Conference calls have become increasingly common in recentyears, yet there is little empirical evidence regarding the effectof conference calls on executive compensation. In this study, weexamine the effect of voluntary disclosures on equity incentives.We hypothesize that voluntary disclosures, as measured by confer-ence calls, affect executive compensation contracts. Using a datasetof 6263 firm-year observations from both conference call and non-conference call firms, our results are consistent with the argumentthat the board of directors substitutes voluntary disclosures formore costly corporate governance mechanisms. Alternatively, infirms where CEOs have less equity incentives, the owners demandmore voluntary disclosures. The results of this study should beof great importance to executives and capital market participantsinternationally, such as investors and analysts, since we provideevidence that conference calls affect incentive based compensationcontracts, which were shown in prior studies to be value relevant.
Over the past two decades equity incentives via stock-based compensation have become a veryimportant feature of the contracting environment between shareholders and executives (Core et al.,2003). The fundamental reason for the use of stock-based compensation is to align the interests ofshareholders and managers by tying executive wealth to stock performance (Frydman and Sacks, 2010;
A. Vlittis, M. Charitou / Research in International Business and Finance 27 (2013) 80– 91 81
Murphy, 1999; Bebchuk and Fried, 2003). Because executive effort is unobservable, compensationrisk is imposed on executives in order to motivate them to take actions that are in the best interestof the shareholders. In addition, the stock price contains information that investors have about themanagers’ actions and so it may be useful for contracting purposes. Therefore, equity incentives aremore important in firms with higher information asymmetry, monitoring difficulties, and agency costs.
Managers, however, make voluntary disclosures of unverifiable private information in the formof management earnings forecasts, press releases, conference calls, and/or published financial state-ments. These disclosures reduce information asymmetry and influence market prices (Beyer et al.,2010; Aboody and Kasznik, 2000; Wang, 2007; Karamanou and Vafeas, 2005; Larcker and Zakolyukina,2012). It is this relationship between information asymmetry, voluntary disclosures, and stock basedcompensation that this study explores.
In this study we hypothesize that voluntary disclosures affect compensation contracts. This hypoth-esis draws from arguments and findings in the corporate governance literature. The main argument inthe literature is that a high level of information asymmetry, and thus monitoring difficulties, requiresthe implementation of more powerful, and possibly more costly, corporate governance mechanismsto help align managers and shareholders interests (La Porta et al., 1998; Berger, 2011; Bebchuk andWeisbach, 2010). Bushman et al. (2000, 2004) provide evidence suggesting that firms with less infor-mative financial statements substitute more costly governance mechanisms, such as more powerfulequity incentives, to compensate for their less useful accounting numbers. Firms with poor financialstatements, however, can compensate for the lack of information by inducing the manager to voluntarydisclose his private information instead of adopting more costly corporate governance mechanisms.Our hypothesis is analogous to the substitutability argument of Bushman et al. (2004). Therefore, wehypothesize that firms can substitute costly corporate governance mechanisms (equity incentives) bycommitting to more forthcoming voluntary disclosure policies.
Managerial incentive contracts and disclosure policies are endogenous and are modeled as a systemof simultaneous equations. The model resembles a demand-supply specification, and can take similarinterpretation. So, an alternative interpretation of our hypothesis is that in firms with lower equityincentives shareholders will demand more voluntary disclosures in order to satisfy their monitoringneeds.
In this study, conference calls proxy for a firm’s voluntary disclosure policy. An important char-acteristic of conference calls is that they constitute ex-ante commitments to voluntary disclosures.A maintained assumption is that firms that host conference calls provide higher quality of voluntarydisclosures to the market (Matsumoto et al., 2011; Kimbrough and Loius, 2011). So the latent variableof interest is the quality of voluntary disclosure that takes the form of conference calls if it exceeds ahurdle. A firm’s compensation policy is proxied by the manager’s portfolio of equity incentives (Coreand Guay, 1999).
Using a dataset of 6263 firm year observations of which 3204 are conference call firms and 3059non-conference call firms, we show that conference calls affect equity based compensation contracts.This study differs from the extant research in the following respects. First, we examine whethervoluntary disclosures can improve contract efficiency or substitute for other more costly corporategovernance mechanisms. Second, we use conference calls to proxy for the voluntary disclosure poli-cies of firms. Conference calls provide a unique setting to study voluntary disclosures because of theircharacteristics: they are unscripted and unregulated, interactive, and unlike earnings forecasts theyconstitute ex-ante commitments to voluntary disclosures.
In Section 2, we provide background, motivation and hypothesis development. Section 3 presentsthe research design. Section 4 presents the empirical results and robustness tests. The final sectionconcludes.
2. Background, motivation and hypothesis
Over the past two decades conference calls have become a popular venue for managers to discloseinformation. The majority of conference calls take place a few days after quarterly earnings announce-ments. A typical call opens with the CEO and the chief financial officer discusses in detail the financialperformance of the company. The discussion is followed by a question-and-answer session, where
82 A. Vlittis, M. Charitou / Research in International Business and Finance 27 (2013) 80– 91
sell-side analysts and institutional investors have the opportunity to gain information that addressestheir concerns. Two of the topics which are commonly addressed are new product developments andmanagers’ guidance on future performance (Matsumoto et al., 2011; Kimbrough and Loius, 2011).
Conference calls differ from other venues of voluntary disclosures in several ways. They areunscripted and unregulated and allow CEOs to communicate large amount of complex information toa group of analysts and institutional investors simultaneously. Conference calls are also interactive,allowing participants to ask specific questions about the company’s financial performance, products,its environment, and its future prospects (Kimbrough, 2005; Matsumoto et al., 2011). This is an impor-tant difference because, unlike earnings forecasts or other written reports, managers do not haveperfect control on what is going to be discussed. Anecdotal evidence suggests that managers spendconsiderable time preparing for the conference call and that they value their ability to appear preparedand confident about their company’s prospects. Finally, conference calls are ex-ante commitments tovoluntary disclosures. Evidence shows that once companies initiate conference calls they tend to con-tinue holding them. This again is a distinctive and important characteristic of conference calls becausemanagers commit to a disclosure policy before they observe any pre or post-decision information.In contrast, there is no evidence showing that firms adopt explicit disclosure policies with regard toearnings forecasts or other types of voluntary disclosures (Hollander et al., 2010; Matsumoto et al.,2011).
In October 2000, the SEC passed Regulation FD that requires firms to disclose any material infor-mation through public disclosures. Conference calls were at the center of the controversy surroundingthe passage of Regulation FD. Before Regulation FD, conference calls were targeted to a select audi-ence usually consisting of sell-side analysts and institutional investors. Managers had tight controlover who had access to the call and who could ask questions. Regulation FD was mainly a response tocomplaints by individual investors and regulators that selective disclosures, such as conference calls,put the general public at a competitive disadvantage in the search for information (Kross and Suk,2012; Mayew, 2008; Chen et al., 2010).
Consistent with the SEC’s concerns with selective disclosure, prior research has documented thaton average, conference calls convey material information to the market. A recent study by Matsumotoet al. (2011) shows that both the presentation and the discussion segments of the conference call haveincremental information content over and above the press release. Kimbrough and Loius (2011) showthat bidders are more likely to hold conference calls at merger announcements when the mergersare financed with stocks. Kimbrough (2005) finds that the initiation of a conference call leads to asignificant reduction in the serial correlation of analyst forecast errors. In a similar vein, Bowen et al.(2002) show that conference calls increase analysts’ ability to forecast earnings accurately and decreasedispersion among analysts. Bushee et al. (2003) document that trade volume and price volatility areelevated during the conference call periods. Brown et al. (2003a) show that firms initiating conferencecalls experience a sustained reduction in information asymmetry. Finally, Kohlbeck and Magilke (2004)show that the information content of earnings is higher in firms hosting conference calls than itis in firms that do not. Overall, the evidence suggests that material information is released duringconference calls.
The argument against closed conference calls adopted by the SEC, and proponents of RegulationFD, is that they increase the information advantage of informed investors over the uninformed andtherefore, increase their trading profits at the expense of the uninformed. If indeed conference callsincrease the information asymmetry between informed and uninformed investors, however, thenuninformed investors will require a risk premium to compensate them for trading with informedinvestors (Easley and O’Hara, 2004; Agrawal et al., 2006; Francis et al., 2006; Heflin et al., 2011).This, in turn, would increase a firm’s cost of capital. Furthermore, it is not obvious that conferencecalls actually increase informed investors’ trading profits. If informed investors observe the sameinformation then competition among them will be intense and this can improve the terms of trade foruninformed traders (Kross and Suk, 2012).
A maintained assumption in this study is that closed conference calls prior to regulation FD allowcompanies to disclose high-quality information. Firms in highly competitive environments may with-hold material information if competitors can exploit it (Verrecchia, 1983; Beyer et al., 2010). Closedconference calls allow firms to disclose more precise information by partially avoiding proprietary
A. Vlittis, M. Charitou / Research in International Business and Finance 27 (2013) 80– 91 83
costs. This is because managers have control over who has access to the information released. Con-sistent with this argument an AIMR survey in 2001 documents that more than 60 percent of sell-sideand buy-side analysts believe that the ‘quality’ of the information disclosed during conference callsby the companies has deteriorated after Regulation FD (AIMR, 2001). However, empirical evidence onthe effects of regulation FD on the ‘quality’ of information disclosed is mixed (Kross and Suk, 2012;Bushee et al., 2004; Sunder, 2002; Brown et al., 2003b).
In the empirical and analytical disclosure literature the decision of voluntary disclosures involvesa trade off between a reduction in the information asymmetry component of cost of capital fromincreased disclosure against litigation and proprietary costs. There is limited empirical evidence thatexamines the relation between voluntary disclosures and corporate governance mechanisms, partlybecause extant literature is based on the two underlying assumptions that managers’ and sharehold-ers’ incentives for voluntary disclosures are perfectly aligned and that agency conflicts are perfectlyresolved through optimal contracts and corporate governance mechanisms.
Various researchers examined the role of disclosure in explaining governance mechanisms. Morespecifically, Aboody and Kasznik (2000), document that managers delay disclosures of good news andaccelerate disclosures of bad news prior to fixed stock option awards, consistent with the argumentthat managers make opportunistic voluntary disclosures that maximize their stock option compensa-tion. An alternative interpretation of the findings is that managers act in the shareholders interests. Forexample, because the incentives to increase stock price volatility created by an in the money optionare lower than those created by an at the money option, firms may wish to award in the money optionsbut would like to avoid the financial costs of such options. To accomplish this, the board of directorsallows managers to time disclosures (Core, 2001; Berger, 2011). This interpretation, however, doesnot explain why the board of directors does not use restricted stocks or cash compensation to achievethis objective instead of options. Also, it implies that the board of directors can predict the stock pricereaction to voluntary disclosures or that the exercise price of stock options is not important as long asthey are in the money, which is unlikely to be the case.
Ajinkya et al. (2004) assume that the managers have private incentives to withhold or to biasearnings forecasts for various reasons, including opportunities for insider trading profits and repu-tational risks of erroneous forecasts. They show that stronger corporate governance mechanisms, asproxied by greater institutional ownership and board of directors’ composition, induce managersto issue earnings forecasts more frequently. They also show that in firms with effective corpo-rate governance, the earnings forecasts are more specific, more accurate, and less optimisticallybiased.
Nagar et al. (2003) assume that managers avoid disclosing because such disclosures reduce theirprivate control benefits. These researchers argue that stock-based compensation is a way to mitigatethis agency problem. The argument is that stock-based compensation motivates managers to issuegood news in order to boost the stock price while the potential negative investor interpretation ofsilence and litigation costs provide incentives to release bad news. Consistent with this predictionthe paper shows that firms’ disclosures as measured by management earnings forecast frequency andanalysts’ subjective ratings of disclosure practice (AIMR scores), are positively related to the proportionof CEO compensation affected by stock price and the value of shares held by the CEO (Bebchuk andWeisbach, 2010; Beyer et al., 2010).
The aforementioned studies basically relax the first underlying assumption of the voluntary dis-closure literature. That is, they relax the assumption that managers and shareholders interests ofvoluntary disclosures are perfectly aligned. Aboody and Kasznik (2000) show that managers have pri-vate incentives to time their disclosure because it increases their compensation. Ajinkya et al. (2004)and Nagar et al. (2003) argue that managers have private incentives to withhold their informationand that corporate governance mechanisms help align these incentives. In essence, the above papersassume that shareholders and managers are engaged in a zero-sum game when it comes to voluntarydisclosures.
Our motivation is based on the substitution argument of Bushman et al. (2000, 2004), who con-jecture that firms whose accounting numbers do a poor job in capturing the current value relevantinformation will substitute into more costly corporate governance mechanism to compensate for theinformation inadequacies. More specifically, they show that firms with lower timeliness of earnings
84 A. Vlittis, M. Charitou / Research in International Business and Finance 27 (2013) 80– 91
Table 1Sample selection and description. In Panel A of this table we present initial dataset and steps followed to reach to our finaldataset. In Panel B, we present sample distribution per conference and non conference call firms.
Panel A: Sample selection criteria
Initial sample from execucomp database 8467Excluding financial firms (SIC codes 6000–6999) 7417Excluding firms with missing compensation data 7243Excluding firms with missing control variables (CRSP, and Compustat information) 6263
Final sample/firm-years 6263Number of unique firms 1681
Panel B: Sample distribution
ccall = 0 ccall = 1 Total
Total 3059 3204 6263Row Pct 48.84 51.16 100
ccall = 0 are non conference call firms, ccall = 1 are conference call firms.
have more costly and powerful corporate governance mechanisms, as measured by the board of direc-tors’ composition, size, and compensation, the CEO’s equity incentives, and the concentration of stockownership by outside shareholders. Our argument is that firms can satisfy the needs of investors andtherefore avoid the costs of more powerful corporate governance mechanisms (i.e. imposing morerisk on a risk-averse agent through equity incentives) by committing to more transparent voluntarydisclosure policies. Thus, the aforementioned discussion leads us to the following hypothesis:
H1. In firms that host conference calls CEOs have less equity incentives.
The null hypothesis is that compensation contracts and voluntary disclosure policies are not relatedbut they are endogenously determined given a firm’s primitive characteristics, like growth opportu-nities and information asymmetry (Core, 2001; Beyer et al., 2010; Bebchuk and Weisbach, 2010).
3. Research design and variable measurement
3.1. Dataset
Conference call data are identified from Thomson Financial First Call, an information provider toinstitutional investors. First Call maintains a dataset of daily conference call schedules since 1995,which we use to identify firms hosting conference calls. Consistent with prior literature, we keptconference calls that took place within a thirty three day window (−3 to 30) surrounding the earningsannouncement date, so as to exclude special conference calls.
The data for executive compensation were collected from three sources. We obtained data on CEOoption and stock holdings, and option and restricted stock grants from Standard and Poor’s Execucompdatabase, stock returns data from CRSP and financial information data from Compustat. We obtainedInformation on institutional and insider holdings, and analyst following, from Compact dataset andI/B/E/S respectively.
Panel A of Table 1 presents sample selection criteria. We begin with an initial sample of 8467firm-years from the Execucomp database. Consistent with prior studies, we exclude 1050 firm-yearobservations that belong to financial firms (SIC 6000–6999) since these firms have different character-istics from other industrial firms. We eliminate 174 firm-year observations with missing compensationdata. Finally, we exclude 980 firm-year observations due to unavailable data on the control variables.The final sample consists of 6263 firm-years (1681 unique firms). Results in panel B show that out ofthe 6263 firms, 51.16% are conference call firms and the rest non-conference call firms.
A. Vlittis, M. Charitou / Research in International Business and Finance 27 (2013) 80– 91 85
3.2. Portfolio equity incentives
We use the one year approximation procedure suggested by Core and Guay (1999) to measure theoverall CEO portfolio equity incentives in a given year. Portfolio equity incentives are measured as thesensitivity of the CEO’s total portfolio of stock and stock options to 1% change in price (“portinc”).1 Thevariable proxies for the dollar change in the value of CEO’s stock and option portfolio for a percentagechange in firm value. An alternative measure is to use the percentage of CEO’s ownership which proxiesthe dollar change in the value of a CEO’s portfolio for a dollar change in firm value. Baker and Hall(1998), and Core and Guay (2002) provide a discussion of the benefits and limitations of each proxy.2
3.3. Control variables for portfolio equity incentives
The rational behind using equity incentives is to align the interests of shareholders and managers bytying executive wealth to stock performance. Because executive effort is unobservable, compensationrisk is imposed on executives in order to motivate them to take actions that are in the best interestof the shareholders. Therefore, we expect that portfolio equity incentives will be more important infirms with higher monitoring difficulties and agency costs.
To control for other determinants of portfolio incentives we include several control variablessuggested by prior research as economic determinants of managerial ownership and stock-basedcompensation.
Prior research shows that firm size is positively related to incentive compensation (Bebchuk andWeisbach, 2010; Smith and Watts, 1992). Larger firms require more talented executives and are moredifficult to monitor, making incentive compensation more important for these firms. Following priorresearch we proxy for firm size using the logarithm of the market value of equity (“lnmv”).3
Similarly, firms operating in noisier environments are more difficult to monitor. Because of thesehigher monitoring costs, firms in noisier environments are more likely to use incentive-based com-pensation in order to align the interests of managers and shareholders (Bebchuk and Weisbach, 2010;Demsetz and Lehn, 1985). Following prior research we use the logarithmic transformation of idiosyn-cratic risk as a proxy for noise (“risk”). We measure idiosyncratic risk as the standard deviation of36-month market adjusted cumulative returns, and expect this variable to be positively related toincentive compensation.
Firms with abundant investment opportunities have a range of possible investment decisionsknown fully only to the top management and the CEO. In these firms, it is very difficult for share-holders to monitor CEO’s actions, and they are more likely to rely on stock-based compensation toalign the CEO’s incentives (Bebchuk and Weisbach, 2010; Smith and Watts, 1992). Furthermore, share-holders want their risk-averse CEOs to accept risky, and value increasing investment projects, and aremore likely to rely on option plans to motivate such behavior. We measure growth opportunities(inversely) using the book-to-market ratio (“bm”) as the ratio of book value of assets to the sum ofmarket value of equity and book value of liabilities and expect that firms with lower book-to-marketratio will have higher equity incentives.
As CEOs approach retirement, they may be inclined to forgo value-increasing projects because suchexpenditures are more likely to reward the successor CEO. To mitigate this “horizon problem” firmswith CEOs nearing retirement are more likely to increase equity incentives (Dechow and Sloan, 1991).We use the logarithm of CEO tenure as a proxy for the horizon problem and we expect it to be positivelyrelated to equity incentives (“ceoten”).
Firms may substitute stock option compensation for cash compensation if they are cash constraint,or in order to avoid financial reporting costs (Yermack, 1995; Murphy, 1999; Bebchuk and Weisbach,2010). This is because, unlike cash compensation, stock options require no contemporaneous cash
1 To proxy for the stock options expected volatility, we use the 60 month stock price volatility variable provided by Execucomp.For detailed information for the calculation of this variable see Core and Guay (1999), Black and Scholes (1973).
2 The results do not change when we also used the percentage of a CEO’s ownership.3 Untabulated results remain the same when we used the logarithm of sales and total assets.
86 A. Vlittis, M. Charitou / Research in International Business and Finance 27 (2013) 80– 91
payout, and they are only disclosed in the footnotes of the financial statements. We follow Core andGuay (1999) and measure the degree of cash shortfall (“cashcon”) as the three-year average of [(com-mon and preferred dividends − cash flow from investing activities − cash flow from operations)/totalassets]. To proxy for financial reporting costs, we follow Dechow et al. (1996) and categorize a firmas dividend constrained (“divcon”) if [(retained earnings + cash dividends and repurchases during theyear)/prior year’s cash dividends and stock repurchases)] is less than two in any of the previous threeyears or if the denominator is zero for all three years. Firms that are dividend constrained are assumedto be more concerned with debt covenant violations.
Finally, we include industry and year indicator variables to control for industry and year effects.Industry controls are based on 2-digit SIC codes. The model for the level of equity incentives is sum-marized as follows:
+ �6 cashconit + �7 divconit + �8 industry controlsit + �9 Year controlst + εit
Since voluntary disclosure policies, as measured by conference calls, and compensation con-tracts are endogenous, a simultaneous equation model emerges, with the difference that one ofthe dependent variables is latent and unobserved. Consistent with prior literature, we formulatethe model following Heckman (1978). To estimate the system of equations we follow the two-stage procedure. More specifically, in the first stage we regress each endogenous variable on allexogenous variables (reduced form). In the second stage, our two equations will be separately esti-mated with the right hand side endogenous variable replaced by its fitted value from the first stageregression.4
4. Empirical results
4.1. Descriptive statistics and correlation analysis
Table 2 presents summary statistics by conference call classification. Comparing results in panels Aand B of Table 2, we observe that conference call firms have higher equity-based incentives than non-conference call firms. The median change in CEO wealth form a 1% change in stock price is $219,000for conference call firms and $125,000 for non conference call firms. Also conference call firms appearto be larger with median market value of equity about $1.6 billion compared to $733 million for nonconference call firms.
Table 3 presents Pearson correlation coefficients. Results show that equity incentives are posi-tive and significantly correlated with size (“lnmv”), growth opportunities (“bm”), and CEO tenure(“ceoten”). Apart from the risk variable (“risk”) which is not significantly correlated with equity incen-tives, the findings are consistent with prior research. Results also show that equity incentives arepositive and significantly correlated with the conference call variable (“ccall”) which is consistentwith our expectations.
4.2. Multivariate analysis
Table 4 presents the results of the determinants of equity incentives assuming that voluntarydisclosures as proxied by conference calls are exogenous. The findings are consistent with priorresearch. Larger firms (“lnmv”), in riskier environments (“risk”), with higher growth opportunities(“bm”), whose CEO approaches retirement (“ceoten”), and that are dividend constrained (“divcon”),use more stock-based incentives. Results also show that the conference call variable (“ccall”), is notsignificantly related to equity incentives (�1 = 0).
4 Since infoqual is a latent variable, (1.1) was estimated using probit both in the first and the second stage.
A. Vlittis, M. Charitou / Research in International Business and Finance 27 (2013) 80– 91 87
Table 2In this table we present descriptive statistics for conference call firms (Panel A) and for non conference call firms (Panel B). Ourdataset consists of 3204 conference call firm-year observations and 3059 non conference call firm-year observations. Variablesare defined as follows: lnportinc: portfolio incentives are measured as the natural logarithm of the sensitivity of the CEO’stotal portfolio of stock and stock options to 1% change in price; ccall: 1 if a firm hosts conference calls during the year, and 0otherwise; lnmv: logarithm of market value of equity; risk: standard deviation of returns, measured as the standard deviationof 36-months market adjusted stock returns prior to the fiscal year; bm: the ratio of book value of assets to the sum of marketvalue of equity and book value of liabilities; ceoten: the logarithmic of CEO tenure; cashcon; three year average of [(commonand preferred dividends − cash flow from investing activities − cash flow from operations)/total assets]; divcon: 1 if [(retainedearnings + cash dividends and repurchases during the year)/prior year’s cash dividends and stock repurchases)] less than 2.
Variable Mean Std dev Lower quartile Median Upper quartile
In this table we present Pearson correlation coefficients for portfolio incentives and their determinants for our final datasetof 6263 firm year observations. Variables are defined as follows: lnportinc: portfolio incentives are measured as the naturallogarithm of the sensitivity of the CEO’s total portfolio of stock and stock options to 1% change in price; ccall: 1 if a firm hostsconference calls during the year, and 0 otherwise; lnmv: logarithm of market value of equity; risk: standard deviation of returns,measured as the standard deviation of 36-months market adjusted stock returns prior to the fiscal year; bm: the ratio of bookvalue of assets to the sum of market value of equity and book value of liabilities; ceoten: the logarithmic of CEO tenure; cashcon;three year average of [(common and preferred dividends − cash flow from investing activities − cash flow from operations)/totalassets]; divcon:1 if [(retained earnings + cash dividends and repurchases during the year)/prior year’s cash dividends and stockrepurchases)] less than 2. Correlations with an absolute value greater than 0.03 are significant at the 0.05 level.
The results in this table should be interpreted with caution because we assume that the decisionto hold conference calls is exogenous. Since the decision to hold conference calls is endogenous andfirms self-select into the conference call/non conference call subsamples the results are likely to sufferfrom endogeneity bias. We address endogeneity bias in the following subsection.
88 A. Vlittis, M. Charitou / Research in International Business and Finance 27 (2013) 80– 91
Table 4Regression results on the determinants of equity incentives.
In this table we present results of the determinants of equity incentives. Our dataset consists of 6263 firm-year observations ofwhich 3204 are conference call firms and 3059 non-conference call firms. Variables are defined as follows: lnportinc: portfolioincentives are measured as the natural logarithm of the sensitivity of the CEO’s total portfolio of stock and stock options to 1%change in price; ccall: 1 if a firm hosts conference calls during the year, and 0 otherwise; lnmv: logarithm of market value ofequity; risk: standard deviation of returns, measured as the standard deviation of 36-months market adjusted stock returnsprior to the fiscal year; bm: the ratio of book value of assets to the sum of market value of equity and book value of liabilities;ceoten: the logarithmic of CEO tenure; cashcon; three year average of [(common and preferred dividends − cash flow frominvesting activities − cash flow from operations)/total assets]; divcon:1 if [(retained earnings + cash dividends and repurchasesduring the year)/prior year’s cash dividends and stock repurchases)] less than 2. The probit’s p-values are based on Wald Chi-square. The goodness of fit test is based on Hosmer–Lemeshow Chi-square. Industry and year indicators are included in themodel as control variables.lnportinc = �0 + �1 ccall + �2 lnmv + �3 risk + �4 bm + �5 ceoten + �6 cashcon + �7 divcon + �8 industry controls + �9 year con-trols + ε.
4.3. Simultaneous equations
Table 5 presents the findings from the simultaneous system of equations using the instruments ofconference calls from the reduced form regression. More specifically, in the first stage we estimatethe probability of holding conference calls by regressing the conference call dummy on exogenousvariables suggested by prior studies to affect the likelihood of holding conference calls. In the secondstage we examine the determinants of equity incentives by replacing the conference call dummy withits fitted value obtained from the first stage regression (Maddala, 1983). The most important changefrom the previous table is the effect of conference calls on equity incentives. After controlling forsimultaneous equation bias, the conference call variable becomes significant and negatively relatedto equity incentives (�1 < 0). This finding provides support for our hypothesis. Even after controllingfor other determinants of equity incentives, it appears that CEOs in firms that host conference callshave less equity incentives. This is consistent with the argument that the board of directors substitutesvoluntary disclosures for more costly corporate governance mechanisms. Alternatively, in firms whereCEOs have less equity incentives, the owners demand more voluntary disclosures.
4.4. Regression diagnostics and robustness tests
To ensure that the results are not sensitive to extreme observations we re-estimate the regressionsafter removing outliers using different techniques. The results still hold and do not seem to be influ-enced by extreme observations. As a robustness test we also perform the analysis using the averageequity incentives of the top five executives instead of just the CEO of the firm. The results again seemto be robust to this specification. In order to ensure that the proxy for equity incentives used in thestudy is not driving the results we re-estimate the regressions using as a dependent variable the CEO’spercentage ownership. Again the results are robust to this formulation.
As a final robustness test we take the averages of all the independent variables over the sampleperiod and perform the same analysis using unique firms instead of firm-years. We also included
A. Vlittis, M. Charitou / Research in International Business and Finance 27 (2013) 80– 91 89
Table 5Results from simultaneous system of equations controlling for the endogeneity of conference calls and equity incentives.
In this table we present results from a simultaneous system of equations by controlling for the endogeneity of conference calls.Our dataset consists of 6263 firm-year observations of which 3204 are conference call firms and 3059 non-conference callfirms. Variables are defined as follows: lnportinc: portfolio incentives are measured as the natural logarithm of the sensitivityof the CEO’s total portfolio of stock and stock options to 1% change in price; ccall: is the fitted value obtained from the firststage regression model explaining the likelihood of holding conference calls; lnmv: logarithm of market value of equity; risk:standard deviation of returns, measured as the standard deviation of 36-months market adjusted stock returns prior to thefiscal year; bm: the ratio of book value of assets to the sum of market value of equity and book value of liabilities; ceoten:the logarithmic of CEO tenure; cashcon; three year average of [(common and preferred dividends − cash flow from investingactivities − cash flow from operations)/total assets]; divcon: 1 if [(retained earnings + cash dividends and repurchases duringthe year)/prior year’s cash dividends and stock repurchases)] less than 2. The probit’s p-values are based on Wald Chi-square.The goodness of fit test is based on Hosmer–Lemeshow Chi-square. Industry and year indicators are included in the model ascontrol variables.lnportinc = �0 + �1 ccall + �2 lnmv + �3 risk + �4 bm + �5 ceoten + �6 cashcon + �7 divcon + �8 industry controls + �9 year con-trols + ε.
industry controls in the conference calls equation in order to show the industry effects on the results.Firms are assigned to the conference call sub-sample if they host conference calls for all five years inthe sample period. The final sample consists of 1681 unique firms of which 272 are conference callfirms.
Table 6 presents the effect of conference calls on equity incentives. Results show that that theconference call variable is significant and negatively related to equity incentives, which is consistentwith the research hypothesis of this study. The finding indicates that firms substitute equity incentiveswith more forthcoming disclosure policies. All other variables (except “cashcon” and “divcon”) aresignificant and consistent with our expectations.
5. Conclusions
In this study we examined the relation between corporate governance mechanisms and voluntarydisclosures. Corporate governance mechanisms are proxied by the CEO’s equity incentives while vol-untary disclosure policies are proxied by conference calls. We hypothesized that firms can substitutecostly corporate governance mechanisms with more forthcoming disclosure policies. Using a datasetof 6263 firm-year observations from both conference call and non-conference call firms, our resultssupport our expectations. Moreover, in order to control for the simultaneous determination of incen-tive compensation and voluntary disclosure policies, we test the predictions using a simultaneoussystem of equations where both incentive-based compensation and voluntary disclosure policies areassumed to be endogenous. The results are largely consistent with the predictions, providing supportfor our hypothesis. In summary, the study contributes to the voluntary disclosure literature by show-ing that firms can use more transparent voluntary disclosure policies to substitute for more costlycorporate governance mechanisms.
This study could be extended in various ways. Future studies may examine the relation betweenvoluntary disclosures and other corporate governance mechanisms. For example, similar predictionscan be made about the relation between voluntary disclosures and board of directors’ composition,
90 A. Vlittis, M. Charitou / Research in International Business and Finance 27 (2013) 80– 91
Table 6Results from simultaneous system of equations by controlling for the endogeneity of conference calls and equity incentives.
In this table we present results on the average values of all variables over the sample period. Firms that host conference callsfor all 5 years in the sample are assigned to the conference call sub-sample. Our dataset consists of 1681 firms of which 272 areconference call firms and 1409 non-conference call firms. Variables are defined as follows: lnportinc: portfolio incentives aremeasured as the natural logarithm of the sensitivity of the CEO’s total portfolio of stock and stock options to 1% change in price;ccall: 1 if a firm hosts conference calls for all 5 years in the sample, and 0 otherwise; lnmv: logarithm of market value of equity;risk: standard deviation of returns, measured as the standard deviation of 36-months market adjusted stock returns prior tothe fiscal year; bm: the ratio of book value of assets to the sum of market value of equity and book value of liabilities; ceoten:the logarithmic of CEO tenure; cashcon; three year average of [(common and preferred dividends − cash flow from investingactivities − cash flow from operations)/total assets]; divcon: 1 if [(retained earnings + cash dividends and repurchases during theyear)/prior year’s cash dividends and stock repurchases)] less than 2. The probit’s p-values are based on Wald Chi-square. Thegoodness of fit test is based on Hosmer–Lemeshow Chi-square. Industry indicator is included in the model as control variable.lnportinc = �0 + �1 ccall + �2 lnmv + �3 risk + �4 bm + �5 ceoten + �6 cashcon + �7 divcon + �8 industry controls + ε.
size, and compensation. Moreover, managers with higher stock based compensation are more exposedto the risk of insider trading violations. In recent years, the SEC has increased enforcement and insidertrading sanctions. In addition, companies appear to adopt policies and procedures to regulate trading inthe stock by its own insiders. Managers intending to trade their stock holdings have incentives to hostconference calls in order to reduce the risk of insider trading violations and to correct any perceivedundervaluation. The insider trading behavior of insiders around conference calls is an interestingavenue for future research as well.
References
Association of Investment Management and Research, 2001. Regulation FD e-survey Summary, January 2001.http://www.cfainstitute.org/pressroom/01releases/regFD surveysum.html.
Aboody, D., Kasznik, R., 2000. CEO stock option awards and the timing of corporate voluntary disclosures. Journal of Accountingand Economics 29, 73–100.
Agrawal, A., Chadha, S., Chen, M., 2006. Who is afraid of RegFD? The behavior and performance of sell-side analysts followingthe SEC’s fair disclosure rules. Journal of Business 79, 2811–2834.
Ajinkya, B., Bhojraj, S., Sengupta, P., 2004. The Governance Role of Institutional Investors and Outside Directors on the Propertiesof Management Earnings Forecasts. Working Paper, University of Florida.
Baker, G., Hall, B., 1998. CEO incentives and firm size. Journal of Labor Economics 22 (4), 767–798.Bebchuk, L., Fried, J., 2003. Executive Compensation as an Agency Problem. Journal of Economic Perspectives 17 (3), 71–92.Bebchuk, L., Weisbach, M., 2010. The State of Corporate Governance Research. The Review of Financial Studies 23, 939–961.Berger, P.G., 2011. Challenges and opportunities in disclosure research—a discussion of the financial reporting environment:
review of the recent literature. Journal of Accounting and Economics 51, 204–218.Beyer, A., Cohen, D.A., Lys, T., Walther, B.R., 2010. The financial reporting environment: review of the recent literature. Journal
of Accounting and Economics 50, 296–343.Black, F., Scholes, M., 1973. The Pricing of Options and Corporate liabilities. Journal of Political Economy 81 (3), 637–654.Bowen, M.R., Davis, K.A., Matsumoto, A.D., 2002. Do conference calls affect analysts’ forecasts? The Accounting Review 77 (2),
285–316.Brown, S., Hillegeist, S., Lo, K., 2003. Conference Calls and Information Asymmetry. Working Paper, Emory University, North-
western University and The University of British Columbia.Brown, S., Hillegeist, S., Lo, L., 2003. Regulation FD and Voluntary Disclosures. Working Paper, Emory University, Northwestern
University, and The University of British Columbia.Bushman, R., Chen, Q., Engel, E., Smith, A., 2000. The sensitivity of Corporate Governance Systems to the Timeliness of Accounting
A. Vlittis, M. Charitou / Research in International Business and Finance 27 (2013) 80– 91 91
Bushman, R., Chen, Q., Engel, E., Smith, A., 2004. Financial accounting information, organizational complexity and corporategovernance systems. Journal of Accounting and Economics 37, 167–201.
Bushee, B., Matsumoto, D., Miller, G., 2003. Open versus closed conference calls: the determinants and effects of broadeningaccess to disclosure. Journal of Accounting and Economics 34 (1–3), 149–180.
Bushee, B., Matsumoto, D., Miller, G., 2004. Managerial and Investor Responses to Disclosure Regulation: The Case of Reg FDand Conference Calls. Accounting Review 79, 617–643.
Chen, X., Cheng, Q., Lo, K., 2010. On the relationship between analyst reports and corporate disclosures: exploring the roles ofinformation discovery and interpretation. Journal of Accounting and Economics 37, 293–314.
Core, J., 2001. A review of the empirical disclosure literature: discussion. Journal of Accounting and Economics 31, 441–456.Core, J., Guay, W., 1999. The use of equity grants to manage optimal equity incentive levels. Journal of Accounting and Economics
28, 151–184.Core, J., Guay, W., 2002. Estimating the value of employee stock option portfolios and their sensitivities to price and volatility.
Journal of Accounting Research 40 (3), 613–630.Core, J., Guay, W., Larcker, D., 2003. Executive equity compensation and incentives: a survey. Federal Reserve Bank of New York
Economic Policy Review 8 (1), 27–50.Dechow, P., Huton, A., Sloan, R., 1996. Economic consequences of accounting for stock-based compensation. Journal of Account-
ing Research 34, 1–20.Dechow, P., Sloan, R., 1991. Executive incentives and the horizon problem: an empirical investigation. Journal of Accounting
and Economics 14, 51–89.Demsetz, H., Lehn, K., 1985. The structure of corporate ownership: causes and consequences. Journal of Political Economy,
1155–1177.Easley, D., O’Hara, M., 2004. Information and the cost of capital. The Journal of Finance 59 (4), 1553–1583.Francis, J., Nanda, D., Wang, X., 2006. Re-examining the effects of regulation fair disclosure using foreign listed firms to control
for concurrent shocks. Journal of Accounting and Economics 41, 271–292.Frydman, C., Sacks, E.R., 2010. Executive compensation: A new look from a long-term perspective, 1936–2005. Review of
Financial Studies 23 (5), 2099–2138.Heckman, J.J., 1978. Dummy endogeneous variables in simultaneous equation systems. Econometrica 46, 931–960.Heflin, F., Kross, W., Suk, I., 2011. The Effect of Regulation FD on the Properties of Management Earnings Forecasts. Working
Paper, Florida State University, SUNY, Buffalo.Hollander, S., Pronk, M., Roelofsen, E., 2010. Does silence speak? An empirical analysis of disclosure choices during conference
calls. Journal of Accounting Research 48 (3), 531–563.Karamanou, I., Vafeas, N., 2005. The association between corporate boards, audit committees and management earnings fore-
casts. Journal of Accounting Research 43 (June).Kimbrough, M., 2005. The effect of conference calls on analyst and market under-reaction to earnings announcements. The
Accounting Review 80 (1), 189–219.Kimbrough, M.D., Loius, H., 2011. Voluntary disclosure to influence investor reactions to merger announcements: an examina-
tion of conference calls. The Accounting Review 86 (2), 637–667.Kohlbeck, M., Magilke, M., 2004. The impact of concurrent conference calls on the information content of eamings announce-
ments. Working paper. University of Wisconsin.Kross, W.J., Suk, I., 2012. Does regulation FD work? Evidence from analysts’ reliance on public disclosure. Journal of Accounting
and Economics 53, 225–248.La Porta, R., Lopez-de-Silanes, F., Shleifer, A., Vishny, R., 1998. Law and finance. Journal of Political Economy 106, 1113–1155.Larcker, D., Zakolyukina, A., 2012. Detecting deceptive discussions in conference calls. Journal of Accounting Research.Maddala, G.S., 1983. Limited-dependent and qualitative variables in econometrics. Econometric Society monographs in quan-
titative economis, vol. 3. Cambridge University Press, Cambridge Cambridgshire; New York.Matsumoto, D., Pronk, M., Roelofsen, E., 2011. What makes conference call useful? The information content of Managers’
presentation and analysts’ discussion sessions. The Accounting Review 84 (4), 1383–1414.Mayew, W., 2008. Evidence of management discrimination among analysts during earnings conference calls. Journal of Account-
ing Research 46 (3), 627–659.Murphy, K.J., 1999. Executive Compensation. In: Ashenfelter, O., Card, D. (Eds.), Handbook of Labor Economics. North Holland,
pp. 2485–2563.Nagar, V., Dhananjay, N., Wysicju, P., 2003. Discretionary disclosure and stock-based incentives. Journal of Accounting and
Economics 34, 283–309.Smith, C., Watts, R., 1992. The investment opportunity set and corporate financing, dividends, and compensation policies. Journal
of Financial Economics 32, 263–292.Sunder, S., 2002. Investor Access to Conference Call Disclosures: Impact of Regulation Fair Disclosure on Information Asymmetry.
Working Paper, New York University.Verrecchia, R., 1983. Discretionary disclosure. Journal of Accounting and Economics 5, 179–194.Wang, I., 2007. Private earnings guidance and its implication for disclosure regulation. The Accounting Review 82, 1299–1332.Yermack, D., 1995. Do corporations award CEO stock options effectively? Journal of Financial Economics 39, 237–269.
