The market reaction to international cross-listings:

Journal of Financial Economics 51 (1999) 103—123

The market reaction to international cross-listings: evidence from Depositary Receipts Darius P. Miller* Lowry Mays College and Graduate School of Business, Texas A&M University, College Station, TX 77843, USA Received 27 October 1997; received in revised form 2 March 1998

Abstract This paper examines the stock price impact of international dual listings. The sample consists of 181 firms from 35 countries that instituted their first Depositary Receipt program over the period 1985—1995. The market reaction to a Depositary Receipt program is larger in magnitude and more pervasive than previously reported. The stock price reaction is related to choice of exchange, geographical location (i.e., emerging or developed markets), and avenues for raising equity capital (i.e., public versus private offerings).  1999 Elsevier Science S.A. All rights reserved. JEL classification: G15 Keywords: Segmentation; Depositary Receipts; Capital barriers

* Corresponding author. Tel.: 409 845 4894; fax: 409 845 3884; e-mail: [email protected].  This paper is based on Chapter 2 of my doctoral thesis at the University of California, Irvine. I am indebted to Philippe Jorion for guidance throughout this project. I would like to thank Geert Bekaert, John Bizjak, Vihang Errunza, Nai-Fu Chen, Charles Cuny, Roberto Gutierrez, Robert Haugen, Neal Stoughton, Eli Talmor, and anonymous referee, the editor, Jerold Warner, and seminar participants at the Berkeley Program in Finance, McGill University, Texas A&M University, University of California at Davis, University of Miami, University of Southern Illinois at Carbondale, and the University of Washington for helpful comments and suggestions I would also like to thank Bernard Johnson and Dori Flannagan from The Bank of New York, as well as Evelyn Walsh from National Quotation Bureau, Inc., for supplying some of the data used in this study. Support from the Center for International Business Studies at Texas A&M University is gratefully acknowledged. Prior drafts of this paper were entitled ‘Why do foreign firms list in the United States? An empirical analysis of the depositary receipt market’. 0304-405X/99/$ — see front matter  1999 Elsevier Science S.A. All rights reserved PII: S 0 3 0 4 - 4 0 5 X ( 9 8 ) 0 0 0 4 5 - 2

104

D.P. Miller/Journal of Financial Economics 51 (1999) 103—123

1. Introduction Over the last decade, the number of foreign firms listing their securities on U.S. capital markets as Depositary Receipts (DRs) has increased dramatically. By the end of 1994, over 800 firms sponsored DR programs, an increase of 756 percent since 1986. In 1994 alone, a total of 289 new Depositary Receipts from 44 countries were established. The increase in DR programs has been particularly large among firms in emerging markets, where direct and indirect barriers to capital flows can be most acute. Two hundred and nine, or 73%, of the non-US companies that established sponsored DR programs in 1994 were from emerging markets. While many theoretical models of asset pricing under barriers to international capital flows predict an increase in security prices upon cross-listing, previous empirical studies find little or no evidence that dual listing increases firm value. This paper pursues two objectives. First, I reexamine the impact of dual listing on firm value, concentrating on the effects around the announcement of a DR program. I find positive abnormal returns around the announcement date, providing evidence that firms benefit from listing shares outside their home market. The second objective is to measure the effect created by barriers such as illiquidity and investor recognition on share value. Using stocks from 35 countries, I exploit the institutional differences in dual listings to examine how these barriers affect a firm’s stock price and therefore its cost of capital. I find that abnormal returns are largest for firms that list on major U.S. exchanges such as NYSE or NASDAQ, rather than on OTC ‘pink sheets’ or PORTAL. This finding is consistent with the literature supporting the hypotheses that more liquidity and a larger shareholder base increase shareholder wealth. Further, in contrast to the existing evidence on U.S. firms, I find that when foreign firms raise new equity capital in a public DR offering, the abnormal returns are positive, while in a private DR offering, the abnormal returns are negative. Dual listing can be rationalized by the existence of market segmentation. Segmentation can arise from direct barriers (e.g., ownership restrictions and taxes) or from indirect barriers (e.g., information availability, differences in accounting standards or liquidity risk). Stapleton and Subrahmanyan (1977) suggest that dual listing on foreign capital markets can circumvent market segmentation, thus increasing firm value and lowering the cost of capital. Therefore, if dual listing reduces segmentation, analyzing the impact of a dual listing provides a measure of the effects of international market segmentation. Tests of market segmentation typically involve modeling a specific barrier to integration and deriving equilibrium returns (see, e.g., Alexander et al., 1987; Errunza and Losq, 1985; Eun and Janakiramanan, 1986; Stulz, 1981). Empirical tests (see, e.g., Hietala, 1989; Jorion and Schwartz, 1986) confront the joint hypothesis of market integration, efficiency in the market and specification of

D.P. Miller/Journal of Financial Economics 51 (1999) 103—123

105

the asset pricing model. Since these are one period models, they are unable to account for changes in barriers through time. Recent research by Bekaert and Harvey (1995) incorporates the time varying nature of barriers to capital flows. Using the event study methodology, this paper examines the impact of barriers to capital flows without the need to specify an asset pricing model. By investigating how abnormal returns vary by institutional and geographical differences, it is also possible to analyze multiple types of barriers. Prior empirical research examining the market reaction to dual listings focuses on the effects around the date the security is listed on the U.S. exchange. Eun et al. (1993) and Domowitz et al. (1995) find abnormal returns are insignificant around the DR listing, while Jayaraman et al. (1993) find that only Japanese firms have positive shareholder gains. Alexander et al. (1988) and Foerster and Karolyi (1996) find abnormal returns are positive before the listing yet become negative once trading begins in the United States. Little evidence of gains to shareholders upon the listing of a DR program exists. I depart from earlier studies in two ways: First, I focus on the date the dual listing is announced. The extensive literature on firms that move from the OTC or NASDAQ to the NYSE demonstrates the importance of utilizing the announcement date rather than the listing date. Van Horne (1970), McConnell and Sanger (1984), Ying et al. (1977), Sanger and McConnell (1986), and Kadlec and McConnell (1994) all document positive abnormal returns to the announcement of an exchange listing. Second, I examine the stock price reaction across each type of DR program. Each DR program trades off varying degrees of liquidity and investor recognition with disclosure requirements. For instance, firms can list their shares on the OTC ‘pink sheet’, PORTAL, NASDAQ, AMEX, or the NYSE markets, but those listing on the NASDAQ, AMEX, or NYSE must reconcile their financial statements to US Generally Accepted Accounting Principles (GAAP). In addition, the establishment of a DR program may also entail raising new equity capital. This may be done in a public offering or via a private placement of DRs. Therefore, using information on each type of DR program provides an opportunity to measure the effects of liquidity, changes to US; GAAP, and international equity offerings on share value. The remainder of the paper is organized as follows: Section 2 gives an overview of the structure of the DR market, while Section 3 describes the data set used in the paper. Section 4 outlines the stock price response to a DR program, and Section 5 presents the differences in the stock price reaction. A summary is given in Section 6.

2. The structure of the Depositary Receipt market Depositary Receipts were created in 1927 by J.P. Morgan as a means for U.S. investors to participate in the London Stock Market. A DR is a negotiable

106

D.P. Miller/Journal of Financial Economics 51 (1999) 103—123

certificate issued by a depositary bank for a number of non-U.S. securities that are held by the depositary’s custodian in the home market of the non-U.S. company. DRs are registered with the SEC and trade like any other U.S. security. They are quoted and pay dividends in U.S. dollars. DRs traded outside the U.S. are called Global Depositary Receipts (GDRs). Since the holder of a DR has the right to redeem the receipt for the underlying share, the DR and the underlying share are virtually perfect substitutes for each other, after adjusting for transactions costs. These costs include fees paid to the Depositary Bank for DR creation or cancellation. Early DR programs were generally initiated at the request of investors, without company authorization. These ‘unsponsored’ programs can have multiple registrar, transfer and paying agents. In the 1950s, several Australian and South African mining companies created the ‘sponsored’ DR program. Under the sponsored program, a company signs an agreement with one depositary to be the sole agent for its DRs. The sponsored program gives more control of DRs to the firm, allowing it to compile ownership characteristics of its investors. In 1983, the Securities and Exchange Commission (SEC) required that all new DR programs must have company approval in order to be established. DRs offer several potential advantages for U.S. investors seeking portfolio diversification: they allow for investment in countries which have restricted access to their primary equity market (for example, Korea); they are denominated, and pay dividends, in U.S. dollars; the depository bank is responsible for the distribution of financial statements to investors; trading costs are lower; settlement occurs in the U.S., which may be faster and more reliable than in the home market, and withholding tax payments may be simpler. In addition, many DR programs can lead to greater company disclosure, such as full SEC reporting according to U.S. accounting standards. Companies have a choice of four types of DR facilities: three levels of public offerings as well as private placement. Table 1 summarizes the characteristics of the DR programs by exchange, accounting standards, SEC registration, capital raised, time to completion and costs. The least costly way for a company to cross-list its shares is to establish a ‘Level I’ DR program. Level I DRs trade in the U.S. over the counter (OTC) ‘pink sheet’ market and on some exchanges outside the United States. By filing a 12g3-2(b) exemption from the 1934 Exchange Act, the company does not have to comply with U.S. GAAP or full SEC disclosure. Fifty-six percent of the approximately 1500 DR programs are classified as Level I. Level II DRs are traded on the NASDAQ, AMEX, or NYSE, and are used by companies seeking greater liquidity and investor recognition. Major exchange DR programs, however, entail greater costs. The initial fee alone can exceed $1 million. Firms that issue Level II DRs must also reconcile to U.S. GAAP, report quarterly and meet the listing requirements of the particular U.S. exchange where they trade. Because of the higher costs and more stringent reporting

Home Country Standards

Exempt Existing shares only (public offering)

10 weeks

)$25,000

Accounting standards

S.E.C. registration Share issuance

Time to completion

Costs $200,000—700,000

10 weeks

Full Registration Existing shares only (public offering)

U.S. GAAP

NYSE, AMEX or NASDAQ

Level II

$500,000—2,000,000

14 weeks

Full registration New equity capital raised (public offering)

U.S. GAAP

NYSE, AMEX or NASDAQ

Level III

$250,000—500,000

16 days

Exempt New equity capital raised (private offering)

Home Country Standards

PORTAL

144a

Foreign securities traded in the U.S. are required to perform periodic reporting under the 1934 Exchange Act, provided that the company’s equity securities are held of record by 500 or more persons, of which 300 or more are U.S. residents. This requires quarterly reporting, filing the form 20-F in U.S. Generally Accepted Accounting Principles. Level I and 144a DRs are eligible for a 12g3-2(b) exemption from this requirement, and only have to supply to the SEC copies of information that the company makes public in its home country. Privately placed Depositary Receipts are also eligible for the 12g3-2(b) exemption. They trade between Qualified Institutional Buyers under SEC Rule 144a, which provides a safe harbor exemption from the registration requirements of the Securities Act of 1933. Source: Global Offerings of Depositary Receipts, A ¹ransaction Guide (The Bank of New York, 1995)

OTC ‘pink sheets’

Primary exchange

Level I

Table 1 Characteristics of Depositary Receipt programs traded in the U.S.

D.P. Miller/Journal of Financial Economics 51 (1999) 103—123 107

108

D.P. Miller/Journal of Financial Economics 51 (1999) 103—123

requirements, many firms choose a Level I DR program instead of a major exchange DR program. While Level I and II DRs are created using existing shares, firms can also tap the U.S. capital markets via a public offering or a private placement of DRs. DRs have evolved into one of the most popular tools for raising international capital. In 1994 alone, over $20 billion dollars was raised through the DR market, including 77% of the total equity capital raised by firms located in emerging economies. Level III DRs raise new equity capital in a public offering and trade on the NASDAQ, AMEX, or the NYSE. The issuer registers the offering under the 1933 Securities Act and reports under the 1934 Exchange Act. The company must meet full SEC disclosure requirements, comply with U.S. GAAP, report quarterly, and meet the listing requirements of the U.S. exchange where it chooses to list. Both Level II and Level III programs require the firm to complete Form 20-F, which is similar to a 10-K report. As of 1 June 1995, there were 329 Level II and Level III programs listed on the NYSE, AMEX, and NASDAQ. In 1994, over $11 billion was raised via Level III DRs, an increase of 633% from 1990. Finally, Rule 144A Depositary Receipts (RADRs), are DRs that raise new equity capital via private placement. Rule 144A was adopted by the SEC in April, 1990 to increase the liquidity of privately placed DRs by allowing Qualified Institutional Buyers (QIBs) to trade among themselves without a holding restriction. The SEC defines QIBs as either institutions that manage at least $100 million in securities or registered broker-dealers owning and investing $10 million in securities of nonaffiliates. Prior to rule 144A, privately placed DRs could not be resold until they had been held by the investor for a three year period. RADRs are traded on PORTAL, a screen based automated trading system developed by NASD to support the secondary trading of Rule 144A securities. Despite these measures, the 144a market has remained illiquid, with the majority of trades occurring in unregulated offshore markets. The major advantage of 144a private placements, however, is that they are allowed an exemption from the 1934 Exchange Act and therefore can be used to raise capital without meeting the reporting and disclosure requirements of a U.S. public offering. In 1994, over $8 billion was raised using 144A Depositary Receipts, an increase of more than 980% over 1990. 3. The data The sample consists of 181 stocks domiciled outside the United States that announced their first DR Program over the period from 1985 to 1995. The sample is constructed from data compiled by the Bank of New York consisting of all existing DR programs. A company must have an identifiable announcement and listing date to be included in the sample. In addition, return

D.P. Miller/Journal of Financial Economics 51 (1999) 103—123

109

data on the underlying stock is required starting 150 days before the announcement date and ending at least 125 days after the listing date. Return data for each stock as well as the corresponding national market index are compiled from the Datastream International data base. Daily closing price and dividend data are used to compute daily total returns for each underlying security while index returns are taken directly from Datastream. Announcement dates are collected from the Lexis/Nexis data base. The search algorithm uses key words found in a preliminary sample of announcements. These include terms for the instrument such as Depositary, 144a, ADR(s), and GDR(s). Since the first announcement often is for the initial application to the SEC, Securities and Exchange Commission and SEC are included. The text and headlines of the Lexis/Nexis articles are searched using the company name and these key words. The earliest press release is taken as the announcement date. The announcement date is taken as the earliest press release written in English. Conversations with Reuters Information Services indicated that the announcement may be reported in English with at most a one day lag to allow translation from the local source. To account for this and any time zone differences, a three day window around the announcement date is examined throughout the analysis. Listing dates for OTC DRs (Level I) are supplied from the National Quotation Bureau Inc., the publishers of the ‘pink sheets’. Listing dates are defined as the first date of market maker activity in the security. For DRs privately placed under Rule 144A, the listing date is taken to be the day the securities are offered. These offering dates are supplied by the Bank of New York. The mean time between the announcement and listing date is 77 days. Table 2 provides summary characteristics of the sample by DR program and country of registration. Of the 181 announcements, 88 (49%) are for Level I, 23 (13%) for Level II, and 30 (16%) for Level III programs. The remaining 40 (22%) are for DRs privately placed under Rule 144A. Of the 35 countries represented in the sample, the United Kingdom, with 29 (16%), has the largest number of firms.

4. Stock price response to a Depositary Receipt program 4.1. Empirical method An event study procedure is used to measure changes in share value around the announcement of a firm’s initial DR program. To measure abnormal  The majority of the announcements came from Reuters News Service. Others came from such sources as the Financial Times, South China Morning Post, Euroweek, LatinFinance, Business Wire, Australian Financial Review, Korea Economic Daily, Business Times (Singapore), and the Regulatory News Service.

110

D.P. Miller/Journal of Financial Economics 51 (1999) 103—123

Table 2 Sample statistics of 181 foreign stocks that initiated depositary receipt programs over the period 1985—1995 Level I DRs are listed on the over-the-counter ‘pink sheet’ market. Level II DRs are listed on the NYSE, AMEX or NASDAQ. Level I and Level II DRs do not raise new equity capital. Level III DRs are listed on the NYSE, AMEX or NASDAQ, and 144a DRs are traded by qualified institutional buyers on the PORTAL system. Both Level III and 144a DRs raise new equity capital. Country

Type and quantity of listing Level I

Level II

144a

Total

Level III

Argentina Australia Austria Brazil Chile China Denmark Finland France Germany Hong Kong India Indonesia Ireland Italy Japan Korea Luxembourg Malaysia Mexico Netherlands New Zealand Norway Philippines Portugal Singapore South Africa Spain Sweden Switzerland Taiwan Thailand Turkey United Kingdom Venezuela

2 5 3 3 0 1 0 1 7 4 8 1 1 1 0 7 0 0 2 3 2 0 2 1 1 2 2 1 1 4 0 2 1 19 1

1 5 0 0 2 0 0 0 0 1 0 0 0 0 1 3 1 0 0 0 2 1 0 0 0 0 0 0 2 0 0 0 0 4 0

1 2 0 0 8 0 1 0 1 0 2 0 0 1 1 0 1 0 0 5 0 0 1 0 0 0 0 0 0 0 0 0 0 6 0

1 1 0 1 2 0 0 1 1 0 0 15 0 0 0 0 5 1 0 2 0 0 0 1 0 1 0 0 0 1 6 0 1 0 0

5 13 3 4 12 1 1 2 9 5 10 16 1 2 2 10 7 1 2 10 4 1 3 2 1 3 2 1 3 5 6 2 2 29 1

Total

88

23

30

40

181

D.P. Miller/Journal of Financial Economics 51 (1999) 103—123

111

returns, I estimate a market model for each firm using local currency daily returns. As a proxy for the market return, I use a market capitalization weighted index for each country from Datastream International. Recent research by Foerster and Karolyi (1996) and Urias (1994) document changes in risk parameters around cross-listing. With the announcement day defined as day 0, the OLS market model coefficients are estimated separately in a pre-announcement (day !150 to day !26) and post-announcement period (day #26 to day #150). Abnormal returns are determined by the prediction errors from the OLS market model. Coefficients from the pre-announcement model are used to calculate abnormal returns from day !25 to day #25, while the post-announcement model is used to compute abnormal returns after day #26. Abnormal returns are then averaged across firms to form the average abnormal return. The standard Brown and Warner (1985) test statistic accounting for cross sectional dependence is used to test the null hypothesis of no abnormal performance. 4.2. Changes in share value at announcement Table 3 presents average abnormal returns for the 50 days surrounding the announcement of the initial DR program. Fig. 1 summarizes the evidence. Consistent with the hypothesis that a dual listing increases firm value, abnormal returns around the announcement period are positive and significant. Between day !1 to day #1, firms announcing a listing in the U.S. via a DR program experience a positive abnormal return of 0.0115 (t"3.87). The day !1, day 0, and day #1 abnormal returns are 0.0054 (t"3.14), 0.0067 (t"3.94), and !0.0007 (t"!0.39). The sign-ranked z-statistics for day !1 to #1 are 2.53, 1.73 and !0.96, respectively. In addition, the increase in share value around the announcement date appears permanent. The cumulative abnormal return between day #2 and day #25 is 0.0071 (t"0.84). For the period including the announcement window through the listing day, the average abnormal return is positive and significant (0.0389 (t"2.40)). Over the 125 day post-listing period, the average abnormal

 To verify the robustness of the results, various methodologies are employed to calculate abnormal returns. For instance, if markets are not completely segmented, the firm’s shares may be priced with respect to its home market as well as the market where it dual lists (Stapleton and Subrahmanyam, 1977). Predicted returns are determined from a multivariate OLS regression of the firm’s domestic market proxy and the S&P 500 index. The results are robust with respect to this change, as well as the interval for estimating market model parameters (e.g., pre-announcement and post-listing periods) and currency denomination (e.g., local currency or US dollars). For example, the 3 day announcement, announcement to listing and 125 day post-listing returns are 0.0141 (t"4.75), 5.94 (t"4.14), 0.0317 (t"1.63) respectively, when the simple market adjusted model is used.

112

D.P. Miller/Journal of Financial Economics 51 (1999) 103—123

Table 3 Average abnormal returns of 181 foreign stocks around the announcement of a Depositary Receipt program Abnormal returns are market model adjusted using parameters estimated over a 125 prelisting period, from day !150 to !26 relative to the announcement date. A national stock market index in each country is used as a proxy for the market portfolio. The sample period is 1985—1995. * and ** indicate significance of the t-statistic at the 0.05, and 0.01 levels, respectively. The percentage of firms with nonnegative abnormal returns are given, with -, and -- indicating significance of the signed rank test at the 0.05 and 0.01 levels, respectively. Event day

Average abnormal return

Percentage nonnegative

Cumulative abnormal return

!25 !20 !15 !10 !5 !4 !3 !2

!0.0009 0.0007 !0.0011 0.0007 !0.0005 !0.0014 0.0014 !0.0005

43 44 46 46 46 44 47 50

!0.0009 0.0025 0.0098 0.0066 0.0106 0.0092 0.0107 0.0101

0.0054** 0.0067** !0.0007

5750 45

0.0155 0.0222 0.0216

0.0023 0.0011 0.0012 0.0006 !0.0001 0.0011 0.0004 !0.0033

46 49 51 49 43 52 46 42-

0.0239 0.0250 0.0262 0.0268 0.0327 0.0335 0.0327 0.0286

!1 0 1 2 3 4 5 10 15 20 25

return is 0.0030 (t"0.16). While Alexander et al. (1988), Eun et al. (1993), and Foerster and Karolyi (1996) find negative abnormal returns after a firm’s DRs begin trading in the United States, our evidence is consistent with the findings of Jayaraman et al. (1993) and Domowitz et al. (1995) who find firms earn a normal rate of return following listing. The positive announcement period effect, taken with the normal post-listing performance, is consistent with the equilibrium models of asset pricing under barriers to capital flows (see, e.g., Stapleton and Subrahmanyam, 1977; Errunza and Losq, 1985; Eun and Janakiramanan, 1986; Alexander et al., 1987; Eun et al., 1993). In these models, the international trading of assets in otherwise segmented capital markets serves to integrate international capital markets.

D.P. Miller/Journal of Financial Economics 51 (1999) 103—123

113

Fig. 1. Cumulative abnormal returns from day !25 before to day #25 after the announcement of a Depositary Receipt program. The daily abnormal returns are market model adjusted for each security. The daily abnormal returns are averaged across firms then cumulated. The sample is for 181 firms that listed shares as Depositary Receipts over the period 1985—1995.

Share value increases and cost of capital decreases as a result of the dual listing. 5. Cross-sectional differences in stock price reactions In this section, market segmentation is examined in two ways. First, the DR exchange is used as a proxy for two specific indirect barriers: investor recognition and illiqudity. Second, firms are grouped by the development of their economy to proxy for both indirect and direct (legal) barriers. In addition, this section documents the stock price reaction to global equity offerings using DRs. 5.1. Univariate test results 5.1.1. The effects of illiquidity and investor recognition Merton’s (1987) extension of the CAPM predicts that an increase in a firm’s shareholder base will lower investors’ required return and increase the market value of the firm’s equity. He suggests that managers have an incentive to broaden their ‘investor recognition’ by listing on a national exchange. A dual listing can also be motivated by the superior liquidity services of a major exchange. For example, Amihud and Mendelson (1986) show expected returns should decrease when firms adopt policies to add liquidity. Given that liquidity problems are often cited as a major barrier to investing in emerging markets as discussed by Chuhan (1992), a dual listing may circumvent this indirect barrier

114

D.P. Miller/Journal of Financial Economics 51 (1999) 103—123

Table 4 Average abnormal returns around the announcement of a Depositary Receipt program by exchange, geographic location and equity offering type Abnormal returns are market model adjusted using parameters estimated over a 125 day prelisting period, from day !150 to !26 relative to the announcement date. A national stock market index in each country is used as a proxy for the market portfolio. The sample period is 1985—1995. The cumulative abnormal residual is given by the sum of the average abnormal residuals around the announcement date (day t"0). Free Emerging denotes firms located in emerging markets without legal barriers to capital flows. Restricted Emerging denotes firms located in emerging markets with legal barriers to capital flows. All market classifications are constructed from the IFC’s Emerging Stock Markets Fact book. * and ** indicate significance of the t-statistic at the 0.05, and 0.01 levels, respectively. t"!1 to #1

t"#2 to #25

0.0257 0.0032 0.0099

!0.0109 0.0127** 0.0263**

!0.0419 0.0265* 0.0118

!0.0002 0.0251

0.0087** 0.0154*

0.0022 0.0141

!0.0002 0.0017 0.0377

0.0087** 0.0269* 0.0092

0.0022 0.0277 0.0067

Public (30) Private placement (40) No capital raised (111)

0.0171 0.0257 0.0027

0.0323** !0.0109 0.0139**

Full sample (181)

0.0101

0.0115**

DR type (N)

t"!25 to !2

Panel A: Exchange PORTAL(40) OTC pink sheets (88) NYSE/NASDAQ(53) Panel B: Geographic location Developed (107) Emerging (74) Panel C: IFC Classification Developed (107) Free emerging (26) Restricted emerging (48) Panel D: Equity offering type 0.0106 !0.0419 0.0238* 0.0071

and increase firm value. While shareholder base and bid/ask spread data are unavailable for most foreign markets, I use the DR exchange to test the joint hypothesis that illiquidity and investor recognition segment international capital markets. This proxy is motivated by the results of Kadlec and McConnell (1994) that U.S. firms which list on the NYSE experience a positive stock price response that is positively related to decreases in the bid/ask spread and increases in shareholder base. Panel A of Table 4 presents average abnormal returns around the announcement of a DR program by exchange. Foreign firms which list on the NYSE or

D.P. Miller/Journal of Financial Economics 51 (1999) 103—123

115

NASDAQ experience the largest stock price response (0.0263, t"6.64). Average abnormal returns are smallest for firms that dual list on PORTAL (!0.0109, t"!1.47). Finally, average abnormal returns for firms listing on the OTC ‘Pink Sheet’ market (Level I DR) are positive and significant (0.0127, t"2.83). Consistent with the ‘superior liquidity and investor recognition’ hypotheses and the US results of Sanger and McConnell (1986), foreign firms experience the largest positive abnormal returns when listing on a major US exchange. The smallest stock price response is for the firms that list on PORTAL, where liquidity and investor awareness are notoriously low. For example, several companies (e.g., Argentina’s Telefonica and Venezuela’s Cormon) have abandoned their 144a DR program in favor of a NYSE listing citing poor liquidity. Changes in the number of shareholders for 144a DRs should also be less than those trading on major exchanges, given the SEC restriction that only QIBs may trade on PORTAL. The difference between the NYSE/NASDAQ and PORTAL firms is 3.72% (t-statistic of difference"6.49). The stock price response of firms listing on the OTC is larger than firms on PORTAL (t-statistic of the difference"4.23). Since the OTC market does not have QIB restrictions like PORTAL, liquidity and investor recognition should be greater. In addition, abnormal returns of OTC firms are smaller than the NYSE/NASDAQ sample (t-statistic of the difference"!3.16). This result also supports the market segmentation hypothesis, since liquidity is recognized to be lower on the OTC ‘Pink Sheets’ than on the NASDAQ or NYSE. The evidence indicates that foreign firms listing in the U.S. experience stock price responses that are positively related to proxies for liquidity and investor recognition. I submit this as indirect evidence that DRs can mitigate the effects of these barriers, thus integrating capital markets as predicted by Stapleton and Subrahmanyam (1977). 5.1.2. Geographic location If share value is influenced by international restrictions to capital flows, then the stock price reaction will differ across markets in ways that are related to the severity of the restrictions. Therefore, firms located in markets where barriers to capital flows are more acute should experience larger abnormal returns upon dual listing. Panel B of Table 4 presents the cumulative abnormal returns for the sample firms by market classification. During the three day announcement window, firms in developed markets experience abnormal returns of 0.0087 (t"2.84). For firms located in emerging markets, the three day announcement period return is 0.0154 (t"2.39). While the point estimate is larger for firms from emerging markets, the difference is not statistically significant (t-statistic"1.41).  See ‘Nasdaq Campaigns to Distance Itself From Firms on OTC Bulletin Board’ in WSJ, 8 August 1997.

116

D.P. Miller/Journal of Financial Economics 51 (1999) 103—123

Classifying markets by segmentation, however, is ambiguous. As Bekaert and Harvey (1995) observe, ‘2(I)n general, it is hard to infer the actual degree of market segmentation from the complex set of capital market restrictions in place in a particular country at any one time’. To proxy for the degree of segmentation, I propose a classification based on the Investment Regulations Summary developed by the International Finance Corporation (IFC). Firms are divided into three groups: developed markets, free emerging markets, and restricted emerging markets. The classification encompasses indirect barriers (developed versus emerging) as well as direct barriers (free versus restricted emerging). An example of a restricted emerging market is Chile, which allows the repatriation of capital only after a one year holding period. Since this is done on an annual basis, it permits time variation in the classification. Important to note is that DRs do not circumvent percent holding restrictions, one of the most prevalent direct barriers. Therefore, this test may understate the actual degree of market segmentation. Panel C of Table 4 presents the cumulative abnormal returns for the sample firms by IFC classification. For firms located in emerging markets with a low level of legal barriers to capital flows (Free Emerging classification), the three day announcement period return is 0.0269 (t"2.22). This represents a difference in abnormal return of 1.82% over the developed market sample, which is statistically significant at the 0.01 critical level. Consistent with the international market segmentation hypothesis, firms located in emerging markets have larger abnormal returns than those domiciled in developed markets. It follows that if markets are further segmented by legal barriers to international flows, firms located in countries that have the highest degree of these barriers should have the largest abnormal returns. For firms located in emerging markets with the largest degree of legal barriers (Restricted Emerging), the three day announcement period returns are 0.0092 (t"1.32). These returns are actually lower than those of firms located in developed markets. While this result appears inconsistent with the international market segmentation hypothesis, several points are worth noting. First, the majority (29 of 48) of the firms classified as Restricted Emerging are 144a Depositary Receipts. As shown earlier, these firms have a negative stock price reaction during the announcement window. Therefore, indirect barriers such as liquidity, poor disclosure and investor recognition may outweigh the legal barriers for these securities. Second, the pre-announcement period abnormal returns for the  The IFC classifies entry into emerging markets as free, relatively free, special classes of shares, authorized investors only or closed. These rankings are published yearly (since 1987) in the IFC Emerging Stock Market Factbook. Sample firms that are classified in the year of listing as free are labeled Free Emerging Market, all others classified as emerging by the IFC are labeled Restricted Emerging Market. Those countries not labeled as emerging by the IFC are classified as Developed Market.

D.P. Miller/Journal of Financial Economics 51 (1999) 103—123

117

Restricted Emerging sample are significant at the 10% level (abnormal returns and t statistic are 0.0377 and 1.92). The !25 to #1 cumulative abnormal returns are 0.0469 (t"2.25). This abnormal return of 4.69 percent is larger than the Developed as well as the Free Emerging classification, supporting the international market segmentation hypothesis. While the positive pre-listing returns may result from sample selection bias, it also could indicate leakage in the pre-announcement period. This could be caused by insider trading or from imprecision in the announcement dates, both of which may be most acute in emerging markets. 5.1.3. International equity offerings While DRs have become the most widely used method for firms in emerging markets to raise equity capital, there is little evidence on how these offerings affect shareholder wealth. I attempt to fill this void by measuring abnormal returns around the announcement of DR equity offerings. Panel D of Table 4 reports abnormal returns for capital raising DRs. As previously reported, foreign firms which raise new equity capital in a private placement on PORTAL (RADRs) experience a stock price reaction of !0.0109 (t"!1.47). Firms announcing a public equity offering (Level III DR) experience a positive and significant stock price reaction of 0.0323 (t"5.67). These results stand in sharp contrast with the U.S. studies of equity offerings. For example, Wruck (1989) finds private placements in the U.S. are accompanied by a 4% increase in shareholder wealth while Masulis and Korwar (1986) find that public equity offerings decrease shareholder wealth by 3%. The evidence supports the hypothesis that non-U.S. managers issue equity based on different considerations than their U.S. counterparts. For example, Kang et al. (1995) and Kang and Stulz (1996) argue the model of Myers and Majluf (1984) is inappropriate for Japanese firms. In support of this hypothesis, Kang et al. (1995) finds a positive stock price reaction to offshore warrant bond issues and Kang and Stulz (1996) find insignificant abnormal returns to seasoned equity offerings for Japanese firms. In tests not reported here, capital raising and noncapital raising DRs trading on major exchanges (Level III and II, respectively) are compared. The three day Level III average abnormal return is 3.23% while the Level II return is 1.83%. The difference of 1.4% is statistically significant at the 0.05 critical level. The positive difference is consistent with the assertion of Kang and Stulz (1996) that managerial incentives are not homogeneous across countries. Important to note is the influence of Level III Chilean stocks. Significant government restrictions exist on the ability of Chilean firms to access international capital markets as discussed by Glen and Pinto (1994). Foreign investors face repatriation restrictions, currency controls, and high withholding taxes. According to the IFC, less than 25% of the Chilean market was accessible to foreign investors at the end of 1993. In fact, Errunza et al. (1997) find that Chile

118

D.P. Miller/Journal of Financial Economics 51 (1999) 103—123

remains one of the few foreign markets that U.S. investors cannot mimic using domestically traded securities. Consistent with the hypothesis that legal barriers cause market segmentation, announcement window abnormal returns for the 8 Level III Chilean firms are a very high 8.23%. Additional tests were performed to determine the influence of the Chilean firms. First, the remaining Level III firms have positive and significant abnormal returns of 1.41%. The difference between Level II and Level III is not statistically significant excluding Chile. Supporting the earlier findings, firms raising equity capital using DRs do not experience a significantly lower return than noncapital raising DRs. Also, since Chilean firms are classified as Restricted Emerging, we can still reject the hypothesis that abnormal returns are equal across Developed, Free Emerging and Restricted Emerging excluding Chile. 5.2. Multivariate regression estimates In order to separate the influence of indirect barriers such as illiquidity and investor recognition from direct (i.e., legal) barriers, a cross sectional analysis of the three day abnormal stock returns is performed. ¹-statistics are computed using heteroskedastic consistent variance estimates (White, 1980). To measure the influence of indirect barriers, Regression 1 (Table 5) includes dummy variables for PORTAL, OTC, and NYSE. The results indicate that abnormal returns are related to the DR exchange. The coefficient for PORTAL is insignificant (!1.09%, t"!1.57) while the OTC and NYSE/NASDAQ coefficients are positive and significant (1.27%, t"2.11 and 2.63%, t"3.83, respectively). The F-test that the coefficients are equal is rejected at the 0.01 critical level. Consistent with the univariate tests of illiquidity and investor recognition, abnormal returns are largest for firms that list on the NYSE or NASDAQ and smallest for those on PORTAL. Regression 2 includes a dummy variable (EMERGE) for market classification to determine if the effects are larger in emerging markets. The coefficient on EMERGE is statistically significant (2.40%, t"1.98), suggesting the influence of indirect barriers is stronger in emerging markets. This result is consistent with indirect barriers serving to segment capital markets, given that emerging economies typically have less liquidity, less investor recognition, and fewer disclosure requirements than developed markets. To evaluate the marginal effect of direct barriers, the dummy variable RESTRICT is added to the model in Regression 3. RESTRICT takes on the value one if the firm is classified as an emerging market with legal barriers to capital flows by the IFC’s Emerging Stock Markets Factbook. The coefficient on RESTRICT is insignificant (!0.33%, t"!0.19). This suggests that after controlling for indirect barriers, direct barriers do not influence abnormal returns. The point estimate on EMERGE is of similar magnitude to the previous model, but it is not statistically significant (2.57%, t"1.45). The loss of

D.P. Miller/Journal of Financial Economics 51 (1999) 103—123

119

Table 5 Regression of three day announcement average abnormal returns on firm characteristics The sample includes 181 foreign firms that announced a DR program between 1985 and 1995. Abnormal returns are market model adjusted using market model parameters estimated over a 125 prelisting period, from day !150 to !26 relative to the announcement date. A national stock market index in each country is used as a proxy for the market portfolio. PORTAL, NYSE/NASDAQ, and OTC denote dummy variables that take on the value one if the firm is cross-listed on that exchange. EMERGE denotes a dummy variable that takes on the value one if the firm is classified as an emerging market by the International Finance Corporation (IFC). RESTRICT denotes a dummy variable that takes on the value one if the firm is classified as an emerging market with legal barriers to capital flows by the IFC’s Emerging Stock Markets Fact book. CAPITAL denotes a dummy variable that takes on the value one if the DR program raises new equity capital. All coefficients are multiplied by 100. ¹-statistics (in parentheses) are computed using heteroskedastic consistant variance estimates. * and ** indicate significance of the t-statistic at the 0.05, and 0.01 levels, respectively. Independent variable

Regressions Reg. 1

Reg. 2

Reg. 3

Reg. 4

!1.09 (!1.57)

!3.19* (!2.52)

!3.10** (!2.66)

!2.49 (1.66)

OTC

1.27* (2.11)

0.73 (1.46)

0.71 (1.48)

1.27* (2.10)

NYSE/NASDAQ

2.63** (3.83)

1.77* (2.45)

1.78* (2.54)

1.83* (2.10)

2.40* (1.98)

2.57 (1.45)

PORTAL

EMERGE RESTRICT

!0.33 (!0.19)

CAPITAL p-value for the model R adjusted p-value for the test that PORTAL"OTC"NYSE/NASDAQ

0.00 0.09 0.00

0.00 0.11 0.00

0.00 0.11 0.00

1.40 (1.05) 0.00 0.09 0.00

significance for EMERGE in Regression 3 seems to be the result of a large standard error, which could be caused by the imprecise ability of geographical location to proxy for capital market restrictions as Bekaert and Harvey (1995) find. Finally, Regression 4 examines the effect of capital raising using depositary receipts. The coefficient on CAPITAL, which takes on the value one if the DR raises new equity capital, is insignificantly positive (1.40%, t"1.05). The nonnegative coefficient provides further support to the univariate results that

120

D.P. Miller/Journal of Financial Economics 51 (1999) 103—123

managerial incentives are not homogeneous across countries. As in Regressions 1—3, the hypothesis that the coefficients of the three exchange variables are equal is rejected at the 0.01 critical level. In sum, the multivariate evidence supports the univariate results that the DR exchange is an important determinate of abnormal returns. The evidence for direct barriers, however, is not significant. These results suggest that indirect barriers are the dominant factor in segmenting markets. An alternative explanation is that the proxy for direct barriers is misspecified, and therefore the test is weak. Since DRs do not circumvent prevalent direct barriers such as percent holding and capital inflow ceilings, geographical location may not capture the complex set of capital market restrictions in place through time.

Table 6 Average abnormal returns of 16 foreign stocks around the announcement of an upgrade from level I to Level II depositary program A level I DR program is defined as a cross-listing on the OTC ‘pink sheet’ market without raising new equity capital. A level II DR program is defined as a cross-listing on a national US exchange (NYSE, AMEX or NASDAQ) without raising new equity capital. The sample period is 1985 to 1995. * and ** indicate significance of the t-statistic at the 0.05, and 0.01 levels, respectively. The percentage of firms with nonnegative abnormal returns are given, with -, and -- indicating significance of the signed rank test at the 0.05 and 0.01 levels, respectively. Event day

Average abnormal return

Percentage nonnegative

Cumulative abnormal return

!25 !20 !15 !10 !5 !4 !3 !2

0.0022 !0.0010 !0.0001 !0.0021 !0.0090 0.0044 0.0052 0.0036

56 63 44 44 31 56 63 75

0.0022 0.0182 0.0359 0.0392 0.0225 0.0269 0.0321 0.0356

56 69 63

0.0434 0.0569 0.0656

7563 63 56 44 63 38 50

0.0701 0.0779 0.0810 0.0874 0.1081 0.0921 0.0965 0.1078

!1 0 1 2 3 4 5 10 15 20 25

0.0077 0.0135** 0.0087 0.0045 0.0078 0.0031 0.0064 0.0047 !0.0080 0.0035 0.0037

D.P. Miller/Journal of Financial Economics 51 (1999) 103—123

121

5.3. Evidence on upgrades in Depositary Receipt programs As a final and perhaps more direct test of the ability of indirect barriers to segment capital markets, an event study is performed on firms that upgrade their DR program from Level I to Level II. The sample of firms that upgraded DR programs was supplied by the Bank of New York. To be included, an identifiable announcement and listing date as well as daily return data for the t"!150 to t"#125 period were required. Recall that Level I DRs trade on the OTC ‘pink sheet’ market while Level II DRs trade on the NASDAQ, AMEX, or NYSE but must reconcile to US GAAP. Therefore, the results of an upgrade from Level I to Level II should provide an additional test of the effects of illiquidity and investor recognition. Table 6 reports the average abnormal returns of 16 foreign stocks that announced an upgrade from a Level I to a Level II DR program. Because of the limited size of the sample, the results must be interpreted with caution. The three day announcement window returns are 0.030 (t"3.20). This result adds to the evidence of Kadlec and McConnell (1994) that the improved liquidity services and investor recognition provided by a major exchange increase shareholder wealth. The positive stock price reaction is also associated with increased disclosure requirements. Therefore, the results are also consistent with the findings of Amir et al. (1993) that reconciliation to U.S. GAAP is value-relevant.

6. Conclusion This paper presents new evidence on the stock price impact of an international dual listing. Unlike previous studies, I measure announcement effects on the stock price to show that the market reaction to a DR program is larger in magnitude and more pervasive than previously reported. Controlling for institutional and geographical differences in DR programs, I also find significant differences in the stock price reactions that are related to barriers to capital flows. Abnormal returns are largest for firms that list on major US exchanges such as NYSE or NASDAQ and smallest for firms that list on PORTAL. This finding is consistent with previous research suggesting that indirect barriers such as liquidity risk and low investor recognition segment capital markets. There is also weak evidence that direct barriers can cause market segmentation. Firms located in Chile, where legal barriers to capital flows are prevalent, have extremely large positive abnormal returns. Finally, contrary to the evidence on U.S. firms, the results presented here suggest that foreign firms that enter US capital markets to raise new equity capital in a public offering experience a positive change in shareholder wealth. Those in a private offering experience a negative change in shareholder wealth. Overall, the results provide empirical support for the hypothesis that dual listing can mitigate

122

D.P. Miller/Journal of Financial Economics 51 (1999) 103—123

barriers to capital flows, resulting in a higher share price and a lower cost of capital.

References Alexander, G., Eun, C., Janakiramanan, S., 1987. Asset pricing and dual listing on foreign capital markets: a note. Journal of Finance 42, 151—158. Alexander, G., Eun, C., Janakiramanan, S., 1988. International listings and stock returns: some empirical evidence. Journal of Financial and Quantitative Analysis 23, 135—151. Amihud, Y., Mendelson, H., 1986. Asset pricing and the bid-ask spread. Journal of Financial Economics 17, 223—249. Amir, E., Harris, T., Venuti, E., 1993. A comparison of the value-relevance of U.S. versus non-U.S. GAAP accounting measures using form 20-F reconciliations. Journal of Accounting Research 31, 230—263. Bekaert, G., Harvey, C., 1995. Time-varying world market integration. Journal of Finance 50, 403—444. Brown, S., Warner, J., 1985. Using daily stock returns: The case of event studies. Journal of Financial Economics 14, 3—31. Chuhan, P., 1992. Sources of portfolio investment in emerging markets. Unpublished working paper. World Bank, Washington, D.C. Domowitz, I., Glen, J., Madhavan, A., 1995. International cross-listings, ownership rights and order flow migration: evidence from Mexico. Unpublished working paper. University of Southern California. Errunza, V., Losq, E., 1985. International asset pricing under mild segmentation: theory and test. Journal of Finance 40, 105—124. Errunza, V., Hogan, K., Hung, M., 1997. Have the gains from international diversification dissappeared? Unpublishd working paper. McGill University. Eun, C., Janakiramanan, S., 1986. A model of international asset pricing with a constraint on the foreign equity ownership. Journal of Finance 41, 897—914. Eun, C., Claessens, S. Jun, K., 1993. International trade of assets, pricing externalities, and the cost of capital, Claessens, S., Gooptu, S. Portfolio Investment in Developing Countries. World Bank Discussion Papers, pp. 287—298. Foerster, S., Karolyi, A., 1996. The effects of market segmentation and illiquidity on asset prices: evidence from foreign stocks listing in the U.S. Unpublished working Paper, University of Western Ontario. Glen, J., Pinto, B., 1994. Debt or equity? How firms in developing countries choose. International Finance Corporation Discussion Paper, 22. Hietala, P., 1989. Asset pricing in partially segmented markets: evidence from the Finnish market. Journal of Finance 44, 697—718. Jayaraman, N., Shastri, K., Tandon, K., 1993. The impact of international cross listings on risk and return: the evidence from American depositary receipts. Journal of Banking and Finance 17, 91—103. Jorion, P., Schwartz, E., 1986. Integration vs. segmentation in the Canadian stock market. Journal of Finance 41, 603—614. Kadlec, G., McConnell, J., 1994. The effect of market segmentation and illiquidity on asset prices: evidence from exchange listings. Journal of Finance 49, 611—636. Kang, J., Kim, Y., Park, K., Stulz, R., 1995. An analysis of the wealth effects of Japanese offshore dollar-denominated convertible and warrant bond issues. Journal of Financial and Quantitative Analysis 30, 257—270.

D.P. Miller/Journal of Financial Economics 51 (1999) 103—123

123

Kang, J., Stulz, R., 1996. How different is Japanese corporate finance? An investigation of the information content of new security issues. The Review of Financial Studies 9, 101—139. Masulis, R.W., Korwar, A., 1986. Seasoned equity offerings: an empirical investigation. Journal of Financial Economics 15, 3—29. McConnell, J., Sanger, G., 1984. A trading strategy for new listings on the NYSE. Financial Analyst Journal 40, 34—48. Merton, R., 1987. Presidential address: a simple model of capital market equilibrium with incomplete information. Journal of Finance 42, 483—510. Myers, S.C., Majuluf, N.S., 1984. Corporate financing and investment decisions when firms have information that investors do not have. Journal of Financial Economics 13, 187—221. Sanger, G., McConnell, J., 1986. Stock exchange listings, firm value, and security market efficiency: the impact of Nasdaq. Journal of Financial and Quantitative Analysis 21, 1—25. Stapleton, R., Subrahmanyan, M., 1977. Market imperfections, capital market equilibrium and corporate finance. Journal of Finance 32, 307—319. Stulz, R., 1981. On the effects of barriers to international investment. Journal of Finance 36, 923—934. Urias, M., 1994. The impact of security cross-listing on the cost of capital in emerging markets. Unpublished Working Paper. Stanford University. Van Horne, J., 1970. New listings and their price behavior. Journal of Finance 25, 783—794. White, H., 1980. A heteroskedasticity-consistent covariance matrix estimator and a direct test for hetroskedasticity. Econometrica 48, 817—838. Wruck, K., 1989. Equity ownership concentration and firm value: evidence from private equity financing. Journal of Financial Economics 23, 3—28. Ying, L., Lewellen, W., Schlarbaum, G., Lease, R., 1977. Stock exchange listings and securities returns. Journal of Financial and Quantitative Analysis 12, 415—432.