empirical strategy follows a growing thread of literature that

takes advantage of the random assignment of judges and variation in judge

interpretation of the law (Kling (2006), Doyle (2007), Chang and Schoar (2013),

Dobbie and Song (2015), Galasso and Schankerman (2015)).

2These circumstances are not fully captured by the standard LBD linkages that link plants

over time. For example, if an auto parts manufacturer, AutoABC, shuts down, and the building is

then occupied by a shoe manufacturer, ShoesXYZ, linkages at the LBD would consider the death

of AutoABC and the birth of ShoesXYZ as two separate incidents. Our linkages connect the two,

by showing that ShoesXYZ replaced AutoABC at this real estate location. For details on how LBD

linkages are constructed, see Jarmin and Miranda (2002). We describe our linkages in detail in

Section III.Aas well as in the Internet Appendix, which may be found in the online version of this

article.