Journal of Financial Economics 102 (2011) 150–166

Contents lists available at ScienceDirect

Journal of Financial Economics

journal homepage: www.elsevier.com/locate/jfec

Creditor rights and corporate risk-taking$

Viral V. Acharya a,b,c,d, Yakov Amihud a,n, Lubomir Litov e
a
New York University, Stern School of Business, United States
b
CEPR, United Kingdom
c
ECGI, Belgium
d
NBER, United States
e
Washington University in St. Louis, Olin Business School, United States

a r t i c l e in f o abstract

Article history: We propose that stronger creditor rights in bankruptcy affect corporate investment
Received 5 March 2008 choice by reducing corporate risk-taking. In cross-country analysis, we find that
Received in revised form stronger creditor rights induce greater propensity of firms to engage in diversifying
5 April 2010
acquisitions that are value-reducing, to acquire targets whose assets have high recovery
Accepted 12 April 2010
Available online 27 May 2011
value in default, and to lower cash-flow risk. Also, corporate leverage declines when
creditor rights are stronger. These relations are usually strongest in countries where
JEL classification: management is dismissed in reorganization and are also observed over time following
G31 changes in creditor rights. Our results thus identify a potentially adverse consequence
G32
of strong creditor rights.
G33
& 2011 Elsevier B.V. All rights reserved.
G34

Keywords:
Bankruptcy code
Corporate reorganization
Investment
Diversification

1. Introduction creditors would enslave their debtors-in-default until

they fully repaid their debt. During some periods in Rome,
Throughout history, defaulting on debt has incurred creditors could use physical punishment for debtors in
harsh punishment. In Biblical times and in ancient Greece, default.1 The United Kingdom had debtor’s prisons until

$
Acharya is C.V. Starr Professor of Finance. Amihud is Ira Leon Rennert Professor of Finance. We acknowledge with gratitude comments and
suggestions that helped improve the paper by Barry Adler, Kenneth Ahern, Reena Aggarwal, Franklin Allen, Heitor Almeida, Meghana Ayyagari, Moshe
Barniv, Bo Becker, Sreedhar Bharath, Bernie Black, Long Chen, Sid Chib, Jonathan Cohn, Jeff Coles, Phil Dybvig, Espen Eckbo, Alex Edmans, Isil Erel, Mara
Faccio, Mike Faulkender, Julian Franks, Radha Gopalan, Todd Gormley, Bill Greene, Todd Henderson, Joel Houston, Kose John, Lutz Johanning, Ohad Kadan,
Sandy Klasa, Anzhela Kniazeva, Diana Kniazeva, William Megginson, Todd Milbourn, Natalie Moyen, Ed Morrison, Holger Mueller, Harold Mulherin, Paige
Ouimet, Troy Paredes, Katharina Pistor, Amiyatosh Purnanandam, Stefano Rossi, Antoinette Schoar, Alan Schwartz, Oren Sussman, Anjan Thakor, Rohan
Williamson, Daniel Wolfenzon, Jeff Wurgler, David Yermack, Bernie Yeung, the seminar participants at Washington University in Saint Louis, NYU
Salomon Center corporate governance seminar, University of Michigan, Tel Aviv University, Bar Ilan Iniversity, Hebrew University, Interdisciplinary
Center in Herzliya, the 2008 Conference on Law and Economics at the University of Pennsylvania, Cornell University’s Empirical Legal Studies Conference,
2008 UNC-Duke Corporate Finance Conference, the University of Gent 2008 Bankruptcy and Reorganization Conference, University of California at San
Diego, University of Arizona, and especially two anonymous referees. We thank Simeon Djankov for providing access to the creditor rights data. Rong
Leng provided excellent research assistance. A part of this paper was completed while Acharya was at London Business School. Acharya is grateful for
research support from the ESRC (Grant no. R060230004) awarded to the London Business School Corporate Governance Research Center.
n
Corresponding author at: New York University, Stern School of Business, 44 West 4th Street, New York, NY 10012, United States.
E-mail addresses: vacharya@stern.nyu.edu (V.V. Acharya), yamihud@stern.nyu.edu (Y. Amihud), litov@wustl.edu (L. Litov).
1
In 450 BC: The Twelve Tablets, Section III, Debt. Penalties ranged from imprisonment to removing parts of the body.

0304-405X/$ - see front matter & 2011 Elsevier B.V. All rights reserved.
doi:10.1016/j.jfineco.2011.04.001
 V.V. Acharya et al. / Journal of Financial Economics 102 (2011) 150–166
their abolition by the 1869 Debtors Act. In contrast, the
modern-day norm of limited liability in bankruptcy
reduces creditor rights by limiting their ability to pursue
debtors who default on promised payments.2 The question
is: how strong should state-mandated creditor rights-in-
default be, given that stockholders and creditors are limited
in their ability to contract around them? Schwartz (1997,
p. 127) points out: ‘‘When a firm becomes insolvent, how-
ever, the state-supplied dispute resolution procedure—the
bankruptcy system and court—is mandatory; parties cannot
contract in the lending agreement for an alternative proce-
dure.’’ Because bankruptcy laws apply uniformly to all firms
and have precedence over private firm-specific contracts,
they may lead to inefficient outcomes for some firms.3
We examine the effect of creditor rights on corporate
risk-taking. In particular, we ask: what effect does the
strength of creditor rights have on firm investment deci-
sions? While a harsh penalty in default reduces fraud and
opportunistic behavior by debtors, it might also discourage
bona-fide risky investment.
This paper henceforth examines the link between
creditor rights and corporate investment policy. It thus
differs from research on creditor rights that mainly
analyzes their effect on financing policies, such as the
suggestion by Djankov, McLeish, and Shleifer (2007) that
stronger creditor rights encourage increase in the supply
of credit.4 We propose that stronger creditor rights induce
risk-reducing investments. Strong creditor rights in
default may lead to inefficient liquidation, which extin-
guishes the continuation option of a firm’s enterprise and
thus hurts shareholder value. Also, creditor rights that
mandate the dismissal of management in bankruptcy
impose private costs on managers. To avoid these costs,
shareholders and managers lower the likelihood of dis-
tress by reducing cash-flow risk. Such risk reduction can
result in value loss due to foregoing profitable invest-
ments, or from undertaking value-decreasing diversifying
investments; strong creditor rights can thereby result in
dead-weight costs to firms and to the economy at large.
Further, while strengthening creditor rights increases the
propensity to lend, it may reduce firms’ demand for credit,
resulting in lower overall levels of corporate debt.
Our empirical analysis examines the effects of creditor
rights on corporate risk-taking. We use variation of
creditor rights across countries as an explanatory variable.
La Porta, Lopez-de-Silanes, Shleifer, and Vishny (1998)
2
Europe still has effective and strict penalties upon bankruptcy, as
noted in The Economist, 23 March 2002, Fear of failure—Europe’s fear of
bankruptcies: ‘‘In Europe, by contrast [to the United States], failed firms
face bigger hurdles. First, in the continent that created the debtor’s
prison, insolvency is still tainted with moral failure. In some countries,
company directors are personally liable for bankruptcy. That steeply
raises the penalties for failure—and so deters entrepreneurs from taking
risks.’’
3 5
This is why Schwartz (1997) proposes that the state makes
available to firms a menu of bankruptcy procedures. Then, allowing
parties flexibility in contracting for preferred bankruptcy procedures
alleviates undesirable effects on investment arising from strong creditor
rights.
4
This analysis contends that stronger creditor rights reduce the
dead-weight cost due to debtor-creditor conflict (Jensen and Meckling,
1976).
151
suggest that creditor rights are largely a function of
national legal origin. While some variation in creditor
rights may be an endogenous response of the legal system
to the characteristics of the national economy, it is
reasonable to assume for our empirical analysis that
creditor rights are predetermined.
We employ several measures of corporate risk-taking
and examine their relation to creditor rights across
countries and over time. Our primary analysis focuses
on the effect of creditor rights on the nature of mergers
and acquisitions (M&A) in a country. We focus on M&A
because these transactions provide an opportunity to
observe a major corporate investment and its effect on
corporate risk—whether the acquisition is diversifying
across industries and thereby risk-reducing, or focusing
within-industry. In M&A, we can also identify whether the
assets in which the firm invests are of high or low
recovery value. Further, corporate investment in the form
of M&A decisions is not tainted by the cross-country
differences in reporting practices that affect other mea-
sures of investment (such as capital expenditures and
research and development (R&D) expenses). We find the
following:
(1) In countries with stronger creditor rights, firms have
greater propensity to do diversifying acquisitions across
industries. This is observed when estimating the like-
lihood of diversifying acquisitions at the level of single
acquisitions as well as at the level of a country or at the
level of an industry in a country. Also, geographically
diversifying (cross-border) acquisitions are more likely in
countries with stronger creditor rights. Further, in coun-
tries with stronger creditor rights, firms have segments
in a greater number of lines of business, a measure of
diversification.
(2) In countries with strong creditor rights, acquirers
whose assets have low recovery value in distress are
more likely to acquire target firms with high-recovery
assets.5 This is because high asset recovery values
enable firms in distress to defer default by liquidating
some assets and by using the proceeds to service debt.
Thus, by acquiring a high-recovery target, a low-recovery
firm reduces or defers the likelihood of default in case of
distress.
(3) The effect of diversifying acquisitions on the acquirer’s
value is more negative in countries with stronger cred-
itor rights. In such countries, acquirer profitability, mea-
sured by return on assets (ROA), is significantly lower
post-acquisition, and the acquirer’s abnormal stock
return is also lower. In contrast, in focusing (same-
industry) acquisitions, there is no adverse effect of
stronger creditor rights on the acquirer’s value.
Assets with high recovery value have lower costs of liquidation.
These incur lower loss of value in distressed sales and, following the
definition of Shleifer and Vishny (1992), have lower specificity in that
they are fungible across industries and hence trade at prices that are
closer to their value in best use. Our measure of high-recovery industries
is based on the realized recovery rates of debts of defaulted firms in
various industries, documented by Acharya, Bharath, and Srinivasan
(2007).
152 V.V. Acharya et al. / Journal of Financial Economics 102 (2011) 150–166
Our results on the choice of corporate risk through
mergers and acquisitions are further corroborated with
alternative measures of corporate risk. We present two
additional findings on the creditor rights-corporate risk
relation:
(4) In countries with stronger creditor rights, firms
choose to reduce cash-flow risk, as measured by the
standard deviation of firm ROA. We obtain this result
at the level of a single firm, a country, or an industry
in a country.
(5) In countries with stronger creditor rights, corporate
leverage is lower. This finding is notable, as it contra-
dicts the argument that stronger creditor rights brings
about a greater level of lending.
Findings (1) and (2) above are generally stronger for
creditor rights that correspond to whether management is
dismissed in bankruptcy, reflecting an effect of manage-
rial self-interest on corporate policies, as well as whether
there is automatic stay in bankruptcy. Finally, all of the
above effects are also observed over time, as a result of
strengthening (and, vice versa, weakening) of a country’s
creditor rights.
To control for a potential dependence of creditor rights
on the country’s characteristics, we include in our models
country-level variables that may be related to creditor
rights, such as legal origin, macroeconomic risk, the extent
of the rule of law, flexibility to fire employees, the size of the
capital market, accounting disclosure standards, and per-
capita gross domestic product (GDP). We also perform
robustness tests using variables that measure the propen-
sity to avoid risk (cultural and religious variables), and we
present direct comparisons between the U.S. and the U.K.
which arguably are similar in economic and financial
development as well as in culture, but are quite different
in their creditor rights (much stronger in the U.K. than in
the U.S.).
Our results on the role of managerial self-interest in the
decision to diversify when faced with stronger creditor rights
is consistent with evidence from Amihud and Lev (1981),
who find that diversifying acquisitions and lower risk of
return on equity are more pervasive in manager-controlled
firms than in stockholder-controlled firms. Further, Tufano
(1996) proposes that managerial risk-aversion induces cor-
porate hedging in public gold mining firms. Tufano (1998)
proposes that hedging reduces the need to access external
capital markets and thus lessens the discipline that market
scrutiny imposes on corporate insiders.
Among theoretical studies on the effects of creditor
rights, Adler (1992) suggests that while strong creditor
rights induce the manager to increase firm risk as the
firm approaches default, their ex ante effect is to reduce
risk to avoid insolvency. Manso (forthcoming) proposes
that penalizing failing entrepreneurs through tough bank-
ruptcy procedures inhibits innovation, and Acharya and
Subramanian (2009b) argue that strong creditor rights
can deter financial leverage and risk-taking in more
innovative industries. Adler, Capkun, and Weiss (2007)
propose that the recent strengthening of creditor rights in
the U.S. induces firms to delay default and to waste assets
in the process. Empirically, Acharya and Subramanian
(2009a) find that in countries with stronger creditor
rights, technologically innovative industries innovate less,
employ lower financial leverage, and grow more slowly.6
Chava and Roberts (2008) and Nini, Smith, and Sufi (2009)
find that restrictive debt covenants and enforcement of
covenant violations, which provide firm-specific creditor
rights, inhibit capital investment. These findings point out
potentially harmful effects of creditor rights with respect
to risky investment.
This paper proceeds as follows. Section 2 discusses the
data, the empirical design, and the test results and Section 3
offers concluding remarks.
2. Hypotheses, data, and empirical design
We test whether stronger creditor rights across coun-
tries cause firms to undertake lower risk. Our main
analysis pertains to reduction of risk via diversifying
acquisitions, which lower the likelihood of default (after
controlling for leverage). We test the following:
Hypothesis Ia. The propensity to undertake diversifying
acquisitions is greater in countries with stronger creditor
rights.
We test this hypothesis by examining diversifying
acquisitions across industries at three different levels:
individual acquisitions, average country-level acquisi-
tions, and average industry-level acquisitions. We also
test the hypothesis by examining geographic diversifica-
tion, measured by cross-border acquisitions.
Acquisitions can help reduce the likelihood of default
when an acquirer whose assets have low recovery value
in bankruptcy acquires a target with high-recovery assets,
which lose less of their value in distressed sales and fetch
prices that are closer to their value in best use (using the
notion of asset specificity from Shleifer and Vishny, 1992).
A firm in financial distress can liquidate some of its high-
recovery assets and use the proceeds to service its debt,
thus deferring default. Also, Eckbo and Thorburn (2003)
suggest that it is in the manager’s interest to increase the
recovery rate of debt in default (which is related to asset
characteristics), because the probability of rehiring man-
agers who are automatically dismissed in bankruptcy in
Sweden is increasing in the recovery rate of the firm’s
debt. Our hypothesis is:
Hypothesis Ib. In countries with stronger creditor rights,
firms in industries whose assets have low recovery value
are more likely to acquire firms in industries whose assets
have high recovery value.
Firms can diversify across industries by means alternate
to acquisitions. We therefore test a related hypothesis:
Hypothesis Ic. The number of lines of business in which a
firm engages is higher in countries with stronger creditor
rights.
6
There is an inverse relation between the strictness of personal
bankruptcy laws and entrepreneurship, as reflected in the extent of self-
employment. See Fan and White (2003) and Armour and Cumming
(2005) for details.
 V.V. Acharya et al. / Journal of Financial Economics 102 (2011) 150–166 153

Next, we conjecture that creditor rights induce corpo-
rate managers to trade off corporate value for lower risk
when undertaking diversifying acquisitions.
Hypothesis II. In countries with stronger creditor rights,
diversifying acquisitions are followed by lower profitability
(measured by return on assets), and their announcement
induces more negative stock price reaction. Creditor rights
have no such effects on firm performance in focusing (same
industry) acquisitions.
Our remaining hypotheses test whether firms reduce
two additional measures of risk—operating risk and
financial risk—in countries with stronger creditor rights:
Hypothesis III. In countries with stronger creditor rights,
firms have lower cash-flow risk, measured by the stan-
dard deviation of the annual EBITDA-to-assets ratio.
Hypothesis IV. Corporate leverage (debt-to-assets ratio)
is decreasing in the strength of creditor rights.
These hypotheses are tested in a cross-section of coun-
tries, as well as over time, accounting for changes in creditor
rights. The data in our analysis include legal and economic
country variables, as well as data on individual companies
and acquisition transactions. Table 1 describes the construc-
tion and the data sources for all variables.
2.1. Creditor rights
We use the data on creditor rights from La Porta,
Lopez-de-Silanes, Shleifer, and Vishny (1998), who record
creditor rights components in a cross-country sample as
of 1994. The variable CRIGHTS is the sum of four provi-
sions: AUTOSTAY, REORG, SECURED, and MANAGES (see
Table 1 for details). Each of these components takes a
value of one if it is present in the country’s bankruptcy
code, or of zero if it is absent. Consequently, the range of
values for CRIGHTS is zero through four. In our 38-country
sample (see Table 2), the mean of CRIGHTS is 2.08 with
standard deviation of 1.28. As a robustness check, we use
the creditor rights data of Djankov, McLeish, and Shleifer
(2007) to re-estimate our models and to examine the
effects of changes in creditor rights.
2.2. Creditor rights and diversification in M&A activity
The tests of risk reduction through cross-industry diver-
sifying acquisitions employ data from the Securities Data
Corporation (SDC) Platinum Mergers & Acquisitions database
for the period 1994–2004. Our sample consists of acquisi-
tions in 38 countries with data on creditor rights as of 1994,
which satisfy the following requirements. We include only
mergers where both the acquirer and the target are under
the same jurisdiction. (Later, we present evidence on the

Table 1
Variable definitions.

Definition Source

Dependent variables
Pr(same-industry Probability of an acquisition being in the same industry. It equals one if acquirer SDC Platinum Mergers & Acquisitions
merger) and target are in the same two-digit industry, and zero otherwise
SAME The ratio of the number of mergers in the same two-digit SIC code to the number of SDC Platinum Mergers & Acquisitions
all domestic mergers in a country or in an industry in the country
ROA EBITDA/Assets. All data are annual. EBITDAj,c,t is earnings before interest, taxes, and Compustat Global Vantage
depreciation and amortization (the sum of data items #14 and #11), which is
unaffected by methods of accounting depreciation that differ across countries, and
Assetsi,c,t is contemporaneous total assets (data item #89). ROA is winzorized at 1%
in both tails
CAR—Cumulative Cumulative abnormal return computed over the seven-day window surrounding Datastream
abnormal return the event date, from day -3 to day þ 3, where day 0 is the announcement day. The
( 3 to þ3) daily abnormal return is the difference between the daily stock return and the
market model expected return, where the market model parameters are estimated
from a regression of weekly stock returns on the country’s stock return index. The
parameter estimation employs up to 105 weeks of data and no less than 52 weeks,
up to nine weeks before the week of the acquisition announcement. The
cumulative abnormal return is calculated over the seven days surrounding the
announcement day. The resulting sample of firms’ CAR is then winsorized at 1% at
both tails
Firm risk (RISK) RISKj,c is the standard deviation of firm j in country c of industry-adjusted ROAj,c,t, Compustat Global Industrial/
where ROAj,c,t ¼ EBITDAj,c,t/Assetsj,c,t. t is the year, and we require at least eight years Commercial Annual Database
of data. Here, ROA is industry adjusted by subtracting in each year the sample-wide
median ROA for the firm’s industry (using 2-digit SIC code). Data are for the period
1992–2005. The entire data of ROAi,c,t are winsorized at 0.5% in both tails to account
for extreme observations. The entire firm sample of RISKi,c is then winsorized at 1%
in both sides of the sample distribution. The measure is similar to the one used in
John, Litov, and Yeung (2008)
Country risk (RISKn) The median of RISKj,c across firms in country c
Leverage (also used Ratio of total debt-to-total assets in book value. Debt is total liabilities minus Compustat Global Vantage
as a control variable) equity and minus deferred taxes. Leverage data are winsorized in the entire
population at 1% in both tails. Industry-adjusted leverage is calculated as the firm’s
leverage minus the respective industry’s sample-wide median leverage (using
two-digit SIC code)
154 V.V. Acharya et al. / Journal of Financial Economics 102 (2011) 150–166

Table 1 (continued )

Definition Source

Creditor-rights variables
Creditor rights An index aggregating creditor rights, following La Porta, Lopez-de-Silanes, Shleifer, La Porta, Lopez-de-Silanes, Shleifer, and
(CRIGHTS) and Vishny (1998). It is the sum of the four indexes that follow. CRIGHTS then Vishny (1998); Djankov, McLeish, and
ranges between zero and four Shleifer (2007)
No automatic stay Equals one if the reorganization procedure does not impose an automatic stay on La Porta, Lopez-de-Silanes, Shleifer, and
(AUTOSTAY) the assets of the firm upon filing the reorganization petition, creditors are able to Vishny (1998)
seize their collateral after the reorganization petition is approved. It equals zero if
such restriction does exist in the law
Reorganization Equals one if the reorganization procedure imposes restrictions, such as creditors’ La Porta, Lopez-de-Silanes, Shleifer, and
(REORG) consent or minimum dividend for a debtor to be able to file for reorganization. It Vishny (1998)
equals zero for countries without such restriction
Secured debt first Equals one if secured creditors are ranked first in the distribution of the proceeds La Porta, Lopez-de-Silanes, Shleifer, and
(SECURED) that result from the disposition of the assets of a bankrupt firm, as opposed to other Vishny (1998)
creditors such as employees or government. Equals zero if non-secured creditors,
such as the government and workers, are given absolute priority
No management Equals one if an official appointed by the court, or by the creditors, is responsible La Porta, Lopez-de-Silanes, Shleifer, and
stay (MANAGES) for the operation of the business during reorganization, that is, management does Vishny (1998)
not retain administration of its property pending the resolution of the
reorganization. Equivalently, this variable equals 1 if the debtor does not keep the
administration of its property pending the resolution of the reorganization process,
and zero otherwise

Control variables
Shareholder rights An index that aggregates shareholder rights. ‘‘The index is formed by adding one Quotation is from La Porta, Lopez-de-
(SHRIGHTS) when: (1) the country allows shareholders to mail their proxy vote to the firm, Silanes, Shleifer, and Vishny (1998)
(2) shareholders are not required to deposit their shares prior to the general
shareholders’ meeting, (3) cumulative voting or proportional representation of
minorities in the board of directors is allowed, (4) an oppressed minorities
mechanism is in place, (5) the minimum percentage of share capital that entitles a
shareholder to call for an extraordinary shareholders’ meeting is less than or equal
to 10% (the sample median), or (6) shareholders have preemptive rights that can be
waived only by a shareholders’ vote. The index ranges from zero to six’’
Log(GDP per capita) Natural logarithm of the average real GDP per capita in US dollars, 1994–2000 Penn World Tables, Version 6.1
Macroeconomic risk The standard deviation of the quarterly growth in real industrial production for International Financial Statistics of
(Macro risk) each country in the period 1990–2004. For some countries, we use instead the International Monetary Fund (IMF)
index of manufacturing production: Argentina, Chile, Greece, Hong Kong,
Indonesia, New Zealand, Peru, Philippines, Singapore, and South Africa. For
Argentina, Canada, Taiwan, and Thailand, data are from the international database
of Global Insight. The variable is measured in decimal points
Rule of law The assessment of the law and order tradition of the country. Calculated as International Country Risk Guide; La
‘‘average of the months of April and October of the monthly index between 1982 Porta, Lopez-de-Silanes, Shleifer, and
and 1995. Scale from zero to ten, with lower scores for less tradition for law and Vishny (1998)
order’’
Legal origins A dummy variable that identifies the legal origin of the Company law or La Porta, Lopez-de-Silanes, Shleifer, and
Commercial Code of each country. The detailed origins are French, German, Nordic Vishny (1998); and the CIA Factbook
(default is Common) 2003
Accounting disclosure An index created by the examination of the annual report in 1994 of companies International Accounting and Auditing
across countries on their inclusion or omission of 90 line items Trends, Center for International
Financial Analysis and Research
Emerging markets Dummy variable equal to one if the country’s per capita GDP (in US$, average over Penn World Tables, Version 6.1
1994–2000) is less than the median for the sample of countries
Flexibility to fire An index of the ease to fire workers based on a study of the employment laws Doing Business Report, 2004, The
(divided by 100) World Bank
Country corporate The annual top corporate tax rate for 1992–2002, per country. It is then applied for World Tax Database at the University
tax rate 2003–2005 when necessary of Michigan
Log(Market cap) The logarithm of the stock market capitalization in U.S. dollars in 1994 World Market Indicators database, The
World Bank
Transaction value The amount paid in U.S. dollars. SDC Platinum Mergers & Acquisitions.
Tangibility Net fixed (tangible) assets/Total assets Compustat Global Vantage

effect of creditor rights on geographic diversification, using
cross-border transactions.) We exclude acquisitions where
the acquirer is in the financial industry (Standard Industry
Classification (SIC) header 6), which include leveraged buy-
outs (LBOs), or a regulated industry (SIC headers 48
and 49), transactions where the acquirer and the target are
the same company (repurchases recorded as acquisitions),
transactions where the acquirer is a mutual company,
investment company, subsidiary, or state-owned enterprise,
and transactions in which the equity ownership stake
acquired from the target is less than 20%. Finally, we include
only countries with at least 50 transactions that satisfy the
above criteria, though additional data requirements on
transaction value reduce the sample size for some countries.
 V.V. Acharya et al. / Journal of Financial Economics 102 (2011) 150–166 155

Table 2
Overall descriptive statistics.
The table describes the total number of domestic mergers in the sample countries for 1994–2004 that enter Table 3 regressions. The sample presented
consists of the countries for which we have La Porta, Lopez-de-Silanes, Shleifer, and Vishny (1998) data on creditor rights. We exclude countries that have
less than 50 qualified transactions in the sample period (before imposing additional data requirements). A transaction is qualified if the percentage of
acquired shares is at least 20%. We exclude financial industry (SIC header 6) and regulated industry companies (SIC headers 48 and 49) from the country
transaction count. The mergers and acquisition data are from SDC Platinum Mergers and Acquisitions database. The year of creditor rights change is the
one from the Djankov, McLeish, and Shleifer (2007) study. We also present data on the median country cash-flow risk proxy, RISKn. Refer to Table 1 for
variable definitions.

Acquirer’s No. of Proportion of same- Cash-flow Shareholder Creditor Macroeconomic $ GDP per
country mergers industry mergers (%) risk proxy rights rights volatility capita
Count SAME RISK* SHRIGHTS CRIGHTS Macro risk GDP

Argentina 66 55.33 0.058 4 1 0.07 $7,801

Australia 1,618 61.72 0.121 4 1 0.04 $20,948
Austria 14 64.52 0.036 2 3 0.09 $26,220
Belgium 49 57.54 0.043 0 2 0.08 $24,649
Brazil 143 70.26 0.070 3 1 0.03 $4,143
Canada 2,071 61.37 0.094 5 1 0.01 $20,647
Chile 41 61.84 0.033 5 2 0.04 $4,604
Denmark 80 56.47 0.049 2 3 0.07 $32,434
Finland 154 54.60 0.054 3 1 0.08 $23,856
France 434 59.79 0.045 3 0 0.10 $24,033
Germany 201 55.31 0.057 1 3 0.04 $26,443
Greece 70 47.22 0.043 2 1 0.06 $11,219
Hong Kong 190 34.11 0.064 5 4 0.13 $23,850
India 236 57.87 0.051 5 4 0.07 $423
Indonesia 39 60.53 . 2 4 0.07 $868
Ireland 92 63.59 . 4 1 0.08 $21,376
Israel 73 45.45 0.075 3 4 0.02 $16,391
Italy 333 53.31 0.038 1 2 0.12 $19,814
Japan 1,771 46.80 0.022 4 2 0.03 $36,616
Malaysia 369 25.27 0.066 4 4 0.05 $3,982
Mexico 82 62.59 0.049 1 0 0.03 $4,421
Netherlands 101 57.80 0.059 2 2 0.11 $24,802
New Zealand 98 57.73 0.073 4 3 0.06 $15,528
Norway 130 58.94 0.079 4 2 0.07 $33,844
Peru 26 68.63 0.058 3 0 0.07 $2,296
Philippines 42 56.00 0.080 3 0 0.18 $1,041
Portugal 56 65.31 0.036 3 1 0.06 $10,782
Singapore 243 32.19 0.064 4 4 0.06 $22,916
South Africa 372 49.84 0.061 5 3 0.02 $3,413
South Korea 198 32.48 0.051 2 3 0.06 $9,545
Spain 338 64.08 0.040 4 2 0.08 $14,535
Sweden 186 58.53 0.067 3 2 0.16 $26,812
Switzerland 38 57.67 0.046 2 1 0.07 $37,908
Taiwan 52 44.90 - 3 2 0.06 $12,580
Thailand 83 43.95 0.065 2 3 0.05 $2,396
Turkey 17 50.00 0.097 2 2 0.07 $2,810
United Kingdom 5,624 58.61 0.071 5 4 0.05 $21,767
United States 17,491 59.07 0.088 5 1 0.01 $30,899

We test Hypothesis Ia by estimating the likelihood of
same-industry acquisition in a country as a function of the
strength of creditor rights, controlling for other variables.
A diversifying acquisition is one where the acquirer and
the target are not in the same industry (using two-digit
SIC code).7 The main explanatory variable in our analysis
is CRIGHTS, the aggregate measure of creditor rights from
La Porta, Lopez-de-Silanes, Shleifer, and Vishny (1998),
and its components, AUTOSTAY, REORG, SECURED, and
MANAGES. We predict that the coefficient of CRIGHTS is
negative; that is, there is lower likelihood of same-industry
mergers in countries with stronger creditor rights.
7
The results are qualitatively similar when we employ industry
classification at the three-digit SIC level.
The univariate relation between the country’s creditor
rights and diversifying mergers is observed in Fig. 1,
which plots the country variable SAMEc, the proportion
of same-industry acquisitions in country c, as a function of
the country’s CRIGHTS. The pattern shows a negative
relation between these two variables; that is, stronger
creditor rights lead to greater diversification.
Next, to test our Hypothesis Ia, we do a detailed
transaction-level analysis in which we estimate the like-
lihood of a same-industry acquisition as a function of
creditor rights and control variables. The country-level
control variables include shareholder rights index,
SHRIGHTS (obtained from La Porta, Lopez-de-Silanes,
Shleifer, and Vishny, 1998); its effect should be positive
if same-industry acquisitions, which produce synergies
and retain the same level of asset risk, are in shareholders’
156 V.V. Acharya et al. / Journal of Financial Economics 102 (2011) 150–166

0.7

Proportion of same-industry mergers

0.6

0.5

0.4

0.3

0.2
0 1 2 3 4
Creditor rights

Fig. 1. The fraction of same-industry mergers and acquisitions (two-digit SIC code) out of all mergers and acquisitions in a country, SAME, and creditor
rights, CRIGHTS. The vertical axis represents the share of same-industry mergers in a country. The horizontal axis represents creditor rights. The fitted line
represents the slope from a Tobit regression of the fraction of same-industry mergers in a country on a constant and the creditor rights index. The slope
coefficient is  0.043 with t-statistic¼ 3.94.

interest (after controlling for creditor rights). To control data limitation. Over 45% of the acquirers in our sample
for the effects of level of development and of efficiency and 88% of the target firms have no accounting informa-
of the capital market (see La Porta, Lopez-de-Silanes, tion. Consequently, leverage data on both acquirer and
Shleifer, and Vishny, 1997), we include the following target are available for only 2,586 transactions, about 8%
variables (details on the construction of these variables of the sample (without the U.S. and the U.K., we have only
are in Table 1): Log(Market cap),8 Accounting disclosure, 746 transactions with leverage data). In addition, leverage
and Rule of law. These variables should positively affect in any country is partly endogenous to the country’s
the likelihood of same-industry acquisitions if the internal creditor rights. We therefore use estimated leverage vari-
capital markets in conglomerates substitute for less- ables, derived from an instrumental variables regression.
efficient outside capital markets. Similarly, the Emerging For all transactions with data on leverage (see Table 1 for
market dummy variable should have a negative coefficient definition), we estimate a regression for both acquirer and
if internal capital markets in conglomerates are valuable. target firm leverage as follows: We regress the acquiring
Flexibility to fire proxies for the efficiency of the labor (target) firm’s leverage on all country-level control vari-
market, which may affect merger type. Legal origin influ- ables and on two exogenous variables, the ranks (in
ences a number of national institutional variables, including quartiles) of U.S. median leverage and U.S. median tangi-
creditor and shareholder rights (La Porta, Lopez-de-Silanes, bility (the ratio of fixed assets-to-total assets) for the
Shleifer, and Vishny, 1998); this variable also interacts with industry of the acquiring (target) firm over the period
the likelihood of bankruptcies (Claessens and Klapper, 1992–2004. The U.S. has a low level of creditor rights
2005). These three legal control variables are obtained (CRIGHTS¼1), which implies under our hypotheses a less-
from Levine and Demirguc-Kunt (2001) and La Porta, constrained choice of leverage, and it has the most data
Lopez-de-Silanes, Shleifer, and Vishny (1998). on all industries, making the estimation more reliable.
The country-level control variables include Macro risk Accordingly, we impute the leverage of an acquirer
and the country’s average real GDP per capita, which is a (target) firm in any industry in any country as the
proxy for the degree of economic development. The effect leverage in that country and industry estimated from a
of Macro risk should be negative if managers in riskier model of the acquirer’s (target’s) leverage as a function of
countries carry out more diversifying mergers. We control two exogenous industry variables, using U.S. data, and of
for GDP per capita because developed and developing the acquirer’s own country’s exogenous control variables.
countries may have different investment opportunity sets. For our main test, we estimate a probit model for each
The transaction-related control variables are Transac- transaction j in country c,
tion value and Leverage of both acquirer and target firms. Prðsame industry mergerÞj,c
Leverage represents the financial distress risk, which the
¼ a0 þ a1  CRIGHTSc þcontrol variables: ð1Þ
acquirer should take into account. This risk induces
diversifying mergers, and therefore, we expect the coeffi- The dependent variable equals one if acquirer and target
cients of leverage to be negative. In estimation, we face a are in the same two-digit SIC industry, and zero other-
wise. Hypothesis Ia implies that a1 o0. The model
includes year dummy variables and estimation-cluster
8
Our results are robust to an alternative definition of capital market standard errors by country.
development, using the ratio of market capitalization-to-GDP as of 1994.
However, by this definition, Malaysia, Hong Kong, and South Africa rank
The results in Table 3 are consistent with Hypothesis
at the top, while the U.S. ranks in eighth place, after Chile; Japan is Ia. The coefficient of CRIGHTS is negative and statistically
ranked thirteenth, after Thailand and the Philippines. significant (column 1), meaning that stronger creditor
 V.V. Acharya et al. / Journal of Financial Economics 102 (2011) 150–166 157

rights are associated with greater propensity to diversify
(lower probability of same-industry merger). The results
are similar when we exclude the U.S. (column 6) or both
the U.S. and the U.K. (column 7), which account for by far
the largest number of acquisitions. All four creditor rights
components have negative and significant coefficients.
The component with the most negative effect is MANAGES,
underscoring the importance of managerial dismissal in
bankruptcy as an incentive to diversify. Based on columns
1 and 7 of Table 3, the marginal effect of CRIGHTS on the
propensity to acquire same-industry target, evaluated at
mean CRIGHTS (‘‘local elasticity’’), is  9.49%, and
16.48% when excluding the U.S. and the U.K.
Notably, the effect of shareholder rights is opposite to
that of creditor rights. The positive and significant coeffi-
cient of SHRIGHTS when excluding the U.S. suggests that
shareholder interests induce focusing acquisitions, after
controlling for the effect of creditor rights. The evidence
on the effect of internal capital markets in conglomerates
is mixed. While the positive coefficient of Log(Market cap)
implies that in countries with developed capital markets
there is a lesser need for conglomerate mergers, this

Table 3
Merger-level analysis: proportion of same-industry mergers.
The table presents the coefficient estimates from probit regressions. The dependent variable equals one if both acquirer and target are in the same
industry, using two-digit SIC code. A country is included in our sample if it has at least 50 qualified transactions over the sample period. A transaction is
included if the percentage of acquired shares is at least 20%. Excluded are transactions where the acquirer is from the financial industry (SIC header 6) or
regulated industry (SIC headers 48 and 49). CRIGHTS are as of 1994. The control variables include shareholder rights, rule of law, macroeconomic risk,
legal origins, the logarithm of the stock market capitalization, the index of flexibility to fire, the quality of accounting disclosure, an emerging market
indicator, the logarithm of average real per capita GDP (1994–2000) in US$, the logarithm of transaction value, and the imputed leverage for the acquirer
and the target (the predictors are the U.S. industry quartile rank of the median leverage and median tangibility, and all exogenous control variables). All
variables are defined in Table 1. The regressions include year fixed effects (not reported). Models (1) through (5) include all countries. Model (6) excludes
the U.S. Model (7) excludes both the U.S. and the U.K. The t-statistics in parentheses are based on robust estimation of standard errors with errors cluster-
adjusted at the country level. nnn, nn, and n indicate significance at the 1%, 5%, and 10% levels, respectively. Sample period is 1994–2004.

Variable All countries Exclude U.S. Exclude U.S. and U.K.

(1) (2) (3) (4) (5) (6) (7)

nnn nnn
CRIGHTS  0.245  0.411  0.420nnn
(6.33) (3.66) (3.56)
AUTOSTAY  0.415nnn
(5.74)
REORG  0.524nnn
(5.18)
SECURED  0.318nnn
(3.78)
MANAGES  0.848nnn
(6.89)
SHRIGHTS 0.022  0.002 0.012 0.029 0.143nnn 0.218nnn 0.112nnn
(0.91) (0.08) (0.46) (1.16) (5.38) (4.00) (2.88)
Log (Market cap) 0.293nnn 0.247nnn 0.266nnn 0.207nnn 0.225nnn 0.134nnn  0.035
(6.48) (5.94) (6.01) (5.56) (6.20) (3.09) (0.87)
Flexibility to fire 0.836nn 1.077nnn 1.37nnn 0.993nn  0.262 0.503  0.449
(2.15) (2.68) (3.32) (2.41) (0.67) (1.17) (0.94)
Accounting disclosure  0.026nnn  0.033nnn  0.030nnn  0.047nnn  0.011nn  0.024nnn  0.035nnn
(4.46) (5.23) (4.82) (6.26) (2.21) (3.09) (4.31)
Emerging market 0.661nnn 0.421nnn 0.505nnn 0.303nnn 0.653nnn 1.932nnn 0.956nn
(5.99) (4.75) (5.27) (4.25) (6.08) (3.42) (2.58)
Rule of law 0.375nnn 0.544nnn 0.362nnn 0.445nnn 0.026 0.952nnn 0.600nnn
(5.79) (7.85) (5.66) (6.87) (0.34) (4.13) (4.33)
French legal origin  0.388nnn  0.189n  0.305nn  0.224nn 0.0004 0.213 0.009
(2.87) (1.71) (2.52) (2.10) (0.01) (1.61) (0.07)
German legal origin  0.613nnn  0.86nnn  0.950nnn  0.968nnn 0.101 1.087nn 0.673n
(9.32) (9.96) (10.58) (10.49) (1.00) (2.35) (1.69)
Nordic legal origin 1.167nnn 0.903nnn 1.245nnn 1.097nnn 1.212nnn 2.458nnn 1.393nn
(5.3) (4.49) (5.33) (5.10) (5.70) (3.33) (2.58)
Macro risk  0.207  2.945nnn  1.841nn  6.312nnn  0.803 4.755nn 1.765
(0.22) (3.18) (1.96) (6.76) (0.97) (2.15) (1.13)
Log(GDP per capita)  0.178nnn  0.199nnn  0.097nn  0.055  0.152nnn  0.423nnn  0.354nnn
(3.37) (3.71) (2.04) (1.20) (3.09) (3.13) (3.35)
Log(Transaction value) 0.086nnn 0.090nnn 0.091nnn 0.096nnn 0.083nnn 0.083nn 0.064nnn
(5.41) (5.47) (5.47) (5.52) (5.30) (2.29) (2.85)
Acquirer’s leverage (imputed) 1.746n 1.755 1.734 1.737 1.754n  0.486  1.376n
(1.69) (1.68) (1.64) (1.64) (1.70) (0.35) (1.75)
Target’s leverage (imputed)  7.647nnn  7.861nnn  8.062nnn  8.251nnn  7.584nnn  13.957nnn  7.700nn
(6.13) (6.14) (6.17) (6.17) (6.09) (3.51) (2.57)

Number of countries 38 38 38 38 38 37 36
Observations 33,221 33,221 33,221 33,221 33,221 15,730 10,106
Chi-squared 4,449.7 4,279.3 1,696.8 1,375.8 5,870.4 1,838.4 2,079.4
158 V.V. Acharya et al. / Journal of Financial Economics 102 (2011) 150–166
coefficient becomes negative and insignificant when
excluding the U.S. and the U.K. While internal capital
markets in diversified firms are claimed to be of greater
value in poorer countries, we obtain that the coefficient of
Emerging market is positive (i.e., greater likelihood of
focusing acquisitions), and that of GDP per capita is
negative (i.e., greater likelihood of diversifying acquisi-
tions in richer countries). Accounting disclosure, an aspect
of a developed capital market, has a negative coefficient,
implying a greater degree of diversification in such
markets. On the other hand, Rule of law, which is asso-
ciated with capital market development, has a positive
coefficient, consistent with the importance of internal
capital markets in countries with underdeveloped
markets.
Target’s leverage has a significant negative effect on the
propensity to do same-industry mergers, suggesting that
higher financial distress risk induces diversifying acquisi-
tions. The elasticity of target leverage (evaluated at the
mean of Target’s leverage) is 2.96, meaning that a 1%
increase in the target leverage ratio reduces the likelihood of
same-industry merger by nearly 3%. Reinforcing this effect,
Acquirer’s leverage also has a negative and marginally
significant coefficient when excluding the U.S. and the
U.K.9 The effects of Flexibility to fire, Macro risk, and French
legal origin and German legal origin are inconclusive, given
that their coefficients switch signs and change significance
depending on whether the U.S. and the U.K. are included.
2.2.1. Robustness tests
We briefly review some robustness tests that we per-
form. Details on these tests are available in an Appendix at
the Journal of Financial Economics (JFE) Web site. First, to
address the concern that both creditor rights and corporate
strategy might be driven by culture-based factors, we add to
the model culture-based variables: the religious composi-
tion of the country’s population, following Stulz and
Williamson (2003), and the Uncertainty avoidance index of
business people in the country, due to Hofstede (2001). Still,
CRIGHTS retains its negative and highly significant coeffi-
cient. We also compare the U.S. and the U.K. alone, which
are similar in culture and financial institutions but differ in
their creditor rights. We find that the probability of diversi-
fying mergers is significantly higher in the U.K., whose
CRIGHTS equals four versus one in the U.S. Second, we
address the concern that the propensity to do same-indus-
try acquisitions is affected by the country’s antitrust laws.
We control for that, using data from Hylton and Deng
(2007), and find that the coefficient of CRIGHTS remains
negative and significant. Third, we test our hypothesis at the
aggregate country level, in which each country—large and
small alike—is one observation. We regress SAMEc (propor-
tion of same-industry acquisitions) on CRIGHTSc and
9
The fact that both target and acquirer leverage are associated with
diversifying acquisitions is consistent with our focus on the managerial
agency problem, and potentially inconsistent with the view of agency
problems between creditors and shareholders being of importance for
corporate risk-taking. The latter would imply that higher acquirer
leverage leads to stronger risk-taking incentives, and thereby, would
be associated with focusing acquisitions.
country-level control variables. The coefficient of CRIGHTS
remains negative and significant.
Geographic diversification: We estimate a model similar
to that in Table 3, where the dependent variable equals
one for domestic acquisition, or zero for cross-border
acquisition. The creditor rights pertain to the acquirer’s
country (among the 38 countries in our sample; the target
firms can be from any country). The coefficient of CRIGHTS
is again negative and statistically significant (  0.111,
t ¼5.23), and remains so even if we exclude the U.S. and
the U.K. (over 60% of the acquisitions). This is also
consistent with our hypothesis that stronger creditor
rights induce diversifying acquisitions.
Industry-level analysis: Because countries differ in the
composition of their industries, we estimate our model at
the industry level, following the methodology in Rajan and
Zingales (1998). We measure the inherent propensity in
an industry to do same-industry acquisitions by the pro-
pensity to do that in the U.S., whose acquisition market is
the most active and whose relatively low level of creditor
rights (CRIGHTS¼1) makes firms less constrained. We
estimate the following model:
SAMEk,c ¼ g0 þ g1  SAMEk,US þ g2  CRIGHTSdm
c  SAMEk,US
þCountry fixed effects: ð2Þ
SAMEk,US is the proportion of same-industry acquisitions
in the U.S. in industry k out of all acquisitions where the
acquirer is in industry k (using two-digit SIC code during
the period 1994–1997), and SAMEk,c is defined similarly
as the proportion of same-industry acquisitions of acquirers
in industry k in country c for the subsequent period,
dm
1998–2004. CRIGHTSc is the demeaned value of CRIGHTSc,
obtained by subtracting the overall sample mean of
CRIGHTS. We include an industry from a given country if it
has at least six qualified transactions during the period
1998–2004. We estimate model (2) by the Tobit method. We
obtain that g1 ¼1.310 (t¼9.83) and g2 ¼  0.263 (t¼7.56).
That is, across industries, strong creditor rights in a country
reduce the tendency to engage in same-industry acquisitions
relative to the inherent tendency in these industries.
2.2.2. Risk reduction in acquisitions and industry
recovery rates
The next test, Hypothesis Ib, examines the effect of
creditor rights on the choice of target in acquisitions in
terms of the recovery rate of its assets in default (hence-
forth, recovery). Recovery is the extent to which the price
of the assets sold in distress is close to the value of
the assets in their best use, following the inverse of the
definition of asset specificity in Shleifer and Vishny (1992).
We propose that a firm with high-recovery assets can
better accommodate financial distress, by partially liqui-
dating such assets, using the proceeds to service debt and
thus deferring default. Such a deferment increases the
value of the call option embedded in the firm’s equity,
increasing its value.10 This assumes that the firm’s
10
Berger, Ofek, and Swary (1996) find that a high recovery value of
assets (imputed from book value items) is particularly valuable for firms
in financial distress.
 V.V. Acharya et al. / Journal of Financial Economics 102 (2011) 150–166
volatility does not appreciably change following the sale
of an asset and that the proceeds from such sale are
used to service debt. If, however, the cash obtained from
liquidating risky assets is retained, for example—to satisfy
a working capital covenant restriction, the firm’s volatility
declines and the net effect on the equity value is uncer-
tain. But such risk reduction may serve the interests of a
risk-averse manager who chooses to reduce volatility in
response to strong creditor rights that may threaten
managerial job security. In this scenario, high-recovery
assets become attractive. This scenario would motivate
managers of bidder firms with low-recovery assets to
acquire targets with high-recovery assets to enhance their
chances of survival.
We test the effect of creditor rights on the type of target
that a low-recovery bidder seeks. Denote the event of low-
recovery acquirer (AL) buying a high-recovery target (TH) by
AL \ TH, and the probability of this event occurring among
all such acquisitions by Pr(AL \ TH9TH). We estimate the
determinants of Pr(AL \ TH9TH) in the model
PrðAL \ TH9THÞj,c ¼ d0 þ d1  CRIGHTSc þcontrol variables: ð3Þ
We hypothesize that d1 40: stronger creditor rights induce
acquirers in low-recovery industries to buy target firms in
high-recovery industries. We assign to firms the recovery
level of the industry in which they operate, using the data in
Acharya, Bharath, and Srinivasan (2007, Table 2), which
employ historical experience on defaults in the U.S. over
the period 1982–1999.11 In the estimation model, the uni-
verse is all targets with high recovery, and the bidders are
either low recovery (dependent variable¼1), or high recov-
ery (dependent variable¼ 0). The control variables are those
from Table 3.
The results in Table 4 support Hypothesis Ib. The coeffi-
cient of CRIGHTS is positive and significant for the entire
sample as well as when excluding the U.S. and the U.K.,
which constitute more than half the sample (columns 6–10).
That is, stronger creditor rights induce greater likelihood of
acquisition of high-recovery targets by low-recovery firms.
All components of creditor rights have positive coefficients,
and except for the component that secured creditors are paid
first (SECURED), they are all statistically significant. The
relatively large coefficient of the managerial dismissal com-
ponent (MANAGES) underscores the evidence in Eckbo and
Thorburn (2003) cited above. Macro risk has a positive effect
on the likelihood of low-recovery firms acquiring high-
recovery firms, which is consistent with the view that such
acquisitions are chosen as means to reduce risk.12 This test
provides additional evidence that creditor rights affect the
11
Low-recovery industries (in terms of two-SIC code headers) include:
transportation (37, 40, 41, 42, 44, 45, 46, 47), high technology and office
equipment (35, 36, 38), consumer/service sector (52, 53, 54, 55, 56, 57, 58,
59, 72, 73, 75, 76, 78, 79), and leisure time/media (27, 48, 70). High-
recovery industries include: energy and natural resources (10, 12, 13, 14,
24), building products/homebuilders (8, 15, 17, 24, 28, 29, 32, 34), and
healthcare/chemicals (28, 80). We alternatively follow Dyck and Zingales
(2004), characterizing as low recovery rate industries the following:
mining, manufacturing, and transportation. Our results are similar.
12
We also estimate the model as a country-level regression, where
the dependent variable is the proportion of all high-recovery targets
in the country acquired by low-recovery bidders (we use logistic
159
nature of corporate investment—the choice of an acquisition
target—in an attempt to mitigate the likelihood of default.
2.2.3. Diversification across industry lines of business
In addition to diversifying through acquisitions, firms
can develop internally a portfolio of lines of business in
different industries. We now test Hypothesis Ic, on the
relation between creditor rights and the number of seg-
ments that firms have. The data, obtained from the
Compustat North America Segment file, are confined to
firms that file such reports by U.S. rules.13 These include
non-U.S. firms whose stock is traded in the U.S. through
American Depository Receipts, or that are listed in the
U.S., or report by U.S. rules to conform to securities laws,
as well as Canadian firms. Surely, this sample is limited to
firms from countries whose characteristics (such as size,
visibility, or type of business) cause their securities to trade
in the U.S., or to non-U.S. firms that choose to be covered by
U.S. financial reporting rules. With this caveat in mind, we
analyze the data on the effect of creditor rights on firms’
tendency to diversify across lines of business.
The sample consists of 836 firms in the manufacturing
industries (SIC codes 1000 through 3999) from 21 coun-
tries, excluding the U.S. (which has overwhelmingly more
firms than any country in this sample), with 4,520 firm-
years of reporting between 1992 and 2005. We also require
that firms have leverage data in either the Compustat North
America or the Compustat Global Vantage databases, and as
before, we include a country only if at least six firms have
data in that country. Because Canadian firms constitute
about half of the firm-years in our sample, we repeat our
estimation for a sample of 20 countries with 2,132 firm-
years that excludes Canadian firms.
We estimate a probit regression, where the dependent
variable equals one if the firm has more than one business
segment, and zero otherwise. CRIGHTS is the main expla-
natory variable in our model, which includes all other
explanatory variables that appear in Table 3, with firm
leverage estimated and instrumented with the same
variables. We hypothesize that the coefficient of CRIGHTS
is positive, i.e., stronger creditor rights induce firms to
diversify across lines of business (internally or by acquisi-
tion). Indeed, the coefficient of CRIGHTS is positive, 0.346
(t¼ 2.94); excluding Canada: 0.236 (t¼ 3.45).
In an ordinary least squares regression, where the
dependent variable is Log(number of segments), the coeffi-
cient of CRIGHTS is 0.110 (t ¼3.72); excluding Canada:
0.106 (t ¼2.51).14 These results further support our
hypothesis that stronger creditor rights induce firms to
reduce their risk through diversification across industries.
(footnote continued)
transformation). The coefficient of CRIGHTS is positive, 0.288, with
t ¼3.37, consistent with the results for the single-acquisition regression.
13
Foreign firms listing in the U.S. are required to file with the
Securities Exchange Commission (SEC) and reconcile with U.S. Generally
Accepted Accounting Principles (GAAP) and Financial Accounting Stan-
dards Board (FASB) rules (in particular, Statements of Financial Account-
ing Standards (SFAS) 131 and its predecessor, SFAS 14, regarding the
reporting of segments data) in their annual reports.
14
Standard errors are clustered by country, and the model includes
year dummy variables.
160 V.V. Acharya et al. / Journal of Financial Economics 102 (2011) 150–166

Table 4
Recovery rates and mergers and acquisitions.
The table presents coefficient estimates of probit models. The dependent variable equals one if Pr(TH\AL9TH) ¼1, i.e., if the target is in a high-recovery
industry and the acquirer is in a low-recovery industry. The universe is all target firms in a high-recovery industry. Included are all transactions where
the percentage of acquired shares is at least 20%. Excluded are transactions involving acquirers that are financial industry (SIC header 6) or regulated
industry companies (SIC headers 48 and 49). The following industries are classified as low recovery (two-SIC code headers): transportation (37, 40, 41, 42,
44, 45, 46, 47), high technology and office equipment (35, 36, 38), consumer/service sector (52, 53, 54, 55, 56, 57, 58, 59, 72, 73, 75, 76, 78, 79), or leisure
time/media (27, 48, 70). The following industries are classified as high recovery (two -SIC code headers): energy and natural resources (10, 12, 13, 14, 24),
building products/ homebuilders (8, 15, 17, 24, 28, 29, 32, 34), or healthcare/chemicals (28, 80). This classification follows Acharya, Bharath, and
Srinivasan (2007). All variables are defined in Table 1. The leverage of acquirer and target is calculated as in Table 3. The sample period is 1994–2004. The
absolute values of the t-statistics are shown in parentheses below the coefficients and are based on robust standard errors that are cluster-adjusted at the
country level. We include year fixed effects (not reported). nnn, nn, and n indicate significance at the 1%, 5%, and 10% levels, respectively.

Variables All countries Excluding the U.S. & U.K.

(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)

CRIGHTS 0.128nn 0.354nnn

(2.45) (4.56)
AUTOSTAY 0.915nnn 0.425nn
(6.30) (2.38)
REORG 0.277n 0.492nnn
(1.93) (2.64)
SECURED 0.124 0.51nn
(0.60) (2.33)
MANAGES 0.619nnn 1.466nnn
(3.50) (5.25)
SHRIGHTS  0.107n 0.016  0.114n  0.118  0.22nnn  0.18nnn  0.191nnn  0.165nn  0.211nnn  0.357nnn
(1.65) (0.22) (1.71) (1.54) (2.71) (3.32) (2.81) (2.33) (3.35) (5.81)
Log (Market cap) 0.083  0.615nnn 0.12n 0.165nn 0.116n 0.283nnn 0.25nn 0.205nn 0.081 0.416nnn
(1.17) (6.49) (1.86) (2.27) (1.90) (2.94) (2.13) (2.05) (0.66) (3.47)
Flexibility to fire  1.250  2.379nn  1.716  1.305  0.274  0.497  1.108  1.496  1.456 1.86nn
(1.34) (2.32) (1.52) (0.99) (0.33) (0.67) (1.12) (1.67) (1.21) (2.16)
Accounting disclosure  0.05nnn 0.018  0.045nnn  0.034nnn  0.056nnn  0.035nnn  0.039 nnn
 0.039 nnn
 0.036nnn  0.048nnn
(3.73) (1.15) (3.38) (3.05) (4.51) (3.23) (3.12) (3.6) (3.45) (5.23)
Emerging markets  0.162  1.138nnn  0.059 0.027  0.259  0.453  0.492  0.345  0.84nn  0.157
(1.02) (5.69) (0.49) (0.21) (1.68) (1.47) (1.44) (0.99) (2.31) (0.49)
Rule of law 0.190  0.56nn 0.225 0.161 0.424nn 0.186  0.209 0.107  0.145 0.909nnn
(1.01) (2.33) (1.18) (0.76) (2.05) (0.81) (0.8) (0.44) (0.55) (3.15)
French legal origin  0.405 0.317  0.418  0.538  0.72nnn  0.330  0.849nn  0.579  0.707nn  0.774nn
(1.44) (1.04) (1.52) (1.59) (2.69) (1.09) (2.49) (1.63) (2.14) (2.57)
German legal origin  0.257nn 1.483 nnn
 0.099  0.084  0.871 nnn
 1.118nnn  1.017nnn  0.578  0.868n  2.025nnn
(2.08) (9.38) (0.94) (0.62) (3.57) (3.65) (2.70) (1.32) (1.90) (6.15)
Nordic legal origin 0.733nn  2.317nnn 0.817nn 0.982nn 0.670n 0.156 0.289 0.112  0.218 0.279
(2.00) (4.46) (2.26) (2.08) (1.79) (0.35) (0.58) (0.25) (0.48) (0.57)
Macro risk 5.734n 9.374nnn 6.488nn 9.048nnn 4.761 3.149 5.129n 5.223nn 5.228n 4.462
(1.94) (3.01) (2.42) (3.92) (1.68) (1.15) (1.7) (1.99) (1.79) (1.6)
Log(GDP per capita) 0.249n 0.600nnn 0.173 0.156 0.272 nn
0.3nn 0.363nn 0.211 0.222 0.249nn
(1.75) (3.47) (1.29) (0.99) (1.98) (2.32) (2.09) (1.57) (1.25) (2.05)
nnn n
Log (Transaction value) 0.015  0.136 0.017 0.019n 0.017n 0.073nnn 0.076 nnn
0.077 nnn
0.074nnn 0.085nnn
(1.49) (6.45) (1.70) (1.95) (1.7) (3.11) (3.23) (3.22) (3.14) (3.65)
Acquirer’s leverage (imputed)  12.99nnn  26.83nnn  12.88nnn  12.71nnn  12.96nnn  15.25nnn  14.80nnn  14.84nnn  14.72nnn  15.17nnn
(9.47) (7.03) (9.25) (9.08) (9.4) (12.8) (12.65) (12.04) (11.71) (12.37)
Target’s leverage (imputed) 5.838nnn 32.688nnn 5.658nnn 5.342nnn 5.762nnn 10.807nnn 10.217nnn 10.046nnn 9.833nnn 10.133nnn
(9.03) (14.58) (8.66) (7.66) (8.83) (7.75) (7.22) (6.62) (6.93) (7.45)

No. of countries 38 38 38 38 38 36 36 36 36 36
Observations 6,495 6,495 6,495 6,495 6,495 2,599 2,599 2,599 2,599 2,599
Chi-squared 28,376.0 6,360.2 43,325.1 13,403.8 12,529.0 27,974.7 6,449.9 15,708.8 9,494.0 13,115.9

2.3. Value effects of creditor rights in diversifying
cross-industry acquisitions
Hypothesis II suggests that stronger creditor rights
induce managers to undertake diversifying acquisitions
even when they harm corporate performance. To some
extent, creditor rights may be desirable because they reduce
the dead-weight cost that results from the conflict of
interests between debtors and creditors (see Jensen and
Meckling, 1976). For example, debtors may choose to
continue the operation of a distressed firm even if it erodes
firm value in order to retain the option of a rebound. Then,
diversification that reduces this conflict may be beneficial.
However, we hypothesize that in countries with stron-
ger creditor rights, diversifying acquisitions are associated
with worse firm performance. In testing this hypothesis,
we present as a benchmark the relation between firm
performance and creditor rights in focusing (same indus-
try) acquisitions, which are likely to be undertaken for
economic reasons, such as synergy. In focusing acquisi-
tions, we do not expect that the strength of creditor rights
is negatively associated with firm performance.
 V.V. Acharya et al. / Journal of Financial Economics 102 (2011) 150–166
The effect of diversification on corporate performance
is subject to debate, but much of the evidence suggests
that diversification is value-reducing.15 In countries with
underdeveloped capital markets, diversified conglomer-
ates may be beneficial by enabling internal capital mar-
kets that substitute for the costly outside capital market.
However, Scharfstein and Stein (2002) point out that
internal capital markets can have a ‘‘dark side’’ if they
cause inefficient subsidization of poorly performing
groups.16
We test the consequences of diversifying acquisitions
as a function of a country’s creditor rights using two
measures of firm performance: (1) the change in the
acquiring firm’s return on assets, ROA, in the three years
after the acquisition compared to the ROA in the year
prior to acquisition, and (2) the acquirer’s stock price
change, measured as the cumulative abnormal return,
CAR, over a seven-day window surrounding the acquisi-
tion announcement. (Details on the calculation of ROA and
CAR are in Table 1.) We estimate the following model
separately for focusing and for diversifying acquisitions:
Performancej,c ¼ Z0 þ Z1  CRIGHTSc þ control variables:
In the first test, Performancej,c is DROA(k)j,c ¼ROAj,c(t þk)
ROAj,c(t  1), the change in the return on assets of
acquirer firm j in country c, k year(s) after the acquisition
compared to one year before its announcement, t is the
effective year of the acquisition, and k ¼1, 2, 3. We use
ROA as a performance measure, following Healy, Palepu,
and Ruback (1992), who suggest that this measure is
unaffected by the method of accounting used in mergers.
The control variables include all the country variables
used in the previous analyses and Log(transaction value).
To control for the endogeneity of the decision to diversify,
we add the inverse Mills ratio from the probit regression
presented in Table 3. The model includes acquirer indus-
try fixed effects to control for industry-wide changes in
profitability, and year fixed effects. We estimate the
model separately for one, two, and three years after the
merger (k¼1, 2, 3).
The coefficient Z1 measures the extent to which the
post-merger change in the acquirer’s ROA is affected by
the country’s creditor rights. Our hypotheses are:
For diversifying acquisitions: Z1 o0, because managers
may sacrifice corporate performance to attain lower
corporate risk.
For focusing (same industry) acquisitions: Z1 Z0.
The results in Table 5, columns 1–6, support our
hypothesis. The coefficient Z1 is negative and highly
15
See (among others) Morck, Shleifer, and Vishny (1990), Berger
and Ofek (1995), Comment and Jarrell (1995), Moeller and
Schlingemann (2005), Campa and Kedia (2002), Villalonga (2004),
Ammann, Hoechle, and Schmid (2008), Laeven and Levine (2007), and
Schmid and Walter (2009).
16
Related literature on the dark side of internal capital markets also
includes Berger and Ofek (1995), Lamont (1997), Scharfstein (1998),
Shin and Stulz (1998), Rajan, Servaes, and Zingales (2000), Lamont and
Polk (2002), Comment and Jarrell (1995), Khanna and Palepu (2000),
Lins and Servaes (2002), and Lee, Peng, and Lee (2008).
161
significant in diversifying mergers for all three post-
merger horizons. That is, the acquirer’s post-merger
performance declines significantly in countries with
stronger creditor rights up to three years after the merger,
relative to its performance in the pre-merger year. For
focusing (same industry) acquisitions, the coefficient Z1
is not significantly different from zero, again consistent
with Hypothesis II. We conclude that diversifying acquisi-
tions that seem to occur in response to stronger creditor
rights are followed by deterioration in corporate
performance.17
When we exclude the U.S. and the U.K. (which have by
far the largest number of observations within the sample),
the results remain qualitatively the same. For the sample
of diversifying acquisitions, the coefficient of CRIGHTS in
model (4) is  0.015 (t ¼3.43) for dROA(1)j,c, for dROA(2)j,c
it is  0.017 (t¼4.07), and for dROA(3)j,c it is  0.013
(t¼ 4.17). In contrast, for focusing (same industry) acqui-
sitions, the coefficients of CRIGHTS are all insignificantly
different from zero.
In the second test, Performancej,c is the acquirer’s CAR
over the seven days surrounding the announcement
ð4Þ
day.18 The results on the effect of creditor rights on the
acquirer’s CAR support our hypothesis and are consistent
with the results on post-acquisition changes in the
acquirer’s ROA. As presented in Table 5 columns 7–8,
diversifying acquisitions lower the acquirer’s value in
countries with stronger creditor rights: the coefficient Z1
of CRIGHTS in model (4) is negative and significant for
diversifying acquisitions (column 8), 0.013 (t ¼5.14),
meaning a value loss of 1.3% upon the announcement of
such deals for any unit of CRIGHTS, whereas for focusing
(same industry) acquisitions (column 7) it is positive and
marginally significant, 0.011 (t¼1.65).19 Excluding the
U.S. and the U.K., the coefficient of CRIGHTS in diversifying
acquisitions is 0.011 (t¼2.40), significant, while in
focusing acquisitions it is 0.002 (t ¼0.26), insignificant.
When estimating the model for the U.S. and the U.K.
alone, which are similar in culture and financial institu-
tions but different in creditor rights (CRIGHTS ¼4 for the
U.K. and one for the U.S), the results are similar. Acquirers
in the U.K. lose significantly more in diversifying acquisi-
tions, whereas in focusing acquisitions they gain more.20
These results suggest that stronger creditor rights induce
17
As a robustness check, we re-estimate the model using the
creditor rights variable from Djankov, McLeish, and Shleifer (2007),
which varies over time for five countries. The results are qualitatively
similar.
18
We use this time window because in some countries, low stock
liquidity causes a slower adjustment of stock prices to information, and
lax enforcement of insider trading rules may cause trading on the
information some days prior to the formal announcement. Our results
are qualitatively similar for smaller event windows, such as  1 to þ 1.
19
The results are qualitatively similar when we use Djankov,
McLeish, and Shleifer (2007) on CRIGHTS. The coefficient of CRIGHTS in
the diversifying acquisitions model is  0.007 (t ¼3.36) and it is 0.004
(t¼ 0.53) for same-industry acquisitions.
20
We estimate a regression of CAR as a dependent variable in a model
similar to that in Table 5, with data from these two countries alone. We set
a dummy variable UK¼ 1 for an acquirer in the U.K., where creditor rights
are stronger, and UK¼0 for a U.S. acquirer, and we exclude country-related
variables. We obtain that for diversifying acquisitions, the coefficient of UK
162 V.V. Acharya et al. / Journal of Financial Economics 102 (2011) 150–166

Table 5
Effects of creditor rights on performance for focusing and diversifying mergers.
The table includes two performance variables: (1) DROA(k) ¼ ROA(tþ k)  ROA(t  1), where ROA is return on assets ¼ EBITDA/Assets, and t is the year of
the merger; (2) Cumulative abnormal returns on the acquirer’s stock, CAR, from three days before the merger announcement to three days after the
merger announcement. Abnormal returns are calculated from a market model whose parameters are estimated from weekly returns and market model
for each country using 105 weeks but not less than 52 weeks, up to nine weeks before the week of the merger announcement. We include year and
industry fixed effects (not reported). To account for the choice of type of acquisition, we add the inverse Mills ratio, computed using probit model (1) in
Table 3 for regressions that include all countries. The t-statistics (in parentheses) are based on robust standard errors cluster-adjusted at the country
level. The nnn, nn, and n indicate significance at the 1%, 5%, and 10% levels, respectively. Refer to Table 1 for variable definitions.

ROA(t þ1) ROA(t  1) ROA(t þ2)  ROA(t  1) ROA(tþ 3)  ROA(t 1) CAR (t  3 to t þ3)

Focusing Diversify Focusing Diversify Focusing Diversify Focusing Div.

(1) (2) (3) (4) (5) (6) (7) (8)

CRIGHTS 0.0003  0.028nnn 0.026  0.031nnn  0.017  0.030nnn 0.011  0.013nnn

(0.02) (8.20) (1.30) (10.28) (0.98) (8.84) (1.65) (5.14)
SHRIGHTS 0.004 0.009  0.008  0.007 0.012  0.003  0.008n 0.002
(0.37) (1.44) (0.61) (1.27) (1.03) (0.45) (1.74) (0.56)
Log(Market cap) 0.001  0.007  0.019  0.003 0.030  0.002  0.017nn 0.004
(0.11) (0.93) (0.90) (0.49) (1.49) (0.26) (2.31) (1.49)
Flexibility to fire  0.124 0.041  0.258 0.042 0.237 0.041  0.173nn  0.033
(0.77) (0.76) (1.03) (0.79) (1.11) (0.79) (2.41) (1.13)
Accounting disclosure 0.001 0.0001 0.002 0.0003  0.004n  0.001 0.002  0.0001
(0.44) (0.06) (0.76) (0.26) (1.79) (0.79) (1.62) (0.18)
Emerging markets 0.025 0.006  0.078 0.008 0.103 0.002  0.051nn 0.022nnn
(0.47) (0.38) (0.94) (0.48) (1.43) (0.13) (2.2) (2.89)
Rule of law  0.021  0.019  0.079 0.019 0.062 0.016  0.058 0.02n
(0.44) (0.69) (1.06) (0.92) (1.02) (0.82) (1.53) (1.87)
French legal origin 0.037  0.026  0.034  0.021 0.065  0.025  0.032 0.019
(0.86) (1.34) (0.50) (1.34) (1.35) (1.39) (1.28) (1.66)
German legal origin 0.055 0.003 0.079 0.005  0.062  0.006 0.049n 0.013nn
(1.35) (0.27) (1.13) (0.44) (1.16) (0.53) (1.96) (2.34)
Nordic legal origin 0.020  0.024  0.145n  0.026 0.093  0.021  0.047nn 0.047nnn
(0.42) (0.85) (1.84) (1.02) (1.23) (0.72) (2.32) (2.98)
Macro risk 0.105  0.169 0.780  0.619nnn  0.427  0.161 0.279  0.058
(0.27) (1.0) (1.09) (4.08) (1.04) (0.92) (1.45) (0.63)
Log(GDP per capita) 0.001 0.013 0.022  0.012  0.006  0.009 0.015  0.02nn
(0.13) (0.85) (1.06) (0.95) (0.4) (0.73) (1.10) (2.22)
Log(Transaction value)  0.002  0.003  0.009  0.003 0.009  0.001  0.005 0.002nnn
(0.37) (0.95) (0.93) (0.72) (1.21) (0.3) (1.32) (2.66)
Dummy: Target is public 0.004  0.005 0.011  0.001 0.006  0.003  0.016nnn  0.017nn
(0.63) (1.57) (1.29) (0.39) (1.25) (0.78) (5.01) (2.48)
Inverse Mills Ratio  0.082 0.040  0.573 0.053 0.618 0.092n  0.434nn 0.014
(0.21) (1.08) (0.9) (1.08) (1.25) (1.96) (2.01) (0.65)

Observations 8,788 5,752 8,198 5,491 7,742 4,770 7,500 5,725

R-squared (%) 8.8 11.7 8.5 15.0 4.2 13.6 1.6 2.0

firms to undertake diversifying acquisitions, even in the
presence of a post-acquisition decline in profitability and
value, perhaps as a response to the threat of default.
2.4. The effects of changes in creditor rights
Thus far, we have analyzed the cross-country relation
between creditor rights and two sets of variables: type of
acquisition (focusing or diversifying) and post-acquisition
performance. We now test the effect of creditor rights on
these variables over time, exploiting seven changes in
creditor rights that occurred in six countries: Indonesia,
Israel, Japan (two changes), Sweden, Thailand, and Russia,21
(footnote continued)
is  0.017 (t¼34.52) while for focusing acquisitions, the coefficient of UK is
0.071 (t¼ 5.89).
21
Russia is included only in this table’s regressions, not in any other
estimation, because it has a unique legal origin. Its inclusion with a
documented in Djankov, McLeish, and Shleifer (2007). These
changes decreased CRIGHTS by one unit—that is, they
weakened creditor rights—except for the 2002 change in
Japan that raised CRIGHTS by one unit.
We estimate the following regression, a variant of
model (1):
DepVarj,c ¼ y0 þ y1  DCRIGHTSc þcontrol variables: ð5Þ
DepVar is the dependent variable in the regression: (i) the
probability of same-industry acquisition, Pr(same indus-
try); (ii) the probability of same-country acquisition,
Pr(same country); (iii) the probability of a low-recovery
acquirer buying a high-recovery target, Pr(AL\TH9TH)j,c;
(iv) the change in acquirer’s ROA k years after the
(footnote continued)
unique dummy variable for its legal origin does not change any of the
reported results.
 V.V. Acharya et al. / Journal of Financial Economics 102 (2011) 150–166 163

acquisition, DROA(k)¼ ROA(tþk)  ROA(t  1); and (v) the
cumulative abnormal return on the acquirer’s stock from
three days before the acquisition announcement to three
days thereafter, CAR (t  3 to tþ 3). In cases where creditor
rights became weaker, we set DCRIGHTSc ¼0 for the years
that follow the year of change and set DCRIGHTSc ¼1 for
the period that precedes it (including the change year).
Analogously, if creditor rights were strengthened,
DCRIGHTSc ¼ 1 during the period following the year of the
change and DCRIGHTSc ¼0 beforehand. As discussed, all
changes in CRIGHTSc during the sample period, but for one
case, made creditor rights weaker. For the 32 countries
in our sample with no changes in creditor rights,
DCRIGHTS¼0 for the entire sample period. The control
variables are Transaction value (in logarithm) and, impor-
tantly, country fixed effects, which account for all coun-
try-based variables, in line with the difference-in-
differences methodology. We also include year fixed
effects and industry fixed effects (except for dependent
variable (iii), where the classification is done by industry).
We estimate the regressions for the dependent variables
(i)–(iii) above by the probit method and for the dependent
variables (iv)–(v) by OLS, with standard errors clustered at
the country level to account for potential within-country
correlation in the residuals.
In accordance with our hypotheses, we expect that
in model (5), y1 o0 for all dependent variables except
for Pr(AL\TH9TH)j,c, where we expect y1 40. The regres-
sion results in Table 6 support our hypothesis, as
follows:
(i) DepVar¼Pr(same industry): The coefficient of
DCRIGHTSc is negative and statistically significant:
y1 ¼ 0.145 (t ¼6.19). This means that weakening of
creditor rights significantly increases the likelihood
of same-industry mergers and reduces the extent of
diversifying, risk-reducing mergers. The marginal
effect from this regression is  0.057, i.e., an increase
in the strength of creditor rights by one (on a scale of
zero to four) is associated with a statistically sig-
nificant drop in the probability of a same-industry
merger by 5.7%.
(ii) DepVar¼Pr(same country): y1 ¼  0.213 (t ¼2.83).
Weakening of creditor rights reduces the likelihood
of cross-country mergers.
(iii) DepVar¼Pr(AL \ TH9TH) (the likelihood of a high-
recovery target being acquired by a low-recovery
firm): y1 ¼0.147 (t ¼1.08). The strengthening of cred-
itor rights increases the propensity of firms with low-
recovery assets to seek and buy target firms with

Table 6
The effect of changes in creditor rights.
The creditor rights change dummy, DCRIGHTS, equals one after the year of change from a period of stronger creditor rights, and zero otherwise, and it
equals zero after the year of change from a period of weaker creditor rights, and one otherwise. It equals zero for the control sample of no change in
creditor rights. Included are all mergers and acquisitions where the acquired percentage shares is at least 20%, the transaction has a disclosed value, and
the time changes in creditor rights are available in Djankov, McLeish, and Shleifer (2007). We exclude transactions where the acquirer is in the financial
industry (SIC header 6) or regulated industry (SIC headers 48 and 49). The sample period is 1994–2004. Models (1), (2) and (3) are estimated by the
probit method and the rest are estimated by OLS. The dummy variable to measure the probability of same-industry acquisition, Pr(Same industry), equals
one when bidder and target are in the same industry. The dummy variable to measure the probability of an acquisition of high-recovery target by a low-
recovery acquirer, Pr(AL\TH9TH), equals one when, among all acquisitions of target firms whose assets have high recovery value, the bidder firm’s assets
have low recovery value. For regression (2) only, we include all cross-country and within-country mergers that meet the requirements above. The post-
acquisition change in return on assets is DROA(k) ¼ROA(t þk)  ROA(t  1), where k ¼1, 2, or 3, calculated for each merger with available data, where t is
the effective year of the merger. CAR is the cumulative abnormal returns on the acquirer stock from three days before the acquisition announcement to
three days after it. The standard errors are cluster-adjusted at the country level. Included (but not reported for brevity) are fixed effects for country, year,
and industry (two-digit SIC code—acquirer’s industry), following the difference-in-differences methodology of Bertrand, Duflo, and Mullainathan (2004).
Model (3) does not include industry fixed effects, in line with Table 4. nnn, nn, and n indicate significance at the 1%, 5%, and 10% levels, respectively. Refer to
Table 1 for variable definitions.

Part I. Mutivariate analysis

Variable Pr(Same industry) Pr(Same country) Pr(TH \ AL9TH) DROA(1) DROA(2) DROA(3) CAR (t  3 to t þ 3)
(1) (2) (3) (4) (5) (6) (7)

DCRIGHTS c,t  0.145nnn  0.213nnn 0.147  0.042nnn  0.051nnn  0.023nnn  0.005nn

(6.19) (2.83) (1.08) (5.78) (9.22) (4.93) (3.31)
Log (Transaction 0.028n  0.07nn  0.057  0.002n  0.002  0.001n  0.001
value) (2.52) (2.22) (7.73) (1.75) (1.03) (1.73) (1.07)
Fixed effects Country, year, Country, year, Country, year Country, year, Country, year, Country, year, Country, year,
industry industry industry industry industry industry
Observations 33,221 52,756 6,495 14,540 13,689 12,512 13,225

Part II. Details of changes

Country Year of law change Detail of change

Indonesia 1998 Change to SECURED ¼ 0

Israel 1995 Change to AUTOSTAY ¼0
Japan 2000 and 2002 2000: Change to SECURED ¼ 0
2002: Change to AUTOSTAY ¼ 1
Russia 1998 1998: Change to MANAGES ¼ 0
Sweden 1995 Change to REORG ¼0
Thailand 1999 Change to REORG ¼0
164 V.V. Acharya et al. / Journal of Financial Economics 102 (2011) 150–166
high-recovery assets. In this case, however, the coef-
ficient is not statistically significant.
(iv) DepVar¼ DROA(1), DROA(2), and DROA(3): The coeffi-
cients of DCRIGHTSc are, respectively, y1 ¼  0.042
(t¼ 5.78), y1 ¼ 0.051 (t ¼9.22), and y1 ¼  0.023
(t¼ 4.93). All coefficients are negative, as expected,
and significant. These results mean that in countries
where creditor rights were weakened, acquirers
made more profitable acquisitions than they did
beforehand. Notably, in this regression we include
all acquisitions, not only the diversifying ones,
because of the self-selection in diversification. When
including only diversifying acquisitions, the three
coefficients of DCRIGHTSc are slightly more negative:
0.062 (t ¼6.87), 0.069 (t¼ 9.24), and y1 ¼  0.027
(t¼ 4.30), all statistically significant.
(v) DepVar¼CAR(t  3 to t þ3): The negative and signifi-
cant coefficient of DCRIGHTSc, y1 ¼  0.005 (t ¼3.31),
means that after creditor rights were weakened,
acquisitions became more value-enhancing, the
improvement being 0.5% of the acquirer’s value.
When estimating the model with diversifying acqui-
sitions only (which are subject to selection), the
coefficient of DCRIGHTSc is again slightly more nega-
tive,  0.008 (t ¼3.08).
In summary, the results on the effects of changes in
creditor rights are consistent with those obtained in the
cross-section analysis and thus support Hypothesis II on
the effects of creditor rights on types of firm investments
and on value.
2.5. Creditor rights and firms’ cash-flow risk
We now test Hypothesis III, whether the level of
corporate cash-flow risk is lower in countries with stron-
ger creditor rights. We test this hypothesis because,
in addition to undertaking diversifying acquisitions
(Hypotheses Ia–Ic), firms can reduce their risk by other
means that are not directly observed. We define RISKj,c as
the standard deviation of the yearly ROAj,c,t of firm j in
country c during the period 1992–2005. (Details are in
Table 1.) The sample includes only firms in the manufac-
turing industries with data for at least eight years and
excludes utilities and financial firms, which are regulated
in many countries. We require at least six firms in a
country, thus having a sample of 5,394 firms in 35
countries. The estimation model is:
RISKj,c ¼ k0 þ k1  CRIGHTSc þcontrol variables:
Our hypothesis implies k1 o0. The control variables are
those used in Table 3, adding firm size (logarithm of
firm’s assets at the beginning of its data in our sample),
which negatively correlates with risk, and industry (one-
digit SIC) dummy variables. In the country-level regres-
sions, the dependent variable is the median for each
country of the variable RISKj,c, and the expalanatory
variables are the country’s creditor rights and all country
variables. The results are as follows: (complete tables are
available in an appendix at the JFE Web site).
Firm-level model Country-level model
CRIGHTS  0.005 (t ¼2.60)  0.010 (t ¼2.95)
AUTOSTAY  0.010 (t ¼1.44)  0.024 (t ¼2.71)
REORG  0.003 (t ¼0.59)  0.003 (t ¼0.36)
SECURED  0.006 (t ¼1.00)  0.005 (t ¼0.56)
MANAGES  0.023 (t ¼ 3.91)  0.036 (t ¼4.10)
These results support our hypothesis: the coefficient
k1 on CRIGHTSc is negative and statistically significant.
As in Table 3, the component of CRIGHTS with the
most negative effect on RISK is managerial dismissal in
reorganization (MANAGES). When we exclude the U.S. and
the U.K., the effect of CRIGHTS is similar,  0.005 with
t ¼2.50.22
We estimate again the model at the industry level.
Following the methodology of Rajan and Zingales (1998)
discussed above, the inherent risk for an industry k is the
median risk level of that industry in the U.S., RISKk,US,
calculated for the 1992–1998 period. RISKk,c is the
median risk in the same industry in country c, calcula-
ted over the subsequent period, 1999–2005. The model is:
RISKk,c ¼ l0 þ l1  RISKk,US þ l2  CRIGHTSdm
c  RISKk,US
þ Country fixed effects: ð7Þ
Hypothesis III implies that l2 o0. We indeed obtain that
l2 ¼  0.158 with t¼ 2.16, supporting our hypothesis.
Creditor rights have a negative effect on corporate cash-
flow risk, relative to the industry’s inherent risk level,
RISKk,US. Also, the result l1 ¼0.862 (t ¼4.49) (R2 ¼30.2%)
supports the use of risk in a U.S. industry to represent the
inherent level of that industry’s risk in other countries.
2.6. Creditor rights and financial leverage
For any given level of corporate activity, corporate
managers can reduce the likelihood of bankruptcy and
its associated costs by lowering financial leverage. Thus,
strong creditor rights should reduce corporate demand
for leverage. On the other hand, La Porta, Lopez-de-
Silanes, Shleifer, and Vishny (1998) suggest that strong
creditor rights increase the supply of credit, which
would raise overall leverage if demand for borrowing is
unaffected.23 By our Hypothesis IV, strong creditor rights
may reduce leverage if the cost imposed by such rights
22
Among the control variables, the instrumented leverage has a
ð6Þ negative and significant coefficient. Large firms have lower risk. Macro
risk has a positive effect on RISKj,c, but is insignificant. Including
Hofstede’s (2001) Uncertainty avoidance index as a control variable, we
obtain that its coefficient is negative and insignificant (  0.0001 with
t ¼  0.032) in the single-firm regression, and  0.0004 (t¼  2.39) in the
country-level regression. In the latter regression, the coefficient of
CRIGHTS is  0.007 (t ¼  4.18), highly significant.
23
Haselmann, Pistor, and Vig (2010) find that the improvement in
enforcement of creditor rights in Central and East European countries,
through creation of a collateral registry, boosted lending. Acharya,
Sundaram, and John (2011) show theoretically and empirically that
the effect of creditor rights on corporate leverage depends on the asset
specificity of the firm. Vig (2007) finds that in India, strengthening
creditor rights reduced leverage.
 V.V. Acharya et al. / Journal of Financial Economics 102 (2011) 150–166
outweighs the beneficial effect on the supply of credit. We
test the effect of creditor rights on leverage via the model:
Leveragej,t,c ¼ m0 þ m1  CRIGHTSc þcontrol variables: ð8Þ
Here, Leveragej,c,t is the industry-adjusted leverage (in
book value) of firm j in country c in year t (details are in
Table 1). The sample of firms is the same as in the tests of
cash-flow risk; it includes 36,237 firm-years with lever-
age data, of which 17,304 are in countries excluding the
U.S. and the U.K. The regression includes the country
variables and firm-specific variables that appear in
Table 3, adding variables that are used in studies of
leverage: country corporate tax rate in each year, firm
asset tangibility, firm size, and firm ROA. The firm control
variables and the country tax rate are lagged one year. The
regression includes year fixed effects, and standard errors
are clustered by country.
Our hypothesis is that in model (8), m1 o0, i.e., stronger
creditor rights induce corporate managers to reduce finan-
cial risk by reducing corporate leverage. We estimate the
model at both the single firm-year level and the country
level. For the country-level regression, we calculate the
yearly average industry-adjusted leverage ratio for the
country and then regress that country’s average annual
leverage on the creditor rights in the country, as well as on
all country variables and on year dummy variables.
The results are as follows (complete tables are available
at the JFE Web site):
Firm-level model Country-level model
CRIGHTS  0.022 (t ¼5.20)  0.019 (t ¼9.32)
AUTOSTAY  0.037 (t ¼2.45)  0.015 (t ¼3.01)
REORG  0.049 (t ¼4.66)  0.027 (t ¼6.92)
SECURED  0.042 (t ¼2.61)  0.032 (t ¼6.06)
MANAGES  0.041 (t ¼3.10)  0.018 (t ¼3.05)
Supporting our hypothesis that stronger creditor rights
reduce leverage, the coefficient of CRIGHTS is negative and
significant at both the single-firm level and the country
level. When we exclude the U.S. and the U.K. in the single-
firm regressions, the coefficient of CRIGHTS is  0.018
with t ¼3.42. All four categories of creditor rights have
consistently negative coefficients that are statistically
significant.24
Finally, we estimate the effect of changes in creditor
rights on the leverage of the firms in our sample. We
estimate model (5) above with DepVar¼Leverage. The
specification of the model is as described in Section 2.4.
We also add leverage-related, firm-specific variables:
Tangibility, Log(assets), and Earnings Before Interest, Tax,
Depreciation and Amortization (EBITDA)/Assets. We expect
24
We obtain similar results in a robustness test that employs
Djankov, McLeish, and Shleifer’s (2007) data on creditor rights. In the
sample of all countries, the coefficient of CRIGHTS is  0.018 with
t ¼3.52, and after excluding the U.S. and the U.K., it is  0.013 with
t ¼4.59. When we add to the single-firm regression the culture-based
Uncertainty avoidance index of Hofstede (2001), its coefficient is negative
but insignificant,  0.0001 (t¼ 0.56), while the negative and significant
effect of CRIGHTS is unaltered: its coefficient is  0.0221 (t ¼ 5.54).
165
the coefficient of DCRIGHTSc to be negative; that is,
weakening of creditor rights increases corporate leverage
because the costs of financial distress to debtors are
lower. Indeed, we obtain that the coefficient of DCRIGHTSc
is 0.035 with t ¼2.53. That is, weakening of creditor
rights causes firms to increase their leverage.
3. Conclusion and discussion
We find that having strong creditor rights in a country
leads firms to reduce risk. They do so by undertaking
diversifying acquisitions, both across industries in the
domestic market and across national borders, and they
increase the number of lines of business in which they
participate. In response to strong creditor rights, firms
also implement policies that reduce cash-flow risk, and
they become reluctant to borrow, which results in lower
leverage in countries with strong creditor rights. Thus,
creditor rights have effects on real as well as financial
corporate policies.
Further, we find that to avoid the higher costs asso-
ciated with stronger creditor rights, firms make diversify-
ing acquisitions even if they hurt corporate performance,
measured by return on assets post-acquisition and by
stock price reaction to the acquisition announcement.
Both measures of performance show declines resulting
from diversifying acquisitions, with the decline being
greater in countries with stronger creditor rights. No such
relation between corporate performance in acquisitions
and creditor rights is observed in focusing (same industry)
acquisitions.
Stronger creditor rights are one way to mitigate stock-
holder expropriation or risk-shifting tendencies that benefit
them at the expense of bondholders; they thereby facilitate
raising external capital. Our findings thus confirm that
creditor rights do what they are expected to do: inhibit
risk-taking by companies. However, it may well be that
stronger creditor rights may induce managers to reduce risk
and to stifle even non-opportunistic risk-taking that would
be beneficial to all claimholders. In this paper, we provide
evidence of this downside from strong creditor rights.
This tradeoff suggests that the existence of stronger
creditor rights is not always desirable. The optimal level
of creditor rights should balance their positive effect on
the supply of credit against their negative effect on
corporate risk-taking and on operating performance, as
well as on the demand for debt. In future work, it would
be interesting to assess directly this important tradeoff.
Appendix A. Supporting information
Supplementary data associated with this article can be
found in the online version at doi:10.1016/j.jfineco.2011.
04.001.
References
Acharya, V., Bharath, S., Srinivasan, A., 2007. Does industry-wide distress
affect defaulted firms? Evidence from creditor recoveries. Journal of
Financial Economics 85, 787–821.
166 V.V. Acharya et al. / Journal of Financial Economics 102 (2011) 150–166
Acharya, V., Sundaram, R., John, K., 2011. Cross-country variations in
capital structures: the role of bankruptcy codes. Journal of Financial
Intermediation 20, 25–54.
Acharya, V., Subramanian, K., 2009a. Bankruptcy codes and innovation.
Review of Financial Studies 22, 4949–4988.
Acharya, V., Subramanian, K., 2009b. Bankruptcy codes and innovation: a
model. Review of Financial Studies online.
Adler, B.E., 1992. Bankruptcy and risk allocation. Cornell Law Review 77,
439–489.
Adler, B., Capkun, V., Weiss, L.A., 2007. Value Destruction in the New Era
of Chapter 11. Unpublished Working Paper, NYU Law School.
Amihud, Y., Lev, B., 1981. Risk reduction as a managerial motive for
conglomerate mergers. Bell Journal of Economics 12, 605–618.
Ammann, M., Hoechle, D., Schmid, M., 2008. Is there Really No Con-
glomerate Discount? Unpublished Working Paper, University of
St. Gallen.
Armour, J., Cumming, D.J., 2005. Bankruptcy Law and Entrepreneurship.
Unpublished Working Paper, University of Cambridge Faculty of
Law.
Berger, P.G., Ofek, E., 1995. Diversification’s effect on firm value. Journal
of Financial Economics 37, 39–65.
Berger, P.G., Ofek, E., Swary, I., 1996. Investor valuation of the abandon-
ment option. Journal of Financial Economics 42, 257–287.
Bertrand, M., Duflo, E., Mullainathan, S., 2004. How much should we
trust differences-in-differences estimates? Quarterly Journal of Eco-
nomics 119, 249–275
Campa, J.M., Kedia, S., 2002. Explaining the diversification discount.
Journal of Finance 57, 1731–1762.
Chava, S., Roberts, M., 2008. How does financing impact investment? The
role of debt covenants. Journal of Finance 63, 2085–2121.
Claessens, S., Klapper, L.F., 2005. Bankruptcy around the world: explana-
tions of its relative use. American Law and Economic Review 7,
253–283.
Comment, R., Jarrell, G.A., 1995. Corporate focus and stock returns.
Journal of Financial Economics 37, 67–87.
Djankov, S., McLeish, C., Shleifer, A., 2007. Private credit in 129 countries.
Journal of Financial Economics 84, 299–329.
Dyck, A., Zingales, L., 2004. Private benefits of control: an international
comparison. Journal of Finance 59, 537–600.
Eckbo, B.E., Thorburn, K.S., 2003. Control benefits and CEO discipline in
automatic bankruptcy auctions. Journal of Financial Economics 69,
227–258.
Fan, W., White, M.J., 2003. Personal bankruptcy and the level of
entrepreneurial activity. Journal of Law and Economics 46,
543–567.
Haselmann, R., Pistor, K., Vig, V., 2010. How law affects lending. Review
of Financial Studies 23, 549–580.
Healy, P.M., Palepu, K.G., Ruback, R.S., 1992. Does corporate performance
improve after mergers? Journal of Financial Economics 31, 135–175.
Hofstede, G.H., 2001. Culture’s Consequences: Comparing Values, Beha-
viors, Institutions, and Organizations Across Nations, second ed. Sage
Publications, California.
Hylton, K., Deng, F., 2007. Antitrust around the world: an empirical
analysis of the scope of competition laws and their effects. Antitrust
Law Journal 74, 271–341.
Jensen, M., Meckling, W., 1976. Theory of the firm: managerial behavior,
agency costs and ownership structure. Journal of Financial Econom-
ics 3, 305–360.
John, K., Litov, L., Yeung, B., 2008. Corporate governance and managerial
risk-taking. Journal of Finance 63, 1679–1728.
Khanna, T., Palepu, K., 2000. Is group affiliation profitable in emerging
markets? An analysis of diversified Indian business groups. Journal
of Finance 55, 867–891.
Laeven, L., Levine, R., 2007. Is there a diversification discount in financial
conglomerates? Journal of Financial Economics 85, 331–367.
Lamont, O., 1997. Cash flow and investment: evidence from internal
capital markets. Journal of Finance 52, 83–109.
Lamont, O., Polk, C., 2002. Does diversification destroy value? Evidence
from the industry shocks. Journal of Financial Economics 63, 51–77.
La Porta, R., Lopez-de-Silanes, F., Shleifer, A., Vishny, R., 1997. Legal
determinants of external finance. Journal of Finance 52, 1131–1150.
La Porta, R., Lopez-de-Silanes, F., Shleifer, A., Vishny, R., 1998. Law and
finance. Journal of Political Economy 106, 1113–1155.
Lee, K., Peng, M.W., Lee, K., 2008. From diversification premium to
diversification discount during institutional transitions. Journal of
World Business 43, 47–65.
Levine, R., Demirguc-Kunt, A., 2001. Financial Structure and Economic
Growth: A Cross-Country Comparison of Banks, Markets and Devel-
opment. MIT Press, Cambridge, MA.
Lins, K., Servaes, H., 2002. Is corporate diversification beneficial in
emerging markets? Financial Management 31, 5–31
Manso, G. Motivating Innovation. Journal of Finance, forthcoming.
Moeller, S.B., Schlingemann, F.B., 2005. Global diversification and bidder
gains: a comparison between cross-border and domestic acquisi-
tions. Journal of Banking and Finance 29, 533–564.
Morck, R., Shleifer, A., Vishny, R., 1990. Do managerial objectives drive
bad acquisitions? Journal of Finance 45, 31–48.
Nini, G., Smith, D., Sufi, A., 2009. Creditor control rights and firm
investment policy. Journal of Financial Economics 92, 400–420.
Rajan, R., Zingales, L., 1998. Financial dependence and growth. American
Economic Review 88, 559–586.
Rajan, R., Servaes, H., Zingales, L., 2000. The cost of diversity: diversifica-
tion discount and inefficient investment. Journal of Finance 55,
35–80.
Scharfstein, D.S., 1998. The Dark Side of Internal Capital Markets II. NBER
Working Paper No. 6352.
Scharfstein, D.S., Stein, J.C., 2002. The dark side of internal capital
markets: divisional rent-seeking and inefficient investment. Journal
of Finance 55, 1537–1564.
Schmid, M.M., Walter, I., 2009. Do financial conglomerates create or
destroy economic value? Journal of Financial Intermediation 18,
193–216.
Schwartz, A., 1997. Contracting about bankruptcy. Journal of Law,
Economics, and Organization 13, 127–146.
Shin, H., Stulz, R., 1998. Are internal capital markets efficient? Quarterly
Journal of Economics 113, 531–552.
Shleifer, A., Vishny, R., 1992. Liquidation values and debt capacity: a
market equilibrium approach. Journal of Finance 47, 1343–1366.
Stulz, R., Williamson, R., 2003. Culture, openness, and finance. Journal of
Financial Economics 70, 313–349.
Tufano, P., 1996. Who manages risk? An empirical examination of risk
management practices in the gold mining industry. Journal of
Finance 51, 1097–1137.
Tufano, P., 1998. Agency costs of corporate risk management. Financial
Management 27, 67–77.
Vig, V., 2007. Access to Collateral and Corporate Debt Structure:
Evidence from a Natural Experiment. Working Paper, London Busi-
ness School.
Villalonga, B., 2004. Does diversification cause the diversification dis-
count? Financial Management 33, 5–27.