-
THE JOURNAL OF FINANCE • VOL. LXIII, NO. 4 • AUGUST 2008
International Cross-Listing, Firm Performance,and Top Management
Turnover: A Test of the
Bonding Hypothesis
UGUR LEL and DARIUS P. MILLER∗
ABSTRACT
We examine a primary outcome of corporate governance, namely,
the ability to iden-tify and terminate poorly performing CEOs, to
test the effectiveness of U.S. investorprotections in improving the
corporate governance of cross-listed firms. We find thatfirms from
weak investor protection regimes that are cross-listed on a major
U.S. Ex-change are more likely to terminate poorly performing CEOs
than non-cross-listedfirms. Cross-listings on exchanges that do not
require the adoption of stringent in-vestor protections (OTC,
private placements, and London listings) are not associatedwith a
higher propensity to remove poorly performing CEOs.
DOES CROSS-LISTING IN THE UNITED STATES IMPROVE the corporate
governance offoreign firms? The “bonding hypothesis” proposed by
Coffee (1999, 2002) andStulz (1999) predicts that after listing on
a major U.S. stock exchange, foreignfirms become subject to
stringent U.S. investor protections that constrain in-siders from
expropriating minority shareholders. Because this hypothesis
hasimportant implications for the effectiveness of U.S. laws and
enforcement aswell as the efficacy of market-based approaches in
improving global corporategovernance, it has attracted the recent
attention of academics and practitionersalike.
To date, empirical support for the bonding hypothesis is
principally drawnfrom the large literature that examines the
economic consequences of cross-listing in the United States.1
However, as Leuz (2006) notes, the evidence inmany of these studies
is fairly indirect, as it is difficult to attribute the
economicconsequences of cross-listing directly to the bonding
hypothesis because many
∗Lel is from the Federal Reserve Board and Miller is from Edwin
L. Cox School of Business atSouthern Methodist University. We thank
an anonymous referee and associate editor; CampbellHarvey (the
editor); Mark Carey, Craig Doidge, Art Durnev, Nandini Gupta, David
Mauer, ChipRyan, Chester Spatt, and Wendy Wilson; and seminar
participants at the 2006 University of NorthCarolina GIA
Conference, the 2006 University of Oregon Corporate Finance
Conference, the 2006Financial Research Association Conference, the
2006 Utah Winter Finance Conference, LouisianaState University, and
the University of Texas at Dallas. We thank Bill Megginson and
MeghannaAyyagari for data access and Charles Murry and Laurel
Nguyen for excellent research assistance.This paper represents the
authors’ opinions and not necessarily those of the Federal Reserve
Board.All errors are the sole responsibility of the authors.
1 Karolyi (1998, 2006) and Benos and Weisbach (2004) provide
comprehensive surveys. We alsodiscuss the literature in Section II
of this paper.
1897
-
1898 The Journal of Finance
theories of cross-listing have similar economic predictions.2
Moreover, the va-lidity of the bonding hypothesis has been called
into question by a number ofrecent studies that document
cross-listed firms’ lack of compliance with certainU.S. laws and
the low number of enforcement actions by U.S. legal
institutions(see, for example, Siegel (2005) and Lang, Raedy, and
Wilson (2006)). Therefore,whether U.S. securities laws and
regulations improve the corporate governanceof cross-listed firms
is under debate as the nascent empirical evidence is pre-dominantly
indirect and yields mixed results.
In this paper we pursue a different approach in testing the
bonding hypothe-sis and examine a direct outcome of corporate
governance: the propensity toreplace poorly performing CEOs. We
argue that if cross-listing actually re-sults in increased
shareholder protections, we should be able to observe spe-cific
outcomes that are consistent with improved corporate governance.
Wefocus on the sensitivity of top executive turnover to performance
since an ex-tensive body of international research shows that a
necessary component ofeffective corporate governance is the ability
to identify and replace poorly per-forming CEOs (see, for example,
Kaplan (1994), Coffee (1999), Murphy (1999),Volpin (2002), Dahya,
McConnell, and Travlos (2002), Gibson (2003), DeFondand Hung
(2004)). We compile a database of 70,976 firm-year observations
from47 countries from 1992 to 2003 to test the hypothesis that CEOs
of cross-listedfirms are more likely to face termination when firm
performance is poor. We findthat the relation between CEO turnover
and poor performance is stronger forcross-listed firms than
non-cross-listed firms, and that the stronger turnover topoor
performance relation for cross-listed firms is concentrated in
firms listedon major U.S. exchanges (for example, Level 2 and 3
American depositary re-ceipts (ADRs)). Firms that list in the
over-the-counter (OTC) market (Level 1),conduct private placements
(Rule 144a), or even list in London do not have asignificantly
different relation between CEO turnover and performance
fromnon-cross-listed firms. Further, we find that the increased
relation betweenCEO turnover and poor performance for cross-listed
firms is strongest in coun-tries with weak investor protections.
Overall, our results are consistent withthe hypothesis that U.S.
securities laws and regulations improve the corporategovernance of
cross-listed firms.
We also investigate several alternative explanations for our
results, includ-ing the potential endogeneities that arise in a
study of cross-listing and gover-nance due to the nonrandom nature
of the decision to list in the United States.For example, we
investigate if our results are due to the notion that
better-governed firms are the ones that self-select to cross-list.
To do so, we examineseveral specifications that measure the
sensitivity of CEO turnover to perfor-mance for cross-listed firms
prior to cross-listing. These tests show that therelation of
turnover to performance is insignificant (significant) in the
pre-cross-listed (post-cross-listed) period, which suggests that
our results are notan artifact of the pre-cross-listed governance
status of our sample firms. We
2 For further details on other theories that have been argued to
generate similar predictions(e.g., market segmentation, investor
recognition, increased liquidity, and better information) seethe
discussion in Doidge, Karolyi, and Stulz (2004a) and Hail and Leuz
(2004).
-
International Cross-Listing 1899
also examine if other potential control changes around
cross-listing such asprivatizations, changes in ownership, M&A
events, and changes in the com-position of board of directions can
explain our results. We find that our re-sults remain robust to
these events. Further, we examine if cross-listing in-duces top
management to leave their jobs to pursue employment in the
UnitedStates where they are likely to be more highly compensated or
if cross-listedfirms terminate poorly performing management because
they are able to ac-cess a more international pool of
top-management candidates. We find thatfirms that change CEOs tend
to replace them with managers from their do-mestic labor pool and
that departing CEOs most often get jobs in the localmarket.
Therefore, shifts in the labor market also do not appear to explain
ourresults.
We subject our tests to a battery of firm- and county-level
robustness testsas well. We find our results are robust to country,
industry, and year fixedeffects in addition to the possible
entrenchment effects of concentrated owner-ship structures. Our
findings are also robust when we exclude countries thatcontain the
largest portion of our sample, remove observations surroundingthe
Asian financial crises, and omit financial and regulated
industries. An im-portant methodological note is that all of our
analyses control for the recentlyrecognized difficulty in
implementing and interpreting interaction effects innonlinear
models (see, for example, Ai and Norton, (2003)).
Our results advance the literature in several ways. First, our
findings addto the debate on whether U.S. securities laws and
enforcement are effective inreaching non-U.S. firms. Second, by
showing that CEOs of cross-listed firmsare more likely to face
termination when firm performance is poor, our resultsalso
contribute to the literature by documenting a specific channel
throughwhich cross-listing improves corporate behavior, something
that is not well-documented in the literature.3 Finally, our
findings also have implications forthe growing literature that
examines how global corporate governance can beimproved (see, for
example, LaPorta et al. (2000) and Coffee (2002)). This re-search
stems from a large number of studies that show that the economic
conse-quences for firms located in countries with poor investor
protections are severe.4
Given the economic impact of poor investor protections and the
correspondingdifficulty in changing a country’s legal structure
(i.e., legal convergence), an im-portant question is whether
market-based approaches (i.e., functional conver-gence), such as
opting-in to a better legal system via cross-listing, can
improvecorporate governance. Our finding that cross-listing in the
United States is as-sociated with improved corporate governance is
consistent with the hypothesisthat the functional convergence of
legal systems to a higher global standard ispossible.
The remainder of the paper proceeds as follows. Section I
discusses relatedliterature. Section II describes the data. Section
III presents the research de-sign. Section IV shows the results and
Section V presents robustness tests.Section VI concludes.
3 See, for example, the discussion in Leuz (2006).4 See, for
example, LLSV (1997, 1998) as well as the survey by Beck and Levine
(2004).
-
1900 The Journal of Finance
I. Related Literature
The bonding hypothesis of Coffee (1999) and Stulz (1999) posits
that firmscross-listed on a major U.S. stock exchange have better
corporate governancethan non-cross-listed firms from the same
country, ceteris paribus, since cross-listed firms are subject to
strong U.S. investor protections.5 For example, cross-listed firms
on U.S. exchanges must adhere to U.S. disclosure practices,
whichrequire them to reconcile their net income and shareholder’s
equity to U.S.generally accepted accounting principle (GAAP),
disclose the identity of ma-jority shareholders (10% or greater),
and follow detailed procedures duringtender offers and going
private transactions. These firms are also subject tofar-reaching
U.S. investor protection laws such as the Foreign Corrupt
Prac-tices Act and, more recently, the Sarbanes Oxley Act.
Cross-listed firms arealso subject to punishment by U.S. law
enforcement, both by the Securitiesand Exchange Commission (SEC) as
well as private investor law suits, andto increased scrutiny from
intermediaries such as financial analysts and debt-rating
agencies.6 In contrast, listing on the OTC market or conducting a
privateplacement allows substantial exemptions from these laws and
regulations.7
Specifically, the bonding hypothesis predicts that, ceteris
paribus: (1) cross-listed firms will have better corporate
governance than non-cross-listed firms,(2) the difference in
governance between cross-listed firms and non-cross-listedfirms
will be greatest in the countries with the weakest investor
protections,and (3) cross-listings that require the most stringent
U.S. investor protections(i.e., on the NYSE, AMEX, or NASDAQ) will
have the largest differences incorporate governance. In this way,
cross-listing in the United States representsa market-based
approach to increased investor protection.
While in theory a cross-listing in the United States should lead
to more ef-fective corporate governance, the ability of a
cross-listing to serve as a bondingmechanism is under debate. On
the one hand, several empirical studies exam-ine the economic
impact of cross-listing in the United States and find evidencethat
is consistent with the bonding hypothesis. This line of research
finds thatcross-listed firms from weak investor protection
countries have larger stock-price reactions (Foerster and Karolyi
(1999), Miller (1999)), higher valuation(Mitton (2002), Doidge,
Karolyi, and Stulz (2004a)), more scrutiny by financialanalysts
(Baker, Nofsinger, and Weaver (2002), Lang, Lins, and Miller
(2003)),lower cost of capital (Errunza and Miller (2000), Hail and
Leuz (2004)), bet-ter information environments (Bailey, Karolyi,
and Salva (2005)), lower votingpremiums (Doidge (2004)), and more
access to external finance (Reese andWeisbach (2002), Lins,
Strickland, and Zenner (2005)). However, ascribing theevidence
contained in many of these studies directly to the bonding
hypothesis
5 It is important to note that while firms may choose to
cross-list for a variety of reasons, oncethey are listed they
become subject to U.S. laws and regulations.
6 Coffee (2002) calls these intermediaries “financial
watchdogs.”7 For example, these firms are not required to register
under the Exchange or Securities acts and
are therefore exempt from most civil liability provisions and do
not have to follow U.S. disclosurepractices (Doidge (2004)).
-
International Cross-Listing 1901
is difficult given the well-known challenge in distinguishing
among the varioustheories of cross-listing and the endogeneity
issues inherent to this literature.8
On the other hand, the evidence in several recent studies
suggests bondingvia cross-listing in the United States is
ineffective. For example, Siegel (2005)finds that the SEC and
minority shareholders have rarely enforced U.S. lawsagainst
cross-listed firms and Lang, Raedy, and Wilson (2006) find that the
ac-counting data of cross-listed firms from weak investor
protection environmentsare of lower quality even though
cross-listed firms are required to follow nom-inally similar
accounting standards as U.S. firms. However, the approaches inthese
papers are not without their drawbacks, as Coffee (2002) and Benos
andWeisbach (2004) suggest that measuring the incidence of legal
actions may un-derstate the deterrent benefit of laws, and Leuz
(2006) argues that disclosurequality differences between
cross-listed and U.S. firms may not be clear evi-dence against
bonding as cross-listed firms are allowed considerable discretionin
preparing their financial statements to U.S. GAAP.
Another challenge researchers face when testing the bonding
hypothesis isthat it is often difficult to assess the quality of
governance from observed mech-anisms of governance because
governance mechanisms often substitute or com-plement one another,
a finding that Doidge, Karolyi, and Stulz (2004b) empha-size is
dependant on the extent of a country’s investor protections.
Further, thisissue is likely to be exacerbated for cross-listed
firms, given the many financialand regulatory changes that take
place around a listing (see, for example, Lang,Lins, and Miller
(2003, 2004)).
In this paper, rather than calculating the stock-price
consequences, legal en-forcement incidents, or changes in
governance mechanisms around a cross-listing to infer improvements
in investor protections, we measure a directoutcome of corporate
governance: the propensity to replace poorly perform-ing CEOs. Why
CEO turnover? Replacing poorly performing CEOs is arguedto be a
necessary condition for good corporate governance (Shleifer and
Vishny(1989, 1997)) and the sensitivity of top executive turnover
to performance as ameasure of the quality of corporate governance
has been supported by a largenumber of studies in the United States
and abroad, including recent researchby Dahya et. al (2002), DeFond
and Hung (2004), Gibson (2003), and Volpin(2002).9
II. Sample Selection and Descriptive Statistics
A. Sample Selection
Our empirical analysis consists of three main parts. First, we
investigatewhether the sensitivity of top executive turnover to
poor firm performance ishigher for cross-listed firms. In these
tests we differentiate cross-listings bytype in order to test
whether cross-listings on major U.S. exchanges, which re-quire the
strongest governance provisions, have the largest effect. Second,
we
8 For example, Sarkissian and Schill (2006) argue that valuation
gains to cross-listing aretransitory.
9 For U.S.-based studies see Hermalin and Weisbach (2003) and
citations contained therein.
-
1902 The Journal of Finance
test if the effect of bonding is greatest for firms that are
located in the countrieswith the weakest investor protection laws.
We do so by examining the sensi-tivity of top executive turnover to
poor firm performance across legal originsand investor protection
laws. Finally, we conduct a battery of tests designedto gauge the
robustness of our results by examining the sensitivity of top
ex-ecutive turnover to poor firm performance in the
pre-cross-listing period, theeffect of other potential governance
changes concurrent with cross-listing, theexclusion of turnover
that occurs in the list year, and the departing and en-tering CEOs’
work history. We also rerun our tests excluding countries
thatcontain the largest portion of our sample firms, omitting firms
with large blockownership, removing observations surrounding the
Asian financial crisis, andexcluding financial and regulated
firms.
To execute this analysis, we gather data on executive turnover
and firm per-formance between 1992 and 2003 from the Worldscope
database.10 The initialsample consists of approximately 38,000
firms from 59 countries. We excludefirms with missing firm-specific
financial and executive data, firms with noidentifiable top
manager, and firms located in countries with missing legal
en-vironment data. We also exclude U.S. firms as the bonding
hypothesis predictsdifferences between cross-listed and
non-cross-listed firms, rather than differ-ences between
cross-listed and U.S. firms (see, for example, Leuz (2006)).
Fi-nally, we exclude firms that are reported in the Worldscope
database only oncebecause we need at least two consecutive years of
nonmissing data on com-pany officers and their titles to compute
CEO turnover. The resulting sampleincludes 70,976 firm-year
observations of 19,091 firms from 47 countries over1992–2003. Every
country in our final sample except Zimbabwe has at least
onecross-listed firm in the United States . A breakdown of the
sample distributionacross countries, cross-listing status, and
years is reported in Table I.
We obtain the list of cross-listed firms using several sources
including theBank of New York, Citibank, NYSE, and Nasdaq and
verify the listing datesusing Lexis–Nexis searches, Form 20-F, etc.
Exchange-traded cross-listings aredenoted as Level 2/3,
over-the-counter cross-listings as Level 1, and privateplacements
as Rule 144a. The data set also takes into account ADR
programupgrades, such as from a Level 1 to a Level 2 program, and
delistings fromthe U.S. market. We also include direct listings.
Most notably, Canadian firmslist their shares on U.S. exchanges
directly without issuing ADRs . Given thatthe increased disclosure
and securities law provisions required in listing on amajor U.S.
exchange are functionally equivalent for ADRs and direct
listings,we classify Canadian firms that are traded on both a
Canadian and a majorU.S. exchange as Level 2/3 ADRs. However, the
exclusion of Canadian firmsfrom the sample does not change our
conclusions.
We follow DeFond and Hung (2004) and use the titles CEO, Chief
ExecutiveOfficer, and Chief Executive to identify the top manager
in each firm. However,
10 We use a total of 37 Worldscope CDROMs during our sample
period. Because of delays byfirms in releasing information and
Worldscope’s backfilling procedure, Worldscope indicated to usthat
multiple CDROMs from each year should be used as they often contain
different numbers offirms.
-
International Cross-Listing 1903
Table IDescriptive Statistics
This table presents the distribution of the sample used in the
regression analysis by country,cross-listing status, and year, and
descriptive statistics for the main firm-level variables. Panel
Adescribes the number of observations, number of firms, and CEO
turnover percentage across coun-tries. Panel B presents the
distribution of the sample over time. Panel C displays the
distributionof the sample by cross-listing status. Panel D presents
the summary statistics for the sample usedin the regression
analysis. The last column in Panel D reports the median differences
of the firmperformance variables between the CEO turnover and
nonturnover observations and the relatedresults from a
nonparametric test on the equality of medians. CL dummy is 1 if the
firm cross-listsin the United States, 0 otherwise. Level 2/3 dummy
is 1 if the firm has a Level 2 or Level 3 ADRprogram, 0 otherwise.
Level 1 dummy is 1 if the firm has a Level 1 ADR program, 0
otherwise. Rule144A dummy is 1 if the firm has a Rule 144A
issuance, 0 otherwise. Lagged Earnings Ratio is the1-year lagged
ratio of earnings before interest and taxes to total assets. Lagged
Excess Returns isthe 1-year lagged total stock returns in excess of
the country average. Total Assets is measured inmillion $U.S. ∗∗∗
indicates significance at the 1% level.
Panel A: By Country
Country No of Obs. No of Firms CEO Turnover %
Argentina 36 18 25.00Australia 2,463 1,001 15.79Austria 530 160
15.09Belgium 711 187 14.21Brazil 348 155 11.78Canada 3,455 1,011
19.28Chile 473 152 18.39China 276 170 19.57Colombia 3 3 33.33Czech
Republic 83 42 15.66Denmark 1,146 283 14.22Finland 766 212
13.71France 3,200 1,019 12.25Germany 3,692 1,047 17.39Greece 517
216 15.28Hong Kong 2,028 797 16.18Hungary 114 36 19.30India 1,268
358 13.17Indonesia 1,090 324 24.13Ireland 418 93 14.83Israel 217 90
21.66Italy 1,192 318 18.04Japan 21,009 3,776 14.15Korea 1,704 675
38.40Luxembourg 59 18 15.25Malaysia 2,185 687 13.27Mexico 304 117
17.76Netherlands 1,034 305 19.05New Zealand 294 90 18.71Norway 356
110 18.82Pakistan 459 111 12.64Peru 87 43 24.14Philippines 674 205
20.47Poland 239 83 19.67
(continued)
-
1904 The Journal of Finance
Table I—Continued
Panel A: By Country
Country No of Obs. No of Firms CEO Turnover %
Portugal 152 62 9.21Singapore 1,304 429 17.02South Africa 1,210
450 16.28Spain 590 163 16.95Sri Lanka 37 14 8.11Sweden 1,359 420
18.91Switzerland 1,413 360 20.74Taiwan 1,241 423 25.56Thailand 898
269 18.04Turkey 316 124 15.19United Kingdom 9,981 2,448
14.59Venezuela 25 11 4.00Zimbabwe 20 6 25.00
Panel B: By Year
Year No of Obs. No of Firms CEO Turnover %
1992 1,602 – 17.851993 4,081 – 17.741994 4,721 – 13.851995 5,369
– 11.511996 6,083 – 12.941997 6,790 – 13.781998 6,359 – 15.851999
6,893 – 14.842000 6,813 – 23.162001 7,712 – 19.852002 9,962 –
17.512003 4,591 – 14.75
Panel C: By Cross-listing Status
Cross-listing Status No of Obs. No of Firms CEO Turnover %
Non-CL firms 65,563 17,729 16.06CL firms 5,413 1,362 19.18Level
2/3 2,088 609 21.50Level 1 2,494 565 16.92Rule 144A 801 188
20.22
Total 70,976 19,091 16.30
Panel D: Summary Statistics
Turnover vs.5th 95th Nonturnover
N Mean Median Percentile Percentile (Medians)
Lagged Earnings Ratio 70,976 0.218 0.052 −0.133 0.202
−0.006∗∗∗Lagged Excess Returns 62,333 −0.011 −0.079 −0.724 0.909
−0.030∗∗∗Total Assets 70,976 259,278 372.651 15.534 494,872 –
-
International Cross-Listing 1905
many countries use other titles for top managers, which vary
across and withincountries. We use two sources to determine the top
manager in the rest of thesample. When available, we use the top
manager titles used by DeFond andHung (2004) and Gibson (2003). For
example, the titles CEO, Chief ExecutiveOfficer, Chief Executive,
and President are used to identify the top managerin Argentina. We
exclude firms in which the top manager title is shared bytwo
officers to prevent a split turnover (Gibson (2003)). For the
remaining 15countries not covered in DeFond and Hung (2004) or
Gibson (2003) (4.68% ofour sample), we use press accounts, country
experts’ opinions, and manual datainspections of manager titles in
each country to determine the top manager title.A list of top
manager titles used in each country is displayed in the
Appendix.11
After the top manager in the firm is identified, we first
compare the last namesand the first letter of first names of top
managers of the firm over time todetermine whether there was a top
manager replacement in any given year.We next hand-check CEO
turnover events for the entire sample given thatDefond and Hung
(2004) find that first names of managers do not consistentlyprecede
their last names in several Asian countries such as Korea and
Japan,and Worldscope infrequently contains typos on executive names
for foreignfirms.
As in DeFond and Hung (2004) and Gibson (2003), we do not know
whether aCEO turnover event is voluntary (for example, due to
retirement) because theWorldscope does not provide information on
CEO age and tenure, and mediacoverage in English for the sample
firms varies substantially across countries.Hermalin and Weisbach
(2003) argue that voluntary turnover is unlikely to berelated to
performance, and hence not distinguishing between voluntary
andforced turnovers events leads to additional noise in the
dependent variable,which only affects standard errors. Consistent
with their assertion, the empir-ical evidence suggests a similar or
more sensitive relationship between CEOturnover and performance for
involuntary (forced) replacements (see, for exam-ple, Huson,
Parrino, and Starks (2001), Dahya et al. (2002), and Kaplan
andMinton (1994)). Therefore, we do not expect this data limitation
to alter ourconclusions.
B. Descriptive Statistics
Panel A of Table I provides summary statistics for the sample
based on afirm’s country of domicile. Turnover ranges from a low of
4% in Venezuela toa high of 38.4% in Korea, with an average of
16.30%. For comparison, overour sample period the U.S. turnover
rate was 12.86%. Similar to other studiesthat employ the Worldscope
database, there is a clustering of observations inJapan and United
Kingdom. Although our analysis is based on fixed countryeffects to
ensure we are comparing CEO turnover differences within
countries,in robustness tests reported later in the paper we remove
observations fromJapan and the United Kingdom and find that our
conclusions are unaffected.
11 We use the terms CEO turnover and top manager turnover
interchangeably.
-
1906 The Journal of Finance
Panel B of Table I shows turnover by year, which ranges from a
low of 11.51% in1995 to a high of 23.16% in 2000. Panel C of Table
I presents turnover by cross-listing status. The panel indicates
that cross-listed firms have higher CEOturnover than
non-cross-listed firms (19.18% vs. 16.06%). Of the
cross-listedfirms, CEO turnover is greatest for Level 2/3 firms,
followed by Rule 144a andthen Level 1 companies (21.50%, 16.92%,
and 20.22%, respectively). Panel Cof Table I also shows that 1,362
foreign firms are identified as cross-listed inour sample, of which
609 are exchange-traded cross-listings (Level 2/3), 565are OTC
cross-listings (Level 1), and 188 are private placements via Rule
144Aissuance (Rule 144a).
We consider various measures of firm performance, including both
operat-ing performance measures and stock-price-based measures. We
augment stockprice data from Worldscope with data from Datastream
International wherepossible. However, we expect the operating
performance measures to be a bet-ter proxy in our international
setting, as both Volpin (2002) and DeFond andHung (2004) find that
stock returns are not related to CEO turnover in coun-tries whose
markets are characterized by high stock-price synchronicity
andilliquidity, attributes that make stock-price-based measures
less informative.12
For our main tests, we focus on the ratio of accounting earnings
before inter-est and taxes (EBIT) to book value of assets (earnings
ratio), and the total stockreturns in excess of the country average
(excess returns). We follow DeFond andHung (2004) and Volpin (2002)
and use EBIT among accounting-based firmperformance measures
because it is not influenced by firms’ capital structurepolicies or
by differential country-level tax regimes. Similar forms of both
vari-ables are used extensively to proxy for firm performance in
studies examiningthe sensitivity of CEO turnover to firm
performance.13 We lag both performancevariables by 1 year to
prevent a possible overlap of the replaced CEO’s perfor-mance with
that of the new CEO. Panel D of Table 1 reports sample statisticsof
the main performance measures and shows that the lagged
performancemeasures are significantly lower in firm-years with CEO
turnover than in non-turnover years. In terms of the depth of the
sample, the mean (median) numberof years a firm is in our
regression analysis is 3.84 (3) years.
We also use sales growth and the change in EBIT to total assets
as alterna-tive accounting-based measures of firm performance and
obtain qualitativelysimilar results. In addition, we recompute our
firm performance measures inwhich industry-adjusted performance is
calculated as firm performance minusthe median value of the
corresponding two-digit SIC global industry and obtainsimilar
results.
12 Harvey (1995) shows that first-order autocorrelations in
emerging markets are positive andsignificant and Lesmond (2005)
finds that liquidity-related transactions costs in countries
withweak legal institutions are higher than in markets with strong
legal systems.
13 See Huson et al. (2001), Mikkelson and Partch (1997), Gibson
(2003), and Kang and Shivdasani(1995) for the accounting-based
measure and Weisbach (1988), Kang and Shivdasani (1995), Defondand
Hung (2004), Huson, Parrino, and Starks (2001), and Hadlock and
Lumer (1997) for the stock-market-based measure.
-
International Cross-Listing 1907
III. Research Design
A. Empirical Model
To test our hypothesis that CEO turnover is more sensitive to
poor perfor-mance for exchange-traded cross-listed firms than
non-cross-listed firms, weestimate a series of probit models that
take the form:
Pr (Turnover) = φ[α + β1(FirmPerformance)] + β2(L23)+
β12(L23∗FirmPerformance) + β3(L1)+ β13(L1∗FirmPerformance) +
β4(R144A)+ β14(R144A∗FirmPerformance) + δX ] (1)
where φ is the standard normal cumulative distribution, L23
refers to exchange-traded cross-listings, L1 refers to OTC
cross-listings, R144A refers to privateplacements, and X is a set
of firm control variables, country controls, industrycontrols, and
year controls. Note that the cross-listed dummies are
time-varyingin that they take the value of 1 in the cross-listing
year and can switch back to0 if the firm delists or changes level
of cross-listing.
We follow previous research and measure turnover as a binary
variable thattakes the value 1 if the top manager is changed in
that year. We include firmsize measured as the natural logarithm of
the book value of total assets inmillions of U.S. dollars. In the
regression analysis, we winsorize the continuousvariables at the 1%
level for each country.
It is also important to note that throughout our analysis, we
include coun-try fixed effects that ensure we are measuring
within-country differencesbetween cross-listed and non-cross-listed
firms as well as controlling for un-observed country effects. In
addition, we include industry dummies using thetwo-digit SIC code
to control for global industry-wide factors that may affectCEO
turnover and firm performance. Finally, our regressions include
indica-tor variables for each year. Our regressions also correct
the standard errorsfor possible serial correlation and
heteroskedasticity by clustering at the firmlevel.
We test our second hypothesis, which posits that the difference
in the sensitiv-ity of top management turnover to performance
between cross-listed firms andnon-cross-listed firms is greatest in
the countries with the weakest corporategovernance, by classifying
countries into strong and weak investor protectionregimes and
comparing coefficients across samples. Alternatively, a
randomcountry effects specification could be employed with
interaction effects, but inour sample this is inappropriate as it
fails the Hausman specification test. Wefocus on three
country-level measures of investor protection. The first
measure,from La Porta, Lopez-De-Silanes, Shleifer, and Vishny
(LLSV) (1997, 1998), iswhether the home country has an English
legal origin, which is an overallmeasure of strong investor
protections. Following Djankov et al. (2005) we usethe
anti-director rights index (ADRI), which is a revised version of
the originalADRI index from LLSV (1997) that addresses the coding
concerns expressed
-
1908 The Journal of Finance
in Pagano and Volpin (2005) and Spamann (2006). The ADRI
represents thedegree of minority shareholder protection. We also
use the anti-self-dealingindex from Djankov et al. (2005), which
measures how difficult it is for minor-ity shareholders to thwart
the consumption of private benefits by controllingshareholders.
Djankov et al. (2005) argue that self-dealing is the central
prob-lem of corporate governance in most countries. In unreported
tests, we alsoexamine other country-level measures of investor
protection from LLSV (1998)and La Porta, Lopez-De-Silanes, and
Shleifer (LLS) (2006), such as the ruleof law, burden of proof,
disclosure, and private law enforcement indexes. In allinstances,
our results are consistent across every measure of high versus
lowinvestor protection. Further, the results are robust to using
Spamann’s (2006)ADRI.
B. Interpretation of Interactions in Probit Models
Recent research by Ai and Norton (2003) and Powers (2005)
emphasizes thedifficulty present in interpreting interactions in
nonlinear models. Strikingly,the interaction effect cannot be
evaluated by looking at the sign, magnitude, orstatistical
significance of the coefficient on the conventional interaction
term.Ai and Norton (2003) show that the interaction effect is
conditional on the inde-pendent variable, and therefore both the
magnitude and statistical significanceof the interaction term can
vary across observations. For example, in our probitspecification
the correct marginal effect of a change in the interaction
variablebetween the L23 dummy and firm performance is
�∂F (u)
∂FirmPerformance�L23
= (β1 + β12) ∗ φ[(β1 + β12) ∗ FirmPerformance + β2 + X δ]− β1 ∗
φ [β1 ∗ FirmPerformance + δX ] (2)
where F (u) = Pr(Turnover), which is given by equation (1) and u
denotes theregression specification. Equation (2) shows that the
marginal effect of theinteraction variable may not be zero even
when β12 is zero. Thus, the standardcoefficient on the interaction
term may have an incorrect magnitude, standarderror, and even sign
relative to the true interaction effect.
To ensure our inferences are correct, we use the methodology
developedby Norton, Wang, and Ai (2004) to compute the correct
marginal effect of achange in the interaction variable between the
respective cross-listed dummyand firm performance. We report both
the marginal effects and their standarderrors and display the
graphs of the distribution of marginal effects and theassociated
z-statistics over the entire range of predicted probabilities for
ourmain models. In tests where our inferences are unambiguous, we
also sum-marize the range of corrected interactions by the mean
interaction effect andits significance.
-
International Cross-Listing 1909
IV. The Effect of Cross-Listing on CEO Turnover
A. By Cross-Listing Type
Table II presents a series of probit regressions that include
interactionsbetween firm performance and cross-listing type to test
the hypothesis thatcross-listed firms have a higher performance to
turnover sensitivity than non-cross-listed firms. Model 1 reports
the results for the accounting-based per-formance measure and Model
2 presents the results for the stock-price-basedperformance
measure. All regression models include country, industry, and
yearfixed effects as well as control for firm size.
Model 1 shows that the interaction between Level 2/3 and lagged
earningsratio is negative and significant (−0.332, t-statistic =
−2.087). In contrast,OTC or Rule 144a cross-listings do not have a
significantly higher propensityto terminate poorly performing CEOs
than non-cross-listed firms’ interaction(coefficient = −0.120,
t-statistic = −0.452 and −0.350, t-statistic = −0.675,
re-spectively). This finding is consistent with the hypothesis that
non-U.S. firmsadopting the strongest governance and reporting
requirements by cross-listingin the United States observe outcomes
consistent with improved governanceover similar firms that are not
cross-listed in the United States. However,given the aforementioned
problems with interpreting simple interaction termsin discrete
choice models, we follow Ai and Norton (2003) and evaluate the
cor-rected marginal effects and their significance at every
predicted probability.Figure 1a shows that for major
exchange-traded cross-listings, the correctedinteraction effects
are overwhelmingly negative across the predicted probabil-ities,
while Figure 1b shows that these interaction effects are also
significant(less than −1.96) for most probabilities. We summarize
the corrected inter-active effect and its significance in the last
row of Table II by reporting themean interaction effect and its
significance (−0.084, t-statistic = −2.082). Interms of economic
significance, the absolute probability of replacing the
CEOincreases by 1.34% for Level 2/3 ADRs when we move from the top
quartileto the bottom quartile of firm performance. For OTC
(Rule144a) cross-listings,the corrected interactive effects,
presented in Figures 1c–f and summarizedin the bottom rows of Table
II, further confirm that the interaction effectis rarely
significant across the range of predicted probabilities (the
mean-corrected effect is −0.030, t-statistic = −0.452, and −0.086,
t-statistic = −0.667,respectively).
For the control variables, we find that firm size is positively
related to CEOturnover.14 Firm performance (lagged earnings ratio)
is negative yet statisti-cally insignificant, a finding that is the
result of pooling countries where firmperformance is unlikely to be
used to evaluate management.15 The coefficienton L2/3 is positive,
indicating that exchange cross-listed firms also have higher
14 In the United States, size is generally thought to capture
the effects of CEO and institutionalstock ownership, board
composition, managerial depth, and formal succession processes
(see, e.g.,Huson et al. (2001)). Gibson (2003) and DeFond and Hung
(2004) also find that firm size is positivelyrelated to CEO
turnover internationally.
15 When we split by investor protection regimes, the coefficient
on lagged earning ratio is negativeand significant in high
protection countries.
-
1910 The Journal of Finance
Table IICEO Turnover and Cross-Listing
This table presents the probit estimates of the relationship
between the probability of CEO turnoverand firm performance
measured by Lagged Earnings Ratio (1-year lagged ratio of earnings
beforeinterest and taxes to total assets) or Lagged Excess Returns
(1-year lagged total stock returns inexcess of the country average
returns). Level 2/3 dummy is 1 if the firm has a Level 2 or Level
3ADR program, 0 otherwise. Level 1 dummy is 1 if the firm has a
Level 1 ADR program, 0 otherwise.Rule 144A dummy is 1 if the firm
has a Rule 144A issuance, 0 otherwise. Log Assets is the naturallog
of total assets measured in million $U.S. The continuous variables
are winsorized at the 1%level for each country. The interaction
effect is defined as the change in the predicted probabilityof CEO
turnover for a change in both the firm performance and the
respective cross-listed dummyusing the methodology of Norton, Wang,
and Ai (2004). The z-statistics appear in parentheses
belowparameter estimates. Robust standard errors are estimated
using the Rogers method of clusteringby firm. ∗∗∗ and ∗∗ indicate
significance at the 1% and 5% level, respectively.
Variable (1) (2)
Log assets 0.023∗∗∗ 0.024∗∗∗[9.939] [9.433]
Firm performance: Lagged earnings ratio −0.002 –[−1.528]
Firm Performance: Lagged excess returns – −0.012[−1.317]
L2/3 0.077∗∗ 0.029[2.226] [0.860]
L2/3 ∗ firm performance −0.332∗∗ −0.092∗∗[−2.087] [−2.115]
L1 0.043 0.076∗∗[1.274] [2.491]
L1 ∗ Firm performance −0.120 0.001[−0.452] [0.107]
R144A 0.049 −0.012[0.704] [−0.202]
R144A ∗ Firm performance −0.350 0.015[−0.675] [0.330]
Constant −0.905∗∗∗ −0.966∗∗∗[−3.245] [−3.302]
Country effects Yes YesIndustry effects (two-digit SIC) Yes
YesYear effects Yes YesObservations 70,976 62,333Pseudo-R2 0.026
0.028
Mean interaction effect for −0.084∗∗ −0.022∗∗L23 ∗ Firm
performance [−2.082] [−2.009]Mean interaction effect for −0.030
0.0002L1 ∗ Firm performance [−0.452] [0.067]Mean interaction effect
for −0.086 0.003R144 ∗ firm performance [−0.667] [0.273]
-
International Cross-Listing 1911
Figure 1. The economic significance of the impact of
cross-listing on the relationshipbetween CEO turnover and firm
performance. The following graphs display the interactioneffects
and corresponding z-statistics on the interaction variable between
the respective cross-listeddummy and firm performance measure
reported in Table II, estimated using Norton, Wang, and Ai(2004).
The interaction effect is defined as the change in the predicted
probability of CEO turnoverfor a change in both firm performance
and the respective cross-listed dummy. Panel A plots thegraphs
associated with the lagged earnings ratio measure and Panel B
depicts the graphs for thelagged excess returns measure. The lines
above and below 0 on the figures located on the right siderepresent
the 5% significance levels (±1.96).
-
1912 The Journal of Finance
Figure 1—Continued.
absolute turnover, a result that is primarily driven by firms
from the UnitedKingdom.16
Model 2 of Table II examines the sensitivity of CEO turnover to
performanceemploying our alternative firm performance measure,
1-year lagged excess
16 Dahya et al. (2002) show that after the Cadbury Act was
passed in 1992, CEO turnoverincreased for U.K. firms.
-
International Cross-Listing 1913
stock market returns. We find that the interaction between L2/3
and stock-price-based performance is negative and significant,
while the interactions be-tween L1 or R144a and stock-price-based
performance are insignificant. There-fore, with this alternative
performance measure we continue to find that cross-listed firms
that are associated with the most stringent U.S. investor
protectionsare more likely to terminate poorly performing CEOs.
Overall, the results contained in Table II provide support for
the hypothesisthat cross-listing on a major U.S. exchange, which
requires the adoption of strin-gent U.S. investor protection laws,
results in a significantly higher propensityto terminate poorly
performing CEOs than their non-cross-listed counterparts.In
addition, firms that cross-list via Level 1 or Rule 144a ADRs do
not have anincreased association between CEO turnover and poor firm
performance.
In untabulated results, we also split our sample into countries
with highand low stock-price informativeness to examine if the CEO
turnover to per-formance sensitivity is higher in countries where
stock prices are more infor-mative about firm-specific performance.
Prior research by DeFond and Hung(2004) and Volpin (2002) argues
that only in countries where stock prices areinformative is CEO
turnover related to stock market performance.17 However,it is
important to note that a significant relation between CEO turnover
andperformance in low informativeness countries is possible if
stock prices be-come more informative about performance due to
cross-listing, something thatFernandes and Ferreira (2006) and
Dasgupta, Gan, Ning, (2005) suggest oc-curs. Consistent with DeFond
and Hung (2004) and Volpin (2002), we find thatthe interaction
between cross-listing types and firm performance is insignif-icant
in countries that have below-median stock-price informativeness,
whilein countries where stock prices are informative, cross-listing
on a major U.S.exchange results in a higher propensity to shed
poorly performing CEOs. Theseresults also suggest that the
increased CEO turnover to performance sensitiv-ity for cross-listed
firms is not driven purely by stock prices becoming moreinformative
for cross-listed firms in certain countries.
Overall, the results in Table II provide support for the bonding
hypothesis.We find that cross-listed firms have outcomes that are
consistent with bettercorporate governance systems than similar
non-cross-listed firms. Further, thefindings suggest that
governance outcome differences are only significant forthose firms
that adopt the strongest U.S. investor protections by listing on
amajor U.S. exchange, rather than an OTC listing or private
placement. Theseresults provide support for the hypothesis that by
cross-listing in the UnitedStates, firms are able to opt-in to
superior corporate governance.
B. The Strength of Bonding for Firms in Low Investor Protection
Countries
Table III tests the third prediction of the bonding hypothesis
that the ef-fect of bonding will be greatest for firms domiciled in
the countries with theweakest investor protections. We test this
hypothesis by splitting the sample
17 Bushman, Piotroski, and Smith (2004) show that corporate
transparency is low in poor investorprotection countries.
-
1914 The Journal of Finance
Tab
leII
IC
EO
Tu
rnov
er,C
ross
-Lis
tin
g,an
dL
egal
En
viro
nm
ent
Th
ista
ble
pres
ents
the
prob
ites
tim
ates
ofth
ere
lati
onsh
ipbe
twee
nth
epr
obab
ilit
yof
CE
Otu
rnov
eran
dfi
rmpe
rfor
man
ceu
nde
rva
riou
sm
easu
res
ofa
cou
ntr
y’s
lega
len
viro
nm
ent.
Fir
mpe
rfor
man
ceis
mea
sure
dby
Lag
ged
Ear
nin
gsR
atio
,wh
ich
isth
e1-
year
lagg
edra
tio
ofea
rnin
gsbe
fore
inte
rest
and
taxe
sto
tota
lass
ets,
orL
agge
dE
xces
sR
etu
rns,
wh
ich
isth
e1-
year
lagg
edto
tals
tock
retu
rns
inex
cess
ofth
eco
un
try
aver
age
retu
rns.
Th
eC
ivil
Law
sam
ple
incl
ude
sfi
rms
loca
ted
inco
un
trie
sw
ith
aF
ren
ch,
Ger
man
,or
Sca
ndi
nav
ian
lega
lsy
stem
.T
he
Com
mon
Law
sam
ple
refe
rsto
firm
slo
cate
din
cou
ntr
ies
wit
hth
eE
ngl
ish
lega
lori
gin
.An
ti-d
irec
tor
righ
tsin
dex
mea
sure
sth
ede
gree
ofm
inor
ity
shar
ehol
der
prot
ecti
on.A
nti
-sel
f-de
alin
gis
anin
dex
ofth
est
ren
gth
ofm
inor
ity
shar
ehol
der
prot
ecti
onag
ain
stse
lf-d
eali
ng
byth
eco
ntr
olli
ng
shar
ehol
der.
All
thes
eco
un
try-
leve
lin
dice
sar
eob
tain
edfr
omD
jan
kov
etal
.(20
05).
Th
em
edia
ns
of3.
5fo
ran
ti-d
irec
tor
righ
tsan
d0.
42fo
ran
tise
lf-d
eali
ng
inde
xu
sed
inD
jan
kov
etal
.(20
05)
are
use
dto
grou
pfi
rms
into
hig
hve
rsu
slo
win
vest
orpr
otec
tion
regi
mes
(low
erth
anor
equ
alto
the
med
ian
refe
rsto
low
gove
rnan
cesu
bsam
ples
).L
evel
2/3
dum
my
is1
ifth
efi
rmh
asa
Lev
el2
orL
evel
3A
DR
prog
ram
,0ot
her
wis
e.L
evel
1du
mm
yis
1if
the
firm
has
aL
evel
1A
DR
prog
ram
,0ot
her
wis
e.R
ule
144A
dum
my
is1
ifth
efi
rmh
asa
Ru
le14
4Ais
suan
ce,0
oth
erw
ise.
Log
Ass
ets
isth
en
atu
rall
ogof
tota
lass
ets
mea
sure
din
mil
lion
$U.S
.Th
eco
nti
nu
ous
vari
able
sar
ew
inso
rize
dat
the
1%le
velf
orea
chco
un
try.
Th
ein
tera
ctio
nef
fect
isde
fin
edas
the
chan
gein
the
pred
icte
dpr
obab
ilit
yof
CE
Otu
rnov
erfo
ra
chan
gein
both
the
firm
perf
orm
ance
and
the
resp
ecti
vecr
oss-
list
eddu
mm
yu
sin
gth
em
eth
odol
ogy
ofN
orto
n,W
ang,
and
Ai(
2004
).T
he
z-st
atis
tics
appe
arin
pare
nth
eses
belo
wpa
ram
eter
esti
mat
es.R
obu
stst
anda
rder
rors
are
esti
mat
edu
sin
gth
eR
oger
sm
eth
odof
clu
ster
ing
byfi
rm.
∗∗∗
and
∗∗in
dica
tesi
gnif
ican
ceat
the
1%an
d5%
leve
l,re
spec
tive
ly.
Pan
elA
:On
e-Ye
arL
agge
dE
arn
ings
Rat
io
Low
An
ti-
Hig
hA
nti
-L
owA
nti
-H
igh
An
ti-
Civ
ilC
omm
onD
irec
tor
Dir
ecto
rS
elf-
Sel
f-L
awL
awR
igh
tsR
igh
tsD
eali
ng
Hig
hD
eali
ng
Var
iabl
e(1
)(2
)(3
)(4
)(5
)(6
)
Fir
m,c
oun
try,
indu
stry
,yea
rco
ntr
ols
Yes
Yes
Yes
Yes
Yes
Yes
Obs
erva
tion
s44
,735
26,2
3741
,007
29,9
6715
,936
55,0
20P
seu
do-R
20.
033
0.02
20.
034
0.02
30.
031
0.02
9M
ean
inte
ract
ion
effe
ctfo
r−0
.341
∗∗∗
0.01
9−0
.333
∗∗∗
0.01
8−0
.328
∗∗∗
−0.0
45L
23∗l
agge
dea
rnin
gsra
tio
[−3.
435]
[0.4
18]
[−3.
339]
[0.3
86]
[−3.
121]
[−1.
023]
Mea
nin
tera
ctio
nef
fect
for
−0.1
120.
040
−0.1
0400
38−0
.023
−0.0
31L
1∗l
agge
dea
rnin
gsra
tio
[−0.
850]
[0.5
40]
[−0.
763]
[0.5
19]
[−0.
154]
[−0.
440]
Mea
nin
tera
ctio
nef
fect
for
−0.1
640.
129
−0.1
830.
058
−0.1
21−0
.076
R14
4∗l
agge
dea
rnin
gsra
tio
[−1.
071]
[0.5
02]
[−0.
975]
[0.3
31]
[−0.
535]
[−0.
497]
-
International Cross-Listing 1915
Pan
elB
:On
e-Ye
arL
agge
dE
xces
sR
etu
rns
Low
An
ti-
Hig
hA
nti
-L
owA
nti
-H
igh
An
ti-
Civ
ilC
omm
onD
irec
tor
Dir
ecto
rS
elf-
Sel
f-L
awL
awR
igh
tsR
igh
tsD
eali
ng
Hig
hD
eali
ng
Var
iabl
e(1
)(2
)(3
)(4
)(5
)(6
)
Fir
m,c
oun
try,
indu
stry
,yea
rco
ntr
ols
Yes
Yes
Yes
Yes
Yes
Yes
Obs
erva
tion
s39
,809
22,5
1936
,552
25,7
7113
,455
48,8
54P
seu
do-R
20.
033
0.02
30.
034
0.02
40.
032
0.03
0M
ean
inte
ract
ion
effe
ctfo
r−0
.044
∗∗∗
0.01
3−0
.054
∗∗0.
007
−0.0
35∗∗
−0.0
09L
23∗l
agge
dex
cess
retu
rns
[−3.
256]
[0.8
01]
[−2.
077]
[0.5
00]
[−2.
492]
[−0.
655]
Mea
nin
tera
ctio
nef
fect
for
0.00
040.
002
−0.0
090.
003
−0.0
002
−0.0
03L
1∗l
agge
dex
cess
retu
rns
[0.1
48]
[0.1
10]
[−0.
250]
[0.7
51]
[−0.
077]
[−0.
200]
Mea
nin
tera
ctio
nef
fect
for
−0.0
002
0.02
60.
004
0.00
4−0
.004
0.01
4R
144
∗lag
ged
exce
ssre
turn
s[−
0.01
5][0
.931
][0
.267
][0
.250
][−
0.25
3][1
.148
]
-
1916 The Journal of Finance
by investor protection regimes and examining the interactions
between cross-listing types and firm performance, for both
accounting- and stock-market-based performance measures (Panels A
and B, respectively). Firm, country,industry, and year controls are
included in all regressions.
The first two columns in Panel A of Table III split the sample
by legal origin, aclassification that proxies for the overall
protection of minority shareholders ina country (see, for example,
LLSV (1998)). Model 1 shows that in countries withCivil Law
tradition, where investor protection is weakest, the corrected
interac-tion between Level 2/3 and Lagged Earnings Ratio is
negative and significant(−0.341, t-statistic = −3.435). Model 1
also shows that the corrected interac-tions between firm
performance and L1 or R144a are insignificant (−0.112,t-statistic =
−0.850, and −0.164, t-statistic = −1.071, respectively).
Therefore,in countries with poor investor protections,
cross-listing on a major U.S. ex-change is associated with an
increased CEO turnover to poor firm performancesensitivity. In
terms of economic significance, the probability of replacing theCEO
increases by 4.36% in absolute terms for Level 2/3 ADRs when we
movefrom the top quartile to the bottom quartile of firm
performance measured inCivil Law countries.18
Model 2 of Panel A presents the results for Common Law
countries, whereinvestor protection is strongest. In these
countries, we find that all the interac-tions between cross-listing
type and the Lagged Earnings Ratio are statisticallyinsignificant.
Further, the difference in the interaction terms between Civil
andCommon Law countries is significant (p-value of lower than
0.01). Therefore,the results indicate that the effect of U.S.
investor protections is most signifi-cant when the firm’s home
country investor protections are weakest. Althoughnot reported, we
find that the Lagged Earnings Ratio coefficient is negative
andsignificant, which is consistent with previous research that
finds accounting-based performance is used for managerial
performance evaluation.
To further test if the extent of bonding is dependent on the
category of pro-tections investors are afforded in a particular
country, we also investigate al-ternative investor protection
indices from Djankov et al. (2005). Models 3 and5 report results
for countries classified as having weak protection of
minorityshareholders and poor safeguards against corporate
tunneling. In both models,we find the interactions (both standard
and corrected) between Level 2/3 andLagged Earnings Ratio are
negative and significant, indicating that in theselow investor
protection countries, cross-listing on a major U.S. exchange is
as-sociated with increased CEO turnover to poor firm performance
sensitivity.Models 4 and 6 report results for the strong investor
protection countries. Inboth these models, the interactions between
cross-listing type (Level 2/3, L1, orR144a) are insignificant,
indicating no difference in the CEO turnover to firmperformance
relation in countries that have strong investor protection.19
18 For comparison, Huson et al. (2001) show that going from the
top quartile to the lowest quartilein EBIT/TA ratio increases the
probability of CEO turnover by 2%.
19 Using alternative measures of investor protection laws from
LLS (2006) and LLSV (1998),such as the burden of proof, investor
protection, private law enforcement, disclosure, and ruleof law
indexes, produces similar conclusions. Further, the results are
robust to using Spamann’s(2006) ADRI.
-
International Cross-Listing 1917
Panel B provides the results for the stock-price-based firm
performance mea-sure. As in Panel A, we find across all weak
investor protection subsamples,the interaction between firm
performance and Level 2/3 is negative and signifi-cant. For
example, the mean-corrected interactive effect in Civil Law
countriesis −0.044 (t-statistic = −3.256). For other cross-listing
types, the interactioncoefficients are insignificant. Further,
Models 2, 4, and 6 show that in stronginvestor protection
countries, cross-listing does not increase the sensitivity ofCEO
turnover to poor firm performance.
Taken together, the results in Table III show that in countries
where investorprotections are weakest, adopting the strongest U.S.
investor protection provi-sions results in significantly greater
propensity to terminate poorly performingCEOs, while in countries
that already have strong investor protections, cross-listing does
not change the CEO turnover to performance sensitivity. Further,the
results show that in weak investor protection countries, only
cross-listingsthat require the most stringent of U.S. investor
protections (exchange-listedcross-listings) are associated with
better governance outcomes. Therefore, theresults provide support
for the second and third main predictions of the
bondinghypothesis.
C. Sensitivity of CEO Turnover to Performance in the
Pre-Cross-Listing Period
The previous analysis establishes that cross-listed firms on
major U.S. ex-changes have outcomes that are consistent with
improved corporate gover-nance over non-cross-listed firms. In this
analysis, the comparison group isall non-cross-listed firms, which
includes firms that may never cross-list aswell as cross-listed
firms in their pre-cross-listing period. However, an alterna-tive
explanation for our findings is that only the better-governed firms
chooseto cross-list on a major U.S. exchange, which would drive our
results. Whileour previous tests employed time-varying
cross-listing indicator variables thatequal one only after the firm
cross-listed, we further examine this issue byperforming two
additional tests in which we examine the sensitivity of CEOturnover
to performance in the pre-cross-listing period.
In the first test, we examine how the sensitivity of CEO
turnover to firmperformance differs before and after cross-listing,
where we restrict the sam-ple to firms that will have a
cross-listing of the same type during our sampleperiod.20 The
advantage of this experiment is that the non-cross-listed
compari-son group consists of firms that will have a cross-listing
of a similar type duringthe sample period (that is, the
pre-cross-listing period of the cross-listed firms).If firms that
pursue exchange-traded cross-listings have better governanceprior
to cross-listing, then we would expect to find little difference
betweencross-listed firms and “to-be” cross-listed firms. Due to
space considerations,in this test we focus on the broad Civil
versus Common Law classificationsusing the lagged earnings ratio
performance measure, but other governance
20 We also examine the comparison groups together with indicator
variables for cross-listingtypes (i.e., Level 2/3, Level 1, and
Rule144a dummies together) and obtain similar results.
-
1918 The Journal of Finance
classifications and performance measures produce consistent
results.21 Model1 in Table IV shows that the sensitivity of
turnover to performance for exchange-traded cross-listings is
significantly larger than that for non-cross-listed firmsthat will
eventually cross-list on a major exchange in low shareholder
protec-tion countries. Therefore, the results suggest that the
difference in governanceoutcomes is driven by the
post-cross-listing period of cross-listed firms. Thecoefficient on
L2/3 also shows that these firms have higher absolute turnoverafter
cross-listing. For Level 1 and Rule 144a cross-listings, the
interactionsbetween cross-listing type and firm performance are not
significantly differentbetween cross-listed and non-cross-listed
firms.
The second way we test if the pre-cross-listing governance of
our samplefirms is driving our results is to exclude all
observations for cross-listed firmsfollowing the cross-listing year
and compare the interactions of “to-be” cross-listed firms’
performance to the full sample of non-cross-listed firms. If the
pre-cross-listing status of cross-listed firms is driving our
results, we might expectto see results in this analysis similar to
our full sample tests presented earlier.An advantage of this test
is that we are able to use the full sample of non-cross-listed
firms. The disadvantage is that there are relatively few
observations inthe prelisting period, which is likely to lower the
power of the test. Table V showsthat when the post-cross-listing
observations are excluded, we no longer findthat cross-listed firms
are more likely to terminate poorly performing CEOs.22
We also find in Table V that Before L1 ∗ Lagged Earnings Ratio
is positiveand significant, which suggests these firms see an
improvement from a verylow level of corporate governance, where
well-performing managers are morelikely to leave, to a low level
where there is no relation between turnover andperformance.
Overall, the results in Tables IV and V show that when compared
to theirpre-cross-listing status, cross-listing on a major U.S.
exchange results in sig-nificantly higher CEO turnover to firm
performance sensitivity. Therefore, theresults indicate that it is
not the pre-cross-listing governance of firms that isdriving the
results. Taken together, the evidence in the preceding sections
sug-gests that in countries with poor investor protections,
cross-listing on a majorU.S. exchange increases the likelihood that
firms will have outcomes consistentwith improved corporate
governance systems.
D. London Listings
Listing in London is an often-cited alternative for firms that
do not wish tosubject themselves to the stringent listing
requirements in the United States.23
21 For example, using the excess returns measure, the
interaction between performance andLevel 2/3 is also statistically
significant (insignificant) in Civil (Common) Law, weak (strong)
anti-director rights, and low (high) anti-self-dealing
subsamples.
22 We also estimate a two-stage model for self-selection that
controls for the decision to list inthe first stage and find that
our results are robust to this specification.
23 See, for example, “London Calling” in Forbes Magazine May 8,
2006.
-
International Cross-Listing 1919
Tab
leIV
CE
OT
urn
over
and
Cro
ss-L
isti
ng:
Fir
ms
wit
ha
Cro
ss-L
isti
ng
ofth
eS
ame
Typ
eT
his
tabl
epr
esen
tsth
epr
obit
esti
mat
esof
the
rela
tion
ship
betw
een
the
prob
abil
ity
ofC
EO
turn
over
and
firm
perf
orm
ance
for
firm
sth
ath
ave
orw
ill
hav
ea
sim
ilar
type
ofA
DR
prog
ram
duri
ng
the
sam
ple
peri
od.T
he
sam
ple
inco
lum
ns
1–2
isli
mit
edto
firm
sth
ath
ave
orw
illh
ave
list
edon
am
ajor
U.S
.exc
han
gedu
rin
gou
rsa
mpl
epe
riod
.Sim
ilar
ly,t
he
sam
ples
inco
lum
ns
3–4
and
5–6
rest
rict
the
sam
ple
tofi
rms
that
list
onth
eO
TC
mar
ket
and
issu
epr
ivat
epl
acem
ents
via
Ru
le14
4aan
ytim
edu
rin
gou
rsa
mpl
epe
riod
,res
pect
ivel
y.L
agge
dE
arn
ings
Rat
iois
the
1-ye
arla
gged
rati
oof
earn
ings
befo
rein
tere
stan
dta
xes
toto
tal
asse
ts.
Th
eC
ivil
Law
sam
ple
incl
ude
sfi
rms
loca
ted
inco
un
trie
sw
ith
aF
ren
ch,
Ger
man
,or
Sca
ndi
nav
ian
lega
lsy
stem
.Th
eC
omm
onL
awsa
mpl
ere
fers
tofi
rms
loca
ted
inco
un
trie
sw
ith
the
En
glis
hle
galo
rigi
n.T
his
clas
sifi
cati
onof
lega
lreg
imes
isob
tain
edfr
omD
jan
kov
etal
.(20
05).
Lev
el2/
3du
mm
yis
1if
the
firm
has
aL
evel
2or
Lev
el3
AD
Rpr
ogra
m,0
oth
erw
ise.
Lev
el1
dum
my
is1
ifth
efi
rmh
asa
Lev
el1
AD
Rpr
ogra
m,0
oth
erw
ise.
Ru
le14
4Adu
mm
yis
1if
the
firm
has
aR
ule
144A
issu
ance
,0ot
her
wis
e.L
ogA
sset
sis
the
nat
ura
llo
gof
tota
las
sets
mea
sure
din
mil
lion
$U.S
.Th
eco
nti
nu
ous
vari
able
sar
ew
inso
rize
dat
the
1%le
velf
orea
chco
un
try.
Th
ein
tera
ctio
nef
fect
isde
fin
edas
the
chan
gein
the
pred
icte
dpr
obab
ilit
yof
CE
Otu
rnov
erfo
ra
chan
gein
both
the
firm
perf
orm
ance
and
the
resp
ecti
vecr
oss-
list
eddu
mm
yu
sin
gth
em
eth
odol
ogy
ofN
orto
n,W
ang,
and
Ai
(200
4).T
he
z-st
atis
tics
appe
arin
pare
nth
eses
belo
wpa
ram
eter
esti
mat
es.R
obu
stst
anda
rder
rors
are
esti
mat
edu
sin
gth
eR
oger
sm
eth
odof
clu
ster
ing
byfi
rm.∗
∗∗,∗
∗ ,an
d∗
indi
cate
sign
ific
ance
atth
e1%
,5%
,an
d10
%le
vel,
resp
ecti
vely
.
Exc
han
ge-T
rade
dA
DR
sO
TC
-Tra
ded
AD
Rs
Pri
vate
Pla
cem
ents
Civ
ilL
awC
omm
onL
awC
ivil
Law
Com
mon
Law
Civ
ilL
awC
omm
onL
awV
aria
ble
(1)
(2)
(3)
(4)
(5)
(6)
Log
asse
ts−0
.011
0.02
8∗−0
.006
0.01
30.
148∗
∗∗0.
122∗
∗[−
0.48
1][1
.770
][0
.299
][0
.769
][3
.028
][2
.299
]L
agge
dea
rnin
gsra
tio
1.40
5−0
.134
−0.0
27−0
.122
−0.8
88−2
.495
[0.9
64]
[−0.
509]
[−0.
837]
[−0.
829]
[−0.
705]
[−0.
668]
L2/
30.
631∗
∗∗0.
021
––
––
[3.2
19]
[0.2
03]
L2/
3∗l
agge
dea
rnin
gsra
tio
−2.8
52∗
0.02
1–
––
–[−
1.90
3][0
.067
]L
1–
–0.
194
−0.0
63–
–[1
.547
][−
0.56
7]L
1∗l
agge
dea
rnin
gsra
tio
––
−0.2
660.
094
––
[−0.
449]
[0.2
50]
(con
tin
ued
)
-
1920 The Journal of Finance
Tab
leIV
—C
onti
nu
ed
Exc
han
ge-T
rade
dA
DR
sO
TC
-Tra
ded
AD
Rs
Pri
vate
Pla
cem
ents
Civ
ilL
awC
omm
onL
awC
ivil
Law
Com
mon
Law
Civ
ilL
awC
omm
onL
awV
aria
ble
(1)
(2)
(3)
(4)
(5)
(6)
R14
4A–
––
–0.
196
−0.7
14[0
.899
][−
1.57
5]R
144A
∗lag
ged
earn
ings
rati
o–
––
–0.
472
3.14
4[0
.303
][0
.796
]C
onst
ant
0.47
3−6
.433
∗∗∗
−11.
514∗
∗∗−6
.577
∗∗∗
−0.7
69−0
.320
[0.6
35]
[−10
.192
][−
40.6
28]
[−18
.563
][−
0.82
8][−
0.42
7]C
oun
try
effe
cts
Yes
Yes
Yes
Yes
Yes
Yes
indu
stry
effe
cts
(tw
o-di
git
SIC
)Ye
sYe
sYe
sYe
sYe
sYe
sYe
aref
fect
sYe
sYe
sYe
sYe
sYe
sYe
s
Obs
erva
tion
s73
71,
567
1,49
61,
656
544
367
Pse
udo
-R2
0.14
70.
061
0.05
90.
067
0.24
20.
171
Mea
nin
tera
ctio
nef
fect
−0.5
65∗
0.00
5−0
.092
0.02
30.
046
0.81
9[−
1.70
2][0
.061
][−
0.63
3][0
.266
][0
.084
][0
.748
]
-
International Cross-Listing 1921T
able
VC
EO
Tu
rnov
erP
rior
toC
ross
-Lis
tin
gT
his
tabl
epr
esen
tsth
epr
obit
esti
mat
esof
the
rela
tion
ship
betw
een
the
prob
abil
ity
ofC
EO
turn
over
and
firm
perf
orm
ance
un
der
vari
ous
lega
len
viro
nm
ents
prio
rto
cros
s-li
stin
g.O
bser
vati
ons
for
cros
s-li
sted
firm
sfr
omth
ecr
oss-
list
ing
year
onar
eex
clu
ded.
Lag
ged
Ear
nin
gsR
atio
isth
e1-
year
lagg
edra
tio
ofea
rnin
gsbe
fore
inte
rest
and
taxe
sto
tota
lass
ets.
Civ
ilL
awsa
mpl
ein
clu
des
firm
slo
cate
din
cou
ntr
ies
wit
ha
Fre
nch
,Ger
man
,or
Sca
ndi
nav
ian
lega
lsys
tem
.Th
eC
omm
onL
awsa
mpl
ere
fers
tofi
rms
loca
ted
inco
un
trie
sw
ith
the
En
glis
hle
galo
rigi
n.A
nti
-dir
ecto
rri
ghts
inde
xm
easu
res
the
degr
eeof
min
orit
ysh
areh
olde
rpr
otec
tion
.A
nti
-sel
f-de
alin
gis
anin
dex
ofth
est
ren
gth
ofm
inor
ity
shar
ehol
der
prot
ecti
onag
ain
stse
lf-d
eali
ng
byth
eco
ntr
olli
ng
shar
ehol
der.
All
thes
eco
un
try-
leve
lin
dice
sar
eob
tain
edfr
omD
jan
kov
etal
.(20
05).
Th
em
edia
ns
of3.
5fo
ran
ti-d
irec
tor
righ
tsan
d0.
42fo
ran
ti-s
elf-
deal
ing
inde
xu
sed
inD
jan
kov
etal
.(2
005)
are
use
dto
grou
pfi
rms
into
hig
hve
rsu
slo
win
vest
orpr
otec
tion
regi
mes
(low
erth
anor
equ
alto
the
med
ian
refe
rsto
low
gove
rnan
cesu
bsam
ples
).B
efor
eL
evel
2/3
dum
my
is1
for
the
peri
odbe
fore
the
firm
cros
s-li
sts
ona
maj
orst
ock
exch
ange
inth
eU
.S.,
0ot
her
wis
e.B
efor
eL
evel
1du
mm
yis
1fo
rth
epe
riod
befo
reth
efi
rmcr
oss-
list
sin
the
OT
Cm
arke
tsin
the
U.S
.,0
oth
erw
ise.
Bef
ore
Ru
le14
4Adu
mm
yis
1fo
rth
epe
riod
befo
reth
efi
rmh
asa
priv
ate
plac
emen
tin
the
U.S
.,0
oth
erw
ise.
Log
Ass
ets
isth
en
atu
rall
ogof
tota
lass
ets
mea
sure
din
mil
lion
$U.S
.Th
eco
nti
nu
ous
vari
able
sar
ew
inso
rize
dat
the
1%le
velf
orea
chco
un
try.
Th
ein
tera
ctio
nef
fect
isde
fin
edas
the
chan
gein
the
pred
icte
dpr
obab
ilit
yof
CE
Otu
rnov
erfo
ra
chan
gein
both
the
firm
perf
orm
ance
and
the
resp
ecti
vecr
oss-
list
eddu
mm
yu
sin
gth
em
eth
odol
ogy
ofN
orto
n,W
ang,
and
Ai(
2004
).T
he
z-st
atis
tics
appe
arin
pare
nth
eses
belo
wpa
ram
eter
esti
mat
es.R
obu
stst
anda
rder
rors
are
esti
mat
edu
sin
gth
eR
oger
sm
eth
odof
clu
ster
ing
byfi
rm.∗
∗∗,∗
∗ ,an
d∗
indi
cate
sign
ific
ance
atth
e1%
,5%
,an
d10
%le
vel,
resp
ecti
vely
.
Civ
ilC
omm
onL
owA
nti
-H
igh
An
ti-
Low
An
ti-
Hig
hA
nti
-L
awL
awD
irec
tor
Rig
hts
Dir
ecto
rR
igh
tsS
elf-
Dea
lin
gS
elf-
Dea
lin
gV
aria
ble
(1)
(2)
(3)
(4)
(5)
(6)
Log
asse
ts0.
022∗
∗∗0.
029∗
∗∗0.
020∗
∗∗0.
029∗
∗∗0.
020∗
∗∗0.
025∗
∗∗[7
.076
][7
.211
][6
.068
][7
.890
][4
.084
][8
.877
]L
agge
dea
rnin
gsra
tio
−0.0
01∗∗
−0.2
63∗∗
∗−0
.001
∗∗∗
−0.2
73∗∗
∗−0
.005
−0.0
01∗∗
[−2.
260]
[−6.
550]
[−2.
912]
[−6.
978]
[−1.
065]
[−2.
334]
Bef
ore
L2/
3−0
.310
0.04
2−0
.342
0.03
8−0
.262
0.02
7[−
1.58
0][0
.478
][−
1.47
1][0
.453
][−
1.32
0][0
.319
]B
efor
eL
2/3
∗lag
ged
earn
ings
rati
o2.
499
0.01
42.
472
0.02
02.
978
−0.1
43[1
.444
][0
.054
][1
.254
][0
.077
][1
.538
][−
0.56
5]B
efor
eL
1−0
.181
−0.0
42−0
.270
∗∗−0
.030
−0.1
31−0
.055
[−1.
509]
[−0.
626]
[−1.
994]
[−0.
471]
[−0.
862]
[−0.
883]
Bef
ore
L1
∗lag
ged
earn
ings
rati
o0.
048∗
∗0.
114
0.05
5∗∗∗
0.10
80.
058∗
∗∗−0
.117
[2.5
36]
[0.8
17]
[2.8
42]
[0.7
81]
[2.7
55]
[−0.
864]
Bef
ore
R14
4A−0
.281
∗0.
352
−0.3
540.
096
−0.4
93−0
.224
[−1.
835]
[1.1
25]
[−1.
558]
[0.5
77]
[−1.
517]
[−0.
940]
(con
tin
ued
)
-
1922 The Journal of Finance
Tab
leV
—C
onti
nu
ed
Civ
ilC
omm
onL
owA
nti
-H
igh
An
ti-
Low
An
ti-
Hig
hA
nti
-L
awL
awD
irec
tor
Rig
hts
Dir
ecto
rR
igh
tsS
elf-
Dea
lin
gS
elf-
Dea
lin
gV
aria
ble
(1)
(2)
(3)
(4)
(5)
(6)
Bef
ore
R14
4A∗l
agge
dea
rnin
gsra
tio
−0.0
01−1
.171
1.50
50.
270∗
∗∗−0
.516
2.28
2[−
1.04
6][−
0.48
2][0
.694
][6
.895
][−
0.19
2][1
.077
]C
onst
ant
−0.9
36∗∗
−0.6
40∗
−0.8
89∗
−1.4
86∗∗
∗−1
.528
∗∗∗
−0.8
69∗
[−2.
048]
[−1.
676]
[−1.
935]
[−2.
822]
[−3.
415]
[−1.
909]
Cou
ntr
yef
fect
sYe
sYe
sYe
sYe
sYe
sYe
sIn
dust
ryef
fect
s(t
wo-
digi
tS
IC)
Yes
Yes
Yes
Yes
Yes
Yes
Year
effe
cts
Yes
Yes
Yes
Yes
Yes
Yes
Obs
erva
tion
s42
,137
23,4
0938
,556
26,9
9214
,869
50,6
62P
seu
do-R
20.
033
0.02
10.
034
0.02
20.
030
0.02
8
Mea
nin
tera
ctio
nef
fect
for
befo
re0.
494
0.00
10.
466
0.00
20.
660
−0.0
34L
23∗l
agge
dea
rnin
gsra
tio
[1.4
21]
[0.0
13]
[1.2
68]
[0.0
38]
[1.4
06]
[−0.
567]
Mea
nin
tera
ctio
nef
fect
for
befo
re0.
010∗
∗∗0.
028
0.01
0∗∗∗
0.02
60.
012∗
∗∗−0
.026
L1
∗lag
ged
earn
ings
rati
o[2
.916
][0
.896
][3
.509
][0
.842
][3
.100
][−
0.87
1]
Mea
nin
tera
ctio
nef
fect
for
befo
re−0
.000
1−0
.353
0.26
50.
063∗
∗∗−0
.068
0.47
9R
144
∗lag
ged
earn
ings
rati
o[−
0.29
6][−
0.49
0][0
.718
][5
.712
][−
0.18
7][1
.134
]
-
International Cross-Listing 1923
Consistent with the hypothesis that a London listing does not
convey the samegovernance commitment as a U.S. listing, Doidge et
al. (2006) find that firmswith high private benefits of control are
more likely to list in London ratherthan on a major U.S. exchange.
Baker et al. (2002) also find that monitoringby financial analysts
is lower in London Stock Exchange (LSE) cross-listingsversus U.S.
cross-listings and Seetharaman, Ferdinand, and Lynn (2002) findthat
U.K. auditors charge higher fees for client firms that are listed
in theUnited States to compensate for the higher risk of
litigation.
To test if cross-listing on the LSE also results in improved
governance out-comes, we gather a sample of 688 firm-year
observations (145 firms from 27countries) that trade in London and
rerun our models to examine the sensitiv-ity of CEO turnover to
performance between firms that cross-listed in Londonand those
firms that did not.24 Since we are examining London listings,
weexclude observations from the United Kingdom. Table VI reports
that the in-teraction between LSE Listing and firm performance is
not significant in anyspecification. In addition, the results are
robust when U.S. firms are included.In unreported tests, we also
find that the difference between listing on a majorU.S. exchange
and on the LSE is statistically significant at the 5% level.
There-fore, unlike a major U.S. exchange cross-listing, we do not
find that listing onthe LSE is associated with better corporate
governance outcomes.
E. Labor Market Effects: The Departing and Replacement CEOs’
Work History
A potential explanation for our results is that cross-listing
also changes thelabor market for top management. For example, one
possibility is that cross-listing may induce top management to
leave their jobs to pursue employment inthe United States where
they are likely to receive higher compensation. Anotherpossibility
is that by cross-listing, firms are able to tap a more
internationalpool of top management candidates and therefore are
more likely to fire poorlyperforming managers. To investigate these
effects, we examine a subsample of150 cross-listed firms that
experienced a CEO turnover after cross-listing. Dueto the labor
intensiveness of this investigation, these CEO turnover events
arerandomly drawn from sample firms, with an equal number for Level
2/3 ADRs,Level 1 ADRs, and private placements.
We first search news articles as well as reexamine the CEO data
on theWorldScope database to find out where the departing CEO found
employment.We are able to find the post-turnover employment
information for 110 CEOs.Of these, only eight went to a U.S. firm,
and only five of these were from amajor exchange cross-listed firm
(Level 2/3). Further, only one of the five leftvoluntarily.
Therefore, this evidence does not appear to be consistent with
alarge number of CEOs leaving to take higher paid jobs in the
United States.
We next investigate the newly appointed CEOs’ previous work
experienceto see if cross-listed firms may be more likely to
terminate poorly performingmanagement when the potential quality of
the labor market is higher. For t