Underpricing of Foreign IPOs in U.S. exchanges: a
comparative study of IPOs from emerging markets and
IPOs from developed markets
MSc International Business Management, specialization International Financial Management
Faculty of Economics and Business University of Groningen May 2010 Supervisor : N. Brunia Student : Y. Wu E-mail : xiaoyarun@hotmail.com Student Number : 1579177
Abstract
Table of contents Page
1. Introduction……… 3
2. Literature review and hypothesis……… 6
2.1. Empirical evidences on IPO underpricing………. 6
2.2. Two principal streams of theories explaining IPO underpricing……….. 6
2.3. Influencing factors on IPO underpricing………... 8
2.4. The “Country of Origin” effect and foreign IPO underpricing in the U.S………. 12
2.5. Hypotheses……… 14
3. Data and Methodology……… 14
3.1. Data……….. 14
3.2. Sample IPOs’ countries of origin………. 18
3.3. Definitions of variables and data sources………. 27
3.4. Descriptive statistics………. 29
3.5. Methodology ……… 37
4. Results……….. 39
4.1. Test result of Hypothesis 1……… 39
4.2. Test results of hypothesis 2……… 39
5. Conclusion……… 52
References……….. 54
Appendices………. 60
1. List of countries (regions) of foreign IPOs went public from 12 April 1996 to 13 March 2009 in the U.S. exchanges……….. 60
2. Data transformations………. 61
3. Test of regression assumptions on transformed variables………. 64
3.1 Test of assumption of linearity………. 65
3.2 Test of assumption of normal residuals……… 66
1. Introduction
The worldwide phenomenon of Initial Public Offering (IPO) has been well documented in the past decades. Most literatures define IPO underpricing as the phenomenon that when issuing companies go public, the offering shares’ closing price of the first-trading day is higher than the offer price (Ritter, 1984; Rock, 1986). To issuing companies, IPO underpricing results in “money left on the table”, which refers to the lost capital that could have been raised if the shares had been offered at a higher price. To investors who were allocated at the offer price, IPO underpricing results in a positive first-day returns (Ritter, 2008).
positive first-day returns without investigation is not that different from the probability of informed investors to get positive first-day returns with investigation. As the uncertainty about the value of the issuing firm increases, the need for security investigation also increases, and the probability for uninformed investors to get positive first-day returns without investigation decreases. Because the direct test of ex ante uncertainty is difficult, various proxies have been adopted. Cater and Manaster (1990) indicate that the more prestigious of underwriters who market the IPO, the less it will be underpriced because prestigious underwriters are associated with lower risk offerings. Corwin and Harris (1999) indicate that IPOs listed in prestigious exchanges, such as NYSE (New York Stock Exchange), are less underpriced because the strict listing requirement of these exchanges tends to make investors believe that the IPOs are less risky.
Motivated by the “country of origin” theory originated from the studies on foreign direct investment and marketing strategies (Roth and Romeo, 1992; Lampert and Jaffe, 1996), which indicates that the generalization and perception about a country influence an actor’s judgments of that country’s products or/and brands, I assume that in addition to the ex ante uncertainties about the issuing firm itself, such as underwriter rank and prestige of listing exchange, there is also a “country of origin” effect on first-day returns of foreign IPOs in the U.S.: the U.S. investors’ perception of a foreign country’s risk level is expected to have influence on his/her evaluation of the country’s IPO, thus affect the first-day returns of the foreign IPO. Foreign IPOs from emerging markets, which are normally associated with higher uncertainties than those of developed markets, are expected to have higher first-day returns than IPOs from developed markets in the U.S..
The sample used in this thesis is composed of 139 foreign IPOs from 25 countries that went public in the U.S. during the period from 12 April 1996 through13 March 2009. Firstly, in order to better examine the role of country risk in affecting first-day returns of foreign IPOs from emerging markets, I use country risk ratings to determine emerging markets and developed markets in the sample. This method is in line with Bruner, et al (2006). Moreover, I compare the performance of three country risk ratings (Euromoney Country Risk Ratings, Institutional Investor Credit Risk Ratings, and Fitch Sovereign Risk Ratings) as classification schemes and find that Euromoney Country Risk Ratings best classifies the sample. Then I conduct Mann-Whitney U test and regression analysis on the sample, and the results indicate that foreign IPOs from emerging markets do not have higher first-day returns than foreign IPOs from developed markets in the U.S. exchanges, and there is no sufficient evidence showing that there is a relationship between first-day returns of foreign IPO and its country of origin.
foreign IPO on its first-day returns has never been done in previous studies of this area. And the comparison among different country risk ratings provides reference for further research on their performance as classification schemes. This thesis also have professional relevance to U.S. investors who are interested in investing foreign IPOs listed in the U.S.. It indicates that whether the foreign IPO comes from emerging market or developed market does not serve as an important predictor of its first-day returns.
2. Literature review and hypothesis
2.1 Empirical evidences on IPO underpricing
The phenomenon of initial public offerings (IPOs) underpricing has been well documented during the past decades. Ibbotson (1975) found an average underpricing of 11.4% in the period of 1960-1969. Studies by Miller and Reilly (1987), and Levis (1990) reported an average IPO underpricing of 20.6 % in the U.S. market and 9.56% in the UK market. Wong and Chiang (1986) and Wang (1998) reported that the underpricing rates are 52% in Japan, 45% in Singapore and Hong Kong, 125% in Malaysia and 579% in China.
2.2 Two principal streams of theories explaining IPO underpricing
the so called “winner’s curse”. In order to keep the uninformed investors in the market, all IPOs need to be deliberately underpriced to make them at least break even. Baron (1982), however, assumes that the underwriter bank is better informed about market demand conditions than the issuer, therefore issuer has the need for underwriter’s advertising and distribution effort, this leads to a principal-agent problem in which underwriter has the incentive to underprice the shares in order to reduce its advertising and distribution effort. Welch (1989) assumes that the issuer is better informed about the company’s true value than investors, higher-valued companies would use underpricing as a signal to reveal its high quality in order to obtain higher price at a seasoned offering.
Theories based on symmetric information include Institutional theory, control theory and behaviour theory, which are not yet mature (Ljungqvist, 2006). Institutional theory claims that firms deliberately sell their shares at a discount to reduce the likelihood of future lawsuits from shareholders disappointed with the post-IPO performance of their shares (Logue, 1973; Ibbotson, 1975). It focuses on three features of the marketplace: litigation, banks’ price stabilizing activities once trading starts, and taxes. Control theory argues that the offer price is determined by the directors of issuing firm with underwriter bank, it is mostly the non-directors of the issuing company who take advantage of IPOs to dispose their shareholdings. The directors tend to underprice the shares which will lead to oversubscription, and oversubscription allows the directors to choose among applicants and achieves a more dispersed ownership from outsiders. In this way the directors can retain control of the firm and the underpricing costs are borne by non-directors who sell shareholdings at the IPO. (Brennan and Franks,1997).
company has the tendency to hold the shares in inventory to prevent price falling, however due to regulatory restriction and market condition, holding back shares is restricted and risky, therefore shares are underpriced to at least break even. However the offer price of underpriced shares is still higher than their fundamental value.
Theories explaining IPO underpricing based on asymmetric information are broadly supported by empirical evidence. Empirical results on symmetric theories are rather mixed. With regard to institutional theory, some studies find relationships between IPO underpricing and the three features of the marketplace, meanwhile underpricing can still be observed in countries where litigation, price stabilization, and taxes play no role in the IPO market (Drake and Vetsuypens, 1993; Lowry and Shu, 2002; Ruud, 1993; Asquith, Jones, and Kieschnick, 1998; Chowdhry and Nanda, 1996; Rydqvist, 1997; Taranto, 2003). Control theory is a relatively new one, which is still subject to empirical testing. Behaviour theory is supported by some evidence, however it is not yet mature. (see review by Ljungqvist, 2006).
2.3 Influencing factors of IPO underpricing
out by informed investors. Faced with this winner’s curse problem, uninformed investors will only submit purchasing orders if, on average, IPOs are underpriced. Rock (1986) assumes that only informed investors cannot take up all the shares on offer even in attractive offerings, therefore the market needs the participation of both types of investors. Rock(1986) assumes that all the IPOs need to be underpriced deliberately in order to make the uninformed investors to at least break even and hence to keep them in the market.
A key implication from Rock’s winner’s curse model is that the degree of IPO underpricing should increase in the ex ante uncertainty about the value of the IPO firm (Beatty and Ritter, 1986). The underlying rationality is that, when the uncertainty about the value of the issuing firm is low, there is not much need for security investigation, the probability for uninformed investors to get positive first-day returns without investigation is not that different from the probability of informed investors to get positive first-day returns with investigation. As the uncertainty about the value of the issuing firm increases, the need for security investigation also increases, and the probability for uninformed investors to get positive first-day returns without investigation decreases. Therefore, Beatty and Ritter, (1986) assumes that in order to submit a purchasing order for shares with greater ex ante uncertainty, the uninformed investors require the shares to be underpriced, in an expected value sense, at a higher degree.
influencing factor to IPO underpricing (Ibbotson et al,1988).
Company age is the number of years of establishment of the issuing company at the time of IPO. For example if an issuing company was established in 1998 and went public in 2008, the company age is 10 years at the time of IPO. One of the difficulties to value a company is how established it is, and company age is normally used as a proxy for this ex ante uncertainty. Young companies which have little operating history are expected to have more uncertainty regarding the pricing per share, and companies have long history are expected to be easier to value (Ritter, 1984). Thus IPOs which are issued by younger companies are expected to have higher first-day returns.
Offering size is equal to the offering price multiplied by the number of shares issued in the IPO. Ritter (1984) documented that offering size has a significant effect on IPO underpricing. IPOs with higher offering sizes are expected to have lower first-day returns. This is based on the understanding that large offerings are normally from big and established companies, whose values are more sound and easier to predict than small companies (B. Francis et al. 2001 ).
Underwriter rank indicates the prestige of the underwriter bank that markets an IPO. Carter, Dark and Singh (1998) assumes that companies use prestigious underwriters to signal that they are high quality companies in order to decrease the level of underpricing, and they provide empirical evidence in their study that underwriter rank has a negative effect on IPO underpricing. Carter and Manaster (1990) provide a ranking of underwriters based on their position in the ‘tombstone’ advertisements in the financial press that follow the completion of an IPO. This ranking has been much used in the empirical IPO literature since it was updated by Jay Ritter. Rank 9 is for the most prestigious underwriters and 0 is for the least prestigious underwriters.
company’s officers and private shareholders (also known as insiders) divided by the total number of shares offered in an IPO in percent. Grinblatt and Hwang (1989) indicate that the percentage of insider offering conveys the company’s value to the market, fewer fractions of shares offered by insiders in an IPO signals a positive outlook to the issuing companies, implying the company of higher quality and less risk. After controlling for other riskiness of the issuing firm, they indicate that IPOs with lower percentage of insider offering are expected to have higher first-day returns.
Underwriter compensation equals to the difference between the offer price and the price per share received by the company divided by offer price in percent (Underwriter compensation = (offer price – price per share received by issuing company) / offer price in percent). Companies with higher risks tend to pay their underwriters more compensation and offer higher underpricing (Beatty and Welch, 1996 ).
Corwin and Harris (1999) indicate that if an IPO is listed in a prestigious exchange which has strict listing requirement, investors tend to believe that the IPO is less risky and of higher quality. NYSE(New York Stock Exchange) is well known to investors for its strict listing requirement for IPOs relative to other listing exchanges (such as Nasdaq, American Stock Exchange, etc.) in the U.S., IPOs listed in NYSE are expected to have lower first-day returns than IPOs that are listed in other exchanges.
High tech industries are defined as the ones having intensive Research& Development (R&D) activities. Chan, Lakonishok, and Sougiannis (2001) suggest that because R&D is intangible assets, they are harder to be evaluated to investors. Thus companies from high tech industries indicate higher uncertainty compared to companies operating in traditional industries. IPOs come from high-tech industries are expected to have higher first-day returns.
Ritter, 1995; Bayless and Chaplinsky, 1996). Ibbotson et al (1988) observe that the hot and cold issue period come in waves. Hot issuing periods are normally associated with higher first-day returns than that of cold issuing periods.
2.4 The “Country of Origin” effect and foreign IPO underpricing in the U.S. When local consumers are not familiar with a foreign product, they tend to evaluate the product not only from the characteristics of the product itself, such as the price, the package, etc., but also from their impression on the country where the product comes from. Empirical evidence supports that the country of origin of a product affects consumers’ evaluations (Han and Terpstra, 1988; Erickson, Johansson and Chao, 1984; Kaynak and Kucukemiroglu, 1992; Hong and Wyer, 1989). Okechuku (1994) investigated the relative importance of the country of origin of a product to consumers in the United States, Canada, Germany and the Netherlands. It was found that the country of origin of a product was one of the most important attributes in consumer preference evaluation, outweighs the brand name, price and other attributes of a product. Respondents in these four developed countries preferred domestically-made products foremost, followed by products made in other developed countries and, lastly, products made in developing countries. Roth and Romeo (1992) and Lampert and Jaffe (1996) call this phenomenon as the “country of origin” effect and define it as the degree to which generalizations and perceptions about a country influence an actor’s judgment of that country’s products and/or brands.
adopting free-market system (Hoskisson, Eden, Lau, & Wright, 2000), and economies of emerging countries (regions) are progressively more integrated into the world economy, some researchers believe that the instability of emerging countries may have even increased (Bordo and Panini Murshid, 2002; Eichengreen and Bordo, 2002). Based on a study done by Martin and Rey (2002), the frequency of financial crashes for emerging countries (regions) who are closed to goods trade is 25% on average during a year, the frequency of financial crashes for emerging countries (regions) who are open to good trade is close to 62% on average during a year, while for developed countries (regions) the number is 9%. This indicates that compared to developed countries (regions), emerging countries (regions) are more prone to crashes than developed countries (Martinez and Santiso, 2003). In addition to the financial crashes, instability caused by political change further aggravates financial vulnerability. Aghion et al. (2001) find that most of institutional and political changes that occurred over a 20-year period in a large sample of 177 countries were concentrated in emerging countries (regions). Santiso (1999a), reports that the three most significant financial crises in Latin America (Mexico in 1994, Brazil in 1999 and Argentina in 2001) took place during a corresponding presidential or parliamentary electoral year. El-Erian (2002) documents in his study that “In most cases, the immediate market reactions to increased political noise has been very pronounced price volatility.” Vaaler and Block (2006) also indicate that market generally views political elections in developing countries (regions) negatively.
differences in underpricing between domestic U.S. IPOs and foreign IPOs listed in the U.S., indicate that, after controlling for the ex ante uncertainties associated with company characteristics, such as company age, underwriter rank, etc., foreign IPOs are more underpriced than domestic IPOs in the U.S. exchanges, due to higher information gathering cost and greater level of uncertainty that investors tend to associate foreign IPOs with. Bell et al (2008) indicate that, in the U.S. market, foreign IPOs come from countries (regions) with higher economic freedom, which are mostly developed markets, have lower level of first-day returns than foreign IPOs from countries (regions) with lower economic freedom, which are mostly emerging markets. This is because the U.S. investors perceive the countries (regions) with higher economic freedom as more legitimate thus less risky, and this may further supports the belief of U.S. investors that these issuing companies are better equipped and supported to succeed in the markets of developed countries (Porter, 1990).
2.5 Hypotheses
Based on the theories and empirical work, I have the following hypotheses:
1. In the U.S. market, foreign IPOs from emerging markets are expected to have higher first-day returns than that of foreign IPOs from developed markets.
2. The country of origin of a foreign IPO listed in the U.S. is expected to have effect on its first-day returns.
3. Data and Methodology 3.1 Data
Using the database of Hoover’s online IPO center, I identified in total 1471 companies went public in the U.S. exchanges (AMEX, NASDAQ, and NYSE) from 12 April 1996 to 13 March 2009, 212 were foreign IPOs from 30 countries (regions) (see Appendix table 1 for country (region) list.).
Table 1: List of countries (regions) included in the sample Argentina Luxemburg Bahamas Mexico Brazil Netherlands Canada Panama China Russia France Singapore
Greece South Africa
Hong kong South Korea
Iceland Switzerland
India Taiwan
Ireland Turkey
Israel UK
Japan No. of countries (regions) : 25
I use country risk ratings to classify whether a foreign IPO in the sample is from emerging market or developed market (Bruner, 2006). I compare the performance of three country risk ratings, Euromoney Country Risk Ratings, Institutional Investor Credit Risk Ratings and Sovereign Risk Ratings, against IMF Advanced economies list 2009 in order to select the best classification scheme.
Institutional Investor Credit risk ratings is also a survey based rating scheme from Institutional Investor magazine. It is an ordinal rating ranging from 0 to 100 points specifically for "country credit" risk. Higher rating points represent lower country credit risk.
Fitch Sovereign Risk Ratings from Fitch ratings is a scale rating scheme which has 9 grades, AAA indicates the lowest risk, and C indicates the highest risk grade. Both Euromoney Country Risk Ratings and Institutional Investor Credit Risk Ratings provide numeric measure of country risk, however Fitch Sovereign Risk Ratings present country risk in letters, for example AAA indicates the country has the lowest risk , C indicates it is a very dangerous country to invest in. For the convenience of calculation, I convert the risk grades represented by letters into numerical values ranging from 1-9, higher values indicate lower country risk. The original expression of Fitch ratings and the numeric values after conversion are presented in table 2 below.
Table 2: Numeric conversion of Fitch ratings Fitch ratings Corresponding value
The IMF Advanced economies list 2009 collected from IMF World Economic Outlook (Oct 2009) is used as a benchmark measuring how well the three rating schemes are in classifying emerging markets and developed markets. The list of countries (regions) identified by the IMF Advanced economies list 2009 is presented below in table 3.
Table 3: List of developed countries (regions) identified by IMF Advanced economies list 2009
IMF advanced economies list 2009
Australia Korea
Austria Luxembourg
Belgium Malta
Canada Netherlands
Cyprus New Zealand
Czech Republic Norway
Denmark Portugal
Finland Singapore
France Slovak Republic
Germany Slovenia
Greece Spain
Hong Kong SAR Sweden
Iceland Switzerland
Ireland Taiwan Province of China
Israel United Kingdom
Italy United States
Japan
3.2. Sample IPOs’ countries of origin
of investors on a country’s risk level. Bruner et al. (2006) in their study on issuing costs of emerging markets IPOs in the U.S. use mainly Euromoney Country Risk Ratings from Euromoeny magazine (Institutional Investor Credit Ratings from Institutional Investor magazine was mentioned but not formally adopted) to classify emerging markets and developed markets, countries (regions) with Euromoney Country Risk Ratings lower than 85 points are classified as emerging markets. There are three drawbacks about the using of this classification scheme in the study. Firstly, whether Euromoney Country Risk Ratings is a good classification scheme was not evaluated. Secondly whether the chosen cut-off rating is appropriate was not tested. Thirdly, how well the performance of Euromoney Country Risk Ratings compared to other classification schemes is not tested. In this thesis I compare the performance of Euromoney Country Risk Ratings together with two other main-stream country risk rating schemes (Institutional Investor Credit Risk Ratings and Sovereign Risk Ratings) against IMF Advanced economies list 2009 in order to select the best classification scheme and to determine the best cut-off point.
The ROC (Receiver Operating Characteristics) curve analysis is a useful way to evaluate the performance of classification schemes that categorize cases into one of two groups (Gönen, 2007) and to determine the cut-off points. The outputs of this analysis include the ROC curve plot, a table presenting area under the curve, and a table presenting coordinates of the curve.
The ROC curve plot is a plot where Y axis indicates the true positive rate (sensitivity), which is the percent of positive cases that are correctly identified as positive by the classification scheme, and X axis indicates the false positive rate (1-specificity), which is the percent of negative cases that are incorrectly identified as positive by the classification scheme.
represents a perfect classification scheme, an area of 90%-100% represents an excellent classification scheme, an area of 80%-90% represents a good classification scheme, and an area of 70%-80% represents a fair classification scheme. An area of 50% represents a worthless classification scheme.
The table presents coordinates of the curve shows all possible cut-off points, and the corresponding true positive rate (sensitivity) and false negative rate (1- specificity). The ROC curve on the plot graphically displays the trade-off relationship between the true positive rate (sensitivity) and the false positive rate (1- specificity) and is used to determine the best cut-off point. The 45- degree diagonal line from lower left to upper right corner of the plot, normally called the chance line, represents all the possible cut-off points that categorize cases by chance, it simply means that using the cut-off points on the chance line to categorize cases is the same as guessing. The performance of the classification scheme is measured by the area under the curve, the further the curve lies above the chance line, the larger the area under the curve, thus the better the performance of the classification scheme. The closer the curve comes to the chance line, the less accurate the classification scheme.
Figure 1: ROC Curve plot of Euromoney Country Risk Ratings
Table 4: Area under the curve for Euromoney Country Risk Ratings
Area Std. Error Asymptotic Sig.a
Asymptotic 95% Confidence Interval Lower Bound Upper Bound
.888 .025 .000 .840 .936
a. Null hypothesis: true area = 0.5
Figure 2: ROC Curve plot of Institutional Investor Credit Risk Ratings
Table 5: Area under the curve for Institutional Investor Credit Risk Ratings
Area Std. Error Asymptotic Sig.a
Asymptotic 95% Confidence Interval Lower Bound Upper Bound
0.780 0.042 0.000 0.698 0.862
a. Null hypothesis: true area = 0.5
Figure 3: ROC Curve plot of Fitch Sovereign Risk Ratings
Table 6: Area under the curve for Fitch Sovereign Risk Ratings
Area Std. Error(a) Asymptotic Sig. a
Asymptotic 95% Confidence Interval Lower Bound Upper Bound
0.728 0.043 0.000 0.645 0.812
a. Null hypothesis: true area = 0.5
Table 7 helps to determine the best cut-off point. It shows the coordinates of the curve of Euromoney Country Risk Ratings, which is a summary of all possible cut-off points and corresponding true positive rate and false negative rate. From table 7, we can see that if the cut-off point is set too low, for example at 62.56, the true positive rate is 100% which indicates that 100% developed markets will be correctly identified as
Table 7: Euromoney Country Risk Ratings’ possible cut-off points and corresponding true positive rate and false negative rate
Coordinates of the Curve
Test Result Variable(s): Euromoney Country Risk Ratings Positive if Greater Than
or Equal To
(True positive rate) Sensitivity
Table 8: Means and medians of first-day returns of emerging markets IPOs and developed markets IPOs
3.3 Definitions of variables and data sources
First-day returns refers to the first-day returns of foreign IPOs in the U.S., it measures the extent of IPO underpricing, higher first-day returns indicates a higher extent of IPO underpricing. First day returns = (first trading day close price – offer price) / offer price in percent. The data are collected from Hoover’s online IPO center.
IPO Country of origin is a predictor of first-day returns. It indicates whether a foreign IPO comes from emerging market or developed market. Countries (regions) have Euromoney Country Risk Ratings lower than 76.72 are classified as emerging markets, while countries (regions) have Euromoney Country Risk Ratings higher than 76.72 are classified as developed markets. IPO Country of origin is a dichotomous variable, value 1= emerging market, value 0 = developed market. The data are collected from Hoover’s online IPO center and Euromoney magazine.
Listing exchange is a predictor of first-day returns. It indicates the prestige of a foreign IPO's listing exchange. NYSE (New York Stock Exchange) is considered as prestigious exchange, other exchanges (NASDAQ, AMEX,etc.) are non prestigious exchanges. It is a dichotomous variable, value 1= NYSE, value 0 = other exchanges. The data are collected from Hoover’s online IPO center.
Offering size is a predictor of first-day returns. The offering size of an IPO is equal to the offering price multiplied by the number of shares issued. The data are collected from Hoover’s online IPO center.
Company age is a predictor of first-day returns. It is the company's age at the IPO. The names of the issuing companies are collected from Hoover’s online IPO center, the age of issuing company is collected from Data Monitor.
Insider offering is a predictor of first-day returns. It is the percentage of shares offered by issuing company's officers and private shareholders. The data are collected from IPO prospectus published on SEC EDGAR website.
Underwriter compensation is a predictor of first-day returns. It is the compensation paid by the issuing companies to the underwriter who markets the IPO. Underwriter compensation = (offer price – price per share received by issuing company) / offer price in percent. The data are collected from IPO prospectus published on SEC EDGAR website.
Industry of issuing company is a predictor of first-day returns. It indicates whether a foreign IPO's issuing company is from high-tech industry or traditional industry. Companies whose SIC (Standard Industrial Classification) codes begin with 737 (Computer programming, software, and services), 283 (Drugs and pharmaceuticals), 357 (Computers and office equipment), 38 (Measuring instruments), 36 (Electrical equipment excluding computers), 48 (Communications), and 37 (Transportation equipment) are classified as high-tech industries, the rest are classified as traditional industries (Chan et al, 2001).It is a dichotomous variable, value 1= high-tech industry, value 0 = traditional industry. The data are collected from IPO prospectus published on SEC EDGAR website.
at least three contiguous months where issue volume fall below the lower quartile. normal periods fall between the upper and lower quartile (Bayless and Chaplinsky, 1996). It is a dichotomous variable, value 1= hot period, value 0 = cold or normal period.
3.4 Descriptive statistics
Figure 4: Histogram of first-day returns
Table 10: Tests of normality on first-day returns Tests of Normalitya
Kolmogorov-Smirnov(a) Shapiro-Wilk
First-day Statistic df Sig. Statistic df Sig.
Returns 0.207 139 .000 0.573 139 .000
For interval predictor offering size, from table 9 we can see that the mean and median are quite different from each other; the high coefficients of skewness and kurtosis indicate that the distribution of offering size is highly positively skewed and much steeper than normal distribution. The histogram of offering size presented in figure 5 also doesn’t look like to fit the normal curve. And the results of normality tests presented in table 11 are all significant. All these indicate offering size is not normally distributed.
Figure 5: Histogram of offering size
Table 11: Tests of normality on offering size Tests of Normalitya
Kolmogorov-Smirnova Shapiro-Wilk
Offering size Statistic df Sig. Statistic df Sig.
.311 139 .000 .487 139 .000
For predictor insider offering, from table 9 we can see that there is a big difference between its mean and median, and the coefficients of skewness and kurtosis are very high. The histogram of insider offering presented in figure 6 shows the distribution doesn’t fit the normal curve. And the normality tests presented in table 12 are all significant.
Figure 6: Histogram of insider offering
Table 12: Tests of normality on insider offering Tests of Normalitya
Kolmogorov-Smirnova Shapiro-Wilk
Insider offering Statistic df Sig. Statistic df Sig.
.334 139 .000 .673 139 .000
Predictor company age is not normally distributed either. From table 9 we can see that the difference between its mean and median is big, and the large coefficients of skewness and kurtosis indicate the distribution of company age is highly positively skewed and is much steeper than normal curve. The histogram of company age presented in figure 7 shows the distribution does not fit the normal curve at all. The significant results of normality tests presented in table 13 also indicate company age is not normally distributed.
Figure 7: Histogram of company age
Table 13: Tests of normality on company age Tests of Normalitya
Kolmogorov-Smirnova Shapiro-Wilk
company age Statistic df Sig. Statistic df Sig.
.337 139 .000 .339 139 .000
The mean and median of predictor underwriter rank are quite close to each other. However the coefficient of skewness is -3.06, which indicates it is highly negatively skewed; and the high coefficient of kurtosis indicates it is much steeper than the normal curve. The histogram of underwriter rank presented in figure 8 shows that its distribution does not fit the normal curve at all. And the normality tests presented in table 14 show significant results. All these indicate underwriter rank is not normally distributed.
Figure 8: Histogram of underwriter rank
Table 14: Tests of normality on underwriter rank
Tests of Normalitya
Kolmogorov-Smirnova Shapiro-Wilk
underwriter rank Statistic df Sig. Statistic df Sig.
.471 139 .000 .417 139 .000
The mean and median of underwriter compensation are equal to each other. However its coefficient of skewness and kurtosis indicate that its distribution is left skewed and steeper than the normal curve. The histogram of underwriter compensation presented in figure 9 confirmed this distribution. The normality tests presented in table 15 are all significant. All these indicate that underwriter compensation is not normally
distributed.
Figure 9: Histogram of underwriter compensation
Table 15: Tests of normality on underwriter compensation Tests of Normality
Kolmogorov-Smirnova Shapiro-Wilk
underwriter Statistic df Sig. Statistic df Sig.
compensation .201 139 .000 .881 139 .000
a. Lilliefors Significance Correction
3.5 Methodology
3.5. 1 Test of Hypothesis 1
To test hypothesis 1, I need to test whether the means or medians of first-day returns of foreign IPOs from emerging markets are higher than foreign IPOs from developed markets in the U.S.. For normally distributed data, mean and variance are typically used to describe the center and spread. However for data that are not normally distributed or contain outliers, they may not be robust enough to accurately describe the data (Pappas and DePuy, 2004, Hollander and Wolfe, 1973). Considering variable first-day returns is not normally distributed, median in this case is a more robust measure of the center of distribution because it is not as heavily influenced by outliers and skewed data as mean and variance are.
In this thesis I use Mann-Whitney U test to compare the medians of first-day returns of foreign IPOs from emerging markets and that of foreign IPOs from developed markets. Mann-Whitney test is often used when you want to test for differences of medians between two groups but your testing variable does not conform to the assumptions of the independent-samples t test (Siegel and Castellan, 1988., Norusis, 2004). A significant result of Mann- Whitney U test indicates that hypothesis 1 is supported. An insignificant result indicates that there is no difference in medians between foreign IPOs from emerging markets and foreign IPOs from developed markets.
3.5.2 Test of Hypothesis 2
regression is appropriate in this case because it is used to model the value of a dependent scale variable based on its linear relationship to one or more predictors. The assumptions of regression analysis are linearity, normality, homogeneity of variance, and independence. Linearity refers to that regression analysis assumes the relationships between the predictors and the dependent variable should be linear. Normality refers to that the regression analysis assumes that the residuals should be normally distributed. Homogeneity of variance (homoscedasticity) refers to that the regression analysis assumes that the residual variance should be constant, an error term with non-constant variance is said to be heteroscedastic. Independence refers to that the regression analysis assumes that the errors associated with one observation are not correlated with the errors of any other observation.
I use scatter plots on first-day returns against each interval variables to check the assumption of linearity. To check the assumptions of normality of the residuals, I use histogram on residuals with superimposed normal curve and P-P plot. If the residuals are normally distributed, the distribution of the residuals displayed in the histogram should fit the normal curve; and in P-P plot, the residuals should closely distributed along the diagonal line. To check the assumption of homoscedasticity and independence, I use scatter plot on residuals against predicted values. If the model is well fitted, there should be no pattern to the residuals plotted against the predicted values. If the variance of the residuals is non-constant across predicted values, then the residual variance is heteroscedastic and not dependent.
4. Results 4.1 Test result of Hypothesis 1
Table 16 shows the test statistics of Mann-Whitney U test on variable first-day returns. The significance value is much higher than 0.05, which indicates that there is no significant difference in medians between foreign IPOs from emerging markets and foreign IPOs from developed markets. Hypothesis 1 of this thesis is not supported. Table 16: Test statistics of Mann-Whitney U test on variable first-day returns.
Test Statisticsa
First-day returns
Mann-Whitney U 2081.500
Wilcoxon W 3209.500
Z -.358
Asymp. Sig. (1-tailed) .360
a. Grouping Variable: country of origin
Note: foreign IPOs are from emerging markets if country of origin=1, foreign IPOs are from developed markets if country of origin=0.
4.2 Test results of hypothesis 2
4.2.1 Testing the assumptions of linearity
Figure 10: Scatter plot on first-day returns by offering size
Figure 11: Scatter plot on first-day returns by underwriter rank
Figure 13: Scatter plot on first-day returns by insider offering
Figure 14: Scatter plot on first-day returns by underwriter compensation
4.2.2 Test of assumption of normal residual
Figure 15: Histogram on regression residualsa
a. Independent variables: offering size, underwriter rank, company age, insider offering and underwriter compensation.
Figure 16: P-P plot of regression standardized residualsa
4.2.3 Test of assumptions of homoscedasticity and independence of residuals To test the assumption of homoscedasticity, I predict first-day returns using multiple predictors including offering size, underwriter rank, company age, insider offering and underwriter compensation and then plot the residuals in figure 17. We can see that the scatterplot shows heteroscedasticity; that is, the variance of the residuals is not constant across values of the predicted values. It also shows the variance of the residuals increases with increasing predicted values. These indicate that the assumptions of homoscedasticity and independence are violated.
Figure 17: Scatter plot of regression residualsa
a. Independent variables: offering size, underwriter rank, company age, insider offering and underwriter compensation.
4.2.4 Selecting significant control variable(s)
variables (Jason, 2002). Underwriter rank and underwriter compensation have negative skewness and high kurtosis, unfortunately none of the transformations seem to work well for these two variables, therefore I keep them as untransformed. The transformed variables and their distributions are presented in appendix 2. Using these transformed variables, I re-examine the assumptions of regression, and find that they are still violated (tests of assumptions are presented in appendix 3). The linearity relationships inspected between transformed first-day returns and transformed predictors are still very weak. Weighted least squares regression(WLS) is normally used to correct heteroscedasticity, however by using offering size, underwriter rank, company age, insider offering, underwriter compensation and the transformed variables respectively as the weight variable, none of them can properly correct the heteroscedasticity. These indicate that it is not suitable to use linear regression to predict first-day returns by offering size, company age, insider offering, underwriter rank and underwriter compensation or their transformed variables.
To further investigate the relationships between first-day returns and offering size, company age, insider offering, underwriter rank and underwriter compensation, I use model fit option in SPSS to re-examine the scatter plots on first-day returns by first-day returns, offering size, company age, insider offering, underwriter rank and underwriter compensation, and scatter plots on transformed first-day returns by transformed predictors to see if any curve pattern can be detected. However, none of the curve patterns fit the data. This indicates offering size, company age, insider offering, underwriter rank and underwriter compensation might be irrelevant predictors for first-day returns of foreign IPOs in the U.S..
Since none of offering size, company age, insider offering, underwriter rank and underwriter compensation are significant, control variable(s) are selected from dichotomous variables including listing exchange, industry of issuing company and issuing period. The included and excluded variable (s) resulted from stepwise regression analysis are presented below in table 19. We can see that only industry of issuing company is qualified to enter. Therefore industry of issuing company will be entered as control variable into the regression on first-day returns. From the model summary of regression on first-day returns by industry of issuing company presented in table 20, we can see that the R square of regression model on first-day returns using industry of issuing company as single predictor is 0.029, which means about 2.9% of variation in first-day returns are explained by this model. The ANOVA table presented in table 21 shows the statistical acceptability of the model. The very small value of regression sum of square shows that only a small amount of variation in first-day returns are explained by this model, the significance level is lower than 0.05, which indicate that the explanation is not due to chance. From table 22 which shows the coefficients of the variables, we can see that the significance value of coefficient of industry of issuing company is lower than 0.05, indicating that it is significant. The coefficient value of 0.133 indicates foreign IPOs from high-tech industries have higher first-day returns than foreign IPOs from traditional industries.
Table 19: Included and excluded variables by stepwise regression analysis Variables Entereda
Method
Industry of issuing company Stepwise (Criteria: Probability-of-F-to-enter <= .050,
Probability-of-F-to-remove >= .100). Excluded Variablesa Beta In t Sig. Partial Correlation Collinearity Statistics Tolerance Listing exchange -.103b -1.20 .232 -.102 .957 Issuing period .013b .155 .877 .013 .974
a. Dependent Variable: first-day returns
Table 20: Model summary of regression on first-day returns by industry of issuing company
Model Summaryb
R R Square Adjusted R Square Std. Error of the Estimate
.170a .029 .022 .380
a. Predictors: (Constant), industry of issuing company b. Dependent Variable: first-day returns
Table 21: ANOVA table of regression on first-day returns by industry of issuing company
ANOVAb Sum of
Squares df Mean Square F Sig.
Regression .590 1 .590 4.090 .045a
Residual 19.775 137 .144
Total 20.365 138
a. Predictors: (Constant), industry of issuing company b. Dependent Variable: first-day returns
Table 22: Coefficients of regression on first-day returns by industry of issuing company Coefficientsa Model Unstandardized Coefficients Standardized Coefficients t Sig. B Std. Error Beta (Constant) .101 .042 2.412 .017
industry of issuing company .133 .066 .170 2.022 .045 a. Dependent Variable: first-day returns
The final regression model is presented blow, where Y is the independent variable first-day returns, country of origin is explanatory variable, and industry of issuing
company is control variable. b1, the coefficient of country of origin, refers to the
IPOs when industries of the issuing companies are the same (i.e. all come from traditional industries or all come from high-tech industries). b2, the coefficient of industry of issuing company, refers to the difference in first-day returns between IPOs from high-tech industries and traditional industries when the IPOs’ countries of origins are the same (i.e. all come from emerging markets or all come from developed markets). a is the value Y is predicted to have when all the predictors are equal to zero, which means a is the first-day returns of developed markets IPOs whose issuing companies come from traditional industries.
Y= first-day returns = a + b1(country of origin) + b2 (industry of issuing company)
Results of the regression model analysis are presented below. The regression model summary presented in table 23 shows the strength of the relationship between the model and the dependent variable first-day returns. R, the multiple correlation coefficient, is the linear correlation between the observed and model-predicted values of first-day returns. The small R value of 0.173 shows a very weak relationship. R Square, the coefficient of determination, is the squared value of the multiple correlation coefficient. The small R square value of 0.030 shows that only 3% of the variation in first-day returns is explained by the model.
Table 23: Model summary of regression on first-day returns by country of origin and industry of issuing company
Model Summaryb
R R Square Adjusted R Square Std. Error of the Estimate
.173a .030 .016 .381
a. Predictors: (Constant), country or origin, industry of issuing company b. Dependent Variable: first-day returns
This indicates that only a very small proportion of variation in first-day returns is explained by the model. The significance value of the F statistic is higher than 0.05, which means that the small amount of variation explained by the model is probably due to chance.
Table 24: ANOVA table of regression on first-day returns by country of origin and industry of issuing company
ANOVAb Sum of
Squares df Mean Square F Sig.
Regression .612 2 .306 2.107 .126a
Residual 19.753 136 .145
Total 20.365 138
a. Predictors: (Constant), country or origin, industry of issuing company b. Dependent Variable: first-day returns
The coefficients table is presented in table 25. It shows whether the coefficients of the predictors in the model are statistically significant and, if so, the direction of the relationship with dependent variable. The significant coefficient of industry of issuing company indicates that the effect of industry of issuing company on dependent variable first-day returns is significant. The positive sign of the coefficient indicates that foreign IPOs come from high-tech industries have higher first-day returns than foreign IPOs come from traditional industries. The coefficient of country of origin is not significant (p=0.700), which indicates that the country of origin of a foreign IPO is not an important predictor in predicting the first-day returns.
Table 25: Coefficients of regression on first-day returns by country of origin and industry of issuing company
Coefficientsa Unstandardized Coefficients Standardized Coefficients t Sig. B Std. Error Beta (Constant) .117 .058 2.006 .047 Country of origin -.027 .070 -.034 -.387 .700
5. Conclusion
In this thesis I find that, during the period from 12 April 1996 through 13 March 2009, the sample foreign IPOs listed in the U.S. are underpriced on average by 15.55%. This is in line with most of the empirical literatures on IPO underpricing.
Hypothesis one of this thesis is that foreign IPOs from emerging markets have higher first-day returns than foreign IPOs from developed markets in the U.S.. The descriptive statistics show that the median first-day returns for foreign IPOs from emerging markets is 4.15%, and the median first-day returns for foreign IPOs is 3.76%. In order to test whether the difference is significant, I use two independent sample Mann- Whitney U test to compare the difference in median first-day returns between foreign IPOs from emerging markets and foreign IPOs from developed markets in the U.S.. The result shows there is no significant difference in first-day returns between foreign IPOs from emerging markets and foreign IPOs from developed markets in the U.S.. This indicates hypothesis one is not supported.
Hypothesis two of this thesis assumes that, in the U.S. market, the country of origin of a foreign IPO affects its first-day returns. In order to test this hypothesis, I firstly select significant control variable(s) from listing exchange, underwriter rank, offering size, company age, insider offering, underwriter compensation, industry of issuing company and issuing period. Diagnostic tests on regression assumptions and results of stepwise regression analysis show that only industry of issuing company is a significant predictor of first-day returns. Then I conduct multiple linear regression on first-day returns by using country of origin as explanatory variable and industry of issuing company as control variable. The results show that country of origin is not a significant predictor of first-day returns, while industry of issuing company is a significant predictor of first-day returns. This indicates that hypothesis two is not supported either.
assumed to have higher first-day returns than foreign IPOs from developed markets. However the test result of hypothesis one indicates that although the difference on first-day returns does exist, it is not significant. And test results of hypothesis two indicate that whether a foreign IPO comes from emerging market or developed market does not have effect on its first-day returns. Secondly, among listing exchange, underwriter rank, offering size, company age, insider offering, underwriter compensation, industry of issuing company and issuing period, which are proved to have effects on IPO first-day returns in previous literatures, only industry of issuing company is proved to be a significant predictor of first-day returns in this thesis, yet with a very small explanatory power.
References
Aghion, P., A. Alesina and F Trebbi, 2001. Endogenous Political Institutions. Working Paper, November. Cambridge, MA: Harvard University.
Samiee, S., 1994. Customer evaluation of products in a global market. Journal of International Business Studies, 25(3), pp. 579–604.
Anderson, P.H. & Chao, P., 2003. Country of origin effects in global industrial sourcing: Toward an integrated framework. Management International Review, 43(4), pp. 339–360.
Andrea L. Demaskey and Gerard T. Olson, 2006, The Role of the Region and Country of Origin on the Initial and Aftermarket Performance of ADR IPOs. Journal of Transnational Management Development, Volume 11, Issue 3, 2006, pp. 39-60. Asquith, D., J.D. Jones, and R. Kieschnick, 1998, Evidence on Price Stabilization and Underpricing in Early IPO Returns, Journal of Finance 53, pp. 1759-1773.
Baron, David P., 1982. A Model of the Demand for Investment Bank Advising and Distribution Services for New Issues. Journal of Finance 37 ,pp.955-976.
Bayless, Mark and Chaplinsky, Susan, 1991, Expectations of security type and the informational content of debt and equity offers, Journal of Financial Intermediation 1, pp.195-214.
Bayless, Mark and Chaplinsky, Susan, 1996, Is There a Window of Opportunity for Seasoned Equity Issuance? The Journal of Finance, Vol. 51, No. 1 (Mar., 1996), pp. 253-278
Beatty, Randolph P., and Ritter, Jay R., 1986, Investment Banking, Reputation, and the Underpricing of Initial Public Offerings. Journal of Financial Economics 15 , pp. 213-232
Beatty, Randolph P., and Welch, Ivo., 1996, Issuer Expenses and Legal Liability in Initial Public Offerings. Journal of Law and Economics, Vol. 39, No. 2 (Oct., 1996), pp. 545-602
Bell, R. Greg, Moore, Curt B. and Al-Shammari, Hussam A.,2008. Country of Origin and Foreign IPO Legitimacy: Understanding the Role of Geographic Scope and Insider Ownership. Entrepreneurship Theory and Practice, Vol. 32, No. 1, pp. 185-202, January 2008.
Benveniste, L.M., Ljungqvist, A., Wilhelm, W.J. Jr., and Yu, Xu, 2003, Evidence of Information Spillovers in the Production of Investment Banking Services. Journal of Finance 58, pp.577- 608.
Bilkey, W.J. and Nes, E., 1982, Country-of-Origin Effects on Product Evaluations, Journal of International Business Studies, No. 13, Spring/Summer , pp. 89-99.
Bordo, Michael D. & Murshid, Antu Panini, 2006. Globalization and changing patterns in the international transmission of shocks in financial markets. Journal of International Money and Finance, vol. 25(4), pp. 655-674
Bradley, Daniel J.; Cooney, John W., Jr.; Dolvin, Steven D.; Jordan, Bradford D., 2006,
Penny stock IPOs. Financial Management, Vol. 35, No. 1 (Spring), pp. 5-30Brennan, M.J., and J. Franks, 1997, Underpricing, Ownership and Control in Initial Public Offerings of Equity Securities in the U.K., Journal of Financial Economics 45, pp. 391-413
Bruner, R., Chaplinsky, S., and Ramchand, L., 2006, Coming to America: IPOs from emerging market issuers. Emerging Markets Review 7, pp. 191-212
Carter, Richard B., Frederick H. Dark, and Ajai K. Singh. 1998. ‘‘Underwriter
Reputation, Initial Returns, and the Long-run Performance of IPO Stocks.’’ Journal of Finance 53 pp.285-311.
Carter and Manaster, 1990, Initial public offering and underwriter reputation. Journal of Finance 45: pp.1045-1067
Chan, Louis K.C., Lakonishok, Josef., and Sougiannis, Theodore., 2001, The Stock Market Valuation of Research and Development Expenditures. The Journal of Finance, Vol.LVI, No.6, pp. 2431-2456
Comstock, A., Kish, R.J., Vasconcellos, G.M., 2003. The post-privatization financial performance of former state-owned enterprises. Journal of International Financial Markets, Institutions and Money 13 (1), pp.19–37
Corwin, S. A. and J. Harris. 1999. The Initial Listing Decisions of Firms that Go Public. Working Paper, University of Georgia.
Chowdhry, B., and V. Nanda, 1996, Stabilization, Syndication, and Pricing of IPOs, Journal of Financial and Quantitative Analysis 31, pp. 25-42.
Demaskey, Andrea L., and Olson, Gerard T., 2006, The Role of the Region and Country of Origin on the Initial and Aftermarket Performance of ADR IPOs. Journal of Transnational Management, 11:3, pp.39 -60
Drake, P.D., and M.R. Vetsuypens, 1993, IPO Underpricing and Insurance against Legal Liability, Financial Management 22, pp. 64-73.
El-Erian, M. 2002a. "Are We There Yet?" PIMCO Emerging Markets Watch, May. Available at http://www.pimco.com.
Erickson, G.M., Johansson, J.K. and Chao, P., 1984, Image Attributes in Multi-attribute Product Evaluations, Journal of Consumer Research, Vol. 11, September, pp. 694-699.
Francis, Bill, Hasan , Iftekhar, and Li, Feng, 2001, Underpricing of Foreing and Domestic IPOs in the U.S. Market: Empirical Evidence. Financial markets, Institutions & Instruments. V.10, No. 1, Feb
Grinblatt, Mark and Hwang, Chuan Yang, 1989, Signalling and the Pricing of New Issues. The Journal of Finance, Vol. 44, No. 2, June, pp. 393-420
Gönen, M., 2007, Analyzing Receiver Operating Characteristic Curves with SAS. Cary, NC: SAS Institute Inc.
Habib, M.A., and A. Ljungqvist, 2001, Underpricing and Entrepreneurial Wealth Losses in IPOs: Theory and Evidence, Review of Financial Studies 14, pp. 433-458. Han, C.M. and Terpstra, V., 1988, Country-of-Origin Effects for Uninational and Bi-national Products, Journal of International Business Studies, No. 14, Summer, pp. 235-255.
Harvey, C.R., 1995. Predictable risk and returns in emerging markets. Review of Financial Studies, pp.773–816.
Hollander, M. and Wolfe, D.A., 1973, Nonparametric Statistical Methods, New York: John Wiley & Sons, Inc.
Hong, S. and Wyer, R.S.,1989, Effects of Country-of-Origin and Product-attribute Information on Product Evaluation: An Information Processing Perspective, Journal of Consumer Research, No. 16, September, pp. 175-187.
Hoskisson, R., Eden, L., Lau, C., & Wright, M., 2000. Strategy in emerging economies. Academy of Management Journal, 43(3), pp. 249–267.
Hymer, S.H., 1960, The international operations of national firms: A study of direct investment. Journal of International Management, 8, pp. 223–240
Ibbotson, R.G., 1975, Price performance of common stock new issues. Journal of Financial Economics 3, 235-272.
Ibbotson, R., Sindelar, J. and Ritter, J., 1988, Initial public offerings, Journal of Applied Corporate Finance, volume1, issue 2, pp. 37-45.
Jason, Osborne., 2002, Notes on the use of data transformations. Practical Assessment, Research & Evaluation, 8(6)
community and landscape ecology, Cambridge University Press
Kaynak, E. and Kucukemiroglu, O., 1992, Sourcing of Industrial Products: Regiocentric Orientation of Chinese Organizational Buyers, European Journal of Marketing, Vol. 26 No. 5, pp. 36-55.
Koh, F., and Walter, T., 1989, A Direct Test of Rock’s Model of the Pricing of Unseasoned Issues. Journal of Financial Economics 23, pp.251-272.
Lampert, S.I. & Jaffe, E.D.,1996, Country of origin effects on international market entry. Journal of Global Marketing, 10 (2), pp.27–52.
Levis, M., 1990, The winner's curse problem, interest costs and the underpricing of initial public offerings, Economic Journal, 100, pp.76-89.
Ljungqvist, Alexander, 2006, IPO underpricing. Handbook of Corporate Finance: Empirical Corporate Finance, Volume A, (Handbooks in Finance Series, Elsevier/North-Holland), Chapter 7
Logue, D., 1973, Premia on Unseasoned Equity Issues, 1965-69, Journal of Economics and Business 25, pp. 133-141.
Loughran, T., and J. Ritter, 1995, The new issues puzzle, Journal of Finance 50, pp.23-52.
Loughran, T., and J. Ritter, 2004, Why has IPO underpricing increased over time? Financial Management, Vol. 33, No. 3 (Autumn), pp. 5-37
Lowry, M., and S. Shu, 2002, Litigation Risk and IPO Underpricing, Journal of Financial Economics 65, pp. 309-335.
Martin, P. and H. Rey, 2002. Financial Globalization and Emerging Markets: With or Without Crash? Working Paper, May. Federal Reserve Bank of New York and Princeton University.
Martinez, Juan and Santiso, Javier, 2003, Financial Markets and Politics: The Confidence Game in Latin American Emerging Economies. International Political Science Review / Revue internationale de science politique, Vol. 24, No. 3, The Political Economy of International Finance. L'Economie politique de la finance international. Jul, pp. 363-395
Michaely, R., and Shaw, W.H., 1994, The Pricing of Initial Public Offerings: Tests of Adverse-Selection and Signaling Theories. Review of Financial Studies 7, pp.279-319.
Nanda, V., and Yun Y., 1997, Reputation and Financial Intermediation: An Empirical Investigation of the Impact of IPO Mispricing on Underwriter Market Value. Journal of Financial Intermediation 6, pp.39-63.
N.J.: Prentice Hall, Inc..
OkechukuChike, 1994, The Importance of Product Country of Origin: A Conjoint Analysis of the United States, Canada, Germany and The Netherlands. European Journal of Marketing, Volume 28, issue 4, pp. 5-19
Pappas, P. A. and DePuy, V., 2004, An Overview of Non-parametric Tests in SAS: When, Why, and How.
Miller, R. and Reilly, F., 1987, An examination of mispricing, returns, and uncertainty for initial public offerings, Financial Management, 16, pp.33-38.
Ritter, J.R., 1984, The Hot Issue Market of 1980. Journal of Business 57, pp. 215-240. Ritter, J.R., 2008, Money left on the table in IPOs by firm.
http://bear.warrington.ufl.edu/Ritter/MoneyOnTheTablebyFirm.pdf
Rock, K., 1986, Why New Issues Are Underpriced, Journal of Financial Economics 15, pp.187-212.
Roth, M. & Romeo, J.,1992, Matching product category and country image perceptions: A framework for managing country of origin effects. Journal of International Business Studies, 23(3), pp.477–498.
Rydqvist, K., 1994, Compensation, Participation Restrictions and the Underpricing of Initial Public Offerings, mimeo, Stockholm School of Economics.
Santiso, J. 1999a. Wall Street and the Mexican Crisis: A Temporal Analysis of Emerging Markets. International Political Science Review, 20(1): 49-72.
Siegel, S., and Castellan, N. J., 1988. Nonparametric statistics for the behavioral sciences. New York: McGraw-Hill, Inc..
Taranto, M., 2003, Employee Stock Options and the Underpricing of Initial Public Offerings, working paper, University of Pennsylvania.
Vaaler, P., Schrage, B., Block, S. 2006. Elections, Opportunism, Partisanship and Sovereign Ratings in Developing Countries. Review of Development Economics, 10: pp.154-170
Wall, M., Heslop, L.A. and Hofstra, G., 1988, Male and Female Viewpoints of Countries as Producers of Consumer Goods, Journal of International Consumer Marketing, No. 1, pp. 1-25.
Wang, K., 1998, The impact of institutional framework, interest rates and demand on IPO underpricing: a theoretical model and empirical test, Doctoral Dissertation, University of Strathclyde, UK, pp.292.
Wong, K. and Chiang, H., 1986, Pricing of new equity issues in Singapore, Asia Pacific Journal of Management, 41, pp.1-10
Websites:
Jay Ritter’s IPO Data website: http://bear.warrington.ufl.edu/RITTER/ipodata.htm Hoover’s online IPO center:
http://www.hoovers.com/global/ipoc/index.xhtml?pageid=1258 SEC EDGAR website:
Appendix
1. Table 1: List of countries (regions) of foreign IPOs went public from 12 April 1996 to 13 March 2009 in the U.S. exchanges
Argentina Israel
Bahamas Italy
Bermuda Japan
Brazil Luxemburg
British virgin island Mexico
Canada Netherlands
Cayman Islands Panama
China Russia
France Singapore
Germany South Africa
Greece South Korea
Hong kong Switzerland
Iceland Taiwan
India Turkey
Ireland UK
