International Review of Law and Economics 37 (2014) 1– 14
Contents lists available at SciVerse ScienceDirect
International Review of Law and Economics
oes high-quality corporate communication reduce insider tradingrofitability?
ebby Van Geyt, Philippe Van Cauwenberge ∗, Heidi Vander Bauwhedehent University, Department of Accountancy and Corporate Finance, Belgium
a r t i c l e i n f o
rticle history:eceived 7 May 2012eceived in revised form 1 March 2013ccepted 10 April 2013
EL classification:101434
a b s t r a c t
Exploring a unique database on insider trading in Belgium, we investigate whether high-quality corpo-rate communication contributes to reducing insider trading profitability and information asymmetry.Using disclosure scores of professional financial analysts as a proxy for communication quality, we finda significant negative association between corporate communication quality and insider trading profit-ability. Closer inspection of different communication channels shows that the quality of annual reports,press releases and investor relation activities is more relevant in explaining insiders’ abnormal returnsthan the quality of corporate websites.
sis and considered additional firm and trade characteristics. For
elgium
. Introduction
This paper examines whether high-quality communication isffective in reducing the profitability of insider trading. Previousesearch has documented that, despite regulations on insider trad-ng, insiders still earn significant abnormal returns from tradingn information asymmetries between insiders and outsiders. Asuggested by analytical work on disclosure (e.g. Diamond, 1985;errecchia, 2001), an important instrument to decrease this asym-etry could be the dissemination of high-quality information. We
herefore hypothesize that high-quality corporate communicationeduces insiders’ abnormal returns.
In an additional analysis this paper also examines whetherhe impact of disclosure quality differs between communicationhannels, i.e. annual reports, press releases, websites and investorelation activities. The different channels and information com-unicated through these channels have specific characteristics
hat might limit or enhance the ability to affect the level of infor-
ation asymmetry, like, for example, timeliness of the disclosed
nformation, time horizon (forward-looking versus backward-ooking), need for external verification and voluntary or mandatory
∗ Corresponding author.E-mail address: [email protected] (P. Van Cauwenberge).
disclosures. We therefore hypothesize that the impact of disclosurequality on insider trading profits and on information asymme-try in general depends on the communication channel. Resultson the effectiveness of the different communication channels andespecially on the difference between semi-mandatory (i.e. annualreports)1 and voluntary channels (i.e. press releases, websites,investor relations) might also be of interest to regulators as theymight indicate potential areas of improvement for mandatorydisclosure requirements. These requirements are imposed by reg-ulators to prevent information asymmetries and stimulate a moreefficient allocation of resources.
Within the extensive literature on insider trading, an importantline of research has focused on the determinants of insider tradingprofitability. Early studies by Jaffe (1974) and Finnerty (1976) iden-tified company risk factors like size and the market-to-book ratio asimportant drivers of insiders’ abnormal returns. Building on thesefindings, later studies attempted to broaden the scope of analy-
example, some researchers examined whether the informationalbenefit of insiders is related to their position within a company (e.g.
1 Information included in annual reports consists of mandatory financial state-ments information possibly supplemented by voluntary disclosures on businesssegments, future prospects, company objectives, etc.
etzer & Theissen, 2009; Fidrmuc, Goergen, & Renneboog, 2006;eyhun, 1986). Others investigated the influence of the debt-to-sset ratio (Aussenegg & Ranzi, 2008), trade size (Cheuk, Fan, &o, 2006; Seyhun, 1986; Wisniewski & Bohl, 2005), trade intensityAussenegg & Ranzi, 2008; Betzer & Theissen, 2009) and cross-isting on foreign stock markets (Chang & Corbitt, 2012; Korczak
Lasfer, 2008). More recently, as researchers and practitionersmphasized the importance of good corporate governance in man-ging the information asymmetry problem, insider trading researchtarted to consider corporate governance related variables. Accord-ngly, previous studies have looked into the effect of ownershiponcentration (Betzer & Theissen, 2009; Del Brio & Perote, 2007;idrmuc et al., 2006), type of controlling shareholder (Betzer &heissen, 2009), board composition (Chang, Hillman, & Watson,005) and executive compensation (Zhang, Cahan, & Allen, 2005).evertheless, while it is generally acknowledged that comprehen-
ive, transparent and timely disclosures are essential elements ofood corporate governance (Bushman & Smith, 2001; Mallin, 2002;itton, 2002; OECD, 2004; Patel & Dallas, 2002), no prior study has
horoughly investigated the effect of the quality of corporate dis-losures on insider trading returns. In this study, we examine thiselation by relating professional analysts’ disclosure scores to therofitability of insider trading.
To address our research question, we use data from Belgianisted companies. La Porta, Lopez-de-Silanes, Shleifer, and Vishny1997) and La Porta, Lopez-de-Silanes, Shleifer, and Vishny (1998)epict Belgium as an insider economy characterized by highlyoncentrated and controlling ownership. In such an environment,inority shareholders are at a disadvantage as large, dominant
hareholders can use their power to privately acquire information,hich makes them less dependent on public communication. As a
onsequence, in Belgium, the role of corporate communication ineducing information asymmetry is potentially very important.
To measure the quality of corporate communication, we use disclosure score granted by the Belgian Association of Financialnalysts (BVFA).2 Each year, the BVFA invites its members to screen
he communication of a number of companies and assign a disclo-ure rating. This rating evaluates several disclosures characteristicsdentified as important attributes of high-quality communica-ion, i.e. preciseness, transparency, timeliness and scope (Brown &illegeist, 2007). Contrary to comparable analyst ratings, like thosessigned by Standards and Poor’s (S&P) and the Association fornvestment Management and Research (AIMR) (Brown & Hillegeist,007; Khanna, Palepu, & Srinivasan, 2004; Patel, Balic, & Bwakira,002; Welker, 1995), the BVFA also evaluates the communicationuality of smaller companies. For these companies, informationsymmetries between insiders and other market participants areotentially more significant, which makes high-quality communi-ation even more relevant (BVFA, 2010a).
To measure the profitability of insider trading, we exploit unique database on insider trading provided by the Belgianinancial Services and Markets Authority (FSMA) and calculatehe cumulative abnormal returns that insiders earn when trad-ng in their own stock. Since the liquidity of some Belgian listedecurities is rather low (Buysschaert, Deloof, & Jegers, 2004), thebnormal returns are estimated either using a standard marketodel (MacKinlay, 1997) or market model adjusted for thin trading
Dimson, 1979) depending on whether stocks are thinly traded orot.
Based on a sample of insider trades that occurred betweenanuary 2006 and August 2010, our results show that high-uality communication reduces the profitability of insider trading.
2 BVFA stands for “Belgische Vereniging van Financiële Analysten”.
f Law and Economics 37 (2014) 1– 14
Furthermore, we find that the quality of annual reports, pressreleases and investor relation activities, is relatively more effec-tive in reducing information asymmetry than the quality corporatewebsites. Investor relation activities, which are used to commu-nicate timely and forward-looking information directly to theinvestor community, appear to be the most effective.
Our research contributes to two streams of literature. First, weadd to the literature on insider trading profitability by examiningthe impact of high-quality communication, as proxied by a com-prehensive measure of disclosure quality assigned by professionalusers of corporate communication. To our knowledge, there areonly a handful of papers that investigate whether corporate com-munication quality influences insiders’ informational benefits. Inaddition, these papers obtain inconclusive results and use indi-rect measures of reporting quality, such as analyst following, newscoverage and value relevance (e.g. Frankel & Li, 2004). In contrastto these studies, we use a more direct and objective measure ofcorporate communication quality assigned by professional usersof corporate communication, i.e. financial analysts and fund man-agers. In addition, our measure includes an individual assessment ofthe quality of annual reports, press releases, websites and investorrelation activities. This allows us to assess whether the effect of thequality of communication differs across alternative communica-tion channels. A general advantage of using externally-developeddisclosure ratings is that these do not involve judgment by theresearcher(s) in question. This facilitates the verification of researchresults and the application of the rating in other research designs(Healy & Palepu, 2001). In addition, researchers only have accessto published information and lack knowledge of disclosures dis-tributed through unpublished channels like analyst meetings andconference calls (Healy & Palepu, 2001). Analysts are also regardedas the primary and most influential users of corporate commu-nication as they communicate with companies on a daily basis(e.g. Hirst, Koonce, & Simko, 1995; IASB, 2005; Revsine, Collins,& Johnson, 2004; Schipper, 1991). This puts them in a privilegedposition to objectively evaluate the quality of corporate disclosures.
Second, our work contributes to the literature examining therelationship between disclosure and information asymmetry byusing an alternative proxy for information asymmetry. Prior workexamined this relation using, for example, bid-ask spreads and theprobability of informed trading as proxies for information asym-metry (e.g. Brown & Hillegeist, 2007; Welker, 1995). By contrast,we proxy information asymmetry by the magnitude of insiders’abnormal returns. The use of this proxy is well-established in theempirical literature (e.g. Chang et al., 2005; Frankel & Li, 2004)and supported by theoretical work (Kyle, 1985). Furthermore, themajority of prior disclosure studies is based on U.S. data (Healy &Palepu, 2001). Re-examining the disclosure–information asymme-try relation for a sample of Belgian listed companies may providevaluable new insights as the Belgian institutional setting differsfrom the U.S. For example, with regard to ownership structures,Belgian listed companies generally have a concentrated and con-trolling ownership (La Porta et al., 1997, 1998). In addition, they areoften controlled by a family or a single controlling owner (Faccio& Lang, 2002). U.S. companies, on the other hand, tend to have dif-fuse ownership and are less (family-) controlled. Regarding theprovision of external capital, Belgian companies primarily raiseexternal capital through bank financing while their U.S. counter-parts generally rely an equity financing (La Porta et al., 1997).Obviously, information needs of both capital providers differ sub-stantially. Other institutional differences include the weaker levelof investor protection (La Porta et al., 1998) and the influence
of corporate law and taxation on financial reporting in Belgium(Vanstraelen, Zarzeski, & Robb, 2003). The above characteristicsof the Belgian institutional environment do on the other handbear a strong resemblance to the economies of other continental
view o
EiePremowtii(sic(eiq
oqciataccmpcttcufidm
ataactsip
2
tulwhoiho2o
asymmetry should be negatively associated with the quality ofcorporate communication (e.g. Diamond, 1985; Verrecchia, 2001).By disclosing more, precise and complete information in a timely
3 With regard to the prohibition of trading on inside information an importantdistinction is made between administrative and criminal sanctions. In particular,in case of trading by insiders themselves, criminal sanctions can only be imposedif there is sufficient proof of a causal connection between the possession of insideinformation and the reprehensible transaction. Administrative fines on the other
D. Van Geyt et al. / International Re
uropean countries with a French-based civil law system. Accord-ng to La Porta et al. (1997, 1998), the Belgium legal and institutionalnvironment is similar to the French, Dutch, Spanish, Italian andortuguese environment. Consequently, we may assume that theesults of our inquiry are, to some extent, generalizable to theseconomies. Prior research on the disclosure–information asym-etry relation in French civil law countries is very limited. The
nly examples that we are aware of are Vanstraelen et al. (2003)hich focused on three European countries including Belgium and
he Netherlands, Aerts, Cormier, and Magnan (2007) which exam-ned disclosure practices in several continental European countriesncluding Belgium, France and the Netherlands and finally, Lakhal2009) which focused on French listed companies. None of thesetudies have however used a comprehensive disclosure score sim-lar to the BVFA-rating that evaluates different communicationhannels. In addition, both Vanstraelen et al. (2003) and Lakhal2009) do not take the quality of disclosures into account. Aertst al. (2007) accounts for disclosure quality by considering the wayn which items are described, i.e. general terms, specific terms oruantitative or monetary terms.
From a regulator’s point of view, we believe that the resultsf our inquiry provide additional insight into the effect of higher-uality communication on information asymmetry. In general, theyonfirm the importance of high-quality communication in reduc-ng information inequities between a company and its stakeholdersnd in preventing unfair enrichment by privileged insiders. In addi-ion, we further deepen the insight into the disclosure–informationsymmetry relationship by examining this relation for differentommunication channels and providing evidence that differenthannels have a different effect on the level of information asym-etry. Interestingly, our results show that, whereas regulators
rimarily focus on annual reports and backward-looking finan-ial statements information, this communication channel is nothe most effective in reducing the level of information asymme-ry. By contrast, investor relation activities, which are used toommunicate timely and forward-looking information on a vol-ntary basis, appear to be the most effective. We believe that thisnding is relevant for regulators and may shed new light on theiscussion concerning the shift toward more or less regulation ofarkets.The remainder of this paper is organized as follows. In Section 2
discussion is provided of the current Belgian legislation on insiderrading. In Section 3 a brief overview of related literature is given,ccompanied by the hypothesis development. The research designnd proxies for information asymmetry and corporate communi-ation quality are discussed in Section 4. Next, Section 5 describeshe data which are used and Section 6 reports some descriptivetatistics. The results of the empirical inquiry are disclosed andnterpreted in Section 7. In Section 8 some sensitivity checks areerformed. Finally, Section 9 concludes.
. Belgian legislation on insider trading
The current Belgian legislation on insider trading is founded inhe 2003 European Directive on insider dealing and market manip-lation (Directive 2003/6/EC), i.e. the Market Abuse Directive. The
egislation is based on the central concept of “inside information”hich is defined as “any information of a precise nature whichas not been made public, relating, directly or indirectly, to oner more issuers of financial instruments or to one or more financialnstruments and which, if it were made public, would be likely to
ave a significant effect on the prices of those financial instrumentsr on the price of related financial instruments” (Law of 2 August002, art. 2). The Belgian legislation formulates three prohibitionsn the use of this inside information (Law of 2 August 2002, art. 25
f Law and Economics 37 (2014) 1– 14 3
and art. 40). First, persons in possession of inside information whoare aware, or should be aware that the information concerned isinside information are prohibited from trading. In particular, theymay not use the information by acquiring or disposing of financialinstruments to which the information relates, or by trying to doso. Second, they may not communicate the inside information tothird parties, except within the framework of the normal exerciseof their job description. Finally, they must also refrain from mak-ing recommendations or induce another person on the basis of theinformation to acquire or dispose of the financial instruments inquestion.
An offender of these legal prohibitions may face administrativesanctions imposed by the FSMA as well as criminal sanctions.3 Inparticular, the FSMA may order the offender to pay an adminis-trative fine between 2500 euro and 2,500,000 euro. If however theoffender obtained a capital gain from the infringement, the maxi-mum fine is raised to twice this gain and, in case of a repeat offense,to three times the capital gain (Law 2 August 2002, art. 36). Withregard to the criminal sanctions, an offender may be condemned toa prison sentence between three months and one year, payment ofa fine between 50 euro and 10,000 euro and/or payment of criminalfine corresponding to a maximum of three times the gain earned,directly or indirectly, by illegal insider trading (Law 2 August 2002,art. 40).
In order to prevent illegal trading by insiders, the Belgian legis-lation has also formulated three preventive measures. First, issuersof financial instruments are obliged to reveal inside informationimmediately. In particular, this information should be published onthe website of the financial market on which the financial instru-ment is listed (Law 2 August 2002, art. 10). Second, issuers mustdraw up a list of persons who have access to inside information.This list must be kept at the disposal of the FSMA for a period offive years (Law 2 August 2002, art. 25bis). Finally, persons who ful-fill an executive function in the issuing company as well as personsclosely related to them, e.g. spouses, partners, children and otherrelatives, are required to report their transactions to the FSMA.The transactions must be reported within five working days aftertheir execution. However, as long as the total sum of the trans-actions during the current calendar year is below 5000 euro, thereporting may be delayed until 31 January of the next calendaryear (Law 2 August 2002, art. 25bis). The FSMA is responsible forpublishing all reported transactions on their website. In case of non-compliance with the above preventive measure, the FSMA has theauthority to impose administrative fines (Law of 2 August 2002,art. 36).
3. Prior literature and hypothesis development
3.1. The impact of corporate communication quality on insidertrading profitability
Theoretical research on disclosure shows that information
hand may be enforced as soon as a person is in possession of inside information andmakes a transaction. No proof is required that a transaction was actually inspired bythe inside information. No distinction between administrative and criminal sanc-tions is made with regard to the prohibitions on tipping inside information to thirdparties and on making trading recommendations (Tison & Ravelingien, 2007).
4 view o
avqercsmMW&d
iotaattd1ti
ctttltl2esafsoigoasoApmhpd
stcoocmbbtf
Ha
D. Van Geyt et al. / International Re
nd transparent manner, companies reduce the amount of pri-ate information while simultaneously increasing the amount anduality of public information available to investors. In general, thexistence of this negative association is supported by empiricalesearch. Using a myriad of proxies for disclosure quality, includingonference call activity and analyst disclosure ratings, studies havehown that a lower level of information asymmetry results intoore informative stock prices (Gelb & Zarowin, 2002; Lundholm &yers, 2002), lower bid-ask spreads (Heflin, Shaw, & Wild, 2005;elker, 1995), less analyst forecast dispersion (Hope, 2003; Lang
Lundholm, 1996), and a lower cost of equity (Botosan, 1997) andept capital (Sengupta, 1998).
Information asymmetry can, however, also affect the profitabil-ty of insider trading. In particular, insider trading research is basedn the presumption that a certain level of information asymme-ry exists between insiders and outside investors as insiders aressumed to have a more in-depth knowledge of a firm’s economicss well as privileged access to private information. If insiders decideo trade upon their informational benefits, prior research indicateshat significant abnormal trading profits can be earned (e.g. Aktas,e Bodt, & Van Oppens, 2008; Fidrmuc et al., 2006; Lin & Howe,990). In addition, supporting theoretical work by Kyle (1985),hese studies have shown that insiders’ profits increase with theirnformational benefits.
Hence, given the above theory and findings that higher-qualityommunication decreases the level of information asymmetry andhat information asymmetry determines the profitability of insiderrading, it can be expected that better communication reduceshe magnitude of insiders’ abnormal returns. However, despite thearge attention given to corporate communication quality by prac-itioners and by researchers in corporate governance and disclosureiterature (e.g. Brown & Hillegeist, 2007; Chen, Chung, Lee, & Liao,007; OECD, 2004; Patel & Dallas, 2002), only few studies havexamined the effect on insider trading profitability. One theoreticaltudy by Baiman and Verrecchia (1996) examined this relationshipnd confirmed that higher-quality disclosures reduce the profitsrom insider trading. Empirically, Frankel and Li (2004) found thatome elements of a firm’s information environment, i.e. the extentf analyst following and the value relevance of financial statements,ndeed mitigate the informational benefits of insiders (i.e. lowerains for and/or less purchase transactions by insiders). However,ther elements of the information environment, i.e. news cover-ge, seem to enhance these informational benefits. A more recenttudy by Betzer and Theissen (2009) used the voluntary adoptionf international accounting standards (i.e. U.S. Generally Acceptedccounting Principles or International Accounting Standards) as aroxy for the informativeness and transparency of financial state-ents. Contrary to their expectations, their results suggested that
igher abnormal insider trading profits are earned in companiesreparing financial statements according to the international stan-ards.
In light of this mixed evidence, we re-examine the relation-hip between the quality of corporate communication and insiderrading profitability using analyst disclosure ratings. This proxy oforporate communication quality has been widely used in previ-us disclosure studies for it provides a comprehensive measuref disclosure quality assigned by professional users of corporateommunication. Testing the relationship between corporate com-unication quality and insider trading profitability, we expect
etter communication to mitigate the information asymmetryetween insiders and outside, uninformed investors and to simul-aneously lower abnormal trading profits. Our test hypothesis is
ormulated as follows:
ypothesis 1. The profitability of insider trades will be negativelyssociated with the quality of corporate communication.
f Law and Economics 37 (2014) 1– 14
3.2. The role of alternative communication channels
The aggregate measure of corporate communication qualityused in this study covers four individual corporate communicationquality ratings. Each rating assesses the quality of communica-tion through a specific communication channel, i.e. annual reports,press releases, corporate websites and investor relation activities.The extent to which the communicated information impacts thelevel of information asymmetry may differ across these commu-nication channels as both the information communicated througheach channel and the channel itself have specific characteristics.Regarding annual reports, for example, the included informationconsists of mandatory financial statements information possiblysupplemented by voluntary disclosures on business segments,future prospects, company objectives, etc. An important charac-teristic of the mandatory financial statements information is thatthis information is verified by an external auditor which enhancesthe level of credibility. Nevertheless, the fact that this mandatoryinformation is subject to international reporting requirements andexternal verification, limits the degrees of freedom for companiesto distinguish themselves regarding the quality of the financialstatements information. Consequently, differences in the qualityof annual report disclosures, if any, are expected to ensue fromdifferences in the quantity and quality of the included voluntaryinformation (Brown & Hillegeist, 2007). Furthermore, despite thefocus of regulators on annual reports and financial statementsin particular, practitioners (i.e. financial analysts and investors)often no longer regard them as the main tool of communicationbecause of their backward-looking nature and lack of timeliness.Both characteristics limit the ability of annual reports to providevaluable, new information to the financial community. Support forthis conjecture was found by several studies which focused on theperception of practitioners in different countries toward annualreports (e.g. Chang & Most, 1985: United States, United Kingdomand New Zealand; Vergoossen, 1993: the Netherlands; AIMR, 2000:United States).
A second potential communication channel is press releases.This communication channel is used by companies to voluntarilydisclose periodic updates of financial results (i.e. quarterly andhalf-year results) as well as information on important events thatcould affect the risk profile of a company (BVFA, 2010b). From ana-lysts’ point of view, the high degree of timeliness has made pressreleases essential for the assessment of companies (BVFA, 2010a).Empirical support for this proposition was found by McNicholsand Manegold (1983) who showed that press releases contain-ing interim financial results pre-empt some information whichis later disclosed through annual reports. In addition, Brown andNiederhoffer (1968) and Brown and Rozeff (1979) provided evi-dence that financial press releases improve forecast accuracy ofannual earnings. A potential limitation of press releases could bethat the disseminated information is unaudited and may thereforebe less credible. Nonetheless, studies by Stocken (2000), Lundholm(2003) and Ball, Jayaraman, and Shivakumar (2011) suggest thatas the credibility of press release disclosures can be subsequentlyverified using audited financial statements information, managersare likely disciplined to be more truthful in their ex ante communi-cations.
A third communication channel used by companies is corpo-rate websites. This communication channel is a permanent sourceof information which is often used to disclose information on, forexample, the company’s history and mission statement, corporategovernance structures and social and environmental issues com-
plementary to the traditional financial information (BVFA, 2010b;Trabelsi, Labelle, & Dumontier, 2008). As such, the informationdisclosed through corporate websites is often also disseminatedthrough other communication channels like annual reports and
view o
piNtioMttEpt2teclr
rwrieDttmami(hra2gm
mclioni
Hmc
4
4
toos
C
wait
D. Van Geyt et al. / International Re
ress releases. This characteristic might potentially limit the abil-ty of web disclosures to affect the level of information asymmetry.evertheless, prior studies on internet reporting agree that volun-
ary web disclosures are taking an increasingly prominent placen corporate communication because of their timeliness and easef access (e.g. Bollen, Hassink, & Bozic, 2006; Jones & Xiao, 2004;arston & Polei, 2004). In particular, the use of internet repor-
ing is more prevalent in North American companies comparedo continental European (Aerts et al., 2007) and especially Easternuropean companies (Bonsón & Escobar, 2006). Nonetheless, com-ared to other European countries, Belgian companies are amonghe top users of voluntary web disclosures (Bonsón & Escobar,002). Focusing on the usefulness of web disclosure in reducinghe level of information asymmetry, Trabelsi et al. (2008) and Aertst al. (2007) found that the extent of voluntary disclosures throughorporate websites is negatively related to the dispersion of ana-yst forecasts. This finding indicates that web disclosures provideelevant information for the evaluation of companies.
A final communication channel used by companies is investorelation activities. The use of this communication channel has beenell established in the U.S. and U.K. for a considerable time. More
ecently, investor relation activities have also in Europe becomencreasingly important in response to the growing reliance on (for-ign) equity financing (Marston & Straker, 2001). A study by Chang,’Anna, Watson, and Wee (2008, p. 378) defined investor rela-
ions as the continuous dissemination of “company information inhe form of annual reports, earnings forecasts, proposed invest-
ents, governance procedures, dividends and financing intentionsnd a wide range of other information, both formal and infor-al”. Accordingly, much of the information communicated through
nvestor relation activities is voluntary, timely and forward-lookingBrown & Hillegeist, 2007). The credibility of the information may,owever, be lower for it is often disclosed verbally and sometimesepresents non-quantifiable and non-verifiable information suchs the degree of optimism held by executives (Brown & Hillegeist,007). Despite these negative characteristics, the BVFA considersood investor relation services as crucial for companies to get infor-ation across to the investor community (BVFA, 2010a).In sum, the above discussion clearly indicates that each com-
unication channel and the included information have specificharacteristics that can enhance or limit the ability to affect theevel of information asymmetry. Given the above findings, wenvestigate whether the effect of corporate communication qualityn insider trading profitability differs across alternative commu-ication channels. The second hypothesis proposed in our study
s:
ypothesis 2. Any relation between the quality of corporate com-unications and insider trading profitability differs between the
ommunication channels.
. Methodology
.1. Research design
To empirically investigate whether high-quality communica-ion reduces information asymmetry – and hence the profitabilityf insider trading – we estimate the following regression usingrdinary least squares and clustered, heteroskedasticity robusttandard errors (Rogers, 1993).
here the dependent variable, CAR(0,20), is the cumulative aver-ge abnormal return over a 21-day event window following eachnsider trade. The test variable, CommunicationQuality, representshe disclosure score awarded by the financial analysts and fund
f Law and Economics 37 (2014) 1– 14 5
managers of the BVFA. The vector x includes a set of controlvariables. In the following subsections, the measurement of theregression variables is explained in detail.
4.2. Measurement of insider trading profitability
To measure the abnormal gains of insider trading, we applyevent study methodology and calculate abnormal returns of insidertrades over a certain period starting from the transaction date ofeach insider trade. However, since the liquidity of some Belgianlisted securities is rather low (Buysschaert et al., 2004), we esti-mate the abnormal returns either using a standard market model(MacKinlay, 1997) or a market model adjusted for thin trading(Dimson, 1979) depending on the liquidity of the underlying shares.The issue is that when a stock is infrequently traded, stock pricesrecorded at the end of a time period may include adjustments tonews events occurring earlier in that period. Consequently, whenusing a standard market model for such stocks, a problem of non-synchronous trading arises due to a mismatch between the returnof these stocks and the return of the market index. To addressthis problem, the aggregated coefficients method of Dimson (1979)includes lagged, leading and contemporaneous beta coefficients inorder to provide unbiased beta estimates for thinly traded securi-ties. Following a suggestion by Friederich, Gregory, Matatko, andTonks (2002), we apply the Dimson-adjustment to stocks with thehighest number of daily zero returns. More specifically, firms arefirst sorted in ascending order based on the number of daily zeroreturns during the estimation and event window. Next, the ordi-nary market model is applied to firms belonging to the first threequartiles (with the lowest number of zero return days), while theDimson-ajdusted model is used to calculate betas for firms in thebottom quartile (with the highest number of zero return days).Applying the adjustment to all stocks would lead to an overestima-tion of the betas of actively traded securities. Following Buysschaertet al. (2004), we add one leading and three lagged coefficients tothe market model for Belgian, thinly traded securities.
Rjt = ˛j + ˇjRmt + εjt Standard market model (2)
Rjt = ˛j ++1∑
k=−3
ˇjkRm,t+k + εjt Dimson-adjusted market model (3)
where Rjt is the daily stock return for firm j on day t adjusted forstock splits, stock dividends and issues; Rmt and Rm,t+k are the dailyvalue-weighed and dividend-adjusted returns of the market indexon day t and day t + k respectively. Our benchmark index Rm is theBrussels All Shares Return Index.
Next, the abnormal return to firm j on day t, ARjt, is calculatedeach day from the insider trading day (day 0) to 20 trading daysafter the event (day + 20):
ARjt = Rjt − ˆ̨ j − ˆ̌jRmt Standard market model (4)
ARjt = Rjt − ˆ̨ j −+1∑
k:=−3
ˆ̌jkRm,t+k Dimson-adjusted market model
(5)
where ˆ̨ j and ˆ̌j(k) are estimated by means of an OLS-regression
over an estimation window of 160 trading days, going from day−160 to day −1. Since our results are reported for a pooled sam-
ple including both purchases and sales, the abnormal returns forinsider sales are multiplied by minus one. Insiders profit whensecurities outperform the market after a buy transaction and whenstocks underperform the market after a sales transaction.
6 view o
fo
C
w2d
4
s((t
lIhi“c
aearavrrisoosaortgqo
aottppscad
t
a(
D. Van Geyt et al. / International Re
Finally, to evaluate the event-specific cumulative abnormal per-ormance from day 0 to day +20, the abnormal returns are summedver the time interval in question:
ARij(0,20) =20∑
t=0
ARjt (6)
here CARij(0,20) represents the cumulative abnormal return over1 trading days for a particular event i of firm j.4 The latter is theependent variable in Eq. (1).
.3. Measurement of corporate communication quality
Corporate communication quality is measured using a disclo-ure rating awarded by the Belgian Association of Financial AnalystsBVFA). In particular, we measure the explanatory variable in Eq.1) as the BVFA-score assigned to company j in the year in whichhe insider trade occurred.
The BVFA is part of the European Federation of Financial Ana-yst Societies (EFFAS) and the Association of Certified Internationalnvestment Analysts (ACIIA). For the past 50 years, this organizationas granted an “Award for the Best Financial Information”. Accord-
ng to the President of the BVFA, the purpose of the award is toreward Belgian listed companies that stand out in terms of finan-ial communication policy, transparency and investor relations”.
Each year, a group of financial analysts and fund managersssigns the disclosure rating. More specifically, financial analystsvaluate companies on four different communication channels, i.e.nnual reports, press releases, corporate websites and investorelation activities. In addition, fund managers also provide anppreciation of the investor relation activities from their point ofiew. Each communication channel is evaluated on different crite-ia. Annual reports are screened on the key numerical items, theeliability and transparency of financial data, and the availability ofnformation on products, services and markets. For press releases,cores are provided on the disclosure of half and full year results,n whether explanations are given on the year-to-year evolutionf these numbers and on the timing of the press releases. Web-ites are judged on the presence of a financial calendar and thevailability and preciseness of the disseminated information, i.e.perational and financial information and information on investorelation activities and corporate governance. Finally, investor rela-ion activities are rated on, among other things, the quality of theuidance, consistency and reliability of the provided information,uickness of response to analysts’ questions, and the organizationf analyst meetings, conference calls and client visits.5
For listed companies to qualify for the BVFA-award, obviously, first criterion to be selected is that there is a sufficient numberf analysts following the company.6 Next, a preliminary ques-ionnaire is sent to the companies themselves. The responses tohe included factual questions are used to underpin the screeningrocess and give a first indication about the willingness to sup-ort financial communication. In a second stage, each company is
creened in detail by financial analysts and fund managers on aompany per company basis. Next, the final results are comparednd discussed within a panel of financial analysts that makes aecision on the final ranking.
4 Obviously, a particular firm can have more than one insider trading event duringhe sample period.
5 Annual valuation grids can be consulted on the website of BVFA for detailed overview of the screening criteria and their respective weightshttp://www.bvfa.be).
6 A minimum of three analysts per company is imposed.
f Law and Economics 37 (2014) 1– 14
The use of analyst disclosure ratings as a measure of com-munication quality is well-established in prior literature. Studiesfocusing on U.S. listed companies generally use the AIMR-ratingwhich strongly resembles the Belgian BVFA-rating (e.g. Botosan &Plumlee, 2002; Brown & Hillegeist, 2007; Healy, Hutton, & Palepu,1999; Lang & Lundholm, 1993, 1996; Nagar, Nanda, & Wysocki,2003; Sengupta, 1998; Welker, 1995). In cross-country studies, theCIFAR (e.g. Bushman & Smith, 2001; Carlin & Mayer, 2003; Hope,2003; Salter, 1998) or S&P index (e.g. Durnev & Kim, 2005; Khannaet al., 2004; Litvak, 2007; Patel et al., 2002) are often applied. Ageneral advantage of using externally-developed disclosure rat-ings is that these do not involve judgment by the researcher(s) inquestion. This facilitates the verification of research results and theapplication of the rating in other research designs (Healy & Palepu,2001). Furthermore, unlike researchers, analysts also have access tounpublished and sometimes informal information disclosed duringanalysts meetings and conference calls (Healy & Palepu, 2001). Ana-lysts are also regarded as the primary and most influential users ofcorporate communication as they communicate with companies ona daily basis (e.g. Hirst et al., 1995; IASB, 2005; Revsine et al., 2004;Schipper, 1991). This gives them the expertise and experience toobjectively evaluate the quality of corporate disclosures.
For studies using researcher-developed instead of externally-developed disclosure indices, two main approaches can bedistinguished, i.e. content analysis (e.g. Aerts et al., 2007; Wallace,Naser, & Mora, 1994) or a dichotomous scoring mechanism wherean item is given a score of one if it is disclosed and a score ofzero otherwise (e.g. Bollen et al., 2006; Trabelsi et al., 2008). Bothapproaches however have several drawbacks. A major issue relatedto content analysis is the determination of the unit of analysis,i.e. words, sentences, paragraphs, etc. (Bravo, Abad, & Trombetta,2009). Studies which have applied content analysis often alsoclaim to not only measure the quantity of communication but alsothe quality as they assume that quantity and quality are posi-tively related. Obviously, a higher number of sentences or wordsdoes not necessarily imply that higher-quality information is pro-vided (Bravo et al., 2009). A potential drawback for studies using adichotomous scoring mechanism is that researchers have to relyon prior studies and/or survey evidence in an attempt to selectitems which are considered useful by investors, financial analystsand standard setters. Furthermore, the disclosure score is obtainedby counting the number of disclosed items. Consequently, againonly the quantity of disclosures is taken into account. Some stud-ies have attempted to incorporate the quality of disclosure byassigning weights. These weights are subjectively determined bythe researcher in question (e.g. Aerts et al., 2007) or based onsurveys among practitioners (e.g. Bollen et al., 2006). Finally, a gen-eral disadvantage of both approaches is the labor-intensity. As aconsequence, studies using these researcher-developed disclosureindices have a tendency to focus on one specific communicationchannel whereas the AIMR and BVFA-rating both evaluate the over-all communication quality by taking multiple disclosure channelsinto account. The latter is an important advantage of both analystdisclosure scores as different communication channels may be usedas complements or substitutes of each other (Botosan & Plumlee,2002; Leuz & Wysocki, 2008).
A potential drawback of analyst disclosure scores is that ana-lysts’ personal motivations may bring bias to the assigned ratings.However, the BVFA is aware of this possibility and imposes severalcontrol mechanisms to enhance to objectivity of the rating. First,companies are individually evaluated by more than one financialanalyst and only summary scores are presented. This reduces the
opportunity and incentives for an individual analyst to provide amore positive evaluation than warranted in order to gain favor withcompany management (Lang & Lundholm, 1993). Second, evalua-tions are based on a checklist of criteria constructed by the BVFA
n consultation with their members. Analysts cannot make theirvaluations capriciously as they have to provide a written justifica-ion for each item where they score a company’s disclosure policybove or below average. To further exclude any errors in the ana-yst ratings, an ex-post and ad hoc verification is performed by BVFAoard members and the top-ranked companies are subjected to andditional evaluation before a panel of financial analysts makes aecision on the definitive ranking.
A specific advantage of the BVFA-rating against other analystatings is that it also evaluates the quality of communication bymaller companies. In 2007, for example, the sample of screenedompanies consisted of 18 members of the Belgian blue-chipndex (i.e. Bel20-index), 18 midcaps and 13 smallcaps. For thesemaller companies, information asymmetries between insidersnd other market participants are potentially larger, which makesigh-quality communication even more necessary (BVFA, 2010a).urthermore, the BVFA-rating has been granted annually between951 and 2010. This differs from CIFAR and S&P ratings whichre only published intermittently making it impossible to evalu-te the year-to-year evolution of a specific company. Furthermore,IFAR and AIMR-ratings were no longer published after 1995nd 1996 respectively. As a consequence, researchers have raiseduestions about the applicability of these scores in contemporarytudies since disclosure requirements have substantially changedver time (Brown & Hillegeist, 2007; Ertimur, 2007; Hussainey,chleicher, & Walker, 2003).
Like the AIMR-rating, BVFA-scores provide a comprehensivevaluation of corporate communication quality, including bothuantitative and qualitative criteria. This is an important advan-age compared to researcher-based indices as well as the CIFARnd S&P-index which generally focus on the number of disclosedtems and not on disclosure content.
Finally, for this country-specific study, the BVFA-rating is pre-erred above cross-country indices as it covers a broad rangef Belgian listed companies. In particular, the BVFA-sample onverage includes 50 companies each year, while only 8 Belgian com-anies were for example included in the S&P index of 2002 (Khannat al., 2004).
.4. Control variables
A number of control variables which are assumed to influencehe profitability of insider trading is included in the regression. Therst control variable is the size of the transaction (TradeSize), which
s equal to the value of the net transaction scaled by the marketalue of the company at the beginning of the fiscal year.7 This con-rol variable is included because larger transactions are assumed toignal stronger beliefs in the future performance of the companyKarpoff, 1987). Thus, if insiders are in possession of higher-qualitynformation, we expect this to be reflected in a larger proportion ofhe firm being traded.
Second, we control for the size of the firm (FirmSize), which iseasured by the log of the market value of equity. A higher poten-
ial for information asymmetry is expected in small companiesBhushan, 1989; Collins, Kothari, & Rayburn, 1987; Grant, 1980).hese firms experience less extensive analyst following (Barth,asznik, & McNichols, 2001; Bhushan, 1989; Lang & Lundholm,996) and media coverage (Fang & Peress, 2009), which makes it
asier for insiders of small companies to have greater informa-ional benefits. Seyhun (1986), Finnerty (1976), and Betzer andheissen (2009), for example, documented that abnormal trading
7 Jenter (2005) argues it is preferable to measure trade size relative to some mea-ure of wealth or total equity instead of using absolute trade size. The former isssumed to be a more relevant measure of trading behavior.
f Law and Economics 37 (2014) 1– 14 7
profits have a negative relation with the size of the company. AlsoLakonishok and Lee (2001) found that insider purchases predictfuture returns only in small companies. Accordingly, we expectabnormal profits to be negatively associated with firm size.
The third control variable we include is the market-to-bookratio (MarketToBook) calculated as the ratio of the market valueof equity divided by the book value of equity, both measured atthe beginning of the fiscal year. This variable controls for a firm’sinvestment opportunity set as firms with a higher market-to-bookratio are assumed to have more unrecognized intangible assetsand valuable research and development projects. As a result, thesegrowth firms are characterized by a greater uncertainty regardingtheir fundamental value, allowing insiders to have greater infor-mational benefits with respect to future prospects and cash flows(Dierkens, 1991; Smith & Watts, 1992). Accordingly, we expect alarger amount of privileged information to be available to insidersof high market-to-book companies. On the other hand, previ-ous studies have also documented that insiders act as contrarianinvestors who take the under- or overvaluation by the marketinto account (e.g. Gregory, Tharyan, & Tonks, 2009; Jenter, 2005;Lakonishok & Lee, 2001; Rozeff & Zaman, 1998). More specifically,these studies assume that low market-to-book values signal under-valuation and thus good future stock market performance, whilehigh market-to-book ratios are associated with bad future shareperformances as they signal overvaluation. If insiders act as con-trarian investors, we expect insiders of low market-to-book firmsto earn higher abnormal profits.
A fourth control variable is the debt-to-asset ratio at the begin-ning of the fiscal year (Leverage). We include this variable to controlfor the proportion of external debt financing in a firm’s capitalstructure. Agency theory assumes that disclosure increases with theamount of external financing (Jensen & Meckling, 1976). In particu-lar, levered firms try to reduce agency costs by disseminating moreinformation, while creditors also produce additional informationabout the borrower in question (Aksu & Kosedag, 2006). In line withthis conjecture, Frankel and Li (2004) argue that firms with moreexternal financing issue more earnings forecasts in order to reduceinformation asymmetry. In addition, Bradbury (1992) found that alarger amount of voluntary segment information was disclosed byhighly-levered firms. Hossain, Perera, and Rahman (1995) extendedthis research and documented a positive association between thetotal amount of voluntary disclosed information and the ratio oflong-term debt to equity. Based on these findings, we argue thatinsiders in highly-levered firms earn lower abnormal returns.
Fifth, we include a dummy variable Sales equal to one if aninsider executes a sales transaction and zero otherwise. In general,previous literature suggests that purchases are more informativethan sales because sales transactions are not always driven by aprofit objective but may also result from diversification or liquidityneeds of the seller (Lakonishok & Lee, 2001). However, in a settingwith highly concentrated ownership, sales might be more profit-able, because controlling shareholders limit sales transactions forfear of losing control and only sell as a result of negative futureprospects. In addition, purchases by insiders of companies withconcentrated ownership probably have a lower information con-tent because they are often control-induced. As a consequence,for Belgian insiders, we expect higher abnormal returns for salestransactions.
Next, we control for the effect of cross-listings (CrossListing)using a dummy variable equal to one if a company is listed on a for-eign stock exchange and zero otherwise. According to the bondinghypothesis, cross-listing subjects companies to domestic as well
as foreign regulatory requirements (Coffee, 1999, 2002). Conse-quently, the additional and potentially stricter regulations likelymitigate the opportunities for insiders and controlling sharehol-ders to exploit their information benefits at the expense of other
8 view o
soigme2p(apga
sfitsiitsIoilh1Bmrsap
5
FmaeoDsbatat
cdtDDla22
iroto
D. Van Geyt et al. / International Re
hareholders (Chang & Corbitt, 2012; Sami & Zhou, 2008). More-ver, cross-listing is expected to further reduce the level ofnformation asymmetry by enhancing firm visibility throughreater analyst following, increased disclosure requirements, aore thorough investor monitoring and an increased media cov-
rage (e.g. Lambert, Leuz, & Verrecchia, 2007; Lang, Lins, & Miller,003; Lang, Lins, & Miller, 2004). Supporting these assumptions,rior studies by Korczak and Lasfer (2008) and Chang and Corbitt2012) found that insiders in cross-listed companies earn lessbnormal returns compared to insiders in domestically-listed com-anies. Consequently, given the above findings, we expect potentialains from insider trading to be lower in companies cross-listed on
foreign stock exchange.Finally, we control for a company’s ownership structure (Owner-
hip) using a variable equal to the percentage of shares held by theve largest shareholders. Generally, supervisory costs are expectedo be lower in companies with a highly concentrated ownershiptructure which increases the incentives of shareholders to mon-tor corporate insiders (Del Brio & Perote, 2007). However, thisncreased monitoring may not result in a lower level of informa-ion asymmetry as the interests of large shareholders and minorityhareholders are not necessarily aligned (Betzer & Theissen, 2009).n particular, controlling shareholders might use their power tobtain representation in the board of directors and acquire insidenformation thereby enlarging the information asymmetry prob-em between managers and controlling shareholders on the oneand and minority shareholders on the other hand (Demsetz,986). Empirically, support for both propositions has been found.etzer and Theissen (2009) provided evidence of larger abnor-al returns in widely held companies, while Demsetz (1986)
eported higher abnormal returns in companies with controllinghare ownership. In light of this mixed evidence, we do not makeny a priori assumptions on the relation between insider tradingrofitability and ownership structure.
. Sample selection
The insider trading data were obtained upon request from theSMA, which is entrusted with the supervision of the Belgian stockarket. Since 2005, insiders are required to report their trans-
ctions to this authority within five business days following thexecution. This legislation is based on the 2003 European Directiven insider dealing and market manipulation (i.e. the Market Abuseirective) and is similar to the requirements on the U.S. and other
tock markets.8 The database includes all insider trades reportedetween January 2006 and August 2010. The fact that transactionsre reported does, however, not guarantee that no illegal transac-
ions are included in our sample as the distinction between legalnd illegal transactions is made based on the fact whether transac-ions are inspired by inside information or not.9
8 Examples of other reporting requirements are New Zealand: continuous dis-losure of trades by all insiders (Tourani-Rad et al., 2003); Poland: 24-h disclosureeadline (Wisniewski & Bohl, 2005); Italy: no disclosure required when total quar-erly cumulative transactions is below D50,000, quarterly disclosure when between50,000 and D250,000, and disclosure within three business days when above250,000 (Bajo et al., 2009); U.K.: insiders must report as soon as possible and no
ater than five business days after the transaction (Fidrmuc et al., 2006), in addition black-out period before earnings announcements is imposed (Betzer & Theissen,009); U.S.: reporting no later than two days following the transaction (Cheng et al.,007).9 Following the European Market Abuse Directive, the Belgian law defines inside
nformation as “any information of a precise nature which has not been made public,elating, directly or indirectly, to one or more issuers of financial instruments or tone or more financial instruments and which, if it were made public, would be likelyo have a significant effect on the prices of those financial instruments or on the pricef related financial instruments” (Law of 2 August 2002, art. 2).
f Law and Economics 37 (2014) 1– 14
The annual BVFA-disclosure scores were gathered from theassociation’s website. For each individual company the yearly totalscore is disclosed as well as the subscores on four individual com-munication channels, i.e. annual reports, press releases, corporatewebsites and investor relation activities. Information on cross-listings and the daily return index for Belgian listed companies wascollected from Datastream. Data on the Brussels All Shares ReturnIndex were obtained from Euronext Brussels.10 Furthermore, dataon company size, market-to-book and debt-to-asset ratios weregathered from Worldscope. The Belfirst database of Bureau Van Dijkwas used to collect data on ownership structure.
The initial insider trading database included 4889 transactionsreported by insiders of 138 different companies from January 2006through August 2010. The database was filtered based on severalsample selection criteria. First, our study focuses only on openmarket purchases and sales. We expect over-the-counter trans-actions to be mainly inter-insider trades, which are not drivenby an informational benefit. Moreover, private transactions lacka market-determined price (Finnerty, 1976), which leads to apotentially large deviation between the negotiated and quotedstock price. Since the calculation of abnormal returns is based onmarket-determined prices, this could introduce a serious bias inthe estimation of insiders’ abnormal gains.
Second, trades involving the acquisition, exercise or conversionof options, warrants, or scripts, were filtered out. We expect thesetransactions to be less plausible to be information-motivated. Forexample, regarding the exercise of stock options, a study by Ofekand Yermack (2000) documented that option exercises are highlycorrelated with the subsequent sale of the underlying securities.Huddart and Ke (2007) therefore claim that the exercise-eventshould be excluded from the sample in order to avoid doublecounting.
Third, transactions were deleted if they were reported beforethe execution. The regulatory objective of this notification duty isto reduce the information asymmetry between insiders and othermarket participants as the knowledge of insider trades providesvaluable information (Givoly & Palmon, 1985). Consequently, oncetransactions are reported, we no longer expect abnormal tradingprofits as prices have already adjusted to the previous release ofthis information.
Fourth, transactions not reported in euro were also deleted. Thisfacilitates the calculation of the relative trade size as market capi-talization is reported in euro by Worldscope.
Fifth, transactions were filtered out if the company involved isnot included in the Brussels All Shares index. This index is usedas the benchmark index in the calculation of abnormal returns.Transactions were therefore removed in order to avoid bias in thecalculation of these returns.
If more than one trade was executed on the same day by thesame or different insiders from the same company, net transactionswere calculated. First, we sum the transaction size of all purchasesand sales respectively. Next, the total value of sold securities isdeducted from the total value of purchased securities.
Furthermore, net transactions less than 20 trading days apartwere deleted from the sample to filter out noise due to successivetrades. If no adjustment for event-clustering is made, the cumu-lative abnormal returns will not only capture the price reactionrelated to the transaction in question, but also to other tradescarried out later within the event window. These adjustments
for netting and event-clustering are consistent with the insidertrading literature (e.g. Betzer & Theissen, 2009; Fidrmuc et al., 2006;Friederich et al., 2002; Jaffe, 1974; Seyhun, 1986).
10 We are grateful to Euronext Brussels for providing the data on the Brussels AllShares Return Index (ISIN: BE0389550956).
D. Van Geyt et al. / International Review of Law and Economics 37 (2014) 1– 14 9
Table 1Sample selection.
Initial sample 4889
Applied filters- over-the-counter transactions 1241- trades not involving buying and selling of common shares 369- trades reported before execution 5- trades not reported in euro 27- trades of companies not included in benchmark 128- net transactions 488- event clustering adjustment 1760- missing stock price data 16- net trade value equal to 0 5- not in BVFA sample 430- missing data on control variables 13
Notes. Descriptive statistics for a pooled sample of net purchases and sales. Abnor-mal returns for sales transactions are multiplied by minus one. CARij(0,20) is equalto the event-specific cumulative abnormal return measured using a standard mar-ket model or a Dimson-adjusted market model depending on the number of zeroreturns. CommunicationQuality, AnnualReport, PressRelease, Website, and InvestorRe-lations respectively represent the disclosure quality score awarded by the BVFA tocompany j in the year of the insider trade for the total disclosure strategy (score on500), annual reports (score on 100), press releases (score on 100), corporate web-sites (score on 100) and investor relation activities (score on 200). TradeSize is equalto the eurovalue of the net transactions divided by the market value of the companyat the beginning of the fiscal year expressed in percentage. FirmSize is equal to thelog of the market value of the company expressed in millions of euros. MarketToBookis equal to the ratio of the market value of the company divided by the book value of
the profitability of sales transactions, our results seem to indicatethat sales (Sales) yield substantially higher abnormal returnscompared to purchases. This result is consistent with other studies
To be included in the sample, companies were also requiredo be listed 160 trading days prior to the event date and 20 dayshereafter in order to prevent missing data problems. Next, we fil-ered out transactions with a net transaction size equal to zero. Inddition, because a disclosure rating is not available for all Belgianisted companies, the sample was further reduced by eliminatingll transactions of companies for which no disclosure rating waseported. Finally, transactions were deleted due to missing dataith regard to the control variables. In Table 1 an overview isrovided of the applied filters and the number of deleted insiderransactions.
. Descriptive statistics
The application of the above filters resulted in a final sam-le of 407 firm-year observations. The sample consists of 199 neturchases and 208 net sales reported by insiders of 52 differentompanies. Additional descriptive statistics are provided in Table 2.his table shows that insider trading is, on average, profitable onhe Belgian stock market. Furthermore, a high standard deviation isbserved with regard to the total BVFA-scores, which indicates thathere is much variety in corporate communication quality acrossompanies. The lowest standard deviation is observed with respecto corporate websites. The quality of web disclosures thus seemso be relatively most uniform across companies. Also, corporateebsites seem to score the highest on the evaluation by financial
nalysts and fund managers. Table 2 further shows that, on aver-ge, only a small proportion of the company (0.05%) is traded bynsiders. The average (median) firm size is equal to 7.37 (7.20) mil-ion euro. Unreported results show that the market value of equityanges from 3.61 to 10.97 million euro, providing evidence of bothmall and large companies being incorporated in the sample. Theebt-to-asset ratio is on average equal to 26.26%, while the medianatio is equal to 24.94%. The average (median) market-to-book ratios equal to 2.65% (1.88%). Finally, the five largest shareholders ofompanies included in our sample on average hold 52.86% of allhares.
Table 3 reports Spearman correlation coefficients of the regres-ion variables. The highly significant negative correlation betweenhe total disclosure score and insiders’ abnormal profits seems tondicate that higher-quality communication contributes to reduc-ng insider trading profitability. Comparing the correlation betweennsider trading profitability and the four separate disclosure scores,
nly press releases and investor relation activities seem to be sig-ificantly correlated with insiders’ abnormal profits. The separateisclosure scores are furthermore all highly correlated with theotal score and positively correlated with each other. The latter
equity at the beginning of the fiscal year expressed in percentage. Leverage is equalto the debt-to-asset ratio at the beginning of the fiscal year expressed in percentage.Ownership is equal to the percentage of shares held by the five largest shareholders.
finding shows a certain consistency within the communicationstrategy of companies. Companies do not seem to devote theirefforts to one particular communication channel, but enhance thequality of all forms of corporate communication.
7. Results
7.1. The impact of the overall corporate communication qualityon insider trading profitability
Table 4 reports the OLS-regression results. These results gen-erally support our expectations regarding the control variables.First, with respect to TradeSize, we expect a positive relationbetween transaction size and insider trading gains (Karpoff, 1987).If insiders put a higher proportion of their stock ownership at stake,they earn higher abnormal profits. This finding is consistent withSeyhun (1986) and Cheuk et al. (2006).11 Second, our results alsoshow that FirmSize is negatively associated with the profitabilityof insider trading.12 This finding seems to confirm that there is ahigher potential for information asymmetry in small companies(Cheuk et al., 2006; Seyhun, 1986). Third, the market-to-book ratio(MarketToBook) is negatively associated with abnormal profits.This finding indicates that investing in undervalued securitiesyields positive abnormal returns, while investing in overvaluedsecurities yields negative abnormal returns. Furthermore, Leverage,which measures the proportion of external financing, does notseem to influence the magnitude of insiders’ profits. With regard to
11 Similar results are obtained when measuring trade size as the percentage of thenumber of shares traded relative to the number of shares outstanding.
12 Similar results are obtained when using alternative measures for company size.
10 D. Van Geyt et al. / International Review of Law and Economics 37 (2014) 1– 14
Notes. Spearman correlations for a pooled sample of net purchases and sales. CARij(0,20) is equal to the event-specific cumulative abnormal return measured using a standardmarket model or a Dimson-adjusted market model depending on the number of zero returns. Abnormal returns for sales transactions are multiplied by minus one. Commu-nicationQuality, AnnualReport, PressRelease, Website, and InvestorRelations respectively represent the disclosure quality score awarded by the BVFA to company j in the yearof the insider trade for the total disclosure strategy, annual reports, press releases, corporate websites and investor relation activities (standardized scores on 100). TradeSizeis equal to the eurovalue of the net transactions divided by the market value of the company at the beginning of the fiscal year expressed in percentage. FirmSize is equal tothe log of the market value of the company expressed in millions of euros. MarketToBook is equal to the ratio of the market value of the company divided by the book valueof equity at the beginning of the fiscal year expressed in percentage. Leverage is equal to the debt-to-asset ratio at the beginning of the fiscal year expressed in percentage.S CrossLe y the
psfoo
TO
NtCembvzS
ales is a dummy variable equal to one for net sale transactions and zero otherwise.xchange and zero otherwise. Ownership is equal to the percentage of shares held b
* Significant at 5%.
erformed in countries with highly concentrated ownershiptructures (e.g. Cheuk et al., 2006). Furthermore, support is also
ound for the bonding hypothesis as cross-listing (CrossListing)n foreign stock exchanges negatively influences the magnitudef insiders’ abnormal returns. Finally, the ownership structure
able 4LS-regression results: total disclosure score.
otes. OLS-regression results for a pooled sample of net purchases and sales. Abnormahe event-specific cumulative abnormal return measured using a standard market modommunicationQuality represents the disclosure quality score on 100 points awarded byurovalue of the net transactions divided by the market value of the company at the begarket value of the company expressed in millions of euros. MarketToBook is equal to the
eginning of the fiscal year expressed in percentage. Leverage is equal to the debt-to-asseariable equal to one for net sale transactions and zero otherwise. CrossListing is a dummero otherwise. Ownership is equal to the number of shares held by the five largest sharignificance levels are two-tailed when ‘Expected sign’ is a ‘?’ and one-tailed otherwise.
* Significant at 10%.** Significant at 5%.
*** Significant at 1%.
isting is a dummy variable equal one if a company is cross-listed on a foreign stockfive largest shareholders.
(Ownership) does not seem to affect the profitability of insidertrading.
With regard to our main research question, model 2 of Table 4shows a highly significant negative coefficient on disclosure quality.This finding implies that high-quality corporate communication is
l returns for sales transactions are multiplied by minus one. CARij(0,20) is equal toel or a Dimson-adjusted market model depending on the number of zero returns.
the BVFA to company j in the year of the insider trade. TradeSize is equal to theinning of the fiscal year expressed in percentage. FirmSize is equal to the log of the
ratio of the market value of the company divided by the book value of equity at thet ratio at the beginning of the fiscal year expressed in percentage. Sales is a dummyy variable equal one if a company is cross-listed on a foreign stock exchange and
eholders. Standard errors are adjusted for firm-clustering and heteroskedasticity.
D. Van Geyt et al. / International Review of Law and Economics 37 (2014) 1– 14 11
Notes. OLS-regression results for a pooled sample of net purchases and sales. Abnormal returns for sales transactions are multiplied by minus one. CARij(0,20) is equal tothe event-specific cumulative abnormal return measured using a standard market model or a Dimson-adjusted market model depending on the number of zero returns.AnnualReport, PressRelease, Website, and InvestorRelations respectively represent the disclosure quality score on 100 for annual reports, press releases, corporate websites andinvestor relation activities awarded by the BVFA to company j in the year of the insider trade. TradeSize is equal to the eurovalue of the net transactions divided by the marketvalue of the company at the beginning of the fiscal year expressed in percentage. FirmSize is equal to the log of the market value of the company expressed in millions ofeuros. MarketToBook is equal to the ratio of the market value of the company divided by the book value of equity at the beginning of the fiscal year expressed in percentage.Leverage is equal to the debt-to-asset ratio at the beginning of the fiscal year expressed in percentage. Sales is a dummy variable equal to one for net sale transactions andzero otherwise. CrossListing is a dummy variable equal one if a company is cross-listed on a foreign stock exchange and zero otherwise. Ownership is equal to the number ofshares held by the five largest shareholders. Standard errors are adjusted for firm-clustering and heteroskedasticity. VIF are the variance inflation factors. Significance levelsa
eo
7c
miwtctia
Hftlrdccoba
lw
miwc
re two-tailed when ‘Expected sign’ is a ‘?’ and one-tailed otherwise.* Significant at 10%.
** Significant at 5%.*** Significant at 1%.
ffective in reducing information asymmetry and the profitabilityf insider trading, consistent with Hypothesis 1.13
.2. The impact of the quality of individual communicationhannels on insider trading profitability
Next, we evaluate whether the individual corporate com-unication channels have a different impact on reducing the
nformational benefits of insiders.14 To investigate this proposition,e first performed four separate regression analyses containing
he control variables and the disclosure score on the respectiveommunication channel (Table 5, models 3–6). It must be notedhat, although the BVFA values investor relation activities twice asmportant as the other communication channels, we standardizedll disclosure scores to a score on 100 for reasons of comparability.
In general, the regression results in Table 5 are consistent withypothesis 2 and show that the influence of disclosure quality dif-
ers across communication channels. First, our results show thathe quality of corporate websites does not seem to affect theevel of information asymmetry and the resulting insider tradingeturns (model 5). Contrary to all other communications, websiteisclosures are not directly and primarily aimed at the investorommunity. In addition, the quality of websites and web dis-losures is rather uniform across companies, which limits the
pportunities to have a decisive impact on insiders’ informationalenefits. Second, models 3 and 4 of Table 5 report significant neg-tive coefficients on AnnualReports and PressReleases. This finding
13 Similar results are obtained when alternative estimation windows for the cumu-ative abnormal returns are used and when no adjustment for overlapping event
indows is applied. Results are available upon request.14 It must be noted that we do not investigate the relation between insiders’ abnor-al returns and the content or quality of a particular disclosure. As such, we do not
nvestigate the reaction of insiders to good or bad news. Our analysis only evaluateshether the average quality of all disclosures through a specific communication
hannel during a fiscal year affects the abnormal returns from insider trading.
suggests that the quality of these communication channels has animpact on the profitability of insider trading. The most importanttool in reducing insiders’ gains and getting valuable informationacross to the investor community, however, seems to be a firm’sinteractions with financial analysts and fund managers as sug-gested by the coefficient on InvestorRelations in model 6. Theinformation communicated through investor relation activities istypically informal, unaudited and not subject to litigation. How-ever, as suggested by Brown and Hillegeist (2007), the credibilityof investor relation activities might be enhanced by reputationalconcerns of managers.
Following the approach of Botosan and Plumlee (2002), we alsoperformed a regression analysis including the disclosure score onall communication channels (Table 5, model 7). The authors foundthat, although the correlation between individual communicationchannels might induce multicollinearity, not controlling for othertypes of disclosure might lead to a correlated omitted variable biasand erroneous conclusions regarding the impact of a particularcommunication channel. However, we do not expect any multi-collinearity problems in the aggregated regression model (model7) as the correlations between the independent variables reportedin Table 3 are below the 0.7 limit identified by Kervin (1992). Inaddition, VIF values are well below the recommended cutoff of 10(Chatterjee & Hadi, 2006). Regression results in model 7 confirmour previous results.
In sum, our regression results seem to indicate that the contentand quality of annual reports and press releases has a comparableimpact on the level of information asymmetry. Furthermore, they
suggest that the disclosure quality of investor relation activities hasthe largest contribution to the reduction of information asymmetryand the resulting abnormal trading profits.15 This communication
15 Using alternative estimation windows for the cumulative abnormal returns doesnot alter our conclusions. Results are available upon request.
12 D. Van Geyt et al. / International Review of Law and Economics 37 (2014) 1– 14
Table 6Robustness checks.
Variables Expected sign Model 8: CARMMij(0,20) Model 9: CARij(0,5) Model 10: CARij(0,10)
Notes. OLS-regression results for a pooled sample of net purchases and sales. Abnormal returns for sales transactions are multiplied by minus one. CARMM ij(0,20) is equal to theevent-specific cumulative abnormal return measured using a standard market model (Model 8). CARij(0,5) and CARij(0,10) are equal to the event-specific cumulative abnormalreturn from day 0 to day 5 or day 10 respectively and are measured using a standard market model or a Dimson-adjusted market model depending on the number of zeroreturns (Models 9 and 10). CommunicationQuality represents the disclosure quality score on 100 points awarded by the BVFA to company j in the year of the insider trade.TradeSize is equal to the eurovalue of the net transactions divided by the market value of the company at the beginning of the fiscal year expressed in percentage. FirmSizeis equal to the log of the market value of the company expressed in millions of euros. MarketToBook is equal to the ratio of the market value of the company divided by thebook value of equity at the beginning of the fiscal year expressed in percentage. Leverage is equal to the debt-to-asset ratio at the beginning of the fiscal year expressed inpercentage. Sale is a dummy variable equal to one for net sale transactions and zero otherwise. CrossListing is equal to the number of foreign exchanges on which a companyis listed in a particular fiscal year. Ownership is a dummy variable equal to one if a company has a concentrated ownership structure and zero otherwise. Standard errors area ed wh
csdnia
8
seaoum8asq
errcwaw
9
cl
w
djusted for firm-clustering and heteroskedasticity. Significance levels are two-tail* Significant at 10%.
** Significant at 5%.*** Significant at 1%.
hannel is aimed directly at the investor community and exhibits aubstantial level of discretion. This provides companies with moreegrees of freedom to customize their disclosures to the investors’eeds. Possibly, the importance of this voluntary disclosure channel
s further enhanced by the broad range of disseminated informations well as the high degree of timeliness.
. Robustness checks
In order to check the robustness of our results, we present someensitivity analyses for the event study methodology. First, we re-xamine the relation between corporate communication qualitynd insider trading profitability using cumulative abnormal returnsnly calculated by means of a standard market model instead ofsing a combination of a standard market model and a marketodel adjusted for thin trading. Results reported in Table 6, model
show that an alternative estimation of the insiders’ gains does notlter our conclusions.16 Consistent with results in Table 4, we find aignificant negative association between corporate communicationuality and the abnormal insider trading profits.
Second, we test whether our conclusions hold if alternativevent windows are used and estimate the cumulative abnormaleturns from day 0 to day 5 and day 10 respectively. Regressionesults are reported in models 9 and 10 of Table 6. They indi-ate that our results are not sensitive to the length of the eventindow. Again, we find support for Hypothesis 1 and find that
bnormal returns to insiders are significantly lower in companiesith a higher communication quality.
. Conclusion
High-quality communication is a key feature of a firm’sorporate governance strategy. Using a sample of Belgianisted companies, this paper investigated whether high-quality
16 Also, results on the impact of different communication channels are similarhen cumulative abnormal returns are calculated using the standard market model.
en “Expected sign” is a “?” and one-tailed otherwise.
communication can reduce insider trading profitability, and thusinformation asymmetry between insiders and outsiders. A uniquefeature of our analysis is that we have proxied the quality of cor-porate communication by disclosure scores that were assigned byfinancial analysts and fund managers who are familiar with thepeculiarities and demands of the companies’ investor community.One of the advantages of using these scores is that they are moreobjective than researcher-assigned scores. Consistent with expec-tations, we have found that high-quality communication limits theprofitability of insider trading. Moreover, we have reported evi-dence on the communication channels that contribute most to thereduction of information asymmetry between insiders and out-siders and the resulting insider trading gains. In particular, wehave documented that, although disclosures in mandatory annualreports have some impact, voluntary disclosure channels, such asinvestor relation programs and press releases, are the most effec-tive channels to reduce information asymmetry between insidersand outsiders. Furthermore, since the quality of corporate web-sites is rather uniform across companies, they cannot explain thevariance in insider trading profitability.
The results of the study are of interest to academics and reg-ulators. From an academic point of view, this study contributes tovarious strands of literature. First, it contributes to the academic lit-erature on insider trading profitability by providing evidence on theimpact of high-quality communication and of the different chan-nels through which companies can communicate on insider tradingprofitability. Second, the study also contributes to the academicliterature on the relation between disclosure quality and informa-tion asymmetry by using insider trading profitability as a proxy forinformation asymmetry rather than, for example, bid-ask spreadsor the probability of informed trading. Furthermore, we are also thefirst to investigate the disclosure–information asymmetry relationin a French civil law country using disclosures scores assigned byprofessional financial analysts and which rate the quality of differ-
ent communication channels.
The results are of interest to regulators for the findings gen-erally underline the importance of high-quality communicationas an instrument to prevent information inequities and unfair
view o
ecsamowmbnl
A
fCRSot(
R
A
A
A
A
A
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
D. Van Geyt et al. / International Re
nrichment by privileged insiders and to stimulate a more effi-ient allocation of resources. Interestingly, however, our resultshow that whereas regulators primarily focus on annual reportsnd backward-looking financial statements information, this com-unication channel is not the most effective in reducing the level
f information asymmetry. By contrast, investor relation activities,hich are used to communicate timely and forward-looking infor-ation on a voluntary basis, appear to be the most effective. We
elieve that this finding is relevant for regulators and may shedew light on the discussion concerning the shift toward more or
ess regulation of markets.
cknowledgements
The authors appreciate the helpful comments and suggestionsrom Millicent Chang and the participants of the 34th Annualongress organized by the European Accounting Association inome, Italy (April 2011). We would like to thank the Financialervices and Markets Authority for making their data available tour study and gratefully acknowledge the financial support fromhe “Special Research Fund” (BOF, Bijzonder Onderzoeksfonds)project: 01J06209) and the “Hercules-project”.
eferences
erts, W., Cormier, D., & Magnan, M. (2007). The association between web-basedcorporate performance disclosure and financial analyst behaviour under differ-ent governance regimes. Corporate Governance: An International Review, 15(6),1301–1329.
IMR. (2000). AIMR corporate disclosure survey: A report to AIMR. Saint-Louis, MO:Fleishman-Hillard Research.
ksu, M., & Kosedag, A. (2006). Transparency and disclosure scores and their deter-minants in the Istanbul stock exchange. Corporate Governance: An InternationalReview, 14(4), 277–296.
ktas, N., de Bodt, E., & Van Oppens, H. (2008). Legal insider trading and marketefficiency. Journal of Banking and Finance, 32(7), 1379–1392.
ussenegg, W., & Ranzi, R. (2008). Corporate insider trading and the short-run priceimpact of private information in continental Europe. Working paper
aiman, S., & Verrecchia, R. E. (1996). The relation among capital markets, finan-cial disclosure, production efficiency, and insider trading. Journal of AccountingResearch, 34(1), 1–22.
ajo, E., Bigelli, M., Hillier, D., & Petracci, B. (2009). The determinants of regulatorycompliance: An analysis of insider trading disclosures in Italy. Journal of BusinessEthics, 90(3), 331–343.
all, R., Jayaraman, S., & Shivakumar, L. (2011). Audited financial reporting and vol-untary disclosure as complements: A test of the confirmation hypothesis. Workingpaper.
arth, M. E., Kasznik, R., & McNichols, M. F. (2001). Analyst coverage and intangibleassets. Journal of Accounting Research, 39(1), 1–34.
etzer, A., & Theissen, E. (2009). Insider trading and corporate governance: The caseof Germany. European Financial Management, 15(2), 402–429.
hushan, R. (1989). Firm characteristics and analyst following. Journal of Accountingand Economics, 11(2–3), 255–274.
ollen, L., Hassink, H., & Bozic, G. (2006). Measuring and explaining the quality ofinternet investor relations activities: A multinational empirical analysis. Inter-national Journal of Accounting Information Systems, 7(4), 273–298.
onsón, E., & Escobar, T. (2002). A survey on voluntary disclosure on the internet:Empirical evidence from 300 European Union companies. International Journalof Digital Accounting Research, 2(1), 27–51.
onsón, E., & Escobar, T. (2006). Digital reporting in Eastern Europe: An empiricalstudy. International Journal of Accounting Information Systems, 7(4), 299–318.
otosan, C. A. (1997). Disclosure level and the cost of equity capital. AccountingReview, 72(3), 323–349.
otosan, C. A., & Plumlee, M. A. (2002). A re-examination of disclosure level and theexpected cost of equity capital. Journal of Accounting Research, 40(1), 21–40.
radbury, M. E. (1992). Voluntary disclosure of financial segment data: New Zealandevidence. Accounting and Finance, 32(1), 15–26.
ravo, F., Abad, C., & Trombetta, M. (2009). Disclosure indices design: Does it makea difference? Revista de contabilidad, 12(2), 253–277.
rown, S., & Hillegeist, S. A. (2007). How disclosure quality affects the level ofinformation asymmetry. Review of Accounting Studies, 12(2–3), 443–477.
rown, P., & Niederhoffer, V. (1968). The predictive content of quarterly earnings.Journal of Business, 41(4), 488–497.
rown, L. D., & Rozeff, M. S. (1979). Univariate time-series models of quarterlyaccounting earnings per share: A proposed model. Journal of Accounting Research,17(1), 179–189.
ushman, R. M., & Smith, A. J. (2001). Financial accounting information and corpo-rate governance. Journal of Accounting and Economics, 32(1–3), 237–333.
f Law and Economics 37 (2014) 1– 14 13
Buysschaert, A., Deloof, M., & Jegers, M. (2004). Equity sales in Belgian corpo-rate groups: Expropriation of minority shareholders? A clinical study. Journalof Corporate Finance, 10(1), 81–103.
BVFA. (2010a). Press release 2010. http://www.bvfa.be/en/fr1.htmlBVFA. (2010b). Grids 2010. http://www.bvfa.be/en/fr1.htmlCarlin, W., & Mayer, C. (2003). Finance, investment, and growth. Journal of Financial
Economics, 69(1), 191–226.Chang, M., & Corbitt, R. (2012). The effect of cross-listing on insider trading returns.
Accounting and Finance, 52(3), 723–741.Chang, M., D’Anna, G., Watson, L., & Wee, M. (2008). Does disclosure quality via
investor relations affect information asymmetry? Australian Journal of Manage-ment, 33(2), 375–390.
Chang, M., Hillman, R., & Watson, I. D. (2005). Are corporate governance mechanismseffective in reducing insider trading profits? Company and Securities Law Journal,23(3), 165–180.
Chang, L. S., & Most, K. S. (1985). The perceived usefulness of financial statements forinvestors’ decisions. University Presses of Florida.
Chatterjee, S., & Hadi, A. S. (2006). Regression analysis by example (4th ed., pp. ). NewYork, NY: Wiley.
Chen, W.-P., Chung, H., Lee, C., & Liao, W.-L. (2007). Corporate governance andequity liquidity: Analysis of S&P transparency and disclosure rankings. CorporateGovernance: An International Review, 15(4), 644–660.
Cheng, S., Nagar, V., & Rajan, M. V. (2007). Insider trades and private information:The special case of delayed-disclosure trades. Review of Financial Studies, 20(5),1833–1864.
Cheuk, M.-Y., Fan, D. K., & So, R. W. (2006). Insider trading in Hong Kong: Somestylized facts. Pacific-Basin Finance Journal, 14(1), 73–90.
Coffee, J. C. (1999). Future as history: The prospects for global convergence in corpo-rate governance and its implications. Northwestern University Law Review, 93(3),641–707.
Coffee, J. C. (2002). Racing towards the top?: The impact of cross-listings andstock market competition on international corporate governance. Columbia LawReview, 102(7), 1757–1831.
Collins, D. W., Kothari, S. P., & Rayburn, J. D. (1987). Firm size and the informationcontent of prices with respect to earnings. Journal of Accounting and Economics,9(2), 111–138.
Del Brio, E. B., & Perote, J. (2007). What enhances insider trading profitability?Atlantic Economic Journal, 35(2), 173–188.
Demsetz, H. (1986). Corporate control, insider trading, and rates of return. TheAmerican Economic Review, 76(2), 313–316.
Diamond, D. W. (1985). Optimal release of information by firms. Journal of Finance,40(4), 1071–1094.
Dierkens, N. (1991). Information asymmetry and equity issues. Journal of Financialand Quantitative Analysis, 26(2), 181–199.
Dimson, E. (1979). Risk measurement when shares are subject to infrequent trading.Journal of Financial Economics, 7(2), 197–226.
Durnev, A. R. T., & Kim, E. H. (2005). To steal or not to steal: Firm attributes, legalenvironment, and valuation. Journal of Finance, 60(3), 1461–1493.
Ertimur, Y. (2007). Discussion of “How disclosure quality affects the level of infor-mation asymmetry”. Review of Accounting Studies, 12(2–3), 479–485.
Faccio, M., & Lang, L. H. P. (2002). The ultimate ownership of western Europeancorporations. Journal of Financial Economics, 65(3), 365–395.
Fang, L., & Peress, J. (2009). Media coverage and the cross-section of stock returns.Journal of Finance, 64(5), 2023–2052.
Fidrmuc, J. P., Goergen, M., & Renneboog, L. (2006). Insider trading, news releases,and ownership concentration. Journal of Finance, 61(6), 2931–2973.
Finnerty, J. E. (1976). Insiders and market efficiency. Journal of Finance, 31(4),1141–1148.
Frankel, R., & Li, X. (2004). Characteristics of a firm’s information environment andthe information asymmetry between insiders and outsiders. Journal of Account-ing and Economics, 37(2), 229–259.
Friederich, S., Gregory, A., Matatko, J., & Tonks, I. (2002). Short-run returns aroundthe trades of corporate insiders on the London stock exchange. European Finan-cial Management, 8(1), 7–30.
Gelb, D., & Zarowin, P. (2002). Corporate disclosure policy and the informativenessof stock prices. Review of Accounting Studies, 7(1), 33–52.
Givoly, D., & Palmon, D. (1985). Insider trading and the exploitation of inside infor-mation: Some empirical evidence. Journal of Business, 58(1), 69–87.
Grant, E. B. (1980). Market implications of differential amounts of interim informa-tion. Journal of Accounting Research, 18(1), 255–268.
Gregory, A., Tharyan, R., & Tonks, I. (2009). Insider trading in glamour and value firms.Working paper.
Healy, P. M., Hutton, A., & Palepu, K. G. (1999). Stock performance and interme-diation changes surrounding sustained increases in disclosure. ContemporaryAccounting Research, 16(3), 485–520.
Healy, P. M., & Palepu, K. G. (2001). Information asymmetry, corporate disclosure,and the capital markets: A review of the empirical disclosure literature. Journalof Accounting and Economics, 31(1–3), 405–440.
Heflin, F. L., Shaw, K. W., & Wild, J. J. (2005). Disclosure policy and market liquidity:Impact of depth quotes and order sizes. Contemporary Accounting Research, 22(4),829–865.
Hirst, D. E., Koonce, L., & Simko, P. J. (1995). Investor reactions to financial analystsresearch reports. Journal of Accounting Research, 33(2), 335–351.
Hope, O. K. (2003). Disclosure practices, enforcement of accounting standards,and analysts’ forecast accuracy: An international study. Journal of AccountingResearch, 41(2), 235–272.
trading under lax law enforcement. International Review of Law and Economics,
4 D. Van Geyt et al. / International Re
ossain, M., Perera, M. H. B., & Rahman, A. R. (1995). Voluntary disclosure in theannual reports of New Zealand companies. Journal of International Financial Man-agement and Accounting, 6(1), 69–87.
uddart, S. J., & Ke, B. (2007). Information asymmetry and cross-sectional variationin insider trading. Contemporary Accounting Research, 24(1), 195–232.
ussainey, K., Schleicher, T., & Walker, M. (2003). Undertaking large-scale disclosurestudies when AIMR-FAF ratings are not available: The case of prices leadingearnings. Accounting and Business Research, 33(4), 275–294.
ASB. (2005). International financial reporting standards. London: InternationalAccounting Standards Board.
affe, J. F. (1974). Special information and insider trading. Journal of Business, 47(3),410–428.
ensen, M. C., & Meckling, W. H. (1976). Theory of the firm: Managerial behav-ior, agency costs and ownership structure. Journal of Financial Economics, 3(4),305–360.
enter, D. (2005). Market timing and managerial portfolio decisions. Journal ofFinance, 60(4), 1903–1949.
ones, M. J., & Xiao, J. Z. (2004). Financial reporting on the Internet by 2010: Aconsensus view. Accounting Forum, 28(3), 237–263.
arpoff, J. M. (1987). The relation between price changes and trading volume: Asurvey. Journal of Financial and Quantitative Analysis, 22(1), 109–126.
ervin, J. B. (1992). Methods for business research. New York, NY: Harper Collins.hanna, T., Palepu, K. G., & Srinivasan, S. (2004). Disclosure practices of foreign
companies interacting with U.S. markets. Journal of Accounting Research, 42(2),475–508.
orczak, A., & Lasfer, M. A. (2008). Does cross-listing mitigate insider trading? Workingpaper.
yle, A. S. (1985). Continuous auctions and insider trading. Econometrica, 53(6),1315–1335.
a Porta, R., Lopez-de-Silanes, F., Shleifer, A., & Vishny, R. W. (1997). Legal determi-nants of external finance. Journal of Finance, 52(3), 1131–1150.
a Porta, R., Lopez-de-Silanes, F., Shleifer, A., & Vishny, R. W. (1998). Law and finance.Journal of Political Economy, 106(6), 1113–1155.
akhal, F. (2009). Does corporate disclosure policy change financial analysts’behaviour? Evidence from France. Journal of Accounting and OrganizationalChange, 5(3), 344–361.
akonishok, J., & Lee, I. (2001). Are insider trades informative? Review of FinancialStudies, 14(1), 79–111.
ambert, R., Leuz, C., & Verrecchia, R. E. (2007). Accounting information, disclosure,and the cost of capital. Journal of Accounting Research, 45(2), 385–420.
ang, M. H., & Lundholm, R. J. (1993). Cross-sectional determinants of analyst ratingsof corporate disclosures. Journal of Accounting Research, 31(2), 246–271.
ang, M. H., & Lundholm, R. J. (1996). Corporate disclosure policy and analyst behav-ior. Accounting Review, 71(4), 467–492.
ang, M. H., Lins, K. V., & Miller, D. P. (2003). ADRs, analysts, and accuracy: Doescross listing in the United States improve a firm’s information environment andincrease market value? Journal of Accounting Research, 41(2), 317–345.
ang, M. H., Lins, K. V., & Miller, D. P. (2004). Concentrated control, analyst following,and valuation: Do analysts matter most when investors are protected least?Journal of Accounting Research, 42(3), 589–623.
euz, C., & Wysocki, P. (2008). Economic consequences of financial reporting anddisclosure regulation: A review and suggestions for future research. Working paper.
in, J.-C., & Howe, J. S. (1990). Insider trading in the OTC market. Journal of Finance,45(4), 1273–1284.
itvak, K. (2007). The effect of the Sarbanes–Oxley act on non-US companies cross-listed in the US. Journal of Corporate Finance, 13(2–3), 195–228.
undholm, R. J. (2003). Historical accounting and the endogenous credibility ofcurrent disclosures. Journal of Accounting, Auditing and Finance, 18(1), 207.
undholm, R. J., & Myers, L. A. (2002). Bringing the future forward: The effect ofdisclosure on the returns–earnings relation. Journal of Accounting Research, 40(3),809–839.
acKinlay, A. C. (1997). Event studies in economics and finance. Journal of EconomicLiterature, 35(1), 13–39.
allin, C. (2002). The relationship between corporate governance, transparencyand financial disclosure. Corporate Governance: An International Review, 10(4),253–255.
arston, C., & Polei, A. (2004). Corporate reporting on the Internet by German com-panies. International Journal of Accounting Information Systems, 5(3), 285–311.
f Law and Economics 37 (2014) 1– 14
Marston, C., & Straker, M. (2001). Investor relations: A European survey. CorporateCommunications: An International Journal, 6(2), 82–93.
McNichols, M., & Manegold, J. G. (1983). The effect of the information environmenton the relationship between financial disclosure and security price variability.Journal of Accounting and Economics, 5(0), 49–74.
Mitton, T. (2002). A cross-firm analysis of the impact of corporate governance onthe east Asian financial crisis. Journal of Financial Economics, 64(2), 215–241.
Nagar, V., Nanda, D., & Wysocki, P. (2003). Discretionary disclosure and stock-basedincentives. Journal of Accounting and Economics, 34(1–3), 283–309.
OECD. (2004). OECD principles of corporate governance. Paris: OECD.Ofek, E., & Yermack, D. (2000). Taking stock: Equity-based compensation and the
evolution of managerial ownership. Journal of Finance, 55(3), 1367–1384.Patel, S. A., Balic, A., & Bwakira, L. (2002). Measuring transparency and disclosure at
firm-level in emerging markets. Emerging Markets Review, 3(4), 325–337.Patel, S. A., & Dallas, G. (2002). Transparency and disclosure: Overview of methodology
and study results—United States. Working paper.Revsine, L., Collins, D. W., & Johnson, W. B. (2004). Financial reporting and analysis.
Upper Saddle River, NJ: Prentice Hall.Rogers, W. (1993). Quantile regression standard errors. Stata Technical Bulletin, 2(9),
19–22.Rozeff, M. S., & Zaman, M. A. (1998). Overreaction and insider trading: Evidence
from growth and value portfolios. Journal of Finance, 53(2), 701–716.Salter, S. B. (1998). Corporate financial disclosure in emerging markets: Does eco-
nomic development matter? International Journal of Accounting, 33(2), 211–234.Sami, H., & Zhou, H. (2008). The economic consequences of increased disclosure:
Evidence from cross-listings of Chinese firms. Journal of International FinancialManagement and Accounting, 19(1), 1–27.
Schipper, K. (1991). Commentary on analysts’ forecasts. Accounting Horizons, 5(4),105–121.
Sengupta, P. (1998). Corporate disclosure quality and the cost of debt. AccountingReview, 73(4), 459–474.
Seyhun, H. N. (1986). Insiders’ profits, costs of trading, and market efficiency. Journalof Financial Economics, 16(2), 189–212.
Smith, C. W., & Watts, R. (1992). The investment opportunity set and corporatefinancing, dividend, and compensation policies. Journal of Financial Economics,32(3), 263–292.
Stocken, P. C. (2000). Credibility of voluntary disclosure. Rand Journal of Economics,31(2), 359–374.
Tison, M., & Ravelingien, F. (2007). Preventie en repressie van misbruik van voorweten-schap in Belgisch recht na de omzetting van richtlijn 2003/6 inzake marktmisbruik.Belgisch kapitaalmarktrecht op Europese leest. Antwerpen/Oxford: Intersentia.
Tourani-Rad, A., Etebari, A., & Gilbert, A. (2003). Profitability of insider trad-ing: New zealand evidence. University of Auckland Business Review, 5(2),1–15.
Trabelsi, S., Labelle, R., & Dumontier, P. (2008). Incremental voluntary disclosure oncorporate websites, determinants and consequences. Journal of ContemporaryAccounting and Economics, 4(2), 120–155.
Vanstraelen, A., Zarzeski, M. T., & Robb, S. W. G. (2003). Corporate nonfinancialdisclosure practices and financial analyst forecast ability across three Europeancountries. Journal of International Financial Management and Accounting, 14(3),249–278.
Vergoossen, R. G. A. (1993). The use and perceived importance of annual reportsby investment analysts in the Netherlands. European Accounting Review, 2(2),219–244.
Verrecchia, R. E. (2001). Essays on disclosure. Journal of Accounting and Economics,32(1–3), 97–180.
Wallace, R. S. O., Naser, K., & Mora, A. (1994). The relationship between the com-prehensiveness of corporate annual reports and firm characteristics in Spain.Accounting and Business Research, 25(97), 41–53.
Welker, M. (1995). Disclosure policy, information asymmetry, and liquidity in equitymarkets. Contemporary Accounting Research, 11(2), 801–827.
Wisniewski, T. P., & Bohl, M. T. (2005). The information content of registered insider
25(2), 169–185.Zhang, W., Cahan, S. F., & Allen, A. C. (2005). Insider trading and pay–performance
sensitivity: An empirical analysis. Journal of Business Finance and Accounting,32(9–10), 1887–1919.
