Signaling, corporate governance, and the equilibrium dividend policy

The Quarterly Review of Economics and Finance 59 (2016) 186–199

Contents lists available at ScienceDirect

The Quarterly Review of Economics and Finance journal homepage: www.elsevier.com/locate/qref

Signaling, corporate governance, and the equilibrium dividend policy夽 Omar A. Esqueda ∗ Tarleton State University, Department of Accounting, Finance, and Economics, 1333 West Washington Street, Stephenville, TX 76402, USA

a r t i c l e

i n f o

Article history: Received 21 June 2014 Received in revised form 20 May 2015 Accepted 22 June 2015 Available online 30 June 2015 Keywords: Cross-listings Dividend policy Emerging markets Signaling Insider ownership Corporate governance

a b s t r a c t The well-documented information content of dividends is contingent on the firm’s corporate governance. Using cross-listing events, we find that firms reach a new equilibrium dividend policy after a shift in the level of shareholder protection and the direction of the dividend adjustment depends on the precross-listing locus of control. Exchange-traded cross-listings can afford to decrease dividend payouts as they substitute dividends with better corporate governance. However, dividend distributions and the likelihood to pay dividends increase when cross-listings are controlled by insiders, supporting the signaling hypothesis. The cross-listing level and ownership structure convey useful information regarding future shifts in dividend payouts. © 2015 The Board of Trustees of the University of Illinois. Published by Elsevier B.V. All rights reserved.

JEL classification: G32 G35

1. Introduction Information asymmetry between corporate insiders and outside shareholders drives widely-held firms to signal their prospects to the financial markets, commonly, through dividend policy. For example, Watts (1973), Black (1976), Miller and Rock (1985), Ambarish, John, & Williams (1987), Noe and Rebello (1996), La Porta, Lopez-de-Silanes, Shleifer, & Vishny (2000), Aivazian, Booth, & Cleary (2003), Asem and Alam (2015) confirm that dividend policy conveys relevant information about the firm. However, interpreting the signals sent by firms is not straightforward as the information content of dividends is contingent on the firms’ corporate governance, in other words, a similar shift in dividend policy may have different connotations for firms with differing levels of shareholder protection (Noe & Rebello, 1996). In addition, evaluating the information conveyed in dividend payouts becomes more

夽 This paper is based on the fourth chapter of my doctoral dissertation. I am grateful to my committee members: Dave Jackson, Thanh Ngo, Judith Swisher, and Cynthia Brown for their valuable input that greatly improved this paper. I thank the editor and an anonymous referee for providing important suggestions that enhanced earlier versions of this manuscript. I am thankful to Rafael La Porta, Karl Lins, Stijn Claessens, Mara Faccio, and Larry Lang for providing ownership structure data relevant for this manuscript. All remaining errors are my own. ∗ Tel.: +254 968 9908; fax: +254 968 9665. E-mail address: [email protected]

challenging as dividends are known to mitigate agency problems (Aivazian et al. (2003); Easterbrook, 1984; Jensen, 1986; La Porta et al., 2000; Mitton, 2004; Rozeff, 1982); hence, dividend policy is endogenous with potential agency problems. This paper examines the implication of the signaling hypothesis and shareholder protection in the discovery of the equilibrium dividend policy. We employ cross-listing events as the trigger for improvements in firm-level corporate governance as documented by Doidge, Karolyi, & Stulz (2004). This study sheds light on the link between signaling and corporate governance and their effect on payout policy. Moreover, we contribute to the cross-listing literature reconciling existing research on the association between cross-listings and dividend policy. Cross-listings offer a unique opportunity to study the effects of firm characteristics on dividend policy as cross-listing events trigger fundamental changes in firms’ capital barriers, agency problems, liquidity, and investor recognition (Karolyi, 2006). La Porta et al. (2000), O’Connor (2006), and Petrasek (2012) have linked shifts in dividend policy to simultaneous improvements in agency costs as implied by the bonding hypothesis, with inconclusive results. In this context, previously tested payout policy models assume homogenous changes in agency costs by cross-listing level. We argue that an equilibrium payout policy is contingent on the firm’s pre-cross-listing locus of control; thereby, control is endogenous with the post-cross-listing shifts in dividend policy. Noe and Rebello (1996) state that increments in

http://dx.doi.org/10.1016/j.qref.2015.06.005 1062-9769/© 2015 The Board of Trustees of the University of Illinois. Published by Elsevier B.V. All rights reserved.

O.A. Esqueda / The Quarterly Review of Economics and Finance 59 (2016) 186–199

dividend payout release positive (negative) signals when the firm is management-controlled (shareholder-controlled). We endeavor to test the bonding hypothesis, the signaling hypothesis, and the liquidity hypothesis of dividends in the context of cross-listing events. The bonding hypothesis (Coffee, 1999, 2002; Stulz, 1999) indicates that cross-listed firms bond themselves to a stock market with stricter rules, hence improving corporate governance. After firms cross-list on U.S. exchanges, firm value increases thereby supporting a “firm-level” bonding hypothesis (Doidge et al., 2004). Payout policy might be affected as firms’ agency costs change and managers act upon this change. La Porta et al. (2000) develop two dividend policy models based on the bonding hypothesis of crosslistings: the outcome model and the substitute model of dividends. The outcome model states that increased corporate payouts are the result of improved corporate governance and pressure by minority shareholders. Hence, improved shareholder protection after cross-listing implies a higher payout ratio. On the other hand, the substitute model of dividends suggests that firms substitute the monitoring mechanism inherent in dividend payments with a corresponding increase in shareholder protection. Dividend payment shows information about the expected future cash flows of the firm (Watts, 1973). Hence, newly cross-listed firms are expected to use payout policies to signal changes in firms’ agency costs. Miller and Rock (1985) recognize that information asymmetry significantly affects the equilibrium level of dividend payouts. Noe and Rebello (1996) state that under information asymmetry, the locus of control is a decisive factor of optimal financial policies. Emerging-market cross-listings are an interesting sample to test this hypothesis as a significantly large proportion of firms from emerging countries tends to be controlled by insiders. When these firms cross-list in the United States, a decrease in information asymmetry is anticipated. Hence, dividend policy signals sent by insider-controlled and management-controlled firms might differ. Insider-controlled firms tolerate higher dividends only if they are pressured from minority shareholders to do so; therefore, an increase in dividend payments signals improvements in corporate governance. Conversely, shareholder-controlled firms are able to send positive signals by reducing payouts as they substitute dividends with improved corporate governance (Noe & Rebello, 1996). Unlike previous studies, we posit that both the outcome model of dividends (insider-controlled firms) and the substitute model of dividends (non-insider-controlled firms) could be supported simultaneously. Mitton (2004) suggests that firms with stronger corporate governance have a higher dividend payout after cross-listing, consistent with the outcome model of dividends. If Mitton’s argument holds, cross-listed firms are expected to increase their dividend payout, consistent with the signaling hypothesis and the outcome model of dividends. We additionally hypothesize that firms with previously low levels of investor protection will have greater dividend increases as managers signal improvements in corporate governance by increasing/initiating dividends.1 Despite the vast amount of research on dividend policy, there is scant evidence on whether there is a shift in dividend policy following a cross-listing event and if so, to what extent. The market segmentation hypothesis suggests that capital barriers are reduced when firms become available to foreign investors thereby decreasing the cost of equity capital (Alexander, Eun, & Janakiramanan, 1988; Errunza & Losq, 1985; Hail & Leuz, 2009).

1 We do not attempt to isolate the individual influence of signaling and agency theories on dividend policy. e.g., Aivazian et al. (2003) indicate that dividend policy serves as both a signaling mechanism and control for managerial opportunism among U.S. firms.

187

Table 1 Description of ADR levels. ADRs level 4 are also known in the literature as ADR Rule 144-A, since this SEC rule allows the private offerings of this securities. ADR type

Capitalraising

Description

Level 1

No

Level 2

No

Level 3

Yes

Level 4

Yes

This level is traded in the U.S. only over the counter (OTC). OTC-ADRs can use either the OTCBB or the Pink Sheets quotation systems. However, this ADR level cannot raise capital in the U.S. ADRs Level 2 are U.S. exchange-traded but are considered non-capital-raising in U.S. equity markets. This ADR level is traded in U.S. exchanges and is allowed to raise capital (issue new shares) in U.S. equity markets. ADRs Level 4 are available in the U.S. only through private offerings for qualified institutional buyers (QIB). This ADR level, also known as PORTAL or Rule 144-A, can raise capital in the U.S. only from QIB.

Nevertheless, the magnitude of the decrease in capital barriers also depends on whether the firm is registered to raise capital (ADRs levels 3 and 4) or not (ADRs levels 1 and 2). We present a detailed description of each cross-listing level in Table 1. Moreover, this hypothesis suggests that dividend payout decreases when firms experience a lower cost of equity capital (Bekaert & Harvey, 2000). Relevant payout policy theories put forward contradictory outcomes as can be seen in the above findings. The association between cross-listing events and dividend policy under varying levels of shareholder protection is not straightforward and warrants further attention. In the context of variations in the dividend policy of cross-listed firms, we test the bonding hypothesis, the signaling hypothesis, and the market segmentation hypothesis to identify the main determinant of dividend policy.2 This manuscript contributes to the literature in several distinct ways. First, we find that following a cross-listing event, the discovery of the new payout policy equilibrium moves in an opposite direction for insider-controlled versus non-insider-controlled firms, thus supporting the signaling hypothesis of dividends. Second, our findings reconcile prior conflicting evidence on the outcome and the substitute model of dividends, thereby supporting the bonding hypothesis. Third, we control for a set of relevant variables that have been omitted in previous research and include a set of firm-level ownership structure variables to proxy for shareholder protection. Fourth, we test the market segmentation hypothesis to clarify shifts in payout policy after cross-listing. Finally, in addition to the panel model with a continuous dividend payout variable, we use pooled logit models that allow for binary dependent variables to test the likelihood of firms being dividend payers. 2. Literature review and hypothesis development Existing dividend literature corroborates a positive market reaction for firms that announce dividend increases/initiations as investors pay a premium for firms that return cash to shareholders.3 When managers decide it is appropriate to return wealth to shareholders, they choose between dividends and share repurchases. However, share repurchases are trivial in emerging markets.

2 Additionally, Petrasek (2012) states that an increase in dividend payments following capital-raising cross-listing events lends support to the liquidity hypothesis of dividends. This hypothesis predicts that capital-raising firms can increase dividends as they are allowed to raise equity capital in the U.S. if necessary. 3 See Allen and Michaely (2003) survey for an extensive review of dividend policy literature.

188

O.A. Esqueda / The Quarterly Review of Economics and Finance 59 (2016) 186–199

Bekaert and Harvey (2000) state that share repurchases are almost non-existent among firms from emerging markets due to the low level of development.4 According to Grullon and Michaely (2002), open-market share repurchases are prohibited and are considered price manipulation in countries, such as Austria, Norway, and Israel. Fama and French (2001) use treasury stock as a measure of stock repurchases. However, treasury stock is rarely used by firms from emerging markets. Given these arguments, we do not include stock repurchases and focus on dividend policy. Dividend yields move in the opposite direction of growth opportunities as companies with high-growth opportunities retain more earnings to invest in positive NPV projects (La Porta et al., 2000). Two widely used proxies for growth opportunities are the market-to-book ratio (Aivazian et al., 2003; Fama & French, 2001; Petrasek, 2012) and sales growth (Petrasek, 2012). DeAngelo, DeAngelo, & Stulz (2006) find that the firm’s life-cycle explains dividend payout better than size, profitability, and growth opportunities, previously believed to be the main factors determining dividend policy.5 Aivazian et al. (2003) and Truong and Heaney (2007) find that the dividend policy of emerging-market firms is influenced by the same factors (profitability, leverage, and investment opportunities) as firms in developed countries, yet there is a clear country-specific variation and a significant influence by large shareholders. Existing research on the dividend policy of foreign firms crosslisted in the United States is scant. O’Connor (2006) determines that the dividend yield decreases after cross-listing events (supporting the substitute hypothesis), whereas Petrasek (2012) points to an increase in dividend yield (consistent with the outcome hypothesis). Furthermore, La Porta et al. (2000) conclude that firms with better corporate governance pay more dividends and shareholders pressure appears to be the trigger for dividend payment increases. Whether cross-listing events affect dividend policy and if so, in what direction, warrants further research as existing evidence is inconclusive. Mitton (2004) suggests that firms with stronger corporate governance have higher dividend payout, consistent with the outcome model of dividends; yet, this relation is limited to countries with strong investor protection. Fenn and Liang (2001) suggest that firms with high managerial ownership and greater potential for agency problems (high free cash flow and few growth opportunities) are more likely to pay dividends. Similarly, Shao, Kwok, & Guedhami (2013) find that firms with a high agency cost of equity tend to pay higher dividends to hinder concerns about expropriation by insiders, consistent with the Noe and Rebello (1996) signaling hypothesis of dividends. Lins, Strickland, & Zenner (2005) confirm that firms dominated by insiders tend to face more agency problems. Therefore, it seems necessary to control for firms’ ownership structure given the reported linkage of agency problems with dividend policy. Does cross-listing influence the payout policy of emergingmarket firms? Payout policy varies according to the level of investor protection as dividends can function as a monitoring mechanism (Easterbrook, 1984; Rozeff, 1982). We test the outcome model and the substitute model of dividends in the context of cross-listing events. Following an increase in shareholder protection, a significant change in dividend payout provides support for the bonding hypothesis. O’Connor (2006) finds significant shifts in dividends for exchange-listed American depository receipts

4 For instance, a Lexis-Nexis search shows only five events for Latin American cross-listings, using key words “share repurchases” and “share buybacks” with country identifiers such as “Brazil” or “Brazilian.” 5 The firm’s life-cycle is measured as the proportion of retained earnings to common equity. This variable is relevant as previous findings that overlook the firm’s life-cycle might overstate the change in dividends of cross-listed firms.

(ADRs) and over-the-counter (OTC) ADRs, but not for level 4 ADRs. These findings lend partial support to the bonding hypothesis as OTC cross-listings do not imply an increase in the level of regulation; therefore, changes in dividend payout are unexpected. The bonding hypothesis suggests changes in agency costs exclusively for exchange-listed firms. Mitton (2004) suggests that firms with stronger corporate governance have higher dividend payouts, consistent with the outcome model of dividends. However, this relationship between corporate governance and dividends is limited to countries with strong investor protection, suggesting that firm-level corporate governance and country-level investor protection are complements rather than substitutes. We hypothesize that there is a shift in dividend payouts for firms experiencing an increase in shareholder protection which can materialize in two ways. First, managers feel increased pressure to distribute cash to shareholders and the improvement in corporate governance is conveyed by increasing/initiating dividends, consistent with the outcome model of dividends. On the other hand, firms could reduce dividend payments following cross-listing since they substitute dividend payout with better corporate governance as the substitute model of dividend suggests. H1. The dividend payout changes significantly after foreign firms cross-list on U.S. exchanges (bonding hypothesis). Whereas, existing dividend policy research on the bonding hypothesis provides conflicting findings, we believe the signaling hypothesis of dividends is able to reconcile previous findings and shed light on the mechanics by which newly cross-listed firms discover an equilibrium payout policy. Noe and Rebello (1996) suggest that firms send positive signals when dividend payout increases (decreases) if firms are management-controlled (shareholder-controlled). Fuller and Blau (2010) state that managers react to outsiders’ perception of firms’ quality level, hence high-quality firms can pay smaller dividends. Insiders are expected to use dividend policy to signal improvements in corporate governance subsequent to cross-listing events. An increase in the likelihood of a firm becoming a dividend payer lends support to the signaling hypothesis. Managers decide to initiate dividends, and avoid omitting dividends, to signal a promising financial future for the firm. Dividend payment decisions from firms with insider ownership are more likely to be driven by the signaling motivation than dividend payment decisions from non-insider-owned firms. Therefore, the signaling hypothesis of dividends implies that insider-controlled firms increase dividends following a cross-listing event. H2. Cross-listings with insider ownership have a higher dividend payout, are more likely to initiate dividends, and are less likely to omit dividend payment (signaling hypothesis). To analyze the signaling hypothesis of dividends (Hypothesis 2), ownership structure data are necessary. However, only exchange-traded ADRs (levels 2 and 3) are required to disclose their ownership structure information to the U.S. Securities and Exchange Commission (SEC). Consequently, the test of the signaling hypothesis of dividends considers only exchange-traded crosslistings. An increase in dividend payments following cross-listing events of firms intending to raise capital in the United States lends support to the liquidity hypothesis of dividends. This hypothesis posits that capital-raising cross-listings (levels 3 and 4) can increase dividends as they are allowed to raise capital in the U.S. equity markets if necessary. Petrasek (2012) suggests that capital-raising firms distribute more wealth to shareholders given the increased access to capital following cross-listing events. Additionally, Boubakri, Cosset, & Samet (2013) find that firms accelerate their equity

O.A. Esqueda / The Quarterly Review of Economics and Finance 59 (2016) 186–199

189

Predicted shifts in dividend policy by relevant theories following U.S. cross-listing events Exchange-traded cross-listings include ADRs levels 2 and 3. OTC are equivalent to ADRs level 1. PORTAL firms equal to ADRs level 4. Insider-owner firms refer to exchange-traded cross-listings only. *A sharper effect is predicted since shareholder protection is expected to improve more substantially. *Exchange-traded only. **Only capital-raising firms, level 3. The predictions of the liquidity hypothesis are based on Petrasek (2012). Predicted effect on

Bonding hypothesis

Dividends

Outcome model

Payout increases

Exchange-traded

Market segmentation

Signaling

Liquidity

hypothesis

hypothesis

hypothesis

Insider-owned*

Exchange-traded**

Substitute model

PORTAL Payout decreases

Exchange-traded

Exchange-traded

Non-insider-owned*

PORTAL OTC No shift

OTC PORTAL

OTC PORTAL

OTC

Fig. 1. Predicted shifts in dividend policy by relevant theories following U.S. cross-listing events Exchange-traded cross-listings include ADRs levels 2 and 3. OTC is equivalent to ADRs level 1. PORTAL firms equal to ADRs level 4. Insider-owner firms refer to exchangetraded cross-listings only. *A sharper effect is predicted since shareholder protection is expected to improve more substantially. *Exchange-traded only. **Only capital-raising firms, level 3. The predictions of the liquidity hypothesis are based on Petrasek (2012).

capital-raising activity after U.S. cross-listings. The commitment to pay dividends is more easily attained if the firm can raise new capital in the United States. Capital-raising ADRs are expected to pay more dividends and are also more (less) likely to initiate (omit) dividend payments. H3. Capital-raising cross-listings have a higher dividend payout and are more likely to be dividend payers (liquidity hypothesis). The market segmentation hypothesis suggests that capital barriers are reduced when firms become available to foreign investors, thereby decreasing the cost of equity capital. Bekaert and Harvey (2000) find that after emerging markets are liberalized, the cost of equity capital decreases and subsequently dividend yields decrease, thus supporting the market segmentation hypothesis. Foreign speculators, through ADRs, closed-end funds, or foreign direct investment cause the cost of equity capital to decrease. Hail and Leuz (2009) find that there is a reduction in the cost of capital for all types of listings in the United States, yet the decrease is more

substantial for exchange-traded cross-listings, consistent with the cost of capital mechanism described by Bekaert and Harvey (2000). Firms that successfully overcome capital barriers, such as firms cross-listed in the United States, have a lesser incentive to distribute retained earnings to investors since the cost of equity capital decreases, hence using equity as a source of capital has become less costly. A significant decrease (omission) in dividend payout supports the market segmentation hypothesis. H4. U.S. cross-listings have a lower dividend payout, are less likely to initiate dividends, and are more likely to omit dividend payments (market segmentation hypothesis). To summarize our hypotheses, Fig. 1 presents a summary of the predicted changes in dividend policy under relevant theories by cross-listing type. There are conflicting predictions by theories; yet, an increase in dividend payout seems to marginally dominate the predictions for exchange-traded cross-listings and insider-owned cross-listings. With the exception of the market segmentation

Table 2 Description of control variables. CASH includes cash and cash equivalents. Short-term investments include assets that can be converted to cash with minimum or zero loss of value as defined by Datastream. Variable

Description

Source

PRETAX CASH Sales Growth MBV SIZE LEVER RE CAPEX Dual-Class

Pre-tax earnings deflated by total assets Cash holdings. Equals the average cash holdings during the year scaled by total assets One year annual growth of sales. Proxies for investment opportunities Market to book value. Investment opportunities. Firm-size. Equals the log of total assets in USD. Leverage equals the ratio of total liabilities to total assets. Firm’s life-cycle is calculated as retained earnings to common equity. Capital expenditures from the income statement deflated by total assets. Dummy equals one if the firm uses dual-class shares.

Datastream Thomson Reuters

Insider Owner Civil Antidirectors TAX LEVEL 1−4

Dummy equals one if the firm is controlled by insiders. Civil indicates legal origin: civil or common law. Country-level shareholder protection indices. Equals one during 2003 and thereafter (represents the enactment of the JGTRRA). Dummy variables that indicates ADR level and capital-raising capabilities. (See Table 2 for ADR level description). Equals one if the firm has an active ADR level-3 or level 4 Cross-listed dummy equals one when firms are cross-listed in the USA. Dummy that equals to one if the firm is an exchange-traded ADR. (ADRs levels 2 & 3). See Table 2 for details. Stock market development equals the ratio of the average country’s market capitalization deflated by its GDP.

Capital-Raising Cross-listed EXCHANGE Stock Market

20-F, La Porta et al. (1999), Claessens et al. (2000), Faccio and Lang (2002), Lins et al. (2005) La Porta et al. (2000), Djankov et al. (2008). www.IRS.gov. Bank of New York, Citibank, and Chase Bank depository receipts websites.

Beck et al. (2000).

190

O.A. Esqueda / The Quarterly Review of Economics and Finance 59 (2016) 186–199

Table 3 Descriptive statistics, N represents the number of firm/year observations. DIVIDENDS is the total amount of dividends paid on common stock by the firm deflated by total assets. YIELD is dividend yield calculated as dividends per share divided by average share price. PAYOUT is the dividend per share divided by earnings per share. PRETAX represents the profits before taxes divided by total assets. LEVER equals total liabilities deflated by total assets. CAPEX are the capital expenditures to total assets. RE is the ratio of retained earnings to total equity. Retained earnings are estimated as the total value of equity capital minus common equity and preferred stock using values from Datastream. RE is multiplied by 1 million given the low average value of this ratio. MBV is the market-to-book ratio. Sales Growth is the firm’s 1 year sales growth. CASH is the amount of cash holdings deflated by total assets. SIZE is the log of total assets in USD. Antidirectors is the home country’s antidirectors rights index. Stock Market is the domestic stock market development. Full sample

Cross-listed firms

Non-cross listed

Variable

N

Mean

S.D.

Min

Max

N

Mean

S.D.

Min

Max

N

Mean

S.D.

Min

Max

DIVIDEND YIELD PAYOUT PRETAX LEVER CAPEX RE MBV SalesGrowth CASH SIZE Antidirectors StockMarket

52,861 52,864 49,640 62,103 62,159 60,900 62,122 57,351 56,758 61,043 62,167 130,158 123,270

0.02 0.03 0.29 0.06 0.49 0.07 -0.02 2.1 0.21 0.15 12.81 3.69 1.4

0.03 0.05 0.48 0.1 0.21 0.07 0.82 2.37 0.46 0.17 1.42 1.53 1.47

0 0 -0.86 -0.3 0.06 0 -5.81 -0.71 -0.61 0 0.12 1 0

0.18 0.32 3.05 0.37 1.14 0.33 0.05 14.93 2.97 1 22.88 5 7.43

7,319 5,855 7,088 7,576 7,595 7,422 7,562 6,661 6,995 6,844 7,599 13,062 12,307

0.02 0.04 0.34 0.07 0.46 0.07 -0.05 1.99 0.23 0.26 14.05 3.73 1.42

0.03 0.05 0.52 0.1 0.21 0.07 0.96 2.25 0.5 0.32 1.65 1.49 1.55

0 0 -0.86 -0.3 0.06 0 -5.81 -0.71 -0.61 0 0.77 1 0

0.18 0.32 3.05 0.37 1.14 0.33 4.58 14.93 2.97 1 19.55 5 7.43

45,542 47,009 42,552 54,527 54,564 53,478 54,560 50,690 49,763 54,199 54,568 117,096 110,963

0.02 0.03 0.29 0.06 0.5 0.06 -0.11 2.12 0.21 0.14 12.64 3.69 1.4

0.03 0.04 0.48 0.1 0.21 0.07 8.04 2.39 0.45 0.13 1.29 1.54 1.46

0 0 -0.86 -0.3 0.06 0 -58.1 -0.71 -0.61 0 0.12 1 0

0.18 0.32 3.05 0.37 1.14 0.33 45.8 14.93 2.97 1 22.88 5 7.43

Table 4 Frequency of categorical variables, N represents the number of firm/year observations available for the corresponding variable. Insider Owner indicates that the firm has an insider owner or group of insider owners with at least 10% claim in the company. Ownership Concentration indicates that there is at least one investor, insider or outsider that owns at least 10% of the firm. Dual-Class equals one if the firm uses more than one class of stock and with different voting rights. PAY is an indicator of whether the firm pays dividends during a given year. Civil law indicates whether the country is ruled by civil or common law. Civil law has weaker legal protection for shareholders. TAX is a dummy variable that represents the decreased tax rate on dividends. TAX equals one during 2003 and thereafter. Variable

N

Yes/No

Insider-Owner

4,515

Ownership concentration

4,599

Dual-class

4,494

Insider Owner × Dual-Class Shares Ownership Concentration × Dual-class shares PAY

N

4,494

Yes No Yes No Yes No Yes

2,184 2,331 4,368 231 840 3,654 567

4,494

No Yes

3,927 819

No Yes No Yes No Yes No

3,675 35,766 17,095 73,038 57,120 49,584 80,574

52,861

Civil Law

130,158

TAX

130,158

% 48% 95% 19% 13%

18%

68% 56% 38%

hypothesis, the bonding, signaling, and liquidity hypotheses do not predict a shift in the dividend policy of over-the-counter ADRs, hence our focus on exchange-traded ADRs. 3. Sample and econometric technique 3.1. Sample The sample of ADRs, their cross-listing level, cross-listing date, country of origin, underlying firm, and sponsorship status are collected from the Citigroup, Bank of New York, and Chase Bank depository receipt websites.6 We focus on emerging-market

6

The corresponding websites are: http://www.adrbnymellon.com/dr directory.jsp, www.citissb.com/adr/www, and https://www.adr.com/DRSearch respectively.

cross-listings as this segment has grown rapidly and has outpaced the number of cross-listings from developed countries in the last two decades.7 In addition, their ownership structure characteristics are of particular interest to our research questions. To be included in the sample, cross-listed firms should have data available on Datastream International (Datastream hereafter). Extant cross-listing literature suggests that, compared with sponsored ADR programs, unsponsored programs do not have the same legal bonding effect on the cross-listed firm. In addition, U.S. exchanges require the ADR program to be sponsored by the foreign-issuer firm. Given the relevant implications for our hypotheses and as is a common practice in related literature, our ADR sample is restricted to sponsored programs. When a firm has multiple ADR programs with different start dates, we consider only the DR program in which a firm is cross-listed as reported by the depository institution. The literature suggests several proxies to quantify dividend payments. We employ the traditional measure of dividend payments used by, among others, O’Connor (2006) and Petrasek (2012), where the payout ratio (PAYOUT) is calculated as dividends per share divided by earnings per share. To verify that the results are not measurespecific, we additionally estimate aggregate dividends to total assets (DIVIDEND), similar to Aivazian et al. (2003). Finally, we create a binary variable that indicates whether or not the firm is a dividend payer (PAY). Current dividend literature suggests numerous control variables relevant for dividend policy models. Table 2 describes the explanatory variables employed in this manuscript along with their source. For country-level shareholder protection, we use the antidirectors rights and judicial efficiency indexes from La Porta et al. (2000) and accounting indexes from Djankov, La Porta, Lopez-deSilanes, & Shleifer (2008). Mitton (2004) employs analogous indices in dividend models. Ownership structure data for cross-listed firms are partially obtained from La Porta, Lopez-de-Silanes, & Shleifer (1999), Claessens, Djankov, & Lang (2000) Faccio and Lang (2002), and Lins (2003). When a firm is not included in the ownership structure datasets indicated above, the data are collected manually from the company websites and the SEC annual reports (20-F). After firms are cross-listed, annual reports are accessed through

7 Comparison at the end of December 2010, based on information provided by the Bank of New York, Citibank, and Chase depository receipts websites discussed above.

O.A. Esqueda / The Quarterly Review of Economics and Finance 59 (2016) 186–199

191

Table 5 Frequency of sample firms by country. Non-cross-listed firms include all single-listed firms available in Datastream with an average of more than $100 million U.S. Dollars in total assets during the sample period. This amount has been considered the bottom threshold required to cross-list in the NYSE and has been used in similar studies (O’Connor, 2009). *Exchange-traded ADR. All firms

No.

Country

Freq.

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25

Argentina Brazil Chile China Colombia Czech Republic Hong Kong Hungary India Indonesia Malaysia Mexico Pakistan Peru Philippines Poland Russia Singapore South Africa South Korea Sri Lanka Taiwan Thailand Turkey Venezuela

44 146 95 1,200 21 10 481 14 1,355 113 230 101 113 53 62 84 189 253 117 611 42 589 129 124 22

Total

6,198

Non-crosslisted %

Freq.

0.71 2.36 1.53 19.36 0.34 0.16 7.76 0.23 21.86 1.82 3.71 1.63 1.82 0.86 1.00 1.36 3.05 4.08 1.89 9.86 0.68 9.50 2.08 2.00 0.35 100

27 68 80 1106 17 10 389 8 1303 109 223 48 109 48 52 79 147 234 81 592 41 553 119 120 13 5,576

% 0.48 1.22 1.43 19.84 0.30 0.18 6.98 0.14 23.37 1.95 4.00 0.86 1.95 0.86 0.93 1.42 2.64 4.20 1.45 10.62 0.74 9.92 2.13 2.15 0.23 100

Level 1 (OTC)

Level 2*

Freq.

Freq.

1 35 0 26 3 0 83 4 4 2 7 12 1 3 5 2 27 18 26 3 0 3 10 1 5 281

% 0.36 12.46 0.00 9.25 1.07 0.00 29.54 1.42 1.42 0.71 2.49 4.27 0.36 1.07 1.78 0.71 9.61 6.41 9.25 1.07 0.00 1.07 3.56 0.36 1.78 100

11 24 12 11 0 0 7 1 4 2 0 22 0 1 3 0 2 0 4 3 0 1 0 0 3 111

Level 3*

%

Freq.

9.91 21.62 10.81 9.91 0.00 0.00 6.31 0.90 3.60 1.80 0.00 19.82 0.00 0.90 2.70 0.00 1.80 0.00 3.60 2.70 0.00 0.90 0.00 0.00 2.70 100

Level 4 (PORTAL) %

3 11 1 52 0 0 2 0 9 0 0 11 0 0 0 0 3 1 2 5 0 7 0 1 0 108

2.78 10.19 0.93 48.15 0.00 0.00 1.85 0.00 8.33 0.00 0.00 10.19 0.00 0.00 0.00 0.00 2.78 0.93 1.85 4.63 0.00 6.48 0.00 0.93 0.00 100

Freq. 2 8 2 5 1 0 0 1 35 0 0 8 3 1 2 3 10 0 4 8 1 25 2 0 1 122

Exchange-traded

%

Freq.

1.64 6.56 1.64 4.10 0.82 0.00 0.00 0.82 28.69 0.00 0.00 6.56 2.46 0.82 1.64 2.46 8.2 0.00 3.28 6.56 0.82 20.49 1.64 0.00 0.82 100

14 35 13 63 0 0 9 1 13 2 0 33 0 1 3 0 5 1 6 8 0 8 0 1 3 219

% 6.39 15.98 5.94 28.77 0.00 0.00 4.11 0.46 5.94 0.91 0.00 15.07 0.00 0.46 1.37 0.00 2.28 0.46 2.74 3.65 0.00 3.65 0.00 0.46 1.37 100

Table 6 Univariate tests of dividend payout of cross-listed firms. Results correspond to the Wilcoxon-paired signed-rank tests of shifts in dividend payout ratios before and after cross-listing. Values are average payout ratios in percentage points. Means are calculated 3 years before cross-listing and 3 years after cross-listing. Panel A shows the results of shifts in payout policy before and after cross-listing for the whole sample of U.S. cross-listed firms, exchange-traded cross-listings, and by insider-ownership category. Panel B tests the mean differences by ADR level. Panel C tests the mean differences of Insider-ownership categories using sub-samples by capital raising intentions. Panel A

All cross-listed firms

All capital-raising

All insider-owner

All Non-insider-owner

Firms After Before Difference

346 27.55 25.39 2.16***

167 19.71 17.32 2.39**

81 17.12 11.28 5.84**

84 27.29 28.84 −1.55

Panel B

OTC

Exchange-traded

Firms After Before Difference

127 34.49 32.06 2.43

Panel C

Insider-owner

PORTAL

Non-Capital-Raising

Non-capital-raising Firms After Before Difference

25 28.86 19.47 9.39*

76 30.69 29.40 1.29

Capital-Raising 91 14.97 12.26 2.71*

52 28.01 26.17 1.84

Non-insider-owner Capital-raising 56 11.88 7.63 4.25

Non-capital-raising

Capital-raising

49 32.56 35.38 -2.82

35 19.92 19.67 0.24

Note: ***, **, * indicate statistical significance at the 0.01, 0.05 and 0.10 level, respectively.

Mergent online. We exclude financial firms as they are highly regulated and highly leveraged, which could potentially bias our dividend models. The data span is from 1990 to 2010, since Datastream has limited accounting data from emerging countries before 1990 (Li, Morck, Yang, & Yeung, 2004) and ADR level 4 were introduced in April 1990 (Boubakri et al., 2013). We control for the maturity of the domestic stock market using the stock market development

variable from Beck, Demirgüc¸-Kunt, & Levine (2000) database.8 The sample includes all firms from countries considered as emerging or developing economies for which Datastream has information. The aforementioned list of countries is similar to Fernandes and

8 Stock market development is defined as the average aggregate market value of publicly listed shares deflated by the country’s GDP (Beck et al., 2000).

192

O.A. Esqueda / The Quarterly Review of Economics and Finance 59 (2016) 186–199

Table 7 Pooled Tobit analysis of dividend policy (full sample), PAYOUT is the dividend per share divided by earnings per share. DIVIDEND represents the amount of total common dividends paid by the firm deflated by total assets. Cross-listed indicates whether the firm is cross-listed in the United States. PRETAX represents the profits before taxes divided by total assets. LEVER equals total liabilities deflated by total assets. CAPEX is the capital expenditures to total assets. MBV is the market-to-book ratio. SalesGrowth is the firm’s 1 year sales growth. CASH is the amount of cash holdings deflated by total assets. SIZE is the log of total assets in USD. Antidirectors is the home country’s antidirectors rights index. StockMarket is the domestic stock market development. Civil law indicates whether the country is ruled by civil or common law. TAX represents the decreased dividend tax rate. TAX equals one during 2003 and thereafter. RE represent the amount of retained earnings to total equity. LEVEL equals one if the firm has an active ADR with the indicated column-level. ADR levels different from the one indicated in the column are not included in the corresponding regression. Standard errors are in parentheses. Panel A. Dependent variable: dividend payout (PAYOUT)

LEVEL

Level 1

Level 1

Level 2

Level 2

Level 3

Level 3

Level 4

Level 4

4.953*** (1.11) -35.06*** (4.32) -101.1*** (10.63)

-4.986** (1.94) -30.48*** (4.64) -85.48*** (10.73)

-14.23*** (2.38) -42.04*** (5.22) -114.1*** (12.05)

145.0*** (21.07)

-17.15*** (2.67) -38.16*** (6.21) -104.5*** (13.96) -5.002*** (0.47) 154.5*** (3.32) -27.27*** (1.41) 24.37*** (3.87) -7.255*** (0.61) 9.450*** (1.99) 4.012*** (0.21) 2.386* (1.45) -0.0995 (0.26) -0.0226 (0.13) 132.5*** (25.04)

1.625 (1.48) -38.89*** (4.92) -109.2*** (11.59)

95.91*** (18.78)

-3.948* (2.10) -27.19*** (5.39) -78.69*** (12.20) -5.011*** (0.47) 154.1*** (3.30) -28.16*** (1.41) 24.91*** (3.87) -7.171*** (0.61) 10.01*** (1.84) 4.155*** (0.21) 2.299 (1.44) -0.0177 (0.27) 0.0195 (0.13) 87.27*** (21.78)

128.6*** (19.89)

1.566 (1.55) -32.16*** (5.79) -93.87*** (13.28) -5.025*** (0.47) 154.5*** (3.30) -27.04*** (1.41) 25.67*** (3.84) -7.285*** (0.60) 12.21*** (2.10) 4.137*** (0.21) 1.821 (1.44) -0.0266 (0.27) -0.0158 (0.13) 106.5*** (23.38)

41,866 Yes Yes Yes 0.0294 -141,725

36,148 Yes Yes Yes 0.0299 -123,955

41,485 Yes Yes Yes 0.0287 -140,244

35,775 Yes Yes Yes 0.0292 -122,534

42,032 Yes Yes Yes 0.0293 -142,107

36,382 Yes Yes Yes 0.0296 -124,646

Const

112.7*** (17.44)

3.965*** (1.18) -27.59*** (5.00) -84.67*** (11.96) -4.736*** (0.44) 152.5*** (3.21) -28.84*** (1.37) 23.81*** (3.78) -6.456*** (0.58) 10.20*** (2.04) 4.206*** (0.20) 1.645 (1.40) 0.0209 (0.26) 0.110 (0.13) 87.38*** (20.16)

N Country effects Industry effects Time effects PseudoR2 LR

43,571 Yes Yes Yes 0.0294 -148,006

37,610 Yes Yes Yes 0.0299 -129,547

Antidirectors Civil StockMarket PRETAX LEVER CAPEX SalesGrowth CASH SIZE TAX × Cross-listed

150.8*** (2.86) -30.31*** (1.26) 20.23*** (3.59) -8.228*** (0.54) 11.95*** (1.90) 4.502*** (0.19) -5.621*** (1.16)

RE MBV

151.2*** (2.94) -30.33*** (1.29) 20.98*** (3.68) -8.866*** (0.56) 10.64*** (1.70) 4.484*** (0.20) -5.129*** (1.19)

151.6*** (2.95) -29.98*** (1.29) 18.83*** (3.67) -9.036*** (0.57) 7.538*** (1.80) 4.257*** (0.20) -4.757*** (1.19)

151.3*** (2.94) -29.20*** (1.29) 21.83*** (3.65) -8.973*** (0.56) 14.21*** (1.96) 4.459*** (0.19) -5.423*** (1.18)

Panel B. Dependent variable: aggregate dividends (DIVIDEND)

LEVEL Antidirectors Civil

Level 1

Level 1

Level 2

Level 2

Level 3

Level 3

Level 4

Level 4

0.321** (0.13) −2.738*** (0.31) −7.642*** (0.77)

0.161 (0.13) −2.236*** (0.35) −6.591*** (0.85) −0.356*** (0.03) 19.67*** (0.23) −2.517*** (0.10) 1.749*** (0.27) −1.140*** (0.04) 1.952*** (0.14) 0.236*** (0.01) 0.192 (0.16) 0.00985 (0.02) 0.232*** (0.01)

−1.165*** (0.19) −2.764*** (0.33) −7.134*** (0.77)

−1.000*** (0.19) −2.360*** (0.37) −6.370*** (0.85) −0.356*** (0.03) 19.66*** (0.23) −2.471*** (0.10) 1.665*** (0.27) −1.173*** (0.04) 1.657*** (0.13) 0.234*** (0.01) 0.247 (0.24) 0.00836 (0.02) 0.226*** (0.01)

−2.506*** (0.30) −3.814*** (0.37) −9.761*** (0.86)

−2.786*** (0.31) −3.290*** (0.43) −8.616*** (0.96) −0.348*** (0.03) 19.52*** (0.23) −2.416*** (0.10) 1.693*** (0.27) −1.188*** (0.04) 1.840*** (0.14) 0.224*** (0.01) 1.165*** (0.34) -0.0012 (0.02) 0.225*** (0.01)

-0.0799 (0.17) -3.480*** (0.35) -9.117*** (0.83)

0.100 (0.17) -2.917*** (0.40) -7.908*** (0.92) -0.355*** (0.03) 19.45*** (0.23) -2.382*** (0.10) 1.820*** (0.27) -1.189*** (0.04) 2.185*** (0.15) 0.231*** (0.01) -0.0819 (0.21) 0.0043 (0.02) 0.226*** (0.01)

StockMarket PRETAX LEVER CAPEX SalesGrowth CASH SIZE TAX*Cross-listed RE MBV

21.47*** (0.21) −2.166*** (0.09) 1.623*** (0.26) −1.231*** (0.04) 2.351*** (0.14) 0.255*** (0.01) 0.0302 (0.16)

21.43*** (0.21) −2.166*** (0.09) 1.552*** (0.26) −1.247*** (0.04) 1.899*** (0.12) 0.252*** (0.01) 0.326 (0.23)

21.31*** (0.22) −2.160*** (0.09) 1.420*** (0.26) −1.268*** (0.04) 1.877*** (0.13) 0.239*** (0.01) 1.195*** (0.33)

21.18*** (0.21) -2.075*** (0.09) 1.727*** (0.26) -1.260*** (0.04) 2.617*** (0.14) 0.253*** (0.01) 0.078 (0.21)

O.A. Esqueda / The Quarterly Review of Economics and Finance 59 (2016) 186–199

193

Table 7 (Continued ) Panel B. Dependent variable: aggregate dividends (DIVIDEND) Level 1

Level 1

Level 2

Level 2

Level 3

Level 3

Level 4

Level 4

Const

8.688*** (1.26)

7.626*** (1.43)

9.006*** (1.34)

8.276*** (1.51)

13.40*** (1.50)

12.06*** (1.72)

11.70*** (1.41)

10.41*** (1.61)

N Country effects Industry effects Time effects PseudoR2 LR

43,825 Yes Yes Yes 0.102 −85,908

37,845 Yes Yes Yes 0.11 −74,080

42,127 Yes Yes Yes 0.101 −82,183

36,383 Yes Yes Yes 0.108 −70,870

41,740 Yes Yes Yes 0.0997 −81,283

36,009 Yes Yes Yes 0.108 −70,015

42,286 Yes Yes Yes 0.101 −82,119

36,617 Yes Yes Yes 0.109 −71,041

Note: ***, **, * indicate statistical significance at the 0.01, 0.05, and 0.10 level, respectively.

Ferreira (2008) and O’Connor (2009). All U.S. cross-listing levels are included in the sample as it is important to study whether the shifts in dividend policy differ across ADR levels. Furthermore, Brown, Liang, & Weisbenner (2007) find that the amount of dividends paid by firms increased following the enactment of the “Jobs and Growth Tax Relief Reconciliation Act” (JGTRRA).9 For individual investors, the after-tax value of dividend increased subsequent to this event (Brown et al., 2007). Table 3 depicts the descriptive statistics. Control variables have similar mean values between cross-listed and non-cross-listed firms. Firms that cross-list have more cash and are slightly larger on average. However, the variables used as proxies for investment opportunities, market to book value and sales growth, suggest that both groups of firms have comparable growth opportunities. Finally, the difference in aggregate dividends (DIVIDEND), payout ratio (PAYOUT), and dividend yield (YIELD) seems negligible overall. Table 4 displays the frequency of categorical variables. Firm ownership concentration seems to be very common in emerging economies as 95% of the cross-listed firms have at least one owner with more than 10% claim on the firm’s cash flows. Hence, it is important to control for ownership concentration interacting with other ownership structure variables. We believe it is more relevant for minority shareholders to determine whether an insider has at least a 10% ownership on the firm’s cash flows (Insider-Owner) as controlling insiders can potentially influence dividend payouts (Noe & Rebello, 1996). Cross-listings with insider-ownership concentration account for about 48% of the firm/year observations. We additionally observe that about 30% of the sampled firms/years are dividend payers. Finally, the frequency of firms by home country and cross-listing level is shown in Table 5. We exclude firms with missing data and firms with less than an average of U.S. $100 million in total assets, similar to O’Connor (2009). There are 5,576 non-cross-listed firms, 281 ADRs level 1 (OTC), 219 ADRs levels 2 and 3 (exchange-traded), and 122 ADRs level 4 (PORTAL). Hereafter, we refer interchangeably to levels 1 and 4 as OTC and PORTAL, respectively. Exchangetraded cross-listings are dominated by Latin American countries with close to half of the number of firms and China with 28% of cross-listings. This fact suggests a strong linkage between firms from these countries and U.S. stock markets. In addition, China, India, and South Korea combined account for over 53% of the number of non-cross-listed firms included in the sample, reflecting the size of their domestic equity markets. 3.2. Econometric Technique

techniques might provide biased results. Additionally, fixed effects are not able to fit the time-invariant regressors included in our model specifications. Tobit regressions are appropriate when at least one tail of the dependent variable’s distribution is censored. Therefore, the Tobit approach fits this study, as about one-third of the sampled firms are non-dividend payers during the sample period. Additionally, when PAYOUT is used, the data are censored at 0 and 100%. In the latter case, we need a “two-limit Tobit.” Tobit models can fit panel data; however, the Hausman (1978) test indicates that the Tobit random effects approach is unfeasible in this study as there is endogeneity of the regressors with the error term. Random effects require that Cov(Xit vi ) = 0 to generate efficient coefficients. Therefore, a panel Tobit model cannot be used. Consequently, we employ pooled Tobit models as recommended by Wooldridge (2002).10 In addition to the Tobit model described above, we employ a logit model to test the binary dependent variable measuring the likelihood of paying dividends (PAY) as applied in previous dividend policy research (DeAngelo et al., 2006; Shao et al., 2013). Panel logit models are not suitable to our data given the Hausman (1978) test results, which reject the feasibility of random effects models. Therefore, we employ pooled logit models. When ownership structure variables are included, the tests are run exclusively on U.S. exchange-listed firms due to data limitations on ownership structure variables for non-U.S. exchange listed firms. Pooled Tobit analysis is a variant of: yit = ˛ +

k 

j

ˇj xit + ˛i + ˛t + vit

(1)

j=1

Where (yit ) takes the form of PAYOUT or DIVIDEND, ˛i and ˛t are firm and year effects, vit is an idiosyncratic error term, and xit is the group of k regressors described in Section 3. In addition, these model specifications include industry and country effects. Pooled Logit analysis employs a variant of:

⎛

PAYit = exp ⎝˛ +

k 

⎞

ˇj xit + ˛i + ˛t + vit ⎠ j

(2)

j=1

Alternatively, Log(PAYit ) = ˛ +

k 

j

ˇj xit + ˛i + ˛t + vit

(3)

j=1

When facing panel data, it is common to employ fixed and random effects; however, in the presence of censored panel data, these

Where PAYit is a binary variable indicating whether the firm is a dividend payer during year i. Variables ˛i and ˛t are firm and year effects, respectively, vit is an idiosyncratic error term, and xit is the

9 The JGTRRA was approved by Congress in 2003. The JGTRRA aims at reducing the maximum tax rate on dividend income from 38.6% to 15%.

10 O’Connor (2006) uses a similar approach to the censored panel data in dividend payout models. Petrasek (2012) employs fixed/random effects models.

194

O.A. Esqueda / The Quarterly Review of Economics and Finance 59 (2016) 186–199

Table 8 Pooled Tobit analysis of dividend policy of exchange-traded cross-listings. PAYOUT is the dividend per share divided by earnings per share. Cross-listed indicates whether the firm is cross-listed in the United States. DIVIDEND represents the amount of total common dividends paid by the firm deflated by total assets. PRETAX represents the profits before taxes divided by total assets. LEVER equals total liabilities deflated by total assets. CAPEX is the capital expenditures to total assets. MBV is the market-to-book ratio. SalesGrowth is the firm’s 1-year sales growth. CASH is the amount of cash holdings deflated by total assets. SIZE is the log of total assets in USD. Antidirectors is the home country’s antidirectors rights index. StockMarket is the domestic stock market development. Civil law indicates whether the country is ruled by civil or common law. TAX represents the decreased dividend tax rate. TAX equals one during 2003 and thereafter. Insider Owner indicates that the firm has at least one insider owner with at least 10% claim in the company. Dual-class equals one if the firm uses more than one stock class. OwnershipConcentration indicates whether at least one investor owns at least 10% of the firm. ConcOwnerDual is the interaction of OwnershipConcentration × DualClass. Capital-Raising equals one if the firm is allowed to raise capital in U.S. exchanges. Standard errors are in parentheses. Panel A: Dividend payout (PAYOUT)

Cross-listed Capital-Raising Antidirectors Civil PRETAX LEVER CAPEX SalesGrowth CASH SIZE TAX StockMarket InsiderOwner InsiderOwner*Cross-listed ConcOwnerDual ConcOwnerDual*Cross-listed

(1)

(2)

(3)

(4)

(5)

6)

-5.200*** (0.72) 4.603*** (0.74) 56.78*** (0.17) 239.5*** (0.75) 70.67*** (3.10) -23.13*** (1.40) 35.87*** (4.78) -13.01*** (0.53) 4.222*** (0.86) 16.67*** (0.05) 19.81*** (0.73) -8.025*** (0.30) 11.59*** (0.89) 6.390*** (0.87) 15.04*** (1.14) 13.40*** (1.15)

-3.884*** (0.71) 3.463*** (0.74) 32.26*** (0.17) 164.4*** (0.74) 71.32*** (3.09) -22.73*** (1.39) 37.75*** (4.76) -13.44*** (0.53) 3.454*** (0.85) 16.40*** (0.05) 19.88*** (0.72) -8.420*** (0.29) 7.729*** (0.88) 10.77*** (0.84)

-3.874*** (0.71) 3.461*** (0.75) 55.20*** (0.17) 233.4*** (0.75) 71.11*** (3.10) -23.00*** (1.40) 38.25*** (4.79) -13.18*** (0.53) 3.665*** (0.86) 16.45*** (0.05) 20.74*** (0.72) -8.376*** (0.30) 7.685*** (0.90) 10.79*** (0.86)

-5.200*** (0.72) 4.603*** (0.75) 45.19*** (0.17) 204.7*** (0.75) 70.67*** (3.09) -23.13*** (1.39) 35.87*** (4.79) -13.01*** (0.52) 4.222*** (0.86) 16.67*** (0.05) 19.81*** (0.73) -8.025*** (0.30)

-10.10*** (0.72) 5.449*** (0.77) 7.048*** (0.18) 91.83*** (0.76) 70.12*** (3.16) -29.50*** (1.40) 59.22*** (4.88) -12.22*** (0.51) 2.227** (0.87) 16.73*** (0.05) 0.213 (0.61) -4.507*** (0.30)

6.390*** (0.80)

8.532*** (0.80)

13.40*** (1.05) 15.04*** (1.13) -460.3*** (0.74)

13.30*** (0.98)

-10.10*** (0.72) 5.449*** (0.76) 28.81*** (0.18) 149.0*** (0.77) 70.12*** (3.19) -29.50*** (1.41) 59.22*** (4.86) -12.22*** (0.51) 2.227** (0.88) 16.73*** (0.05) 0.213 (0.61) -4.507*** (0.30) 8.109*** (0.90) 8.532*** (0.88) 21.77*** (1.12) 13.30*** (1.15)

-255.4*** (0.75)

-356.2*** (0.75)

(4) 1,655 Yes Yes Yes No 0.0863 -5587

(5) 1,655 Yes Yes No Yes 0.0834 -5604.6

(6) 1,655 Yes Yes No Yes 0.0834 -5604.6

-518.2*** (0.75)

-391.1*** (0.74)

23.08*** (1.07) -507.4*** (0.74)

(1) 1,655 Yes Yes Yes No 0.0863 -5587

(2) 1,665 Yes Yes Yes No 0.0857 -5633.8

(3) 1,655 Yes Yes Yes No 0.086 -5588.7

InsiderOwner*DualClass Const

N Firm effects Country effects Time effects Industry effects Pseudo R2 LR Panel B: Aggregate dividends (DIVIDEND)

Cross-listed Capital-Raising Antidirectors Civil PRETAX LEVER CAPEX SalesGrowth

(1)

(2)

(3)

(4)

(5)

(6)

−0.245*** (0.05) 0.874*** (0.06) 1.592*** (0.01) 4.952*** (0.05) 13.11*** (0.26) −0.832*** (0.09) −2.052*** (0.36) −1.106*** (0.04)

−0.133*** (0.05) 0.781*** (0.06) −0.0430*** (0.01) -0.0749 (0.05) 13.15*** (0.26) −0.818*** (0.09) −1.849*** (0.36) −1.121*** (0.04)

−0.138*** (0.05) 0.782*** (0.06) 1.459*** (0.01) 4.442*** (0.05) 13.17*** (0.26) −0.807*** (0.09) −1.864*** (0.36) −1.117*** (0.04)

−0.245*** (0.05) 0.874*** (0.06) 1.817*** (0.01) 5.626*** (0.05) 13.11*** (0.26) −0.832*** (0.09) −2.052*** (0.37) −1.106*** (0.04)

−0.617*** (0.05) 1.040*** (0.06) −1.401*** (0.01) −2.030*** (0.05) 12.94*** (0.27) −1.247*** (0.09) 0.678* (0.38) −1.022*** (0.04)

−0.617*** (0.05) 1.040*** (0.06) −1.298*** (0.01) −1.226*** (0.05) 12.94*** (0.27) −1.247*** (0.09) 0.678* (0.38) −1.022*** (0.04)

O.A. Esqueda / The Quarterly Review of Economics and Finance 59 (2016) 186–199

195

Table 8 (Continued ) Panel B: Aggregate dividends (DIVIDEND)

CASH SIZE TAX StockMarket InsiderOwner InsiderOwner × Cross-listed ConcOwnerDual ConcOwnerDual × Cross-listed

(1)

(2)

(3)

(4)

(5)

(6)

−0.173*** (0.07) 0.316*** (0.00) 1.611*** (0.05) −0.262*** (0.02) −0.225*** (0.06) 0.597*** (0.06) 0.858*** (0.08) 1.088*** (0.07)

−0.217*** (0.06) 0.293*** (0.00) 1.689*** (0.05) −0.288*** (0.02) −0.539*** (0.06) 0.954*** (0.06)

−0.219*** (0.07) 0.298*** (0.00) 1.685*** (0.05) −0.289*** (0.02) −0.541*** (0.06) 0.955*** (0.06)

−0.173*** (0.07) 0.316*** (0.00) 1.611*** (0.05) −0.262*** (0.02)

−0.183*** (0.07) 0.399*** (0.00) 0.533*** (0.04) 0.213*** (0.02)

0.597*** (0.06)

0.761*** (0.06)

1.088*** (0.07) 0.858*** (0.07) -12.39*** (0.05)

1.166*** (0.07)

−0.183*** (0.07) 0.399*** (0.00) 0.533*** (0.04) 0.213*** (0.02) −0.495*** (0.07) 0.761*** (0.06) 0.103 (0.08) 1.166*** (0.08)

1.244*** (0.05)

0.234*** (0.05)

1,663 Yes Yes Yes No 0.191 -3074.2

1,663 Yes Yes No Yes 0.181 -3112.2

1,663 Yes Yes No Yes 0.181 -3112.2

InsiderOwner × DualClass Const

-11.26*** (0.05)

-2.783*** (0.05)

1.510*** (0.07) -10.38*** (0.05)

N Firm effects Country effects Time effects Industry effects Pseudo R2 LR

1,663 Yes Yes Yes No 0.191 -3074.2

1,674 Yes Yes Yes No 0.19 -3096.6

1,663 Yes Yes Yes No 0.19 -3076.7

Note: ***, **, * indicate statistical significance at the 0.01, 0.05 and 0.10 level, respectively.

group of k regressors described above. The Logit models include industry and country effects. 4. Results and discussion Table 6 shows the outcome of univariate tests of shifts in dividend policy of cross-listed firms. We compare the average dividend payout (PAYOUT) 3 years before cross-listing with the average PAYOUT 3 years after cross-listing. In the full sample tests, it appears that there is a substantial increase in the average amount that cross-listed firms distribute to shareholders. Additionally, firms with insider ownership and capital-raising firms seem to pay more dividends than non-inside-owner firms, whereas PORTAL and OTC cross-listings do not appear to significantly increase dividend payments. Table 7 Panel A portrays the results of pooled Tobit models where the dependent variable is PAYOUT. In Panel B, the dependent variable is switched to DIVIDEND. Both of these models include separate tests for each of the four cross-listing levels. The column name indicates the ADR level being tested versus the sample of non-cross-listed firms. ADR levels different from the one indicated in the column are not included in the corresponding regression. Panel A indicates that LEVEL2 and LEVEL3 (exchange-traded) have a negative coefficient, suggesting that following cross-listing on a U.S. exchange, firms tend to decrease the amount of dividend payments compared with what would be expected given firms’ characteristics. These firms appear to substitute dividend payments with improved corporate governance. Conversely, following an OTC cross-listing firms are more likely to be dividend payers as managers attempt to signal better shareholder protection.11 Moreover,

11 The shift in dividend payout of OTC firms is surprising given the lack of a significant increase in dividends in the univariate tests presented in Table 6. However, alternative univariate t-tests show a significant increase in dividend payout at the 10% significance level for OTC cross-listings. It appears that managers of OTC crosslistings attempt to signal better corporate governance and thereby attract investors following a U.S. cross-listing event.

PORTAL firms do not seem to have a significant change in dividend payout after a cross-listing event similar to O’Connor (2006). Our results remain robust to the change of the dependent variable, DIVIDEND. Panel B shows support to the previous conclusions as the coefficients for LEVEL1−LEVEL4 variables have the same sign. Table 8 depicts the results of model specifications including exchange-traded firms only. These models include firm-level ownership structure variables for exchange-traded cross-listings. Panel A employs PAYOUT as the dependent variable. Exchangetraded cross-listings decrease dividend payments as they substitute dividends with increased shareholder protection. We find that exchange-traded capital-raising firms pay more dividends than non-capital raising firms, as the coefficient for Capital-Raising is positive and significant at the 1% level in most model specifications, supporting the liquidity hypothesis. Capital-Raising shows that it is not only important to know whether the firm cross-lists (on a U.S. exchange), but also whether it is allowed to raise capital. Similarly, ownership structure is a strong predictor of the shift in dividend policy. The interaction Insider-Owner × Cross-Listed is positive and significant, suggesting that following cross-listing events, insidercontrolled firms shift to a higher equilibrium dividend payment as stated by Noe and Rebello (1996). Similarly, the interaction of DualClass × Cross-Listed is positive and significant further reinforcing the argument that firm-level weak corporate governance improves the most from a U.S. cross-listing. Insider ownership and dual-class share structures appear to pay more dividends, as indicated by the coefficient of the interaction Insider-Owner × Dual-Class. Yet, having only one of these characteristics appears to increase the payout ratio. Control variables at the country-level and firm-level included in the model specifications have the expected sign and coefficient. The positive coefficient of TAX shows that firms pay more dividends after the dividend tax rate was reduced, following the enactment of the JGTRRA in 2003.12 To confirm the results are not specific to one

12 Control variable TAX has positive and negative coefficients in Table 7 Panel B and A, respectively. The aforementioned mixed results might be due to the sample

196

O.A. Esqueda / The Quarterly Review of Economics and Finance 59 (2016) 186–199

Table 9 Pooled Tobit analysis of dividend policy (full sample). PAYOUT is the dividend per share divided by earnings per share. DIVIDEND represents the amount of total common dividends paid by the firm deflated by total assets. Cross-listed indicates whether the firm is cross-listed in the United States. PRETAX represents the profits before taxes divided by total assets. LEVER equals total liabilities deflated by total assets. CAPEX is the ratio of capital expenditures to total assets. MBV is the market-to-book ratio. SalesGrowth is the firm’s 1-year sales growth. CASH is the amount of cash holdings deflated by total assets. SIZE is the log of total assets in US$. Antidirectors is the home country’s antidirectors rights index. StockMarket is the domestic stock market development. Civil law indicates whether the country is ruled by civil or common law. TAX represents the decreased dividend tax rate. TAX equals one during 2003 and thereafter. Standard errors are in parentheses. Panel A: Dependent variable: dividend payout (PAYOUT).

Cross-listed Antidirectors Civil

(1)

(2)

(3)

(4)

(5)

-0.0114 (0.13) -0.0589 (0.13) -0.678 (0.43)

-0.0281 (0.16) 0.0336 (0.13) -0.498 (0.41)

-0.002 (0.15) -0.0668 (0.13) -0.69 (0.43)

21.75*** (2.48) −2.480*** (0.42) 1.26 (1.07)

20.45*** (2.47) −2.827*** (0.35) 1.227 (1.11) 0.0242 (0.02) 0.217*** (0.06) −1.171*** (0.20) 0.825* (0.45) 0.211*** (0.05)

-0.0396 (0.14) -0.066 (0.15) -0.74 (0.48) -0.0135 (0.08) 21.18*** (2.46) −2.608*** (0.41) 1.316 (1.09) 0.0162 (0.02)

-1.711** (0.81)

-1.358 (1.15)

-0.0766 (0.15) 0.0109 (0.14) -0.609 (0.47) -0.0427 (0.07) 19.82*** (2.38) −2.978*** (0.35) 1.258 (1.09) 0.0214 (0.02) 0.217*** (0.06) −1.141*** (0.20) 0.755* (0.40) 0.233*** (0.04) 0.11 (0.32) -1.667* (0.97)

44,628 Yes Yes -88,785

43,037 Yes Yes -86,051

40,545 Yes Yes -80,823

47,315 Yes Yes -94,372

(1)

(2)

(3)

(4)

(5)

0.114 (1.93) −2.122* (1.24) -5.94 (3.85)

0.101 (2.31) −2.490** (1.25) -5.895 (3.81)

0.511 (2.12) −2.352* (1.26) -6.275 (3.92)

142.4*** (19.68) −30.78*** (5.33) 17.1 (14.52)

145.0*** (21.68) −29.88*** (5.07) 21.45 (15.11) 0.321 (0.24) 0.133 (0.48) −6.694*** (2.00) 1.259 (4.31) 3.412*** (0.82)

-0.039 (2.16) -2.549 (1.82) -7.207 (4.58) -0.335 (1.14) 136.0*** (20.41) −32.39*** (5.45) 17.92 (15.11) 0.149 (0.27)

47,050 Yes Yes −169,120

StockMarket PRETAX LEVER CAPEX RE MBV SalesGrowth CASH SIZE TAX × Cross-listed Const N Country effects Time effects LR

−1.223*** (0.21) 0.898** (0.39) 0.227*** (0.06) -0.0432 (0.33) -1.369 (0.95) 47,150 Yes Yes -94,103

−1.196*** (0.21) 0.794** (0.38) 0.242*** (0.06)

21.92*** (2.48) −2.431*** (0.41)

−1.205*** (0.22) 0.878** (0.40) 0.226*** (0.06)

-1.302 (0.97)

Panel B: Dependent variable: aggregate dividends (DIVIDEND)

Cross-listed Antidirectors Civil

-8.239 (13.88)

-10.09 (17.97)

-0.812 (2.17) -2.526 (1.75) -6.305 (4.45) −0.0327 (0.98) 142.8*** (20.65) −30.62*** (5.11) 20.62 (14.45) 0.228 (0.26) 0.121 (0.46) −6.160*** (1.98) 0.526 (4.27) 3.745*** (0.68) -3.565 (3.77) -8.628 (15.16)

44,377 Yes Yes −161,152

42,794 Yes Yes −154,357

40,308 Yes Yes −146,844

StockMarket PRETAX LEVER CAPEX RE MBV SalesGrowth CASH SIZE TAX × Cross-listed Const N Country effects Time effects LR

−7.806*** (1.99) 0.138 (3.73) 3.801*** (0.83) -4.495 (4.17) -11.32 (14.29) 46,885 Yes Yes −168,635

Note: ***, **, * indicate statistical significance at the 0.01, 0.05, and 0.10 level, respectively.

−7.400*** (2.06) -1.272 (3.71) 3.852*** (0.88)

143.2*** (19.86) −30.40*** (5.23)

−7.538*** (2.23) -0.465 (3.89) 3.651*** (0.93)

-10.36 (15.59)

O.A. Esqueda / The Quarterly Review of Economics and Finance 59 (2016) 186–199

197

Table 10 Pooled logit analysis of dividend policy of exchange-traded cross-listings. Dependent variable: dividend payer (PAY). PAY is an indicator of whether the firm pays dividends during a given year. Cross-listed indicates whether the firm is cross-listed in the United States. PRETAX represents the profits before taxes divided by total assets. LEVER equals total liabilities deflated by total assets. CAPEX is the capital expenditures to total assets. MBV is the market-to-book ratio. SalesGrowth is the firm’s 1-year sales growth. CASH is the amount of cash holdings deflated by total assets. SIZE is the log of total assets in USD. Antidirectors is the home country’s antidirectors rights index. StockMarket is the domestic stock market development. Civil law indicates whether the country is ruled by civil or common law. TAX represents the decreased dividend tax rate. TAX equals one during 2003 and thereafter. Insider Owner indicates that the firm has at least one insider owner with at least 10% claim in the company. Dual-class equals one if the firm uses more than one stock class. OwnershipConcentration indicates whether at least one investor owns at least 10% of the firm. ConcOwnerDual is the interaction of OwnershipConcentration × DualClass. Capital-Raising equals one if the firm can raise capital in U.S. exchanges. Standard errors are in parentheses.

Cross-listed Capital-Raising Antidirectors Civil StockMarket PRETAX LEVER CAPEX SalesGrowth CASH SIZE TAX InsiderOwner InsiderOwner × CrossListed ConcOwnerDual ConcOwnerDual × CrossListed

(1)

(2)

(3)

(4)

(5)

(6)

-1.530*** (0.54) 2.023*** (0.74) 0.677 (1.80) 7.921 (6.14) -0.278 (0.32) 6.626*** (1.77) -2.375* (1.25) 7.679*** (2.91) -0.722 (0.48) 0.959* (0.57) 2.210*** (0.40) 0.559 (0.85) -1.970 (1.83) 2.302** (1.02) 0.281 (2.58) -1.913** (0.94)

-1.580*** (0.53) 2.073*** (0.75) 0.907 (1.78) 8.608 (6.08) -0.274 (0.32) 6.565*** (1.79) -2.394* (1.24) 7.693*** (2.90) -0.726 (0.48) 0.931 (0.58) 2.208*** (0.39) 0.547 (0.86) -1.440 (1.75) 1.775** (0.85)

-1.530*** (0.54) 2.023*** (0.74) 2.648* (1.38) 13.83*** (4.90) -0.278 (0.32) 6.626*** (1.77) -2.375* (1.25) 7.679*** (2.91) -0.722 (0.48) 0.959* (0.57) 2.210*** (0.40) 0.559 (0.85)

-1.983*** (0.51) 1.847*** (0.66) 1.995 (2.76) 21.20** (8.66) -0.155 (0.29) 5.547*** (1.72) -2.932*** (1.14) 10.78*** (2.54) -0.508 (0.38) 0.50 (0.54) 2.063*** (0.39) 0.478 (0.30) -2.063 (1.83) 2.175** (0.86) 0.893 (1.28) -1.582* (0.86)

-2.037*** (0.50) 1.891*** (0.66) 2.809 (2.02) 21.65*** (7.73) -0.153 (0.29) 5.462*** (1.73) -3.001*** (1.13) 10.79*** (2.52) -0.515 (0.38) 0.475 (0.55) 2.065*** (0.38) 0.451 (0.30)

-1.284*** (0.41) 1.074** (0.53) 1.135 (2.36) 13.19* (7.93) -0.0975 (0.20) 7.125*** (1.46) -2.367** (0.97) 9.300*** (2.13) -0.389 (0.30)

1.716** (0.72)

1.831*** (0.28) 0.16 (0.25) -1.061 (1.48) 1.007* (0.55)

-48.30*** (16.63)

-50.30*** (13.74)

-36.21*** (13.82)

(4) 858 85% Yes Yes No Yes 0.441 -317

(5) 858 85% Yes Yes No Yes 0.439 -317.8

(6) 1,051 83% Yes Yes No Yes 0.401 -415.7

-35.23*** (11.44)

-1.587 (2.45) -36.29*** (11.28)

-1.913** (0.94) 0.281 (2.58) -45.08*** (7.85)

(1) 858 86% Yes Yes Yes No 0.466 -302.6

(2) 858 86% Yes Yes Yes No 0.464 -303.5

(3) 858 86% Yes Yes Yes No 0.466 -302.6

InsiderOwner × DualClass Const

N Correctly Classified Firm effects Country effects Time effects Industry effects Pseudo R2 LR

2.302** (1.02)

Note: ***, **, * indicate statistical significance at the 0.01, 0.05, and 0.10 level, respectively.

measure of dividends, the dependent variable is switched to DIVIDEND in Panel B. The main conclusions persist, providing support to our previous findings. Table 9 shows results for the tests of whether cross-listing at any level is a good predictor of shifts in dividend policy. Panel A employs PAYOUT as dependent variable and Panel B tests the same models using DIVIDEND. Both results show that the cross-listing event alone is not relevant to the dividend payment decision. However, it appears that the level of cross-listing conveys relevant information for changes in payout policy. Firm-level control variables have the expected sign overall in both Panels A and B. As a robustness

used in the model, which includes foreign firms not cross-listed in the U.S. These firms are unaffected by the JGTRRA.

check, we further test a binary-dependent variable (PAY), indicating whether the firm is a dividend payer during a given year. We present pooled logit models that confirm our main findings about exchange-traded cross-listings and support our insiderownership conclusions from Table 8. In Table 10, we provide evidence that confirm our earlier findings that exchange-traded firms are less likely to be dividend payers during a given year. Conversely, capital-raising firms have a higher likelihood to become dividend payers than non-capital-raising firms. After cross-listing, firms with insider-ownership increase the likelihood to be dividend payers, whereas firms using dual-class share structure decrease the propensity to pay dividends. The results regarding ownership structure are robust for insider-controlled firms given the coefficient of the interaction InsiderOwner × Cross-Listed, which consistently has a positive and significant effect on dividend payments as it does in our previous Tobit models (Table 8). Contrary to our expectations,

198

O.A. Esqueda / The Quarterly Review of Economics and Finance 59 (2016) 186–199

the interaction Dual-Class × Cross-Listed shows a negative coefficient implying that firms with dual-class share structure are less likely to pay dividends following cross-listings. Moreover, the variable TAX shows a positive but insignificant coefficient. Cross-listed firms appear to increase the amount of dividend payment after the dividend tax rate was reduced, but there is no indication that this event has increased the propensity to become a dividend payer for exchange-traded cross-listed firms. Finally, our logit models shown in Table 10 are able to predict whether the firm is a dividend payer with an average accuracy above 85%. Overall, our main findings remain robust to several techniques and measures of the dependent variable and provide strong evidence supporting the signaling hypothesis as insider-owned cross-listings increase dividend payouts and non-insider-owned cross-listings decrease dividend payments. In addition, the bonding hypothesis is supported as our results show evidence for both the outcome model of dividends (for insider-owned cross-listings) and the substitute model of dividends (for non-insider-owned crosslistings). Our results support the market segmentation hypothesis only for exchange-traded ADRs. The aforementioned results are intuitive since foreign firms listed on U.S. exchanges are expected to overcome capital barriers better than firms listed as OTC or PORTAL. Finally, we observe that the liquidity hypothesis is supported only if the firm with intentions to raise capital is exchange-traded, suggesting that PORTAL cross-listings do not rely on capital raised privately as a means to support a constant stream of dividend payments to investors.

5. Concluding remarks We study the shifts in dividend policy of emerging-market firms cross-listed in the United States and we employ the bonding, signaling, market segmentation, and liquidity hypotheses to explain changes in the equilibrium dividend policy of foreign firms cross-listed in the United States. We conclude that foreign firms allowed to raise new capital on U.S. exchanges pay larger dividends and are more likely to be dividend payers during a given year following a cross-listing event than exchange-traded ADRs not allowed to raise capital in the United States, thereby supporting the liquidity hypothesis of dividends. Furthermore, we conclude that foreign firms cross-listing on U.S. exchanges overcome capital barriers as they pay smaller dividends and are less likely to be dividend payers compared with non-exchange-traded cross-listings. Firms with the potential for managerial entrenchment (insiderscontrolled) appear more likely to increase or initiate dividends following cross-listing than firms without insider-owners. The new equilibrium dividend payout is contingent on the firm’s pre-crosslisting locus of control as management signals differ by ownership structure. Supporting the substitute model of dividends, exchangetraded cross-listings decrease payouts as they substitute dividends with better corporate governance. However, the implication of the outcome model of dividends is supported in that when crosslistings are controlled by insiders, dividend distributions and the likelihood to pay dividends increase following U.S. cross-listings. It appears that insiders can use dividends as a mechanism to affect share prices, rather than as a channel for wealth distribution. Investors should not focus on a U.S. cross-listing event by itself as a precursor to higher dividends as it does not signal future shifts in dividend policy. However, the cross-listing level and ownership structure at the time of cross-listing convey information on future dividend payouts. Our paper provides relevant insights for portfolio managers and individual investors interested in dividend distributions from emerging-market firms. We show the feasibility of evaluating and predicting the potential shift in dividend payout based on the ADR

level and intensity of shareholder protection. We suggest that firms where insiders can influence the decision-making process are more likely to use dividend increases/initiations to signal an optimistic financial future. Our paper is relevant for academics as these findings complement previous research on shareholder protection and dividend signaling mechanisms. References Aivazian, V., Booth, L., & Cleary, S. (2003). Do emerging market firms follow different dividend policies from U.S. firms? Journal of Financial Research, 26, 371–387. Alexander, G. J., Eun, C. S., & Janakiramanan, S. (1988). International listings and stock returns: Some empirical evidence. Journal of Financial and Quantitative Analysis, 23, 135–151. Allen, F., & Michaely, R. (2003). Payout policy. In G. Constantinides, M. Harris, & R. Stulz (Eds.), Handbook of the Economics of Finance (pp. 337–430). Amsterdam: North-Holland. Ambarish, R., John, K., & Williams, J. (1987). Efficient signaling with dividends and investments. The Journal of Finance, 42, 321–343. Asem, E., & Alam, S. (2015). Market movements and the excess cash theory. The Quarterly Review of Economics and Finance, 55, 140–149. Beck, T., Demirgüc¸-Kunt, A., & Levine, R. (2000). A new database on the structure and development of the financial sector. The World Bank Economic Review, 14, 597–605. Bekaert, G., & Harvey, C. R. (2000). Foreign speculators and emerging equity markets. The Journal of Finance, 55, 565–613. Black, F. (1976). The dividend puzzle. The Journal of Portfolio Management, 2, 5–8. Boubakri, N., Cosset, J., & Samet, A. (2013). International cross-listings and subsequent security-market choices: Evidence from ADRs. The Financial Review, 48, 311–341. Brown, J. R., Liang, N., & Weisbenner, S. (2007). Executive financial incentives and payout policy: Firm responses to the 2003 dividend tax cut. The Journal of Finance, 62, 1935–1965. Claessens, S., Djankov, S., & Lang, L. H. (2000). The separation of ownership and control in East Asian corporations. Journal of Financial Economics, 58, 81–112. Coffee, J. (1999). The future as history: The prospects for global convergence in corporate governance and its implications. Northwestern University Law Review, 93, 641–708. Coffee, J. (2002). Racing towards the Top?: The impact of cross-listings and stock market competition on international corporate governance. Columbia Law Review, 102, 1757–1831. DeAngelo, H., DeAngelo, L., & Stulz, R. M. (2006). Dividend policy and the earned/contributed capital mix: a test of the life-cycle theory. Journal of Financial Economics, 81, 227–254. Djankov, S., La Porta, R., Lopez-de-Silanes, F., & Shleifer, A. (2008). The law and economics of self-dealing. Journal of Financial Economics, 88, 430–465. Doidge, C., Karolyi, G. A., & Stulz, R. M. (2004). Why are foreign firms listed in the U.S. worth more? Journal of Financial Economics, 71, 205–238. Easterbrook, F. H. (1984). Two agency-cost explanations of dividends. The American Economic Review, 74, 650–659. Errunza, V., & Losq, E. (1985). International asset pricing under mild segmentation: Theory and test. The Journal of Finance, 40, 105–124. Faccio, M., & Lang, L. H. (2002). The ultimate ownership of Western European corporations. Journal of Financial Economics, 65, 365–395. Fama, E. F., & French, K. R. (2001). Disappearing dividends: Changing firm characteristics or lower propensity to pay? Journal of Financial Economics, 60, 3–43. Fenn, G. W., & Liang, N. (2001). Corporate payout policy and managerial stock incentives. Journal of Financial Economics, 60, 45–72. Fernandes, N., & Ferreira, M. A. (2008). Does international cross-listing improve the information environment. Journal of Financial Economics, 88, 216–244. Fuller, K., & Blau, B. M. (2010). Signaling, free cash flow and “nonmonotonic” dividends. The Financial Review, 45, 21–56. Grullon, G., & Michaely, R. (2002). Dividends, share repurchases, and the substitution hypothesis. The Journal of Finance, 57, 1649–1684. Hail, L., & Leuz, C. (2009). Cost of capital effects and changes in growth expectations around U.S. cross-listings. Journal of Financial Economics, 93, 428–454. Hausman, J. A. (1978). Specification tests in econometrics. Econometrica, 46, 1251–1271. Jensen, M. C. (1986). Agency cost of free cash flow, corporate finance, and takeovers. American Economic Review, 76, 323–329. Karolyi, G. A. (2006). The world of cross-listings and cross-listings of the world: Challenging conventional wisdom. Review of Finance, 10, 99–152. La Porta, R., Lopez-de-Silanes, F., & Shleifer, A. (1999). Corporate ownership around the world. The Journal of Finance, 54, 471–517. La Porta, R., Lopez-de-Silanes, F., Shleifer, A., & Vishny, R. W. (2000). Agency problems and dividend policies around the world. The Journal of Finance, 55, 1–33. Li, K., Morck, R., Yang, F., & Yeung, B. (2004). Firm-specific variation and openness in emerging markets. The Review of Economics and Statistics, 86, 658–669. Lins, K. V. (2003). Equity ownership and firm value in emerging markets. Journal of Financial and Quantitative Analysis, 38, 159–184. Lins, K. V., Strickland, D., & Zenner, M. (2005). Do non-U.S. firms issue equity on U.S. stock exchanges to relax capital constraints? Journal of Financial and Quantitative Analysis, 40, 109–133.

O.A. Esqueda / The Quarterly Review of Economics and Finance 59 (2016) 186–199 Miller, M. H., & Rock, K. (1985). Dividend policy under asymmetric information. The Journal of Finance, 40, 1031–1051. Mitton, T. (2004). Corporate governance and dividend policy in emerging markets. Emerging Markets Review, 5, 409–426. Noe, T. H., & Rebello, M. J. (1996). Asymmetric information, managerial opportunism, financing, and payout policies. The Journal of Finance, 51, 637–660. O’Connor, T. G. (2006). Cross-listing in the U.S. and domestic investor protection. The Quarterly Review of Economics and Finance, 46, 413–436. O’Connor, T. G. (2009). Does cross-listing in the USA really enhance the value of emerging market firms? Review of Accounting and Finance, 8, 308–336. Petrasek, L. (2012). Do transparent firms pay out more cash to shareholders? Evidence from international cross-listings. Financial Management, 41, 615–636.

199

Rozeff, M. S. (1982). Growth, Beta and agency costs as determinants of dividend payout ratios. Journal of Financial Research, 5, 249–259. Shao, L., Kwok, C. C. Y., & Guedhami, O. (2013). Dividend policy: Balancing shareholders’ and creditors’ interests. Journal of Financial Research, 36, 43–66. Stulz, R. M. (1999). Globalization, corporate finance, and the cost of capital. Journal of Applied Corporate Finance, 12, 8–25. Truong, T., & Heaney, R. (2007). Largest shareholder and dividend policy around the world. The Quarterly Review of Economics and Finance, 47, 667–687. Watts, R. (1973). The information content of dividends. The Journal of Business, 46, 191–211. Wooldridge, J. M. (2002). Econometric analysis of cross section and panel data. Cambridge, MA: MIT Press.