International diversification and conference calls

This article was downloaded by: [University of North Carolina]On: 13 November 2014, At: 14:31Publisher: RoutledgeInforma Ltd Registered in England and Wales Registered Number: 1072954 Registeredoffice: Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK

Asia-Pacific Journal of Accounting &EconomicsPublication details, including instructions for authors andsubscription information:http://www.tandfonline.com/loi/raae20

International diversification andconference callsChen-Lung Chin a , Yu-Ju Chen b & Jia-Wen Liang aa Department of Accounting , National Chengchi University ,Taipei City , Taiwanb Department of Business Education , National ChanghuaUniversity of Education , Changhua , TaiwanPublished online: 02 Oct 2012.

To cite this article: Chen-Lung Chin , Yu-Ju Chen & Jia-Wen Liang (2013) Internationaldiversification and conference calls, Asia-Pacific Journal of Accounting & Economics, 20:3, 297-314,DOI: 10.1080/16081625.2012.719856

To link to this article: http://dx.doi.org/10.1080/16081625.2012.719856

PLEASE SCROLL DOWN FOR ARTICLE

Taylor & Francis makes every effort to ensure the accuracy of all the information (the“Content”) contained in the publications on our platform. However, Taylor & Francis,our agents, and our licensors make no representations or warranties whatsoever as tothe accuracy, completeness, or suitability for any purpose of the Content. Any opinionsand views expressed in this publication are the opinions and views of the authors,and are not the views of or endorsed by Taylor & Francis. The accuracy of the Contentshould not be relied upon and should be independently verified with primary sourcesof information. Taylor and Francis shall not be liable for any losses, actions, claims,proceedings, demands, costs, expenses, damages, and other liabilities whatsoever orhowsoever caused arising directly or indirectly in connection with, in relation to or arisingout of the use of the Content.

This article may be used for research, teaching, and private study purposes. Anysubstantial or systematic reproduction, redistribution, reselling, loan, sub-licensing,systematic supply, or distribution in any form to anyone is expressly forbidden. Terms &Conditions of access and use can be found at http://www.tandfonline.com/page/terms-and-conditions

http://www.tandfonline.com/loi/raae20

http://www.tandfonline.com/action/showCitFormats?doi=10.1080/16081625.2012.719856

http://dx.doi.org/10.1080/16081625.2012.719856

http://www.tandfonline.com/page/terms-and-conditions

http://www.tandfonline.com/page/terms-and-conditions

International diversification and conference calls

Chen-Lung China*, Yu-Ju Chenb and Jia-Wen Lianga

aDepartment of Accounting, National Chengchi University, Taipei City, Taiwan; bDepartment ofBusiness Education, National Changhua University of Education, Changhua, Taiwan

(final version received 12 January 2012)

This paper explores the impact of international diversification on the decision to holdconference calls. In addition, we examine the mitigating roles of legal environmentand agency problems in holding conference calls. Using a sample drawn from Tai-wan, we find that greater corporate internationalization (INT) is associated with ahigher likelihood of holding conference calls. We also find that firms are more likelyto conduct conference calls when companies invest in a higher proportion of com-mon law countries. We also find that the positive association between the likelihoodof holding conference calls and corporate INT is less pronounced when the controldivergence of controlling owners increases.

Keywords: internationalization; conference calls; legal system; control rights; volun-tary disclosure

JEL Codes: F23, G14, G32

1. Introduction

A firm’s international diversification is one potentially important strategy to follow inpursuit of sales and earnings growth. Over the past decade, a growing number of listedfirms in developed and emerging markets have substantially expanded their operationsabroad.1 As companies become more internationally diversified, their operations natu-rally become more multifaceted and increase the complexity of information processingfor investor (e.g. Thomas 1999), managers (Birkinshaw et al. 2001), and analysts(Tihanyi and Thomas 2005). Despite the increased information asymmetry induced byINT, there is surprisingly little evidence on the effect of international diversification onvoluntary disclosure, particularly in the context of emerging market. This paper exploresthe association between the extent of a firm’s international diversification and voluntarydisclosure, as measured by conference calls and examine whether firms investing in ahigher proportion of common law countries are more likely to hold conference calls.Next, we examine if the association between the likelihood of holding conference callsand INT is less pronounced for firms with worse ownership structure.

The first question to be addressed in this paper is whether firms are more likely tohold conference calls and hold them more frequently when firms expand more of theiroperations abroad. Expansion into international markets increases organization complex-ity and, in turn, the complexity of information processing for investors (Thomas 2000)and analysts (Duru and Reeb 2002, Tihanyi and Thomas 2005, Herrmann et al. 2008).

*Corresponding author. Email: [email protected]

Asia-Pacific Journal of Accounting & Economics, 2013Vol. 20, No. 3, 297–314, http://dx.doi.org/10.1080/16081625.2012.719856

� 2012 City University of Hong Kong and National Taiwan University

Dow

nloa

ded

by [

Uni

vers

ity o

f N

orth

Car

olin

a] a

t 14:

31 1

3 N

ovem

ber

2014

For example, relative to domestic earnings, foreign earnings encompasses countriesfrom around the world differing drastically in terms of economic conditions, competi-tive forces, political stability, growth opportunities, governmental regulations, etc.(Thomas 2000). Therefore, the levels of information asymmetries among investors andmanagers increase with international diversification. Prior studies document the positiverelation between information asymmetry and cost of capital (e.g. Leuz and Verrecchiain press). Given that conference calls would reduce information asymmetries (Brownet al. 2007), we expect that INT is positively associated with the likelihood of holdingconference calls.

Next, we examine if diversified firms are more likely to hold conference calls wheninvesting in a higher proportion of common law countries. Prior studies (e.g. La Portaet al. 1997, 2000) show that the common law countries (e.g. the USA and the UK)offer the best investor protection while the French-based civil code countries offer theleast protection. Better law protection limits insiders’ ability to acquire private controlbenefits and reduces their incentives to mask a firm’s performance and manage earnings(Leuz et al. 2003). More importantly, in market-oriented common law countries, theretypically is a larger and more diverse base of shareholders and information asymmetryis more efficiently resolved though public disclosure. Hence, there is a larger demandfor accounting quality. In contrast, in planning-oriented code law countries, informationasymmetry is more likely to be resolved by “insider” communication with stakeholderrepresentatives (Ball et al. 2000, 2003). Therefore, we expect that the pervasiveness ofholding conference calls increases when diversified firms invest in a higher proportionof common law countries.

Finally, we examine the role of agency problems in the decision to hold conferencecalls. The primary agency problem in most countries outside the USA is reflected in theconflict of interest between controlling owners and minority owners (La Porta et al.1999, Haw et al. 2004), and the former generally possess control rights in excess ofcash flow rights via stock pyramids and cross-ownership structures. When control diver-gence increases, the controlling owner’s ability and incentive to expropriate minorityinvestors increases also.2 Insiders (such as controlling owners or managers) have incen-tives to conceal private control benefits from outsiders because, if these benefits aredetected, outsiders will take disciplinary actions against them (Leuz et al. 2003). In thecontext of INT, organizational complexity creates a higher level of informational asym-metry and managers may exploit this additional information asymmetry to expropriateminority investors.3 Therefore, controlling owners and managers of multinational firmstend to be less likely to voluntarily disclose private information in an attempt to masktrue firm performance and to conceal their private control benefits from outsiders (Leuzet al. 2003, Haw et al. 2004). As a consequence, we hypothesize that ownership struc-ture with greater control divergence reduces the incentive for managers in diversifiedfirms to hold conference calls.

We focus our analysis on the efficacy of using conference calls as a medium forvoluntary disclosure to the public for the following reasons. First, conference calls havebecome an increasingly common and important means of voluntary disclosure in bothdeveloping and developed countries (Bushee et al. 2003, Chin et al. 2005). Second,one of the purported benefits of conference calls to managers is the flexibility it pro-vides to respond to analysts’ (Frankel et al. 1999), particularly in the context of interna-tional diversification with increased information complexity. Third, conference callsgenerally convey material information to the market as evidenced by significantincreases in trading volume and return volatility during the call periods (Frankel et al.

298 C.-L. Chin et al.

Dow

nloa

ded

by [

Uni

vers

ity o

f N

orth

Car

olin

a] a

t 14:

31 1

3 N

ovem

ber

2014

1999, Bushee et al. 2003). Finally, Brown et al. (2007) show that there is substantialvariation in conference call activities, and this high level of variation allows for rela-tively powerful tests of the association between disclosures and information asymmetry,particularly in the context of INT.

The motivations for using Taiwanese data stem from the following reasons. First,INT is an indispensable strategy for firm growth in the emerging countries. Taiwan ischaracterized by its heavy reliance on exports, small stock markets, and high owner-ship concentration. Thus, the effect of INT on conference calls in Taiwanese firmsshould be more pronounced. Second, the relationship between ownership structureand financial reporting has been less studied in a concentrated ownership context likethat in Taiwan. High concentration of ownership nullifies the principal–agent problembetween owners and managers as well as the related role of accounting-based mana-gerial contracts. It is interesting to see how ownership structure interacts with theextent of conference calls, using Taiwanese data. Third, prior work focuses mainly onthe effects of legal protection on a firm’s earnings quality and investigated how thelegal protection in different regimes affects a firm’s earnings quality. Unlike priorstudies, this paper examines how voluntary disclosure (i.e. conference calls) of a mul-tinational firm is affected by a proportion of foreign investees with better legal protec-tion environment (i.e. common law countries). We believe that using Taiwanese firmsis appropriate to examine whether foreign legal systems have an impact on the deci-sion to hold conference calls.

Our evidence supports the prediction that in the context of an emerging market, Tai-wan, international diversification is positively associated with conference calls, evencontrolling for previously identified determinants of conference calls. Second, wheninvesting in a higher proportion of common law foreign firms, diversified firms aremore likely to hold conference calls. Third, diversified firms with severe agency prob-lem, measured by the control divergence, are less likely to conduct conference calls andconduct them less frequently than firms with less agency problems.

This paper contributes to several streams of the literature. First, our study contrib-utes to the literature on INT and conference calls. To the best of our knowledge, ourpaper is the first attempt to explore the effect of international diversification on confer-ence call likelihood and frequency. Second, we contribute to the literature on INT andcorporate governance by documenting the effects of the controlling owners’ controldivergence of multinational firms on conference calls likelihood and frequency. Thefundamental agency problem in most countries stems from the conflict of interestbetween controlling owner and minority. In combination with the fact that multinationalfirms possess a higher level of information asymmetry, such agency conflict might havean adverse effect on conference calls likelihood and frequency. Hence, our paper pro-vides further insights into the effect of diversified firm’s ownership structure on confer-ence calls. Finally, we contribute to the literature on the association of INT anddisclosures. Unlike prior studies (e.g. Ball et al. 2000, 2003), which explore the differ-ence in properties of accounting earnings between common law and code law countries,our paper investigates whether firms are more likely to hold conference calls wheninvesting in a higher proportion of common law countries.

The rest of this paper is organized as follows. Section 2 introduces the corporateownership structure in Taiwan. Section 3 reviews the literature and presents the hypoth-eses. Section 4 describes our data, sample selection procedure, and research design.Section 5 presents the empirical results, followed by a conclusion in Section 6.

Asia-Pacific Journal of Accounting & Economics 299

Dow

nloa

ded

by [

Uni

vers

ity o

f N

orth

Car

olin

a] a

t 14:

31 1

3 N

ovem

ber

2014

2. Corporate ownership structure in Taiwan

The corporate governance system in East Asia, including Taiwan, is typified by a highconcentration of ownership, family-controlled business groups, and a relative lack ofmajor institutional investors (Claessens et al. 2000). On average, various family groupscontrol 78% of listed companies on the Taiwan stock exchange (TSE). These familiescontrol the lion’s share of the financial resources available through various means andexert substantial control over board decisions and agendas in stockholders’ meetings.

The Taiwanese listed companies are required to file with the securities regulatorsand announce to the public the class and numbers of the shares held or transferred byits directors, supervisors, managers, and shareholders holding more than 10% of thetotal shares of the company.4 Thus, in attempt to reduce other investors’ perception ofdepriving private control benefits at the expense of minority shareholders, controllingowners of Taiwanese listed firms tend to have an incentive to take a variety of measuresto avoid meeting the requirements of disclosing shareholdings. For example, the con-trolling owner may establish a nominal investment company funded by the listed com-pany they control, and then invest in stock issued by those companies using theaccounts of the nominal investment company (what we call cross-shareholding) (Yehet al. 2001).

Another characteristic of the Taiwanese environment is the lack of major institu-tional investor interest in corporate governance. While the role of institutional investorsin corporate governance is receiving increasing attention around the world, institutionalinvestors in Taiwan are not able to play an effective role in corporate governance. Indi-vidual investors constitute 90% of trades, with the remaining trades originating fromcorporate or institutional investors. Since there are regulatory limits as to the percentageof capital that banks and insurance companies can have in public companies, ownershipby institutional investors constitutes only a minor portion of total ownership, approxi-mately 5%. Such market discipline tools as takeovers and hostile bids are not effectivemechanisms for corporate control in Taiwan, primarily because of the predominance ofconcentrated family ownership.

3. Literature and hypotheses

Earlier studies explore the determinants of conference calls. Frankel et al. (1999) findthat firms that hold conference calls tend to be larger and more profitable. Besides,these firms tend to have a larger analyst following and lower accounting quality (Tasker1998). Brown et al. (2007) find that conference calls would reduce information asym-metry among investors, and thus lead to a reduction in cost of capital.5 While confer-ence calls would reduce information asymmetry, few studies address the effect of INTon the decision to hold conference calls. This paper extends the prior literature byexamining the association between the decision to hold conference call and the degreeof INT.

It is well documented that expansion into international markets increases the overallorganizational complexity, and in turn complexity of information processing for investors(Thomas 2000) and analysts (Duru and Reeb 2002, Tihanyi and Thomas 2005, Herrmannet al. 2008). In addition, with increased geographical dispersion of firm assets, corporateinternational diversification also leads to an increase in managers’ discretion over operat-ing decision. Kogut (1983) argues that expansion into international markets increasesfirms’ operational flexibility and allows firms to change value by exploiting the increased


Dow

nloa

ded

by [

Uni

vers

ity o

f N

orth

Car

olin

a] a

t 14:

31 1

3 N

ovem

ber

2014

uncertainty of the international environment.6 Thus, managers enjoy more discretion as afirm increases operations abroad, which in turn exacerbate information asymmetrybetween managers and investors. With increased geographical dispersion of firm assets, itis presumably more difficult for investors or even analysts to carefully scrutinize the firm’searnings reports and make accurate assessment of foreign operations. Thomas (2000)shows that investors underestimate the persistence of foreign earnings. Tihanyi and Tho-mas (2005) and Duru and Reeb (2002) further find that international diversification leadsto less accurate analyst earnings forecasts. Thus, the degree of information asymmetryincreases with the extent of corporate international diversification.

One of the more often cited reasons why information asymmetry is particularlyimportant to firms is because it increases cost of equity (e.g. Leuz and Verrecchia inpress). The disclosure literature also indicates that there are three types of capital marketeffects for firms that make extensive disclosures: improved stock liquidity (e.g. Kimand Verrecchia 1994, Leuz and Verrecchia in press), reduced cost of capital (e.g. Boto-san and Plumlee 2002), and increased information intermediation (e.g. Healy andPalepu 2001).

Extant empirical studies document that higher levels of disclosure might reduceinformation asymmetries (Percy 2000, Lobo and Zhou 2001, Gu and Li 2003, Leuz andVerrecchia in press). A recent study by Brown et al. (2007) further shows that voluntarydisclosure, in particular conference calls, reduces information symmetries among inves-tors. Therefore, we predict that managers of multinational firms are more likely to con-duct conference calls in attempt to mitigate this additional level of informationasymmetry and in turn capital market effects, than otherwise would be the case if thefirm were not diversified. These discussions lead to the first testable hypothesis:

H1: Firms are more likely to conduct conference calls when expanding more operationabroad.

An important difference between common law and code law countries is the manner ofresolving information asymmetry between managers and potential users of accountinginformation. In market-oriented common law countries, there typically is a larger andmore diverse base of shareholders, and information asymmetry is more efficientlyresolved through public disclosure. On the other hand, in planning-oriented code lawcountries, information asymmetry is more likely to be resolved by “insider” communi-cation with stakeholders, so there is a lower demand for high-quality public financialreporting and disclosure (Ball et al. 2000, 2003).

Prior studies show that strong shareholder protection in the marketplace shouldattenuate management opportunism (Jensen and Meckling 1976) and in turn reducetheir incentives to mask private control benefits by manipulating earnings (Jensen andMeckling 1976, Hung 2001, Leuz et al. 2003). Stated differently, managers are lesslikely to behave opportunistically and manipulate earnings in common law countrieswith strong protection than in code law countries with weak protection.7 Thus, manag-ers tend to hold less conference call to hide the expropriation behavior when they investmore in code law countries. We expect that firms are more likely to hold conferencecalls when investing in a higher proportion of common law countries.

However, an alternative view is that the pervasiveness of holding conference callsincreases when firms invest more in code law countries. As mentioned earlier,international diversification leads to a higher degree of information asymmetry and theadverse capital market effects (e.g. increased cost of capital, etc.). Given that conference


Dow

nloa

ded

by [

Uni

vers

ity o

f N

orth

Car

olin

a] a

t 14:

31 1

3 N

ovem

ber

2014

calls provide a unique two-way form of corporate communication (Brown et al. 2007),the firms are more likely to hold conference calls to effectively reduce informationasymmetry when they invest more in code law countries. To the extent that investing incode law countries positively affects the diversified firms’ incentives to disclose, itwould bias our results against the hypothesis that firms are likely to hold conferencecalls when investing in a higher proportion of common law countries. The followinghypothesis summarizes our expectations.

H2: Firms are more likely to conduct conference calls when companies invest in a higherproportion of common law countries.

As mentioned previously, ownership concentration exits in many developed and devel-oping countries (La Porta et al. 1999, Haw et al. 2004). The fundamental agency prob-lem then stems from the conflict of interest between minority shareholders andcontrolling owners. The latter typically exercise control power in excess of their cashflow rights via stock pyramids and cross-ownership structures. Since a smaller fractionof the firm’s cash flow rights relative to voting rights fails to align controlling ownerincentive with those of minority shareholders, controlling owners thus possess incen-tives and abilities to extract private control benefits (expropriation of the firm’s assetsand opportunities) that are not shared by minority shareholders in proportion to theshares owned. International diversification creates additional organizational complexityand information asymmetry, which might exacerbate agency problem. Thus, managersof multinational firms may exploit this additional information asymmetry to derive pri-vate control benefits at the expense of minority shareholders.

Extracting private control benefits, if detected, is likely to invite external interven-tion by minority shareholders, analysts, stock exchanges, or regulators (Haw et al.2004). The desire to avoid external monitoring and loss of reputation induces insidersto mask or conceal their private control benefits by managing reported accounting earn-ings (Fan and Wong 2002, Leuz et al. 2003, Haw et al. 2004).8 Besides the mandatoryfinancial reports, corporate governance structure also has an impact on corporate volun-tary disclosure policy (Daniel et al. 2011). In the presence of control-cash flow rightsdivergence, managers of multinational firms are more reluctant to dissimilate moreinformation about the firm’s economic performance by holding conference calls. Fur-thermore, another characteristic of conference calls is that managers must answer ques-tions from hundreds of professional analysts and money managers simultaneously andon a face-to-face basis, which allow minority shareholders and other stakeholders imme-diate and broad access to information about the firm’s economic performance and pri-vate control benefits. We thus hypothesize that the mitigating role of poor agencyproblem in decision to hold conference calls is more pronounced for firms with greaterinternational diversification. Hence, we propose the third hypothesis as follows:

H3: The positive association between conference calls likelihood and corporate INT is lesspronounced when the control divergence of controlling owners increases.

4. Research design

4.1. Sample

Our sample includes Taiwanese publicly listed firms. The sample years ranged from2000 to 2005. We collected dates and times of conference calls held during the period


Dow

nloa

ded

by [

Uni

vers

ity o

f N

orth

Car

olin

a] a

t 14:

31 1

3 N

ovem

ber

2014

Table 1. Sample selection and descriptive statistics.

Panel A: sample selection

Analysis of conference calls

Number of company yearsduring 2000–2005 period(TEJ)

7733

Less: observations in theregulated industries (e.g.finance and banking)

(428)

Less: observation with foreigninvestment equal to zero

(616)

Less: observations without the following dataForeign investment (1992)Other financial data (1818)Ownership structure (234)

Number of company years in thefinal sample

2645

Firm-year with conference call 478Firm-year without conference

call2167

Total sample 2645

Panel B: descriptive statistics for firms with and without conference calls

Variablea

With and without conferencecalls (N= 2645)

MeandifferenceT-value2

MediandifferenceZ-value2

Conferencecall firm-year(N= 478)

Non-conferencecall firm-year(N= 2167)

Mean Median Mean Median

INT 0.73 0.49 0.66 0.49 3.182⁄⁄⁄ 2.509⁄⁄

LAW 0.49 0.05 0.48 0.05 1.338 0.279OS 0.05 0.02 0.05 0.01 0.909 3.953⁄⁄⁄

LOGASSET 9.90 9.75 9.55 9.52 10.952⁄⁄⁄ 10.374⁄⁄⁄

DEPLAN 0.94 1 0.70 1 16.122⁄⁄⁄ 10.817⁄⁄⁄

SG 0.25 0.21 0.10 0.06 6.002⁄⁄⁄ 10.904⁄⁄⁄

MB 1.98 1.75 1.30 1.05 11.403⁄⁄⁄ 15.462⁄⁄⁄

LDE 0.19 0.14 0.21 0.10 �0.379 2.155⁄⁄

MGTOWN 0.02 0.01 0.02 0.003 3.333⁄⁄⁄ 6.021⁄⁄⁄

ROA 0.11 0.10 0.03 0.04 15.903⁄⁄⁄ 16.291⁄⁄⁄

FOLLOW 7.47 1 0.92 0 11.269⁄⁄⁄ 17.399⁄⁄⁄

DA 0.02 0.01 �0.02 �0.02 8.066⁄⁄⁄ 10.599⁄⁄⁄

aCC_D: equal one if firms have ever provided conference calls during the year, and 0 otherwise. CC_F:the number of conference calls held by a firm in a year. INT: the entropy measure of the internationaldiversification. LAW: the ratio of foreign assets in common law countries to total foreign assets. OS:control rights- cash flow rights. LOGASSET: the logarithm of the total assets. DEPLAN: equal one if afirm raises any capital by bonds, common or preferred stock in the current or year following the yearthe conference calls were held, and 0 otherwise. MB: market value divided by book value. SG: (Sale-sit-Salesit-1)/Salesit-1. LDE: he ratio of long-term liability over stockholders equity. MGTOWN: thepercentage of shares held by the CEO. ROA: earnings divided by total assets. FOLLOW: the numberof analyst following the firm. DA: discretionary accruals based on Modified Jones models.⁄, ⁄⁄, ⁄⁄⁄, signify that the conference call and nonconference call distribution of the given variable aresignificantly different at the 1, 5, and 10%, respectively (one-tailed).


Dow

nloa

ded

by [

Uni

vers

ity o

f N

orth

Car

olin

a] a

t 14:

31 1

3 N

ovem

ber

2014

of 2000–2005 by using a proprietary database, “Information Winner Series,” compiledby the China Times. The corporate ownership structure data, including control rights,cash flow rights, and cross-shareholding information, were collected from the TaiwanEconomic Journal (TEJ) database, company prospectuses, and Business Groups in Tai-wan. The latter is a book published annually by the “China Credit Information Service,LTD.” The financial and stock data were collected from the TEJ database. Analysts’forecasts data are obtained from I/B/E/S database.

To enhance the information transparency of listed companies in Taiwan, TSErequires the listed firms to disclose information about their overseas investmentsthrough the Market Observation Post System.9 The information to be disclosed includesthe amount invested, foreign investee locations, and the profit or loss from these invest-ments. Further, all domestic and foreign investments must be reported in the firm’sannual report. Accordingly, foreign investment data are collected from the MarketObservation Post System and the annual reports.

Our sample firm-years meet the following requirements: First, corporate ownershipstructure data had to be available from the TEJ database, company prospectuses, or thebook Business Groups in Taiwan. The TEJ database was used to identify all publiclylisted firms for which complete accounting data was available, as needed by this study.Conference calls held by start-up firms for the purpose of raising capital are excludedfrom our sample.

Panel A of Table 1 summarizes the sample selection process. There are 7733 firm-years with shares traded in the TSE during our sample period. First, we exclude 428firm-year observations in the regulated industries (e.g. finance and banking) because oftheir dissimilar nature to other industries. We delete 616 firm-year observations withforeign investment equal to zero because the entropy measure of INT will become inva-lid in the regressions for these observations. We also delete 1992 firm-year observationsdue to missing foreign investment data and 1818 firm-year observations due to insuffi-cient financial data. Further, we delete 234 firm-year observations due to missing own-ership structure data. The selection procedure yields 2645 firm-year observations forour conference calls analyses. Panel A of Table 1 also shows that the 2645 firm-yearsare classified into two groups: firm-years with conference calls (N= 478) and firm-yearswithout conference calls (N = 2167). A firm-year for firms that held conference calls isdefined as any year in which a firm held at least one conference call. For example, if afirm held conference calls in 2000, 2001, and 2005, that firm would appear in the data-set as having three firm-years.

4.2. Variables

4.2.1. Internationalization (INT)

Prior studies have recommended a variety of measures to gage international diversifica-tion. Among them, the entropy approach has also been commonly used to measureinternational diversification, which is counted mostly by the regional sales (Hitt et al.1997). This measure has been found to enjoy more validity than some of the othermeasures (Chatterjee and Blocher 1992, Hoskisson et al. 1993).

We measure INT by entropy and it is calculated as follows:


Dow

nloa

ded

by [

Uni

vers

ity o

f N

orth

Car

olin

a] a

t 14:

31 1

3 N

ovem

ber

2014

INT ¼Xn

1

Pi ln1

Pi

� �;

where Pi is the assets of the ith foreign country divided by the firm’s total foreign assetsand n is the number of foreign countries.

4.2.2. Law protection (LAW)

La Porta et al. (1999) argued that countries with common law systems have betterprotections for investors than those with civil law systems. Therefore, based on theexistence of distinct categories of legal systems in different countries, we categorizecountries into common law and civil law countries, and then define the measure of legalprotection, LAW, as the ratio of foreign assets in common law countries to total foreignassets. The greater the ratio, presumably, the greater the protection that the sample firmshave for their foreign investments.

4.2.3. Ownership structure (OS)

To construct a proxy for ownership structure, we proceed in four steps. First, followingClaessens et al. (2000) and La Porta et al. (1999), we identified the ultimate owners viaa determination of each share of voting and cash flow rights.10 Second, following LaPorta et al. (1999), we assume that indirect voting rights were equal to what they calledthe link in the first level of the chain of shares held by firms or organizations that werein turn under the control of the ultimate owner. Third, with respect to cash flow rights,we again followed the Claessens et al. methodology. For example, if firm X owned20% of firm Y and firm Y owned 10% of firm Z, then firm X had the right to 2% ofthe cash flow of firm Z. Finally, to test hypothesis 3, we define ownership structure(OS) as the difference between the ultimate controlling owner’s voting rights and cashflow rights to proxy for corporate governance. That is: there will be more seriousagency problems arising from conflicts between the interests of controlling owners andminority shareholders as OS becomes larger.

4.3. Model

We use the following Probit (or Ordered Probit) regression to test our three hypotheses.

PrðCCÞ ¼ b0 þ b1INTþ b2LAWþ b3INT�OSþ b4LOGASSETþ b5DEPLAN

þ b6MBþ b7SGþ b8LDEþ b9MGTOWN þ b10ROAþ b11FOLLOW

þ b12DAþ e; ð1Þ

where CC is set equal either CC_D or CC_F. When Probit regression is used, the con-ference calls likelihood (CC_D) as a dummy variable is one if firms have ever providedconference calls during the year and zero otherwise. The conference calls frequency(CC_F) is coded with the number of conference calls held by a firm in a year whenOrdered Probit model is used to test our expectation. INT is the entropy measure of theinternational diversification. LAW is the ratio of foreign assets in common law countriesto total foreign assets; OS is defined as the difference between control rights and cashflow rights. We include firm size, measured by the logarithm of the total assets (LOG-


Dow

nloa

ded

by [

Uni

vers

ity o

f N

orth

Car

olin

a] a

t 14:

31 1

3 N

ovem

ber

2014

ASSET), because firm size is related to disclosure practices (Tasker 1998, Frankel et al.1999). We expect that large firms tend to have more frequent conference calls to havebetter communication with the public. Newly offering firms tend to host moreconference calls to lower their costs of capital (Tasker 1998, Frankel et al. 1999). Wethus construct the indicator variable (DEPLAN) set equal to one if a firm raises anycapital by bonds, common, or preferred stock in year t or year t+ 1 and 0 otherwise.

Following Tasker (1998) and Frankel et al. (1999), we include market-to-book ratio(MB) and sales growth (SG) in our model to control for low accounting quality. Weinclude long-term debt-equity ratios (LDE), measured by the ratio of long-term liabilityover stockholders equity, in our model because the management of firms with greaterdebt ratio may wish to diminish potential agency costs by releasing more informationto the investing public (Frankel et al. 1999, Chin et al. 2005). Prior empirical evidence(e.g. Ruland et al. 1990) also indicates that managerial ownership is negatively relatedto disclosure, we thus include MGTOWN, measured by the percentage of shares heldby the CEO, in our regression.

We include firms’ profitability (ROA), measured by earnings divided by total assets,as another control variable because firms with higher profitability are more likely tohold conference calls (Frankel et al. 1999). We also control for analyst followingbecause it plays a role in firms’ choice of an informal information dissemination chan-nel (Tasker 1998). We control for earning quality discretionary accruals (DA), measuredby DA calculated by Modified Jones model in our regression since earnings qualityaffects firm’s incentive to hold conference calls (Tasker 1998).

To be consistent with our hypotheses, we predict that the coefficient estimates onINT are positive (H1), the coefficient estimates on LAW are positive (H2), and the coef-ficient estimates on INT⁄OS are negative (H3).

5. Empirical results

5.1. Descriptive statistics

Panel B of Table 1 reports descriptive statistics for the variables on which our analysesare based, tabulated by firm-years for firms with and without conference calls. The tablealso presents the results of statistics analyses from t-tests and Wilcoxon Z-tests for dif-ference in means and medians between the two types of firms. Panel B indicates thatthe mean and median INT for conference call firms (n = 478) are significantly higherthan those of nonconference call firms (n = 2167). As predicted, it indicates that firmswith the higher degree of INT are more likely to host conference calls relative to thosewith lower degree of INT. The mean LAW for conference call firms is slightly greaterthan those for nonconference call firms. Finally, the median OS for conference callfirms are significantly higher than those of nonconference call firms, suggesting thatfirms with the greater control divergence are more likely to host conference calls, com-pared to those with less divergence, contrary to our predictions.11

Regarding the control variables, Panel B of Table 1 shows that, as predicted, themean (median) firm size (LOGASSET) of conference call firms is 9.90 (9.75) com-pared to 9.55 (9.52) for nonconference call firms. As predicted, the mean (median)DEPLAN of conference call firms is significantly larger than those of nonconferencecall firms, suggesting that there is a positive association between access to capital anddecision to hold conference calls. It further reveals that conference call firms exhibitrelatively rapid growth: average (median) SG is appropriately 25% (21%) for confer-


Dow

nloa

ded

by [

Uni

vers

ity o

f N

orth

Car

olin

a] a

t 14:

31 1

3 N

ovem

ber

2014

ence call firms, compared to 10% (6%) for nonconference call firms, and the differ-ence is significant at the 1% level. Consistent with our prediction, the mean (median)MB for conference call firms is significantly greater than that for nonconference firms,suggesting that growth firms are more likely to hold conference calls. These differ-ences suggest that financial statements may not completely reflect the firms’ futureprospect if a firm has a substantial growth potential, leading to further asymmetricinformation. Thus, management tends to hold more conference calls to present posi-tive perspectives to the market.

The mean (median) and the percentage of shares held by the CEO (MGTOWN) forconference call firms are significantly higher than those for nonconference call firms.Regarding the profitability variables, conference call firms are larger and more profitablethan other firms: mean (median) ROA is 11% (10%) for conference call firms comparedto 3% (4%) for nonconference call firms.

The mean (median) FOLLOW for conference calls firms is significantly larger thanthose for of nonconference call firms at the 1% level, suggesting that firms with greateranalyst following are more likely to hold conference calls. It further reveals that confer-ence call firms exhibit relatively higher discretionary accruals, compared to nonconfer-ence call firms, and the difference is significant at the 1%. The mean (median) DA forconference calls firms is significantly larger than those for of nonconference call firms

Table 2. Analysis of the likelihood of conference calls (Probit model).

Variablea Predict sign

Model (1) Model (2)

coefficient Marginal effectb coefficient Marginal effect

INT + 0.093⁄ 0.019 0.202⁄⁄ 0.043LAW + 0.261⁄⁄⁄ 0.055INT�OS � �1.295⁄ �0.274OS � �0.620⁄⁄ �0.127 0.023 0.005CEOOWN + 2.326⁄⁄ 0.477 2.398⁄⁄ 0.508LOGASSET + 0.307⁄⁄⁄ 0.063 0.298⁄⁄⁄ 0.063LDE + �0.162⁄⁄⁄ �0.033 �0.124⁄⁄ �0.026SG + 0.065⁄ 0.013 0.123⁄⁄ 0.026MB + 0.140⁄⁄⁄ 0.029 0.123⁄⁄⁄ 0.026DEPLAN + 0.593⁄⁄⁄ 0.105 0.516⁄⁄⁄ 0.095ROA + 2.635⁄⁄⁄ 0.540 3.688⁄⁄⁄ 0.781FOLLOW + 0.045⁄⁄⁄ 0.009 0.045⁄⁄⁄ 0.010DA + 0.479 0.098 0.383 0.081CONSTANT �4.947⁄⁄⁄ �5.061LR statisticd 599.61 579.04Probability 0.000⁄⁄⁄ 0.000⁄⁄⁄

Pseudo-R2c 21.87% 23.17

aSee Table 1 for variable definitions.bThe marginal is calculated as β′X/(1 + eβ′X)2, where β′X is computed at the mean values of the inde-pendent variables. For the indicator variable, the marginal effects is the difference in probability whenthe variable equals one vs. when it equals zero, evaluated at the mean of the other variables.cPseudo R2 is McFadden’s measure of goodness of fit, computed as 1–(Lu/Lc), where Lu denotes theunconstrained log-likelihood of the (full) model and Lc denotes the constrained log-likelihood of theconstrained model.dLR statistic and the corresponding p-value are given for the joint test of significance of the modelcoefficients.⁄, ⁄⁄, ⁄⁄⁄ signify that the coefficients of variables are significantly different from zero at the 1, 5, and10%, respectively (one-tailed).


Dow

nloa

ded

by [

Uni

vers

ity o

f N

orth

Car

olin

a] a

t 14:

31 1

3 N

ovem

ber

2014

at the 1% level, suggesting that firms with greater discretional accruals are more likelyto hold conference calls.

5.2. Multivariate analysis

5.2.1. Probit regression results

Table 2 presents the summary results for the Probit regression. In evaluating the effectof our independent variables on conference calls, we found that our models approxi-mately explained 23.17% of the variation in the determinants of conference calls. Theexplanatory power of the models was significant at the 1% level.

Model (1) reflects how the decision to hold conference calls is affected by thedegree of INT. The results indicate that the coefficient of INT is positive, as predicted,and significant at 10% significant level. This is consistent with our predictions thatfirms are more likely to hold conference calls when increasingly expanding operationsabroad. Columns (2) in Model (1) also show the marginal effects of each variable. Themarginal effects are analogous to the slope coefficients in the OLS regression.12 Themarginal effect for INT in Model (1), 0.019, when multiplied by the interquartile range(0.6931472–0.6930569), suggesting that moving from the first to third quartile of INT,increases the probability of holding conference calls by 0.0002%. In sum, Model (1) of

Table 3. Analysis of the likelihood and frequency of conference calls (ordered probit model).

Variablea Predicted sign

Dependent variable (CC_FREQ)

Model (1) Model (2)

Coefficient Z-value Coefficient Z-value

INT + 0.123⁄⁄ 1.86 0.173⁄⁄ 2.00LAW + 0.353⁄⁄⁄ 3.71INT�OS � 0.219 0.25OS � �0.320 �0.96 �0.567 �0.89CEOOWN + 0.804 0.69 0.327 0.27LOGASSET + 0.354⁄⁄⁄ 5.54 0.331⁄⁄⁄ 4.87LDE + �0.170⁄⁄⁄ �2.77 �0.139⁄⁄ �2.13SG + 0.093⁄⁄ 2.14 0.158⁄⁄⁄ 2.84MB + 0.146⁄⁄⁄ 6.65 0.130⁄⁄⁄ 2.84DEPLAN + 0.599⁄⁄⁄ 6.67 0.529⁄⁄⁄ 5.42ROA + 2.218⁄⁄⁄ 5.89 2.952⁄⁄⁄ 6.83FOLLOW + 0.051⁄⁄⁄ 11.67 0.051⁄⁄⁄ 11.31DA + 0.435 1.17 0.391 0.98LR stat.c 781.09 744.99Probability 0.000⁄⁄⁄ 0.000⁄⁄⁄

Pseudo-R2b 18.52% 19.08%

aSee Tables 1 and 2 for variable definitions.bPseudo R2 is McFadden’s measure of goodness of fit, computed as 1 – (Lu/Lc), where Lu denotes theunconstrained log-likelihood of the (full) model and Lc denotes the constrained log-likelihood of theconstrained model.cLR statistic and the corresponding p-value are given for the joint test of significance of the modelcoefficients.Twelve intercepts produced by ordered probit equation has been omitted and not reported here for sim-plicity.⁄, ⁄⁄, ⁄⁄⁄ signify that the coefficients of variables are significantly different from zero at the 1, 5, and10%, respectively (one-tailed).


Dow

nloa

ded

by [

Uni

vers

ity o

f N

orth

Car

olin

a] a

t 14:

31 1

3 N

ovem

ber

2014

Table 2 generally provides supporting evidence of the positive association between thelikelihood of holding conference calls and the level of corporate INT.

Consistent with hypothesis 2, Model (2) of Table 2 reveals that the enhancing roleof law protection on the decision to hold conference calls will be more pronouncedwhen companies invest in a higher proportion of common law countries. The coeffi-cients on law protection (LAW) are significantly positive, as predicted, at a 1% signifi-cance level. The results provide strong evidence supporting the notion that firms aremore likely to conduct conference calls when companies invest in a higher proportionof common law countries. This leads to H2 being supported.

Table 2 also shows that the positive effect of firm’s international diversification onconference calls likelihood will be decreased when the controlling owner’ control diver-gence increases. The coefficients on the intersection between OS and the degree of INTfrom Models (2) is �1.295, significantly negative at a 10% level. The results suggestthat internationally diversified firms are less likely to host conference calls when thedivergence of controlling owners’ control-cash rights increases. This leads to H3 beingsupported.

Turning to the control variables, Table 2 reveals that the coefficients of CEOWN,LOGASSET, SG, MB, DEPLAN, ROA, and FOLLOW in two models are all positiveand significant at the 5–1% level, consistent with prior studies.

Table 4. Analysis of the likelihood of conference calls (probit model).

Variablea Predict sign

Model (1) Model (2)

coefficient Marginal effectb coefficient Marginal effect

INT + 0.150⁄⁄ 0.031 0.179⁄⁄ 0.038LAW + 0.175⁄⁄ 0.037UKUS + 0.515⁄⁄⁄ 0.105 0.473⁄⁄⁄ 0.010INT�OS � �1.516⁄⁄ �0.309 �1.113 �0.235OS � 0.182 0.037 �0.101 �0.021CEOOWN + 2.409⁄⁄ 0.491 2.169⁄⁄ 0.458LOGASSET + 0.310⁄⁄⁄ 0.063 0.300⁄⁄⁄ 0.063LDE + �0.165⁄⁄ �0.034 �0.124⁄⁄ �0.026SG + 0.067⁄ 0.014 0.123⁄⁄ 0.026MB + 0.138⁄⁄⁄ 0.028 0.122⁄⁄⁄ 0.026DEPLAN � 0.581⁄⁄⁄ 0.103 0.511⁄⁄⁄ 0.094ROA + 2.632⁄⁄⁄ 0.536 3.716⁄⁄⁄ 0.784FOLLOW + 0.045⁄⁄⁄ 0.009 0.044⁄⁄⁄ 0.009DA + 0.549 0.112 0.399 0.084CONSTANT �5.026 �5.049⁄⁄⁄

LR stat.d 614.26 587.00Probability 0.000⁄⁄⁄ 0.000⁄⁄⁄

Pseudo-R2c 22.41% 23.49%

aSee Table 1 for variable definitions.bThe marginal is calculated as β′X /(1 + eβ′X)2, where β′X is computed at the mean values of the inde-pendent variables. For the indicator variable, the marginal effects is the difference in probability whenthe variable equals one vs. when it equals zero, evaluated at the mean of the other variables.cPseudo R2 is McFadden’s measure of goodness of fit, computed as 1–(Lu/Lc), where Lu denotes theunconstrained log-likelihood of the (full) model and Lc denotes the constrained log-likelihood of theconstrained model.dLR statistic and the corresponding p-value are given for the joint test of significance of the modelcoefficients.⁄, ⁄⁄, ⁄⁄⁄, signify that the coefficients of variables are significantly different from zero at the 1, 5, and10%, respectively (one-tailed).


Dow

nloa

ded

by [

Uni

vers

ity o

f N

orth

Car

olin

a] a

t 14:

31 1

3 N

ovem

ber

2014

5.2.2. Ordered Probit regression results

Many sample firms in our study provide at least two conference calls in a year. There-fore, we also use Ordered Probit regressions to investigate the association between thefrequency of conference calls, the proxies for INT, the proxies for ownership structure,and the proxies for legal system. Table 3 presents the summary results for the OrderedProbit regression. The dependent variable, the conference calls frequency (CC_F), iscoded with the number of conference calls held by a firm in a year. The Ordered Probitresults are broadly comparable to those produced by the Probit method. Consistent withthe results in Table 2, firms are likely to conduct conference calls more frequently whenexpanding more of their operations abroad. Secondly, firms conduct conference callsmore frequently when companies invest in a higher proportion of common law coun-tries. However, we do not find the moderating effect of the divergence between control-ling owner’s voting rights and cash flow rights on the association between the degreeof INT and the likelihood of holding conference calls to be significantly different fromzero. Thus, H3 is not supported under the Ordered Probit model. This suggests that themoderating role of ownership structure only impacts the decision to hold conferencecall or not, but not the frequency of conference calls.13

5.3. Further analyses

Leuz et al. (2003) argue that the UK and the USA are among outsider economies withlarge stock markets, dispersed ownership, strong investor rights, and strong legalenforcement. Besides, our sample shows that foreign investment in common law coun-tries primarily comprises investment in these two countries. Thus, we further focus ouranalyses on investment in common law countries and explore whether firms are morelikely to conduct conference calls when investing in a higher proportion of the UK andthe USA than investing in a higher proportion of other common law countries.

Thus, we further categorize common law countries into (1) the UK and the USAand (2) other common law countries, and then define the first measure of legal protec-tion, UKUS, as the ratio of foreign assets in the UK and the USA to foreign assets inthe common law countries. The greater this ratio, presumably, the greater the protectionthat the sample firms have for their foreign investments. We add this variable in ourEquation (1) and rerun the regression. Table 4 presents the summary results. The coeffi-cients of UKUS in the models is positive and significant (p < 0.05), consistent with ourpredictions. The results indicate that there is a significant difference in conference calllikelihood across these two common law countries clusters. Due to the most strong lawenforcement in the UK and the USA, firm investing more in the UK and the USA aremore likely to conduct conference calls than firms investing more in other common lawcountries.

5.4. Sensitivity analysis: panel data

Our sample time period ranges from 2000 to 2005. The empirical results may generatespurious results because of our use of panel data. To further examine the robustness ofour empirical results, we choose one randomly selected firm-year observation for eachfirm (Greene 2003). The final simplified sample includes 994 observations. Untabulatedanalyses indicate that the results are broadly comparable to those in Table 2 with theexception that the coefficient of LAW is marginally significant (p < 0.1, one-tailed).


Dow

nloa

ded

by [

Uni

vers

ity o

f N

orth

Car

olin

a] a

t 14:

31 1

3 N

ovem

ber

2014

Thus, the results are not attributable to the frequencies with which a company appearsin the analyses. Again, the empirical results are robust.

6. Conclusion

In this paper, we examine the effects of INT on the likelihood of holding conferencecalls and enrich our understanding of the factors that influence the relationship. UsingTaiwanese listed firms from 2000 to 2005, we document that greater internationaldiversification is associated with higher conference calls likelihood. The findings holdtrue even after controlling for previously identified determinants of conference calls.Thus, we conclude that firm international operations are a critical determinant of thedecision to hold conference calls.

We further reveal that firms investing in a higher proportion of common law coun-tries are more likely to hold conference calls and hold them more frequently. The resultsare consistent with prior studies on the effect of legal system disclosure and reporting.For example, in market-oriented common law countries, there typically is a larger andmore diverse base of shareholders and information asymmetry is more efficientlyresolved though public disclosure. Hence, there is a larger demand for accounting quality(Ball et al. 2000, 2003). Finally, we examine the association between ownership structureand whether to hold conference calls. Ownership structures separating voting rights fromcash flow rights are common outside the USA, less so in the USA. In this paper, we thususe the divergence between controlling owners’ voting rights and cash rights to proxyfor agency problem. We hypothesize and document that the positive association of inter-national diversification with conference calls likelihood is reduced for diversified firmswith more serious agency problems. A limitation of this study is that our results are justbased on data from Taiwan. Overall, however, we believe that our results will be robustin other “insider economies” (Leuz et al. 2003), at least in the East Asian area.

NotesWe appreciate comments of the workshop participants at National Chengchi University, National

Central University, and National Taichung Institute of Technology. We also appreciate theparticipants of 2009 Annual Meeting in American Accounting Association and acknowledgethe financial support from the National Science Committee.

1. For example, according to a 2005 report by Accenture, businesses with revenues exceedingone billion dollars expect foreign sales to grow to an average of 42% of their total revenueover the next three years, up from 35% in 2005 and 26% in 2002 (Accenture, October 2005).Cross-border investments of the US firms have grown by more than 700% (World TradeOrganization 2000), and S&P 500 firms report that foreign sales account for more than 24%of total sales. Besides, of the 10 largest US companies listed on the New York StockExchange, almost one-half of their revenues are generated from foreign operations. In thecase of Taiwan, the volume of international trade and foreign direct investment approach 50%of gross national product in the recent years.

2. Claessens et al. (2000) and La Porta et al.’s (1999) state that in East Asia, corporate controlcan be achieved while holding much less than an absolute majority of the stock. Claessenset al. (2000) reported that the average voting rights held by controlling shareholders in Tai-wan was about 18.96%, while the average cash flow rights held by controlling shareholdersin Taiwan was about 15.98%.

3. Fan and Wong (2005) show that tight control by a few shareholders or a related group ofshareholders creates an entrenchment problem and leads to excess perquisite consumption andthe transfer of corporate assets to the controlling owner’s families or selected others. Hawet al. (2004) also indicate that controlling owners thus possess incentives and the ability toextract private control benefits that are not shared by minority shareholders in proportion to


Dow

nloa

ded

by [

Uni

vers

ity o

f N

orth

Car

olin

a] a

t 14:

31 1

3 N

ovem

ber

2014

the shares owned. Furthermore, a controlling owner who also serves as the management ofthe company may grant himself/herself excess compensation at the cost of minority share-holders.

4. See Articles 25 and 38 of the Securities and Exchange Acts.5. In addition, Frankel et al. (1999) found significantly increasing returns volatility during the

conference call period. Bushee et al. (2003) also find a high level of trading activity andreturns volatility during the conference call period.

6. For example, global manufacturing gives managers additional opportunities to exercise discre-tion by shifting production to lower or higher cost locations.

7. For example, there are many mechanisms for oppressed shareholders to make legal claimsagainst directors in the USA, while there are few such mechanisms in Germany (La Portaet al. 1998). The US managers who materially misstate earnings generally face shareholders’class action suits and securities regulators’ investigations, but German managers rarely facesuch consequences. As a consequence, German managers are more likely to exhibit suchbehavior, compared to the US managers, due to the higher cost of opportunistic behavior (LaPorta et al. 1998, Hung 2001).

8. For instance, Fan and Wong (2002) find that earnings figures are less informative when con-trolling owners possess high voting rights and when voting rights substantially exceed cashflow rights. Haw et al. (2004) also document that earnings management increases as the con-trol divergence of the controlling shareholders increases.

9. The website of the Market Observation Post System is available at http://newmops.tse.com.tw/.10. Again consistent with the methods used by La Porta et al. (1999) and Claessens et al. (2000),

we collected information on share cross-holdings. We also identified family groupings thatcontrolled firms in our sample. These procedures closely followed Claessens et al. (2000) andLa Porta et al. (1999). The resulting understanding of the network of the relationshipsenabled us to delineate the possessors of cash flow vs. voting rights.

11. In our multivariate analysis in Section 5.2, we will further examine the relation between OSand whether to hold conference calls by controlling for other variables well documented to berelated to the likelihood of holding conference calls.

12. We compute the marginal effects as eβ′X/(1 + eβ′X)2, where β′x is computed at the mean val-ues of the independent variables (Greene 2003). For the indicator variable, the marginaleffects is the difference in probability when the variable equals one vs. when it equals zero,evaluated at the mean of the other variables.

13. To examine the multicollinearity between the variables in our regressions, we estimate usingOLS and compute variance inflation factors (VIFs). We find no instances of VIFs greater than3.672, suggesting that our results are not affected by harmful multicollinearity. We also sequen-tially estimate Probit model with one variable omitted each time. The results remain qualitativelythe same. As a result, multicollinearity is unlikely to be a problem in our sample.

References

Accenture, 2005. Information Technology Association of America, ITAA E-Letter. Available from:http://www.itaa.org/isec [Accessed October 2005].

Ball, R., Kothari, S., and Robin, A., 2000. The effect of international institutional factors on prop-erties of accounting earnings. Journal of accounting and economics, 29, 1–52.

Ball, R., Robin, A., and Wu, J., 2003. Incentives versus standards: properties of accountingincome in four East Asian countries. Journal of accounting and economics, 36 (1–3),235–270.

Birkinshaw, J., Toulan, O., and Arnold, D., 2001. Global account management in multinationalcorporations: theory and evidence. Journal of international business studies, 32, 231–248.

Botosan, C.A. and Plumlee, M., 2002. A re-examination of disclosure level and the expected costof equity capital. Journal of accounting research, 40, 21–40.

Brown, S., Hillegeist, S., and Lo, K., 2007. Conference calls and information asymmetry. Journalof accounting and economics, 37, 343–366.

Bushee, B.J., Matsumoto, D.A., and Miller, G.S., 2003. Open vs. closed conference call: thedeterminants and effects of broadening access to disclosure. Journal of accounting and econom-ics, 34, 149–180.


Dow

nloa

ded

by [

Uni

vers

ity o

f N

orth

Car

olin

a] a

t 14:

31 1

3 N

ovem

ber

2014

http://newmops.tse.com.tw/

http://www.itaa.org/isec

Chatterjee, S. and Blocher, J.D., 1992. The continuous measurement of firm diversification: is itrobust? The academy of management journal, 35 (4), 874–888.

Chin, C., Chi, S., and Lin, Y., 2005. On the association between mandatory forecasts and confer-ence calls. Journal of management, 21, 175–197 (in Chinese).

Claessens, S., Djankov, S., and Lang, L.H.P., 2000. The separation of ownership and control inEast Asian corporation. Journal of financial economics, 58, 81–112.

Daniel, S.J., Cieslewicz, J.K., and Pearson, T., 2011. Corporate governance practices, earningsmanagement, and disclosure. Asia-Pacific journal of accounting and economics, 18 (2),624–638.

Duru, A. and Reeb, D.M., 2002. International diversification and analysts’ forecast accuracy andbias. The accounting review, 77 (2), 415–433.

Fan, J.P.H. and Wong, T.J., 2002. Corporate ownership structure and the informativeness ofaccounting earnings in East Asia. Journal of accounting and economics, 33, 401–425.

Fan, J.P.H. and Wong, T.J., 2005. Do external auditors perform a corporate governance role inemerging markets? Evidence from East Asia. Journal of accounting research, 43, 35–73.

Frankel, R., Johnson, M., and Skinner, D.J., 1999. An empirical examination of conference callsas a voluntary disclosure medium. Journal of accounting research 37 (Spring), 133–150.

Greene, W.H., 2003. Econometric analysis. 5th ed. Upper Saddle River, NJ: Prentice-Hall.Gu, F. and Li, J.Q., 2003. Disclosure of innovation activities by high-technology firms. Asia-Paci-

fic journal of accounting and economics, 10 (2), 1–24.Haw, I.M., et al., 2004. Ultimate ownership, income management, and legal and extra-legal insti-

tutions. Journal of accounting research, 42, 423–426.Healy, P.M. and Palepu, K.G., 2001. Information asymmetry, corporate disclosure, and the capital

markets: a review of the empirical disclosure literature. Journal of accounting and economics,31, 405–440.

Herrmann, D., Hope, O., and Thomas, W.B., 2008. International diversification and forecast opti-mism: the effects of Reg FD. Accounting horizons, 22 (2), 179–197.

Hitt, M.A., Hoskisson, R.E., and Kim, H., 1997. International diversification: effects on innova-tion and firm performance in product diversified firms. Academy of management journal, 40,767–798.

Hoskisson, R.E., et al., 1993. Construct validity of an objective (entropy) categorical measure ofdiversification strategy. Strategic management journal, 14 (3), 215–235.

Hung, M., 2001. Accounting standards and value relevance of financial statements: an interna-tional analysis. Journal of accounting and economics, 30, 401–420.

Jensen, M.C. and Meckling, W.H., 1976. Theory of the firm: managerial behavior, agency costsand the ownership structure. Journal of financial economics, 3, 305–360.

Kim, O. and Verrecchia, R., 1994. Market liquidity and volume around earnings announcements.Journal of accounting and economics, 17, 41–68.

Kogut, B., 1983. Foreign direct investment as a sequential process. In: C. Kindelberger, ed. Themultinational corporations in the 1980s. Cambridge, MA: MIT Press.

La Porta, R., et al., 1997. Legal determinants of external finance. Journal of finance, 52 (3),1131–1150.

La Porta, R., et al., 1998. Law and finance. Journal of political economy, 106, 1113–1155.La Porta, R., et al., 2000. Investor protection and corporate governance. Journal of financial eco-

nomics, 58, 3–27.La Porta, R., Lopez-de-Silanes, F., and Shleifer, A., 1999. Corporate ownership around the world.

Journal of finance, 54, 471–517.Leuz, C., Nanda, D., and Wysocki, P., 2003. Earnings management and institutional factors: an

international comparison. Journal of financial economics, 69 (3), 505–527.Leuz, C. and Verrecchia, R., 2003. The economic consequences of increased disclosure. Journal

of accounting research, 38.Lobo, G.J. and Zhou, J., 2001. Disclosure quality and earnings management. Asia-Pacific journal

of accounting and economics, 8 (1), 1–20.Percy, M., 2000. Financial reporting discretion and voluntary disclosure: corporate research and

development in Australia. Asia-Pacific journal of accounting and economics, 7 (1), 1–31.Ruland, W., Tung, S., and George, N.E., 1990. Factors associated with the disclosure of manag-

ers’ forecasts. The accounting review, 65 (3), 710–721.


Dow

nloa

ded

by [

Uni

vers

ity o

f N

orth

Car

olin

a] a

t 14:

31 1

3 N

ovem

ber

2014

Tasker, S., 1998. Bridging the information gap: quarterly conference calls as a medium for volun-tary disclosure. Review of accounting studies, 3, 137–167.

Thomas, W., 1999. A test of the market’s mispricing of domestic and foreign earnings. Journal ofaccounting and economics, 28, 243–267.

Thomas, W., 2000. The value-relevance of geographic segment earnings disclosures under SFAS14. Journal of international financial management and accounting, 11, 133–155.

Tihanyi, L. and Thomas, W., 2005. Information-processing demands and international expansion.Journal of business research, 58, 285–292.

World Trade Organization, 2000. WTO Annual Report. Geneva, Switzerland: WTO Publications.Yeh, Y.H., Lee, T.S., and Woidtke, T., 2001. Family control and corporate governance: Evidence

from Taiwan. International review of finance, 2, 21–48.


Dow

nloa

ded

by [

Uni

vers

ity o

f N

orth

Car

olin

a] a

t 14:

31 1

3 N

ovem

ber

2014

International diversification and conference calls

Documents

Transcript of International diversification and conference calls