Printer Friendly

Why do US firms choose global equity offerings?

This study examines the economic motivation for global seasoned equity offerings made by US firms. We find that firms announcing global offerings have significantly less negative market reactions than had they limited the issues to domestic only. The extent of the reduced price drop at issue announcement is found to be negatively associated with pre-announcement price run-up, firm size, and market-to-book equity, but positively associated with unsystematic risk. We also find that global issuing firms outperform their domestic counterparts for up to three years following the offerings.

**********

Chuck C.Y. Kwok (*)

As an increasing number of foreign companies have raised equity capital in the US capital market, many US companies are seeking global dissemination of their new equity through a procedure that commonly sets aside a proportion of the total shares for targeted foreign investors (global tranche) with simultaneous issue of the remaining shares in standard registered form in the US domestic market (domestic tranche). Although the two tranches are marketed separately by two syndicates of underwriters, the global tranche is usually led by the overseas affiliates of the lead US underwriters of the domestic tranche to ensure tight control over the allocation of shares. The offer price is the same to all investors, regardless of nationality, because the Rules of Fair Practice of the National Association of Securities Dealers (NASD) do not allow price discrimination.

Since there is generally no obstacle to foreign investors purchasing shares directly in the US market, one might wonder about the motivations for and potential benefits of global issues. We develop three hypotheses regarding the motivations of global offerings.

The name recognition and accessibility hypothesis suggests that global issues enhance investor recognition and participation in both the primary and the secondary markets. The alleviation of asymmetric information hypothesis posits that the ability to issue global shares may serve to validate firm quality, reducing the information asymmetry between insiders and investors. The window of opportunity hypothesis postulates that firms may switch to global offerings when domestic demand for their new shares is weak.

We examine the economic effect of global issues by comparing a sample of global seasoned equity offerings made by US industrial companies from 1985 through 1995 and a domestic control sample that includes all purely domestic offerings during the same period.

Our study is closely related to Chaplinsky and Ramchand (2000). (1) They compare the announcement effect of 349 global offerings and a sample of 459 purely domestic offerings made by US companies from 1986 through 1995. They find that the negative price reaction that typically accompanies equity issuance is reduced by around 0.8 percentage point when some shares are sold abroad through a global tranche.

Our study is different from Chaplinsky and Ramchand (2000) in several ways. First, we include all domestic seasoned equity offerings in the control sample, thus considerably increasing the number of domestic offerings for comparison. To enhance the robustness of our results, we also match each global issue by a purely domestic issue based on firm size.

Second, we use Heckman's (1979) two-stage procedure to control for potential self-selection bias. We use the maximum-likelihood method to estimate the model so that the results are both consistent and efficient. We find that firms announcing global issues can reduce the negative market reaction by about one percentage point, as compared to purely domestic issues.

Third, we explore other possible motivations of global offerings. Chaplinsky and Ramchand (2000) attribute the benefit of global issues to an increased number of foreign shareholders following global issues. Our results indicate that, while firms announcing global issues on average have less negative market reactions, the extent of the reduced price drop at issue announcements is negatively associated with pre-announcement price run-up, firm size, and market-to-book equity, but is positively associated with unsystematic risk.

Lastly, we also examine the long-run post-offering performance for the two types of issuers. We find that global issuing firms outperform their purely domestic counterparts for up to three years after the offerings.

I. Institutional Background and Motivation for Global Equity Offerings

In the following subsections, we present some institutional background about global equity offerings. We also develop three hypotheses regarding the motivations for and the economic effect of global equity offerings.

A. Institutional Background

US companies that participate in global equity offerings typically set aside a fraction of the total shares to be offered to foreign investors (global tranche), and they simultaneously issue the remaining shares in standard registered form in the US domestic market (domestic tranche). Registration with the Securities and Exchange Commission (SEC) is required since these offerings involve the sale of new equity in the domestic market. Inclusion of a global tranche and the number of shares allocated to the tranche are set out in the filing statements.

The two tranches are underwritten by two separate syndicates of underwriters, although in most cases, the global tranche is led by the overseas affiliates of the lead US underwriters of the domestic tranche to ensure tight control over the allocation of shares. The NASD does not allow price discrimination, so the offering price is the same to all investors, regardless of their nationality.

Offering prospectuses are prepared for both the domestic tranche and the global tranche. The international prospectus provides the same information as the domestic prospectus. (2) Global shares are not usually offered to the general public, but instead are largely targeted to selected groups of foreign institutional investors. Global offerings may or may not be accompanied by a cross-listing of shares in foreign exchanges. In the case of cross-listing, overseas financial institutions register themselves as shareholders and then convert the registered stocks into bearer shares in the form of depositary receipts (DRs). These depositary receipts can be traded in local exchanges or over the counter in the same way as American Depositary Receipts (ADRs) are traded in the US market. (3) If shares are not cross-listed, foreign holders of US shares can trade their shares only on the US exchange where the shares are listed.

B. Motivation for Global Equity Offerings

If international markets are integrated and investors hold the world market portfolio, it would make no economic impact whether a firm issues its equity at home or abroad. Under market segmentation, however, global participation might have positive wealth effects. We develop three hypotheses regarding the motivations for choosing global equity offerings, given a global market structure.

1. The Name Recognition and Accessibility Hypothesis

Despite the potential benefit of international diversification and easy access to the US equity market, foreign ownership of US shares is still much lower than one would expect in the absence of barriers to cross-border investment. (4) Merton's (1987) model of capital market equilibrium assuming incomplete information provides both an explanation for this phenomenon and a rationale for global offerings.

Merton (1987) argues that investors are aware of only a subset of available securities and that financial intermediaries would be unable to remedy this inherent restriction because of the costs associated with marketing themselves or their securities. As a result, a firm's specific risk may not be fully diversified away and must be reflected in expected returns. The effect of this incomplete diversification on expected returns is greater the higher the firm's specific or unsystematic risk and the greater the weight of the security in the investors' portfolio.

According to Merton's (1987) model, all else equal, an increase in the size of a firm's investor base lowers investors' expected returns, leading to a lower cost of equity and increasing the market value of the firm's shares. This result implies that managers of firms with few shareholders have an incentive to undertake activities to make potential investors aware of the firms' shares. One way to increase visibility is to cross-list shares on a foreign exchange. According to a complete survey of the cross-listing literature by Karolyi (1998), a positive effect is often documented for foreign firms that cross-list their shares in the US markets, while the effect is insignificant for US firms cross-listing their shares abroad. A global equity offering to targeted foreign investors represents a direct effort to enhance visibility. Pre-marketing roadshows allow for direct and effective communication with potential foreign investors, reducing what it costs foreign investors to learn about US shares.

In the US, new shares are typically offered to institutional investors in the primary markets. Benveniste and Spindt (1989) argue that when investment banks sell new shares repeatedly to the same investors, they collect information about new share demand from these regular investors. In a global issue, shares allocated to the global tranche are underwritten by a group of foreign banks while the domestic shares are marketed by a separate syndicate of underwriters. More investment banks are, thus, employed in a global offering, and, since each bank presumably has its own regular client base, more investors have access to an issue.

The benefit of increased investor accessibility is recognized in studies such as Mauer and Senbet (1992). Maner and Senbet (1992) provide a theoretical model to demonstrate how limited investor access to the primary markets (due to distribution restrictions) leads to underpricing of new equity offerings.

2. The Alleviation of Asymmetric Information Hypothesis

Many theoretical explanations of the market reaction to equity announcements focus on the negative information conveyed by the announcements (Myers and Majluf, 1984 and Miller and Rock, 1985). A global offering can potentially mitigate the problem that arises from information asymmetries between managers and investors. It is generally believed that only high-quality issues can be offered in the global markets, largely because international underwriters are perceived as reluctant to underwrite low-quality issues. Thus, the willingness of investment banks to underwrite a global offering, together with its acceptance by foreign investors, can serve as a certification of firm value. Global offerings also draw greater financial analyst coverage and more public attention and monitoring, reducing potential information asymmetries.

Under this hypothesis, the benefit of global participation in new issues is positively related to the degree of asymmetric information.

3. The Window of Opportunity Hypothesis

Since Myers and Majluf (1984), it has been accepted that issuing equity is more expensive when information is asymmetric between firm insiders and outside investors. Bayless and Chaplinsky (1996) document that sometimes announcement-period abnormal returns are less negative, on average, than in other periods. They suggest that there are windows of opportunity when otherwise identical firms receive more favorable prices for new seasoned equity. Loughran and Ritter (1995) also suggest that managers are adept market-timers, issuing new securities at the most opportune time. This market-timing interpretation implies that managers are able to determine when the market may be willing to overpay for their stock and that they take advantage of these opportunities to issue new equity.

When domestic market conditions seem unfavorable to new equity issues or domestic demand for its shares is lower, firms may switch from the domestic market to the global equity market rather than postpone or cancel an issue. If so, US firms are more likely to participate in global offerings when domestic market conditions do not support the issuance of new shares. (5)

II. Data and Summary Statistics

The following subsections describe our data and present summary information of firm and offer characteristics of global issuers versus purely domestic issuers.

A. Data

We identify all seasoned equity offerings made by US firms from 1985 through 1995 in data obtained from the Securities Data Corporation (SDC). We exclude financial institutions, utility companies, and real estate investment trusts (REITs). Rights offerings and unit offerings are also excluded. We further restrict our sample to firms whose returns are available from the Center for Research in Security Prices (CRSP) for at least 180 business days before the announcement date and to firms whose financial data are available on either COMPUSTAT or Disclosure.

This selection procedure yields a total of 354 global seasoned equity offerings and 1,715 purely domestic offerings. Sample distributions and aggregate proceeds by calendar year, industry group, and stock exchange of listing are reported in Table I.

The number of global offerings increased in the 1990s. Global issues do not seem to be concentrated in any particular industry. The majority of offerings (243 issues) are listed on the New York Stock Exchange; 19 are on the American Stock Exchange, and 92 are on the NASDAQ.

B. Characteristics of Global Issuers vs. Domestic Issuers

Major characteristics of global issuers and their domestic counterparts are reported in Table II. Global issuing firms, on average, allocate 20.78% of total shares to be offered exclusively to overseas investors. Global tranche offering percentages range from 6.99% to 66.67%, with a median of 20%.

Global issues are typically large in issue size and made by large firms. The mean offer size is $204.9 million for global offerings, compared to $45.2 million for purely domestic issues. Firms participating in global issues, on average, have market capitalization--measured as the total number of share outstanding times the market price one day prior to the announcement date--of $2267.1 million, more than five times the size of domestic issuers ($425.4 million). Global issuers ($4572.4 million) have more than seven times the mean total assets of domestic issuers ($592.3 million).

Although greater in offer size, global offerings are lower in terms of relative offer size, defined as the total number of new shares to be offered as a percentage of total shares outstanding. The relative offer size, on average, is 19.73% for the global sample, compared to 23.19% for the domestic control sample. Global offerings are not limited to large firms or large issues, however. Gross proceeds from global issues start at $9 million, with a median of $124.7 million.

The market-to-book equity ratio, calculated as the market capitalization divided by the book value of common equity for the most recent fiscal year ending before the offering, is 4.19 for global issuers, which is significantly lower than the 4.91 for purely domestic issuers. The standard deviation of the market model residuals is significantly lower for global offerings than for purely domestic ones.

Table II also provides information about market conditions at the time issues are announced. Pre-announcement issuers' own price run-up is measured as the cumulative abnormal returns from the market model from -60 days to -2 days prior to the announcement day (day 0). The average 60-day price run-up is 2.50% for global offerings and 0.12% for purely domestic offerings.

The US stock market run-up is defined as the cumulative abnormal market returns from -250 to -2, relative to the average market return (from -390 to -251) prior to the announcement date, where market return is the CRSP equally weighted return. We use Morgan Stanley Capital International's EAFE index (Europe, Australia, and Far East) to proxy for the non-US world market return. World market run-up is thus defined as the cumulative abnormal return from -250 to -2 prior to the announcement date relative to the average world market return from -390 to -251.

The exchange rate is the trade-weighted exchange rate of the US with respect to ten major currencies. The weight of each currency is computed from the International Monetary Fund's multilateral exchange rate model, which is based on 1972-1976 trade flows between the US and ten industrialized countries. (6) To be consistent with the two equity market run-ups, we measure the strength of the dollar as the cumulative abnormal daily change in the weighted exchange rate from -250 to -2 prior to the announcement date, compared to the average daily change from -390 to -251.

The results show that domestic market run-up, on average, is negative for both global and domestic issues, but more negative for domestic offerings than for global offerings. The US dollar is significantly stronger than other major currencies when global offerings are made.

Contrary to our expectations, global issues do not take longer to register with the SEC. The average length of the registration period is approximately 37 days for both samples.

III. Empirical Results of Announcement Effects

This section presents our empirical results of announcement effects. We use the traditional market model to compute abnormal returns around issue announcements. To ensure robust results, we match the two samples on a one-to-one basis by firm size. Potential self-selection bias is corrected using the Heckman's two stage procedure that is estimated with the maximum likelihood method. Further tests are conducted to relate the three hypotheses developed in the previous section to the empirical results.

A. Abnormal Returns Surrounding Equity Announcement Dates

To measure the short-term impact of global issues on firm value, we use the traditional market model to compute abnormal returns around new equity announcements. The announcement date (t = 0) is defined as the date the offer is filed with the Securities and Exchange Commission (SEC). The market model is estimated wiith 120 daily returns from day -180 to day -61 prior to the announcement date. The equally weighted CRSP return is taken as the market return.

The results are presented in Table III. Consistent with the existing literature for seasoned equity offerings (e.g., Eckbo and Masulis, 1995 and Masulis and Korwar, 1986), both types of issuers experience significantly negative market reactions. Global issuing firms on average, however, have less negative market reactions than purely domestic issues. The mean announcement-period cumulative abnormal return, or CAR (-1, +1), is -2.37% (median of -2.34%) for the global issues, compared to -2.94% (median of -2.76%) for the domestic sample. The difference is significant at the 10% level by a simple difference-in-the-means test.

B. Robustness Check Using Matching Firm Approach

To ensure robust results, we match the two samples on a one-to-one basis by firm size. For each global equity issue announced in year t, we identify all purely domestic issues announced in the same year. Limiting issues to a single year allows for more homogeneity in issuance conditions. From this set, we select the firm that has a market capitalization closest to that of the global issue to be matched. We further require that the matching domestic issue has a market value that is within 25% of that of the global issue. (7) There are 274 global issues matched in this way. The average market capitalization of these global SEOs is $1124 million, compared to $1130 million for the domestic control sample.

The univariate result that global offerings have less negative market reactions is robust for the two samples matched by firm size. The mean CAR(-1, +1) is -2.32% for the global sample and -3.01% for the size-matched domestic control sample. The difference is statistically significant at the 10% level.

C. Statistical Model and Results

If the choice of global offerings is a random decision, then we can use ordinary least squares (OLS) to estimate the cross-sectional regression of abnormal returns on a set of exogenous variables and a global indicator variable that equals one for global offerings and zero otherwise. The coefficient of the indicator variable represents the cross-sectional difference between the two samples and would be a consistent measure of the announcement effect of global offerings.

We cannot treat the indicator variable as exogenous if the decision of an individual firm to make a global offering is based on individual self-selection, and there are reasons to suspect the presence of a self-selection bias. Surely firms that expect to benefit from global issues will engage in them. In the presence of self-selection, the OLS estimates on the indicator variable would be inconsistent. Consequently, evaluations based on OLS estimates may either overestimate or underestimate the effectiveness of global equity offerings, depending on the nature of self-selection.

Suppose high-quality firms are more likely to make global offerings. These issuers would fare better than other firms, whether they issue new shares at home or overseas. If this is the case, the OLS result would overestimate the benefit from global offerings. If, on the other hand, firms that would experience poorer market reactions had they chosen purely domestic offerings are more likely to participate in global offerings, then the OLS results would underestimate the effect of global participation.

The conventional way to deal with a self-selection bias is to treat the indicator variable as endogenous and use a probit equation to model a firm's selection to switch from one type to another. We use the following model to evaluate the announcement effect of global offerings:

CA[R.sub.i] = [alpha] + [delta][I.sub.i] + [[beta].sub.1] [W.sub.i] + [[mu].sub.i] (1)

[I.sup.*.sub.i] = [gamma][Z.sub.i] + [[epsilon].sub.i] (2)

[I.sub.i] = 1 if [I.sub.i.sup.*] > 0

= 0 if [I.sub.i.sup.*] [less than or equal to] 0

where CA[R.sub.i] is the announcement-period abnormal return, and [W.sub.i] represents a vector of exogenous variables that are related to CA[R.sub.i]. In the probit model, [I.sub.i.sup.*] represents a firm's unobserved sentiment of choosing global offerings; [I.sub.i] is the observed actual choice, which equals one for global offerings and zero for purely domestic offerings. [Z.sub.i] is a vector of variables that influence a firm's choice of offering procedure. We use maximum-likelihood method (MLE) to jointly estimate the two equations. Under this method, the coefficient estimate of the global indicator variable is both consistent and efficient. (8)

The independent variables include market capitalization, issue size, total assets, market-to-book equity (ME/BE), standard deviation of the market model residuals (STDR), and gross spread (GS). As firm size and issue size are highly correlated (with a simple correlation coefficient of 0.86), we scale issue size by firm size and refer to it as relative offer size (RELSIZE). For the same reason, total assets are also scaled by firm size and are referred to as leverage ratio (LEVERAGE).

ME/BE is often used as a proxy for growth opportunities of the issuer. STDR measures the unsystematic risk that cannot be explained by the market model. Gross spread represents total compensation paid to the underwriting syndicate, expressed as a percentage of the gross proceeds. Lee, Lochhead, Ritter, and Zhao (1996) and others suggest that underwriter spread is largely related to offer size, reflecting economies of scale in the issuance of new equity. Investment banks also take into consideration factors beyond size, such as the riskiness of the issue, so we take gross spread as a proxy for the riskiness of an issue.

We include three market-related variables to control for issue conditions: 1) issuer's own stock price run-up (RUNUP); 2) US stock market performance relative to the rest of the world (RELMKT); and 3) strength of the US dollar relative to other major currencies (DOLLAR). It is likely that foreign demand for US shares depends on the relative performance of the US stock market. Brennan and Cao (1997) demonstrate that when domestic investors have a cumulative information advantage over foreign investors about their domestic market, they tend to purchase foreign assets when the return on foreign assets is high. The strength of the US dollar may also affect foreign demand for US shares, which are priced in dollars.

The results are presented in Table IV. In all regressions, the industry dummies (the first two digits of the SIC code) are included to control for potential industry effects, but these specific results are not reported for the sake of space. We discuss the probit model results first. The probit model reveals what kinds of firms are more likely to participate in global offerings and under what circumstances. The results, reported in the first column of Table IV, are generally consistent with those in the summary statistics analyses. That is, all else equal, larger firms and firms issuing more shares as a proportion of current shares outstanding are more likely to choose global offerings. Moreover, firms are more likely to choose global offerings when the US dollar is stronger than the other major currencies.

In the first three regressions for the entire sample, the coefficient for the global indicator variable is significantly positive in each specification, ranging from 0.83% to 1.03%. The result suggests that firms announcing global offerings have significantly less negative market reactions than their purely domestic counterparts.

In regression (4), as a direct test of the name recognition and accessibility hypothesis, we add both the global dummy variable and the relative size of the global tranche (%_GLOBAL). The coefficient of %_GLOBAL is, although positive, statistically insignificant, suggesting that the relative size of the global tranche has no significant influence on the market reaction to announcements of global equity offerings. (9)

To check the robustness of the results, we also present the regression results for the size-matched samples in Table IV. The result is robust when the two samples are matched by firm size. The coefficient estimate for the global dummy variable is 0.9% when all explanatory variables are included (p-value < 0.05).

In summary, all else equal, firms announcing global issues have significantly less of a price drop than had they limited an issue to domestic only.

D. Further Tests and Discussions

To distinguish the motivations for global issues, we add variables that allow the global indicator variable to interact with each of the independent variables in Equation (1):

CA[R.sub.i] = [alpha] + [delta][I.sub.i] + [[beta].sub.1] [W.sub.i] + [[beta].sub.2]([I.sub.i]*[W.sub.i]) + [[eta].sub.i] (3)

This specification allows us to test for coefficient differences, as represented by [[beta].sub.2], for each independent variable between the two samples. Equations (3) and (2) are jointly estimated using the maximum-likelihood method to control for self-selection bias. The net effect of global participation for firm i with attribute [W.sub.i] is measured by ([delta] + [[beta].sub.2][W.sub.i]). A positive [[beta].sub.2], for example, implies that the benefit from global issue increases with [W.sub.i], and vice versa.

When the whole sample is used, the results (not reported) indicate that only three interaction variables have statistically significant coefficient estimates: I*log (ME), I*STDR, and I*RUNUP. The coefficients of I*log (ME) and I*RUNUP are negative, while the coefficient of I*STDR is positive. For the size-matched samples, the interaction variables with significant coefficients are I*STDR (positive), I*ME/BE (negative), and I*RUNUP (negative). These results suggest that, the lower price drop at issue announcements is negatively associated with firm size, growth opportunities, and price momentum, but positively associated with unsystematic risk.

To interpret these results, it is reasonable to associate the alleviation of the asymmetric information problem with variables that proxy for firm quality such as firm size and ME/BE. The idea is that the willingness of investment banks to underwrite an issue together with its acceptance by foreign investors can attest to firm quality, reducing the degree of information asymmetry.

This hypothesis is intuitively appealing, as most theoretical explanations of the negative market reaction at issue announcement rely on the negative information conveyed to the market by the announcements (e.g., Myers and Majluf, 1984 and Miller and Rock, 1985). In our case, firms with more asymmetric information are expected to benefit more from global offerings than other firms, all else equal. Thus, the negative coefficients for I*log(ME) and I*ME/BE are consistent with alleviation of asymmetric information.

In Merton's (1987) model of incomplete information, a firm's unsystematic risk is not fully diversified away, because investors tend to invest only in securities of which they are aware. As a result, unsystematic risk is reflected in expected returns. The effect of this incomplete diversification on expected returns is greater the higher the firm's specific risk and the greater the weight of the security in the investors' portfolio. Under the recognition hypothesis, global issues improve name recognition and attract more investors, producing a large investor base. Thus, the issuers' unsystematic risk can be diversified away through global participation, which reduces expected returns. The announcement effect from global offerings is expected to be positively related to a firm's specific or unsystematic risk. Our empirical result of a positive coefficient for I*STDR is consistent with the recognition hypothesis.

The window of opportunity hypothesis implies that US firms are more likely to participate in global offerings when domestic demand for their new shares is low in the domestic market.

The negative coefficient for I*RUNUP suggests that the lower the price run-up prior to announcement, the less of a price drop global issuing firms experience. Since low price run-up implies less likelihood of overvaluation, this result can be interpreted as consistent with the window of opportunity hypothesis of global offerings.

IV. Comparison of Post-Offering Performance

The empirical evidence presented so far indicates less negative market reactions to announcements of global issues. Recent studies by Loughran and Ritter (1995), Spiess and Affleck-Graves (1995), and Jegadeesh (2000) indicate that, on average, firms completing seasoned equity issues underperform the market in the long run. However, Foerster and Karolyi (2000) document that non-US firms that raise equity capital in the US markets experience better post-issue long-run performance. To examine long-run stock price performance after offerings, we calculate the holding-period (buy-and-hold) return from purchasing shares at the closing price on the day of the offering to the end of the appropriate holding period.

The buy-and-hold return is defined as:

Buy-and-Hold Return = [[[PI].sup.[T.sup.2].sub.[T.sub.1]](1 + [r.sub.it])] (4)

where [T.sub.1] is the date of the first post-issue closing price; [T.sub.2] is the ending date of the holding period; and [r.sub.it] is the return for firm i on day t. We use 1, 2, 3, 6, 12, 24, and 36 months as holding intervals, with 21 trading days in each month.

For each holding period, we also compute the market-adjusted buy-and-hold return as:

Market-Adjusted Buy-and-Hold Return = [[[PI].sup.[T.sup.2].sub.[T.sub.1]](1 + [r.sub.it])]- [[[PI].sup.[T.sup.2].sub.[T.sub.1]](1 + [r.sub.mt])] (5)

where [r.sub.mt] the CRSP value-weighted market return.

Post-offering holding-period returns are presented in Table V. Both global and domestic seasoned equity offerings outperform the market in the short run (up to six months following the offerings), but trail the market in the long run. The mean buy-and-hold return for global issuers is consistently higher than for strictly domestic issuers in each holding period.

In general, the gap in performance widens over time, ranging from 0.51 percentage point (in one month) to 16.3 percentage points (up to three years). According to simple t-tests, global offering firms significantly outperform purely domestic issuers in three years. When adjusted by the CRSP value-weighted market return, global issuers still outperform domestic issuers up to three years after offerings. The performance difference is statistically significant for the 6-, 12-, and 36-month holding periods.

Let us note a number of methodological concerns with the measurement of long-run abnormal returns. Barber and Lyon (1997), Kothari and Warner (1997), and Lyon, Barber, and Tsai (1999) suggest that significant biases can arise in benchmarking solely with market indices. To ensure robust results, we compute post-offering performance using the matching-firm approach. When the two samples are matched one-to-one by firm size, global issuers outperform purely domestic issuers by 11.46 percentage points (11.95 percentage points after adjusting for market returns) up to two years, both significant at the 10% level. The performance difference in the three-year buy-and-hold returns is 10.66 percentage points (11.53 percentage points after adjusting for market returns). The difference is not significant statistically.

To account for the possibility that the observed performance difference is attributable to differences in firm characteristics between the two samples, we regress the market-adjusted return on the global dummy variable, announcement-period abnormal return, the natural log of firm size, and the market-to-book equity ratio. The regression results are presented in Table VI.

The coefficient estimate for the global dummy variable, which measures the cross-sectional difference in performance between the two types of offerings, is positive and significant at conventional levels across all regressions. Global issuing firms outperform their domestic counterparts by 9.7 percentage points up to two years and 13.6 percentage points up to three years after offerings. When the two samples are matched by firm size, the performance differences are, respectively, 12.9 percentage points and 13.1 percentage points for the two-and three-year holding periods.

The significant difference in performance between global and domestic issues is puzzling for two reasons. First, it is not quite clear theoretically how global equity offerings and foreign equity ownership affect a firm's long-run post-offering performance. Second, under market efficiency, any positive impact from a global offering should be recognized and fully absorbed by the market at issue announcement.

Inspection of Table VI reveals that announcement-period abnormal return has some weak predictive power for long-run performance. Its coefficient is positive, but significant only for the three-year holding period and for the firm size-matched sample.

V. Summary and Discussions

This study examines the economic impact of global seasoned equity offerings. We find that firms announcing global equity offerings have significantly less negative market reactions by about one percentage point than what would have been expected had they limited their issues to the domestic market. The extent of the lower price drop is found to be negatively associated with firm size, ME/BE, and price run-up, and positively associated with unsystematic risk.

We interpret these results as consistent with the hypothesis that global participation can potentially mitigate the asymmetric information problem between insiders and investors. The willingness of investment banks to underwrite global issues and acceptance by foreign investors can be a form of certification of firm quality.

The results support the name recognition hypothesis that global issues enhance investor recognition and participation. A larger investor base reduces or eliminates the burden of compensation for a firm's unsystematic risk. The results are consistent with the window of opportunity hypothesis, which posits that firms switch to global markets when the domestic demand for their new issues is weak. We also find that global issuers outperform purely domestic issuers for up to three years after offerings, although both trail the market in the long run.

Given the empirical result of a positive wealth effect, why shouldn't all equity issues be global? The answer presumably lies in an optimization of the trade-off between the desire for greater investor participation and security distribution challenges. Preparing prospectuses for targeted investors, overseas roadshows by senior executives, and the legal and administrative complexities of issuing securities in the international markets, make global equity offerings laborious and time-consuming.

Added to these direct costs, there is a potential negative effect on the domestic tranche if the international tranche does not work as expected. Usually, the international tranche precedes the domestic tranche. US companies usually conduct roadshows abroad first before returning home to present to domestic investors. Negative sentiment from the Euromarket might jeopardize a transaction that would otherwise go smoothly on a purely domestic basis. Some firms have even had to cancel an international tranche owing to low interest in the overseas market.

Global equity offerings also represent a long-term commitment. Companies need to communicate regularly with their foreign shareholders; they cannot take investors' money and then forget them. Otherwise, shares will flow back in to the domestic market, frustrating the firm's efforts to broaden its shareholder base. Sale of a large amount of overseas shares would also exert downward pressure on the issuer's stock price in the domestic market.

Another possible explanation for relatively few global issues by US companies is a lack of incentives for investment banks. In our sample of global issues, lead managers are predominantly US investment banks. Underwriting a global offering is more time-consuming and requires global coordination, yet the underwriting fees are no higher than fees generated from purely domestic issues. Underwriters, thus, have few incentives to promote global offerings. Global competition among investment banks might, in the future, open up more new equity issues to global investors.
Table 1

Sample Distribution by Year, Industry, and Exchange

This table reports sample distributions and aggregate proceeds by
calendar year, industry group, and stock exchange of listing. The sample
includes 354 global SEOs and 1,715 purely domestic SEOs made by US
industrial firms between 1985 and 1995 obtained from the Securities Data
Company's New Issue database. Aggregate proceeds (in $ million) are the
gross proceeds from the issues. The two-digit SIC code is the first
two-digit industry code as classified by Standard & Poor's.

Panel A. Sample Distribution and Aggregate Proceeds by Calendar Year

 Global SEOs Domestic SEOs
 Aggregate Aggregate
Year Number Proceeds Number Proceeds

1985 2 280.0 221 9215.7
1986 7 1815.8 162 6123.6
1987 14 1358.0 118 5770.2
1988 12 1886.4 48 1555.2
1989 12 818.4 97 2968.2
1990 18 2476.8 61 2580.3
1991 51 9516.6 204 8976.0
1992 56 12504.8 162 5848.2
1993 83 17778.6 237 10309.5
1994 41 9384.9 153 8246.7
1995 58 14703.0 252 15850.8

Total 354 72523.3 1715 77444.4
Panel B. Sample Distribution and Aggregate Proceeds by Industry

 Global SEOs Domestic SEOs

Two-Digit Aggregate Aggregate
SIC Industry Group Number Proceeds Number Proceeds

10-14 Mining 30 6645.0 119 7699.3
20-39 Manufacturing 189 37762.2 795 34582.5
40-48 Transportation and 37 10829.9 124 7254.0
 Communications
50-51 Wholesale 7 1029.7 82 2419.0
52-59 Retail Trade 41 8249.2 191 7773.7
70-89 Services 40 7044.0 325 14657.5
Others Miscellaneous 12 1369.2 79 3096.8
Panel C. Sample Distribution and Aggregate Proceeds by Exchange

 Global SEOs Domestic SEOs
Stock Aggregate Aggregate
Exchange Number Proceeds Number Proceeds

NYSE 243 60774.3 417 33693.6
AMEX 19 1915.2 207 5961.6
NASDAQ 92 9834.8 1091 37857.7


Table II. Descriptive Statistics

This table presents summary information of firm and offer characteristics. %_GLOBAL is the number of shares allocated to the global tranche as a percentage of total shares to be offered. SIZE is the gross proceeds from the issue. ME is market capitalization one day prior to issue announcement. RELSIZE is total number of new shares to be offered divided by number of pre-offer shares outstanding. LEVERAGE is defined as market equity/total assets, where total assets are taken from the most recent fiscal year ending before the offering. ME/BE is market-to-book equity ratio, measured as market value of equity divided by book value of equity for most recent fiscal year ending before the offering. STDR is standard deviation of market model residuals. GS represents underwriter spread, or compensation paid to underwriters as a percentage of total gross proceeds. PRUNUP is issuer's own price run-up prior to announcement, defined as cumulative abnormal return from -60 to -2. DMRUNUP is US stock market run-up, defined as cumulative abnormal market returns from -250 to -2 relative to average market return (from -290 to -251) prior to the announcement date, where market return is CRSP equally weighted market return. WMRUNUP is non-US world market run-up, defined as cumulative abnormal market returns from -250 to -2 relative to the average market return (from -290 to -251) prior to announcement date, where market return is return on the MSCI EAFE index. RELMKT is difference between DMRUNUP and WMRUNUP. DOLLAR measures strength of the US dollar, defined as cumulative abnormal daily change in exchange rate from -250 to -2 relative to average daily change from -290 to -251 prior to announcement date, where exchange rate is trade-weighted exchange rate of the US dollar with respect to 10 major currencies. DAY is the number of days in registration with the Securities and Exchange Commission. Difference in the means is based on t-test.
 Global SEOs Domestic SEOs
 [N = 354] [N= 1715]

Variable Mean Med. Std Dev Mean Med. Std Dev

%_GLOBAL 20.78% 20.0% 5.42% - - -
SIZE ($mil.) 204.9 124.7 256.3 45.2 29.3 58.15
ME($mil.) 2267.1 897.8 5899.2 425.4 149.3 1314.3
ASSET($mil.) 4572.4 747.8 19379.1 592.3 81.4 5804.8
RELSIZE 19.73% 14.17% 21.57% 23.19% 19.11% 22.24%
LEVERAGE 1.588 0.992 2.034 1.03 0.653 1.34
ME/BE 4.19 2.932 4.41 4.912 3.338 6.53
STDR 0.0246 0.0224 0.01 0.0318 0.0293 0.0145
GS 4.04% 4.0% 0.96% 5.482% 5.5% 1.49%
PRUNUP 2.501% 3.671% 24.22% 0.122% 0.69% 28.86%
DMRUNUP -2.266% -2.19% 40.6% -4.615% -2.84% 38.94%
WMRUNUP 6.043% 3.52% 25.35% 8.06% 9.32% 26.05%
RELMKT -8.37% -8.18% 51.12% -12.66% -16.77% 49.64%
DOLLAR 3.688% 7.78% 16.73% -0.537% 4.0% 19.19%
DAY 36.5 30 31.8 37.03 28 55.63



 Difference in
Variable Means

%_GLOBAL -
SIZE ($mil.) 159.7 ***
ME($mil.) 1841.7 ***
ASSET($mil.) 3980.1 ***
RELSIZE -3.46%
LEVERAGE 0.553 ***
ME/BE -0.72 **
STDR -0.0072 ***
GS 1.422% ***
PRUNUP 2.379%
DMRUNUP 2.34%
WMRUNUP -2.017
RELMKT 8.40% *
DOLLAR 4.225 *
DAY -0.53

*** Significant at the 0.01 level.

** Significant at the 0.05 level.

* Significant at the 0.10 level.
Table III

Mean Abnormal Returns Surrounding Announcement Dates

This table reports the mean abnormal returns surrounding announcement
dates. CAR(-1, 0) and CAR(-1, +1) are average two- and three-day
announcement-period cumulative abnormal return estimated by using the
market model. ME is the market capitalization one day prior to issue
announcement. N is number of issues in sample. Medians are in
parentheses. Difference in means is based on the simple t-test.

 Whole Sample Firm Size-Matched
 Global Domestic Global Domestic
 SEOs SEOs Difference SEOs SEOs
Variable [N = 354] [N = 1715] in Means [N = 274] [N = 274]

CAR(-1, 0) -1.678% -1.90% 0.22% -1.671% -2.369%
 (-1.775%) (-1.61%) (-1.742%) (-2.332%)
CAR(-1, +1) -2.371% -2.943% 0.57% * -2.322% -3.011%
 (-2.339%) (-2.758%) (-2.323%) (-2.998%)
ME($mil.) 2267.1 425.4 1841.7 *** 1124.1 1130.1
 (897.8) (149.3) (633.3) (637.1)

 Firm
 Size-Matched

 Difference
Variable in Means

CAR(-1, 0) 0.70% *

CAR(-1, +1) 0.69% *

ME($mil.) 6.0


*** Significant at the 0.01 level.

* Significant at the 0.10 level.
Table IV

Regression Results of Abnormal Returns

This table reports the results of the maximum likelihood method (MLE)
estimates of Equations (1) and (2). Standard errors are calculated using
the asymptotic variance matrix developed in Greene (1981) and Heckman
(1979) to correct for heteroskedasticity. Asymptotic t-statistics are in
brackets. Independent variables are defined in Table II. Probit model
results presented in the first column are based on the whole sample. In
all regressions, the industry dummies are included, but thier results
are not reported for brevity. Ordinary least squares estimate of the
global indicator variable, I, is presented in the last row.

Independent Probit Whole Sample
Variables Model (1) (2) (3)

Constant -4.677 -0.0376 -0.0444 -0.0062
 [-7.628] *** [-9.064] *** [-6.617] *** [-0.305]

I 0.0093 0.0083 0.0103
 [3.014] *** [2.262] ** [1.831] *

%_GLOBAL


Log(ME) 0.67 0.0013 -0.0024
 [10.75] *** [1.358] [-1.197]

RELSIZE 1.284 -0.012 -0.011
 [5.995] *** [-2.506] ** [-1.504]

LEVERAGE 0.0315 0.0011 0.0003
 [0.978] [1.438] [0.338]

GS -0.375 -0.287
 [-0.605] [-1.758] *

STDR -8.702 -0.236
 [-1.901] * [-2.407] **

ME/BE -0.014 0.0003
 [-1.325] [1.677]

RUNUP 0.225 -0.0019
 [1.398] [-0.466]

RELMKT 0.0856 -0.0042
 [0.997] [-1.756] *

DOLLAR 1.108 -0.0166
 [4.415] *** [-2.395] **

OLS Estimate 0.0062 0.0038 0.0059
of 1 [2.310] ** [1.123] [1.642] *

Independent Whole Sample Firm Size-Matched
Variables (4) (5) (6)

Constant -0.0069 -0.037 -0.03
 [-0.34] [-3.635] *** [-0.875]

I 0.0044 0.0082 0.009
 [0.275] [2.115] ** [2.085] **

%_GLOBAL 0.0256
 [0.337]

Log(ME) -0.0023 0.0036
 [-1.162] [1.069]

RELSIZE -0.011 -0.0048
 [-1.508] [-0.255]

LEVERAGE 0.0003 -0.009
 [0.347] [-0.688]

GS -0.285 -0.67
 [-1.742] * [-2.239] **

STDR -0.0235 0.272
 [-2.392] ** [1.171]

ME/BE 0.0003 0.003
 [1.801] [0.604]

RUNUP -0.0019 -0.012
 [-0.458] [-2.028] **

RELMKT -0.0043 0.0019
 [-1.779] * [0.461]

DOLLAR -0.017 0.0081
 [-2.399] *** [0.564]

OLS Estimate 0.0079 0.009
of 1 [2.233] ** [2.278] **

*** Significant at the 0.01 level.

** Significant at the 0.05 level.

* Significant at the 0.10 level.
Table V

Post-Offering Buy-and-Hold Returns

This table presents post-offering buy-and-hold returns from purchasing
shares at the closing price on the offer day to the end of a holding
period up to three years. The buy-and-hold return is calculated as:

[[[PI].sup.[T.sub.2].sub.[T.sub.1]] (1 + [r.sub.it)]

The market adjusted return reported, in parentheses, is calculated as:

[[[PI].sup.[T.sub.2].sub.[T.sub.1]] (1 + [r.sub.it])] - [[[PI].sup.
[T.sub.2].sub.[T.sub.1]] (1 + [r.sub.mt])]

where market return, [r.sub.m], is the value-weighted CRSP return.
Difference in means is based on the simple t-test. Return figures are
expressed in percentage.

Holding Whole Sample Firm Size-Matched
Period Global Domestic Difference Global Domestic Difference

1-month 4.04 3.53 0.51 3.91 2.87 1.04
 (3.29) (2.46) (0.84) (3.25) (2.15) (1.10)

2-month 5.62 4.73 0.89 5.51 3.72 1.79
 (3.91) (2.62) (1.30) (4.10) (2.06) (2.04)

3-month 6.66 6.05 0.61 6.60 4.80 1.80
 (3.63) (2.64) (1.0) (3.87) (2.12) (1.75)

6-month 9.38 7.29 2.09 9.08 3.42 5.66 **
 (3.67) (6.72) (3.00 *) (3.56) (-2.26) (5.82 **)

12-month 13.17 12.13 1.04 12.07 9.62 2.45
 (0.32) (-3.24) (3.50) (-0.79) (-3.90) (3.11)

24-month 26.56 19.1 7.46 23.74 12.28 11.46 *
 (-6.45) (-15.65) (8.97 **) (-10.03) (-21.98) (11.95 *)

36-month 40.4 24.10 16.3 ** 32.55 21.89 10.66
 (-16.49) (-30.05) (13.16 **) (-24.87) (-36.4) (11.53)

** Significant at the 0.05 level.

* Significant at the 0.10 level.
Table VI

Cross-Sectional Regressions of Post-Offering Stock Performance

This table presents the cross-sectional regression results of
post-offering stock performance. The dependent variable is the
market-adjusted buy-and-hold return in each holding period. Independent
variables are defined in Table II. In all regressions, the industry
dummies are included, but not reported for brevity. Ordinary least
squares are used in the regressions. t-statistics (in brackets) are
computed using White's (1980) heteroskedasticity-corrected standard
errors.

 Whole Sample Firm Size-Matched
Independent
Variable 24-month 36-month 24-month 36-month

Constant -0.162 -0.242 -0.222 -0.38
 [-1.608] [-1.758] * [-1.042] [-1.27]
I 0.097 0.136 0.129 0.131
 [1.716] * [1.798] * [2.003] ** [1.685] *
CAR(-1, +1) 0.343 0.876 0.759 2.322
 [0.84] [1.498] [0.904] [1.646] *
Log(ME) 0.0034 0.0016 0.0019 0.0097
 [0.20] [0.069] [0.063] [0.236]
ME/BE -0.0004 -0.0084 0.013 0.0021
 [-0.094] [-1.496] [0.197] [0.218]
[R.sup.2] 0.002 0.006 0.01 0.02
F-Statistic 1.11 2.46 1.37 2.53
p-value 0.35 0.04 0.24 0.04

** Significant at the 0.05 level.

* Significant at the 0.10 level.


(1.) Extensive theoretical and empirical studies exist in the literature of international market segmentation and cross-border listing. Errunza and Losq (1985) and Alexander, Eun, and Janakiramanan (1987) demonstrate how companies from segmented markets that issue overseas can lower their cost of capital and increase the market value of their shares. A positive effect is often documented for foreign firms that cross-list their shares in the US markets. See Karolyi (1998) for a complete survey of the cross-listing literature. A study by Miller (1999) finds that market reactions to equity offerings made by foreign firms in the US through an American Depositary Receipt (ADR) program are positive. This result stands in sharp contrast to findings that document a 3% reduction in stock price around issue announcements (Masulis and Korwar, 1986).

(2.) The domestic and international prospectuses are more or less identical. Main differences are in covers that tend to be a "wraparound" for each jurisdiction to lay out required legal disclaimers. Prospectus details have to be the same for disclosure purposes.

(3.) American Depositary Receipts are negotiable certificates issued by a US bank to represent the underlying shares of non-US stock, which are held in trust at a custodian bank in the home market of the non-US company. ADRs arc registered with the SEC and trade like any other US stock. Depositary receipts traded outside the United States are often referred to as Global Depositary Receipts (GDRs). These can be public or private offerings and are settled through Euroclear and CEDEL.

(4.) There is a line of research related to corporate international diversification. For instance, empirical findings in Agmon and Lessard (1977) and Errunza and Senbet (1981, 1984) indicate that US investors have been able to capture gains from international diversification via investing in US multinational corporations.

(5.) In global offerings, the majority of shares are offered in the domestic market. This implies that if the sole reason for issuing equity is to take advantage of market overvaluation, a global equity offering is only a partial solution, since most capital is still raised in the domestic market where shares are not overvalued. On the other hand, if a company decides to issue equity due to positive NPV projects, the global equity offering will sometimes be the only option as that might be the only way to raise enough capital at a fair value. We thank an anonymous referee for pointing this out.

(6.) One potential problem with the weighted exchange rate is that it may fail to reflect the relative importance of specific currencies used by targeted foreign investors. We use the weighted exchange rate since we do not know the nationalities of targeted foreign investors. The EAFE index is used to proxy for the non-US world stock market return for the same reason.

(7.) Firm size is chosen as the matching criterion, because global issues are typically made by large firms. We also match the two samples by market-to-book equity ratio (ME/BE) as well as firm size. Specifically, we rank all domestic issues in a year by ME/BE. We identify those issues with an ME/BE ratio that is within 25% of that of the global issue to be matched. Then, the firm that has a market value closest to that of the global issue is selected. If the market value of this domestic issue is too low (less than 50% of the global issue), then we decide there is no feasible match. The major results arc not materially changed when the two samples are matched either way.

(8.) The Heckman (1979) two-stage procedure is used to estimate the starting values for the MLE. In the first step, the probit equation is estimated, and the results are used to calculate a self-selection adjustment variable called the inverse Mills ratio as follows:

MR = [phi] ([gamma]Z) / [PHI] ([gamma]Z) (for global offerings)

= -[phi] ([gamma]Z) / [1 - [PHI] ([gamma]Z)] (for domestic offerings)

where [phi](.) and [PHI] (.) are, respectively, the density and distribution function of the standard normal. In the second step, the estimated inverse Mills ratio is added to regression (1), which is estimated using ordinary least squares (OLS):

CA[R.sub.i] = [alpha] + [delta][I.sub.i] + [beta]W + [gamma] M[R.sub.i] + [[eta].sub.i]

where [lambda] is the covariance of [mu] and [epsilon]. Evidence of any selection bias is represented by [lambda]. The standard errors of the coefficient estimates are calculated using the asymptotic covariance matrix developed in Greene (1981) and Heckman (1979) to correct for heteroskedasticity. The Heckman estimator is consistent, although not fully efficient. The MLE estimates are both consistent and fully efficient. The software is LIMDEP.

(9.) When both are present as independent variables, the coefficient for I alone does not represent the cross-sectional difference in announcement-period abnormal returns between the two samples. The correct measure should be: [delta] + [beta](%_GLOBAL), where [delta] and [beta] are, respectively, the coefficients for I and %_GLOBAL. With a potential measurement problem associated with %_GLOBAL, however, the results should be interpreted with caution. According to practitioners we contacted, there needs to be some clarification between "filed" amount versus actual "allocation." Regulators require an indication of the percentage of shares to be sold into any jurisdiction; twenty percent is sometimes a rough approximation that gives the banks a reasonable amount of flexibility and satisfies local regulators. In our sample of 354 global issues, 205 issues set the global tranche at 20% precisely. Moreover, conditional on there being a global issue, a larger global tranche may indicate less domestic interest. In t his case, the size of the global tranche is not expected to be associated with the price drop.

References

Agmon, T. and D. Lessard, 1977, "Investor Recognition of Corporate International Diversification," Journal of Finance 32, 1049-1055.

Alexander, G., C. Eun, and S. Janakiramanan, 1987, "Asset Pricing and Dual Listing on Foreign Capital Markets: A Note," Journal of Finance 42, 151-158.

Barber, B. and J. Lyon, 1997, "Detecting Long Run Abnormal Stock Returns: The Empirical Power and Specification of Test Statistics," Journal of Financial Economics 41, 359-399.

Bayless, M. and S. Chaplinsky, 1996, "Is There a Window of Opportunity for Seasoned Equity Issuance?" Journal of Finance 51, 253-278.

Benveniste, L. and P. Spindt, 1989, "How Investment Bankers Determine the Offer Price and Location of New Issues," Journal of Financial Economics 24, 343-362.

Brennan, M.J. and H.H. Cao, 1997, "International Portfolio Investment Flows," Journal of Finance 52, 1851-1879.

Chaplinsky, S. and L. Ramchand, 2000, "The Rationale for Global Equity Offering," Journal of Finance 55, 2767-2790.

Eckbo, E.B. and R.W. Masulis, 1995, "Seasoned Equity Offerings: A Survey," in R. Jarrow, V. Maksimovic, and W.T. Ziemba, Eds., Handbooks in Operations Research and Management Science, Vol. 9, New York, NY, Elsevier Science, 1017-1072.

Errunza, V. and E. Losq, 1985, "International Asset Pricing under Mild Segmentation: Theory and Test," Journal of Finance 40, 105-124.

Errunza, V. and L.W. Senbet, 1981, "The Effects of International Operations on the Market Value of the Firm: Theory and Evidence," Journal of Finance 36, 401-417.

Errunza, V. and L.W. Senbet, 1984, "International Corporate Diversification, Market Valuation, and Size-Adjusted Evidence," Journal of Finance 39, 727-743.

Foerster, S.R. and G.A. Karolyi, 2000, "The Long-Run Performance of Global Equity Offerings," Journal of Financial and Quantitative Analysis 35, 499-529.

Greene, W.H., 1981, "Sample Selection Bias as a Specification Error: Comment," Econometrica 47, 795-798.

Heckman, J.J., 1979, "Sample Selection Bias as a Specification Error," Econometrica 47, 153-161.

Jegadeesh, N., 2000, "Long-Term Performance of Seasoned Equity Offerings: Benchmark Errors and Biases in Expectations," Financial Management 29, 5-30.

Karolyi, G.A., 1998, "What Happens to Stocks that List Shares Abroad? A Survey of the Evidence and its Management Implications," New York University Salomon Bros. Center Monograph Series, Volume 7, No. 1.

Kothari, S.P. and J.B. Warner, 1997, "Measuring Long-Horizon Security Price Performance," Journal of Financial Economics 43, 301-340.

Lee, I., S. Lochhead, J. Ritter, and Q. Zhao, 1996, "The Costs of Raising Capital," The Journal of Financial Research 19, 59-74.

Loughran, T. and J.R. Ritter, 1995, "The New Issues Puzzle," Journal of Finance 50, 23-52.

Lyon, J., B. Barber, and C. Tsai, 1999, "Improved Methods for Tests of Long-Run Abnormal Returns," Journal of Finance 54, 165-201.

Masulis, R.W. and A. Korwar, 1986, "Seasoned Equity Offerings: An Empirical Investigation," Journal of Financial Economics 15, 3-29.

Mauer, D.C. and L.W. Senbet, 1992, "The Effect of the Secondary Market on the Pricing of Initial Public Offerings: Theory and Evidence," Journal of Financial and Quantitative Analysis 27, 55-79.

Merton, R.C., 1987, "Presidential Address: A Simple Model of Capital Market Equilibrium with Incomplete Information," Journal of Finance 42, 483-510.

Miller, D., 1999, "The Market Reaction to International Cross-Listing: Evidence from Depositary Receipts," Journal of Financial Economics 51, 103-123.

Miller, M.H. and K. Rock, 1985, "Dividend Policy under Asymmetric Information," Journal of Finance 40, 1031-1051.

Myers, S.C., and N.S. Majluf, 1984, "Corporate Financing and Investment Decisions when Firms have Information that Investors Do Not Have," Journal of Financial Economics 13, 187-222.

Spiess, D.K. and J. Affleck-Graves, 1995, "Underperformance in Long-Run Stock Returns Following Seasoned Equity Offerings," Journal of Financial Economics 38, 243-276.

White, H., 1980, "A Heteroskedasticity-Consistent Covariance Matrix Estimator and a Direct Test for Heteroskedasticity," Econometrica 48, 817-838.

The authors thank two anonymous referees; McKinley Blackburn, LeRoy Brooks, Scott Harrington, Donald Lessard, and Bernard Yeung; and participants at the 1999 Academy of International Business Annual Meeting for their valuable comments. Derrick Chow, Vice President of Equity Capital Markets, Credit Suisse First Boston Limited, provided useful insights from a practitioner's perspective. We also gratefully acknowledge the support of the Center for International Business Education and Research (CIBER) at the University of South Carolina.

* Congsheng Wu is an Associate Professor at the University of Bridgeport. Chuck C.Y. Kwok is a Professor of International Business at the University of South Carolina.
COPYRIGHT 2002 Financial Management Association
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2002 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Wu, Congsheng; Kwok, Chuck C.Y.
Publication:Financial Management
Geographic Code:1USA
Date:Jun 22, 2002
Words:9697
Previous Article:Managerial optimism and corporate finance.
Next Article:How does strategic competition affect firm values? a study of new product announcements.
Topics:


Related Articles
Why do firms issue convertible debt?
REED ELSEVIER LAUNCHES $1.5 BILLION GLOBAL BOND.
Do firms time equity offerings? Evidence from the 1930s and 1940s.
Ernst & Young: access to achieve your potential and make a difference.

Terms of use | Privacy policy | Copyright © 2021 Farlex, Inc. | Feedback | For webmasters |