IPO Returns and Seasoned Equity Offerings
Amsterdam Business SchoolName Bruno Munnikhof
Number 10268669
BSc/MSc in Business Economics Specialization Finance and Organization Supervisor dr. I.J. Naaborg
Completion 02-01-2015
ECTS 12 points
Abstract
Many different theories have been proposed to explain the underpricing of initial public offerings (IPOs). Many of these theories are not able to explain underpricing for more than a few percent. One of the theories that has a prominent role in explaining the underpricing of IPOs, is the signaling hypothesis (Welch, 1989). However, evidence in favor of this hypothesis is at best mixed. One of the possible explanations for the lack of evidence is that only some firms value underpricing as a signaling device. This paper tests the signaling hypothesis by examining a sample of firms that went public in the U.S. between 2005 and 2010. A distinction is made between domestic and foreign IPOs, since Francis et al. (2010) state that foreign firms are more likely to engage in a signaling strategy. The results show that neither domestic firms nor foreign firms in general deliberately leave money on the table at their IPO. For both groups of firms, there is no significant relation between the degree of underpricing and the probability of a seasoned equity offering (SEO). For domestic firms, there is a positive and significant relation between aftermarket returns and the probability of a seasoned equity issue. This indicates that, for domestic firms,
the market-feedback hypothesis is better than the signaling hypothesis in explaining SEO activities.
Statement of Originality
This document is written by Student Bruno Munnikhof who declares to take full responsibility for the contents of this document.
I declare that the text and the work presented in this document is original and that no sources other than those mentioned in the text and its references have been used in creating it.
The Faculty of Economics and Business is responsible solely for the supervision of completion of the work, not for the contents.
1
Table of Contents
1. Introduction ... 2
2. Literature review ... 3
2.1. Theories based on symmetric information ... 4
2.2. Theories based on asymmetric information ... 5
2.2.1. Investors are better informed ... 5
2.2.2. Issuers are better informed ... 6
2.3 Concluding conclusion of the literature ... 8
3. Methodology and hypotheses ... 9
3.1. Methodology ... 9
3.2. Hypotheses ... 11
3.2.1 Full sample of IPOs ... 11
3.2.2. Domestic IPOs ... 12
3.2.3. Foreign IPOs ... 12
4. Data and descriptive statistics... 13
4.1. Data construction ... 13 4.2. Descriptive statistics ... 14 5. Analysis ... 15 5.1. Empirical results ... 15 5.1.1. Full Sample ... 16 5.1.2. Domestic IPOs ... 19 5.1.3. Foreign IPOs ... 19 5.2 Robustness check ... 20
6. Conclusion and discussion ... 22
6.1. Conclusion ... 22
6.2. Discussion ... 23
2
1. Introduction
Stocks typically close above the offering price on the first-day of the trade. Liu and Ritter (2011) report an average increase of the stock price on the first-day of trade of 24% for the U.S. market in the period 1993-2008. Recent examples of firms that experienced a large increase of their share price on their first day of trade are Twitter (2013) and Alibaba (2014). They were underpriced by 73% and 38%, respectively. Many different theories have been proposed to explain the underpricing of initial public offerings (IPOs).
Boulton, Smart, and Zutter (2011) state that many of these theories assume asymmetric information among participants involved in the process of IPOs. Among others, Welch (1989), Jegadeesh, Weinstein and Welch (1993) and Francis et al. (2010) assume that issuers are more informed than investors. Based on this assumption, they state that high-quality firms can signal their quality to investors by underpricing their IPO. Using underpricing as a signaling device is known as the signaling hypothesis of IPO underpricing. Under the signaling hypothesis, firms deliberately leave money on the table in order to raise additional capital through seasoned equity offerings (SEOs) under more favorable terms (Garfinkel, 1993).
However, Ritter and Welch (2002) and Francis et al. (2010) argue that the evidence in favor of the signaling hypothesis is weak and at best mixed. Jegadeesh, Weinstein and Welch (1993) find an alternative hypothesis; the market-feedback hypothesis. This hypothesis is based on the assumption that the market is better
informed. Under the market-feedback hypothesis, share price appreciations on the IPO date and share price appreciations in the post-IPO period inform the issuer that he underestimated the marginal return of the firm’s projects (Garfinkel, 1993). The issuer uses this information and raises additional capital through SEOs to increase the scale of its projects (Jegadeesh, Weinstein and Welch, 1993).
Francis et. Al (2010) argue that there is a lack of evidence in favor of the signaling hypothesis because researchers are unable to identify firms that value underpricing as a signaling device. They state that firms that experience higher
information asymmetry and have stronger needs for external capital are more likely to use underpricing as a signaling strategy.
3 relationship between the degree of underpricing and the probability of a seasoned equity offering (SEO). If firms deliberately underprice their IPO to signal their quality, they need to recoup the cost of this strategy in some way. One way to do so is to issue seasoned equity within a short period of time of the IPO. Therefore, the question whether a high degree of underpricing will lead to an increase of the probability of a seasoned equity issue within three years of the IPO will be answered.
In addition, the alternative market-feedback hypothesis proposed by Jegadeesh, Weinstein and Welch (1993) will be tested. Under this hypothesis, underpricing does not play a unique role in predicting SEOs. Instead, returns in the post-IPO period will have equal or greater explanatory power for the probability of SEOs (Garfinkel, 1993).
The signaling hypothesis and the alternative market-feedback hypothesis will be tested using a sample of 389 IPOs in the U.S. market between 2005 and 2010. A
distinction is made between foreign IPOs and domestic IPOs, since Francis et al. (2010) state that foreign firms are more likely to engage in a signaling strategy because of a stronger need for external capital and because they possibly face a higher degree of information asymmetry. Splitting the full sample up into a subsample of foreign IPOs and a subsample of domestic IPO will allow to determine whether some firms are more likely to value underpricing as a signaling device than other firms are, as proposed by Francis et al. (2010).
The remainder of this paper is organized as follows. In chapter 2, the main existing theories with respect to IPO underpricing will be explained. In section 3, the methodology used in this paper is explained and hypotheses are formulated based on the existing literature. In chapter 4 , the data used in the sample will be described and descriptive statistics are given for firms included in the final sample. In chapter 5, results of the estimated logit-model are reported. Conclusions will be formulated based on the results and the existing literature in section 6.
2. Literature review
Ibbotson (1975) shows that the first-day return of initial public offerings is positive on average. Ritter and Welch (2002) state that approximately 70 percent of all IPOs is underpriced. These IPOs end the first day of trading at a closing price greater than the offering price. There are many different theoretical explanations that focus on the
4 reason for issuing companies to sell their stock at a price lower than investors are willing to pay (Brennan and Franks, 1997). The different theories can be categorized into two broad categories (Ritter and Welch, 2002). The first set of theories assumes there is symmetric information, whereas the other group of theories assumes there is asymmetric information. The second group can be classified into theories based on the assumption that investors are better informed than issuers and theories for which the opposite holds. Ellul and Pagano (2006) point out that IPO underpricing is mostly explained using theories based on the assumption of asymmetric information.
In addition to theories based on the assumption of symmetric and asymmetric information, research is currently being done with respect to behavioral explanations for IPO underpricing. However, Ljungqvuist and Wilhelm (2005) point out that
behavioral theories often lack the structure for simple econometric exercises that control for important other factors, such as relevant firm characteristics and economic conditions. Therefore, the focus of this paper will be on traditional theories based on symmetric and asymmetric information.
First, the main existing theories based on symmetric information will be explained. Second, theories based on asymmetric information will be discussed. The second group of theories is further split up into theories that assume that investors are better informed and theories that assume that issuers are better informed. Finally, a concluding conclusion of the literature will be given.
2.1. Theories based on symmetric information
The first group of theories is based on the underlying assumption that the same information is available to investors, issuers and underwriters. Ibbotson (1975)
suggests that underpricing might function as a form of insurance against potential legal suits. For example, a positive initial return might reduce potential lawsuits in case of an error in the prospectus. The lawsuits avoidance hypothesis is supported by evidence found by Tinic (1988). In contrast, Drake and Vetsuypens (1993) and Keloharju (1993) provide evidence that underpricing of the IPO did not protect firms from being sued in the period following the IPO. Based on these different studies, Ritter and Welch (2002) state that the avoidance of law suits is not the primary determinant of underpricing.
Another theory based on symmetric information is the after-market liquidity theory. Krigman, Shaw and Womack (1999) show that trading volume in the
5 aftermarket increases when the degree of underpricing is larger. Ellen and Pagano (2006) argue that liquidity risks and expected liquidity in the after-market are
important determinants of IPO underpricing. However, Ritter and Welch (2002) state that it is not clear how the issuing firm benefits from underpricing with increased after-market liquidity as underlying reason. This can only be an important determinant of underpricing in case the increased liquidity is persistent (Booth and Chua, 1996).
2.2. Theories based on asymmetric information
Theories based on asymmetric information assume that one of the parties involved is better informed. First, theories based on the assumption that investors are better informed will be discussed. Then theories based on the assumption that issuers are better informed will be discussed.
2.2.1. Investors are better informed
If investors have more information than the issuer, the issuer does not know how to price the IPO and faces an unknown demand (Ritter and Welch, 2002). Under the assumption that there is a group of investors that is better informed than other
investors, Rock (1986) states that the issuing firm has to offer the shares at a discount to attract uninformed investors. The informed investors will not buy shares in case of a bad issue and the uninformed investors will receive full allocations. However, in case of a good issue, informed investors do desire the shares. This results in excess demand for the shares and the allocation will be rationed, leaving the uninformed investors with only a partial allocation. The fear of only receiving full allocations when the issue is bad is also known as the winner’s curse. For the uninformed investors to break even, the IPO needs to be underpriced.
Welch (1992) argues that subsequent investors might imitate earlier investors when an IPO is sold sequentially. In case the IPO is priced a little too high, this might cause the early investors to abstain. This can lead to complete failure of the IPO, because the other investors abstain simply because they only request shares when they believe an offering is hot (Ritter and Welch, 2002). When the IPO is priced a little too low (underpriced) this will cause early and subsequent investors to request shares, leading to a heavily oversubscribed IPO. Amihud, Hauser and Kirsch (2003) find evidence supporting the theory of Welch (1992).
6 Benveniste and Spindt (1989) give another possible explanation for the
underpricing of IPOs. They show that underpricing can function as a compensation for informed investors that reveal their private information. Underwriters set an initial price range and then gauge demand by gaining information from investors. If potential investors show a willingness to pay a high price, this will cause the eventual offer price to increase. To induce investors to reveal positive information, they need to be
compensated through underpricing and a favorable allocation of shares. This information-gathering theory is supported by evidence from Cornelli and Goldreich (2002). However, Ritter and Welch (2002) state that this theory is not likely to explain underpricing for more than a few percent.
2.2.2. Issuers are better informed
In case issuers are more informed than investors, a lemons problem might exist; investors do not have enough information to distinguish between firms of high quality and firms of low quality. Therefore, investors would only be willing to pay the average-quality price. For that price, only worse than average-average-quality firms would want to sell (Ritter and Welch, 2002). This problem would drive high-quality firms out of the market, leaving only ‘’lemons’’ in the market. However, the high-quality firms can, to distinguish themselves from low-quality firms, signal their type by underpricing their IPO (Allen and Faulhaber, 1989). This can be a credible signal, since investors know that only high-quality firms will be able to recoup the cost of underpricing the initial
offering.
Welch (1989) states that a higher price at a seasoned equity issue can
compensate high-quality firms for the low price at the initial public offering. Since the true quality of a firm may be discovered by the market between the IPO and the
seasoned equity offering, low-quality firms will not be able to imitate high-quality firms. Therefore, high-quality firms can deliberately underprice their IPO, to signal their quality to investors. The high-quality firms can then, after signaling their quality, raise additional capital under more favorable terms in the future. This is known as the signaling hypothesis.
In their empirical investigation of IPO returns and subsequent offerings, Jegadeesh, Weinstein and Welch (1993) find evidence that supports the signaling
7 hypothesis. They examine the relationship between the probability of a seasoned equity issue and underpricing of the IPO by estimating a logit regression. Their sample
includes 1985 IPOs in the 1980-1986 period. They find a positive relation between the degree of underpricing and the probability and size of a seasoned equity offering. However, they also find that returns in the 40-day period following the IPO date are strongly positively related to the probability and size of a seasoned equity offering. This indicates that that the initial return on the date of the IPO does not play a unique role with respect to the prediction of future equity offerings. Therefore, the evidence consistent with the signaling hypothesis can be considered to be weak.
Instead, Jegadeesh, Weinstein and Welch (1993) find an alternative hypothesis termed the market-feedback hypothesis. Hovakimian and Hutton (2010) state that market-feedback refers to the hypothesis that stock price changes can provide information about the profitability of the firm’s projects. In contrast to the signaling hypothesis, the market-feedback hypothesis is based on the assumption that the market is better informed. High returns in the period following the IPO encourages managers to increase the firm’s investment because high returns imply that, in the market’s view, the managers underestimated the marginal return of the firm’s projects. In line with the market-feedback hypothesis, Jegadeesh, Weinstein and Welch (1993) find that returns in the period following the IPO (aftermarket returns) are strongly positively related to the probability of a seasoned equity issue. This indicates that firms use the information provided by stock price changes to decide whether or not to increase the firm’s
investment.
In line with the findings of Jegadeesh, Weinstein and Welch (1993), both Ritter and Welch (2002) and Francis et al. (2010) state that the evidence in favor of signaling theories is weak and at best mixed. However, Francis et al. (2010) argue that it is possible that not all issuers are willing to apply a signaling strategy. It is not necessary for high-quality firms to apply a signaling strategy in case they are not wealth constraint to a certain extent. In addition, they state that firms facing higher information
asymmetry are more likely to use a signaling strategy. They point out that the inability of researchers to identify firms that value underpricing as a signaling device can be the problem that causes the evidence for the signaling hypothesis to be weak.
8 than domestic IPOs. Therefore, they consider it more likely for foreign IPOs, everything else being equal, to engage in a signaling strategy. For that reason, they study a group of foreign IPOs in U.S. capital markets. Their final sample includes 413 foreign IPOs in the U.S. market and covers the period 1985 through 2000.
The foreign IPOs in their sample are categorized based on the level of financial market integration in the home country of the firms. The level of financial market integration refers to the degree to which countries are integrated into world capital markets (Bekaert and Harvey, 1995). Francis et al (2010) define markets as integrated if a country has a fully integrated financial market. They define all other markets as segmented. The two resulting groups of foreign firms are firms from integrated markets and firms from segmented markets. Firms from segmented markets have less access to foreign capital, and thus face more difficulties in raising capital. In addition, firms from segmented markets experience higher information asymmetry. Therefore they expect firms from segmented markets to use a signaling strategy.
To test whether the firms in their sample actually use a signaling strategy, Francis et al. (2010) examine the relation between the probability of a seasoned equity issue and the degree of underpricing by estimating a logit regression similar to
Jegadeesh, Weinstein and Welch (1993). A positive relationship would indicate that firms use IPO underpricing as a signaling strategy. In line with their expectations, firms from segmented markets, that experience a high degree of underpricing, are more likely to issue seasoned equity. However, the probability of a second equity issue for firms from financially integrated markets is strongly related to the returns in the period following the IPO. This is in line with the market-feedback hypothesis (Jegadeesh, Weinstein and Welch, 1993). Based on these results, the signaling hypothesis can be seen as an important determinant of IPO underpricing for firms that have a greater need to access external capital and face higher information asymmetry (Francis et al. (2010)).
2.3 Concluding conclusion of the literature
There are many different theories proposed in order to explain underpricing of IPOs. Ibbotson (1975) and Tinic (1988) suggest that underpricing might function as a form of insurance against potential lawsuits. In addition, Krigman, Shaw and Womack (1999) state that underpricing of the IPO increases the after-market liquidity. These theories
9 are based on the assumption that parties involved in the IPO process have symmetric information. However, most theories developed to explain underpricing are based on the assumption of asymmetric information.
Assuming some investors are more informed, Rock (1986) states that issuing firms have to offer the shares at a discount in order to attract uninformed investors. Based on the same assumption, Welch (1992) argues that early investors might be imitated by subsequent investors when an IPO is sold sequentially.
In contrast, Jegadeesh, Weinstein and Welch (1993) assume that the issuer is better informed. They argue that issuers try to signal their quality by underpricing the IPO, known as the signaling hypothesis. However, the evidence in favor of the signaling hypothesis is considered to be weak and at best mixed. Francis et al. (2010) state that the inability of researchers to identify firms that are likely to engage in a signaling strategy is the reason for the lack of strong evidence. They find that some firms actually value underpricing as a signaling device and that the degree of underpricing is
positively related to the probability of a seasoned equity issue for those firms. However, for other firms, the aftermarket returns are better in predicting SEO activities. This is in line with the market-feedback hypothesis (Jegadeesh, Weinstein and Welch, 1993).
3. Methodology and hypotheses 3.1. Methodology
The objective of this thesis is to test whether firms use IPO underpricing as a signaling device. This is done by examining the relationship between the probability of a SEO within three years of the IPO and the degree of underpricing of the IPO for firms going public in the U.S. market in the 2005-2010 period. Since Francis et al. (2010) point out it is important to identify firms that actually value underpricing as a signaling device, the sample is split into two subsamples; domestic IPOs and foreign IPOs. In addition, the alternative market feedback hypothesis (Jegadeesh, Weinstein and Welch, 1993) is tested. This is done by examining the relationship between the probability of a SEO within three years of the IPO and the returns in the post-IPO period (aftermarket returns).
10 To examine whether the probability of a seasoned equity is related to either the degree of underpricing or the aftermarket returns, the following logit model (similar to Francis et al. (2010)) is estimated:
𝑃𝑖 = 1+𝑒𝑒𝑘𝑖𝑘𝑖,
Where,
𝑘𝑖 = 𝛼 + 𝛽1∙ 𝑈𝑛𝑑𝑒𝑟𝑝𝑟𝑖𝑐𝑖𝑛𝑔𝑖+ 𝛽2∙ 𝐴𝑀𝑟𝑒𝑡1𝑖 + 𝛽3∙ 𝐴𝑀𝑟𝑒𝑡2𝑖+ 𝛽4∙ 𝑙𝑛𝐼𝑃𝑂𝑠𝑖𝑧𝑒𝑖 + 𝛽5∙ 𝐹𝑟𝑎𝑐𝑡𝑖𝑜𝑛𝑆𝑜𝑙𝑑𝑖 + 𝛽6−10 ∙ 𝑌𝑒𝑎𝑟𝑖+ 𝛽11−16∙ 𝐼𝑛𝑑𝑢𝑠𝑡𝑟𝑦𝑖 + 𝑢𝑖,
Where, 𝑃𝑖 is the probability the firm issues seasoned equity within three years of the IPO. The three independent variables of main interest are Underpricing, AMret1 and AMret2. Underpricing is defined as the percentage change of the price of the share on the offer date:
𝑈𝑛𝑑𝑒𝑟𝑝𝑟𝑖𝑐𝑖𝑛𝑔 = (𝑝1−𝑝𝑜
𝑝0 ) ∙ 100.
AMret1 (Aftermarket Returns 1) is defined as the cumulative abnormal return for trading days 1-20 following the IPO date. Similarly, AMret2 (Aftermarket Returns 2) is defined as the cumulative abnormal return for trading days 21-40 following the IPO date. Abnormal returns are estimated as the difference between the actual return and the predicted return. Predicted returns are measured as beta times the market return. Beta is estimated from a market model regression fitted over days 41-140 following the IPO date. As a market proxy, either the NYSE index or the NASDAQ index is used,
depending on which exchange the IPO is listed.
lnIPOsize and Fraction Sold are included as control variables. lnIPOsize is
defined as the natural logarithm of the proceeds (in million USD) of the IPO. lnIPOsize is included as control variable since the amount of capital that is raised at the IPO is likely to influence SEO activities (Jegadeesh, Welch and Weinstein, 1993). Francis et al (2010) and Jegadeesh, Weinstein and Welch (1993) find a positive and significant relationship between lnIPOsize and the probability of a seasoned equity.
11 Fraction Sold at the IPO measures the proportion of the firm sold at the IPO and is calculated as the ratio of the number of shares sold at the IPO and the total number of shares outstanding following the IPO. Fraction Sold at the IPO is included as a control variable since firms that sell only a small fraction of their total shares at the IPO are more likely to issue seasoned equity (Francis et al. (2010)). The lower the fraction sold at the IPO, the higher the probability of a seasoned equity offering is expected to be.
Various industry and year dummies are included to control for potential
differences that are caused by the industry the firm is in or the year the IPO takes place.
3.2. Hypotheses
Since the sample is split up into two subsamples (foreign and domestic IPOs),
hypotheses are formed for the full sample and the two different subsamples separately. First, the hypotheses are given for the full sample. Second, the hypotheses for the subsample of domestic IPOs are given. Finally, the hypotheses are given for the subsample of foreign IPOs.
3.2.1 Full sample of IPOs
Francis et al. (2010) find that only some firms value underpricing as a signaling device. They argue that the lack of strong evidence in favor of the signaling hypothesis is caused by the inability of researchers to identify firms that are likely to use a signaling strategy. In the full sample, no difference is made between firms that are likely to value
underpricing as a signaling device and firms that are not. Therefore, the relationship between underpricing and the probability of a SEO is expected to be weak for the full sample. This would result in an insignificant coefficient of underpricing.
Jegadeesh, Welch and Weinstein (1993) find that the returns in the period immediately following the IPO are better able to predict SEO activities. They find a positive and significant relationship between aftermarket returns and the probability of a SEO. Based on their results, a positive relationship is expected between the
aftermarket returns and the probability of a SEO for the full sample. Therefore, the coefficients of aftermarket returns 1 and 2 are expected to be significant and positive.
12
3.2.2. Domestic IPOs
Although the evidence Jegadeesh, Weinstein and Welch (1993) find for the signaling hypothesis is considered to be weak, they do find a positive relationship between the degree of underpricing and the probability of a seasoned equity offering. However, Francis et al. (2010) find an insignificant relationship between underpricing and the probability of second equity offering for firms from financially integrated markets. Instead, Francis et al. (2010) find that those firms’ returns in the period following the IPO are better able to predict subsequent seasoned equity offerings. Bekaert, Harvey and Lundblad (2005) state that the U.S. financial market is fully liberalized and the U.S. can therefore be considered to have a financially integrated market.
Based on the findings of Francis et al. (2010), the market-feedback hypothesis (Jegadeesh, Welch and Weinstein, 1993) is expected to be a better predictor of
subsequent SEOs for domestic firms. A positive relationship between aftermarket returns and the probability of a seasoned equity offering is therefore expected for domestic IPOs, whereas the relationship between underpricing and the probability of a SEO is expected to be weak or nonexistent for domestic IPOs. Therefore, the coefficients of aftermarket return 1 and 2 are expected to be positive and significant and the
coefficient of underpricing is expected to be not significantly different from zero.
3.2.3. Foreign IPOs
Francis et al. (2010) state the lack of empirical support for the signaling hypothesis is caused by the inability of researchers to identify the group of firms that actually value underpricing as a signaling device. Firms with higher information asymmetry are more likely to use a signaling strategy. Furthermore, they state that foreign firms are more likely to engage in a signaling strategy than domestic firms.
Because it is more likely for foreign firms to use a signaling strategy, the
signaling hypothesis is expected to be a better predictor for their SEO activities than the market-feedback hypothesis. Therefore, a positive relationship between IPO
underpricing and the probability of a seasoned equity issue is expected for foreign IPOs. This would results in a positive and significant coefficient of underpricing. In contrast, the relationship between aftermarket returns and the probability of a SEO is expected to be weak or nonexistent for foreign IPOs, leading to an insignificant or slightly significant positive coefficient of aftermarket returns 1 and 2.
13
4. Data and descriptive statistics 4.1. Data construction
The sample includes firms that went public on the either the New York Stock Exchange or the NASDAQ Stock Market between 2005 and 2010. SEOs within three years of the IPO are included. Moreover, only the first SEO is included for each firm.
All data concerning the IPOs and the SEOs is provided by the Thomson One database. This database includes the IPO date, offer price, first-day return, year the IPO took place in, exchange where the IPO is listed, proceeds of the IPO, industry SIC codes for the firms, the fraction of the total number of shares outstanding that is offered at the IPO (fraction sold) and the country of origin for each firm. With respect to SEOs, the date and the proceeds of the SEOs are provided by the Thomson One database.
In line with Boulton, Smart and Zutter (2011) financial firms, rights offerings, unit offerings, closed-end funds, investment trusts, limited partnerships and depository receipts are excluded. This results in an initial sample of 556 initial public offerings extracted from the Thomson One database.
Stock price data for the firms included in the sample is obtained from
DataStream for 141 days following the IPO date. In addition, DataStream provides daily values of the two market proxies, the NYSE composite index and the NASDAQ composite index.
Finally, offerings are excluded for which any data that is needed is missing. The variables for which a substantial amount of data is missing are; first-day return, stock prices in the 141 day period following the IPO and the fraction of the total number of shares outstanding that is offered at the IPO. The final sample consists of 389 IPOs.
Table 1. Distribution of IPOs and SEOs by domestic/foreign and exchange Table 1 shows the distribution of IPOs on the NYSE and NASDAQ from 2005 to 2010 and SEOs issued within 3 years of the IPO. Since only some firms may actually value underpricing as a signaling device, the sample is separated into two subsamples based on whether the firm is from the U.S. (domestic) or from outside the U.S. (foreign).
No. of IPOs No. of SEOs
Domestic firms 351 143
Foreign firms 38 10
14
4.2. Descriptive statistics
Table 1 gives on overview of the number of IPOs and SEOs for the full sample and the two different subsamples separately. Of the 389 total IPOs in the sample, 38 (9.77%) are foreign. Of the 389 firms, 153 (39.33%) issued seasoned equity within three years of the IPO. 10 (26.32%) out of 38 foreign firms issued seasoned equity within three years of the IPO. With respect to domestic firms, 143 (40.74%) out of 351 firms issued seasoned equity within three years of the IPO.
Table 2 presents descriptive statistics for the IPOs. Underpricing is calculated as (P₁ - P₀)/P₀ x 100, where P₁ is the closing price on the first day of trade and P₀ is the initial offering price. Consistent with the existing literature, firms from both subsamples experience underpricing on average. For the full sample, the average underpricing is 12.57%. The subsample of foreign IPOs has an average underpricing of 11.85% and the subsample of domestic IPOs has an average underpricing of 12.65%. There is a lot of variation in underpricing among firms, as can be seen from the standard deviation reported in table 2. For the full sample, the standard deviation is around 21%.
Aftermarket Returns 1 and Aftermarket Returns 2 are the cumulative abnormal returns for days 1-20 following the IPO and days 21-40 following the IPO. Consistent with the findings of Jegadeesh, Welch and Weinstein (1993), cumulative abnormal returns in the 40 days following the IPO are close to 0 on average for the samples. The average aftermarket returns are 0.23% and 0.79% for the full sample. Domestic firms experience average cumulative returns in the two 20-day periods following the IPO of 0.36% and 0.52%. For the foreign sample, these returns are -0.97% and 3.21% on
average. The positive average returns of foreign firms in the second period following the IPO is caused by a few firms with high abnormal returns in this period, raising the
average.
In addition, table 2 shows the average proceeds of the IPOs for the different samples. Proceeds is defined as the amount of capital a firm raises with its IPO. Finally, statistics of the fraction of the firm sold at the IPO are given in table 2. The Fraction Sold at the IPO is calculated as the number of shares sold at the IPO divided by the total number of shares outstanding following the IPO.
15
5. Analysis
5.1. Empirical results
Table 3 presents the results of the estimated logit model. The model examines the relation between the probability of a seasoned equity issue within three years of the IPO (dependent variable) and stock returns at and around the IPO date. The three main independent variables of interest are Underpricing, Aftermarket Return 1 and
Aftermarket Return 2. In addition, the natural logarithm of the size of the IPO and the fraction of the firm that is sold at the IPO are included as control variables. In some of the models, industry and year fixed effects are included to control for potential differences across industries and years. Results are presented for the full sample
Table 2. Descriptive statistics of the IPOs
Table 2 presents descriptive statistics for foreign and domestic IPOs separately, and for all IPOs combined.
Underpricing is defined as (P₁ - P₀)/P₀ x 100, Where P₁ is the first-day closing price and P₀ is the initial offering price of the IPO. Aftermarket return 1 is the cumulative abnormal return for the first 20 days following the IPO date. Similarly, Aftermarket return 2 is the cumulative abnormal return for days 21-40 following the IPO date. Proceeds is the amount of capital raised through the IPO. Finally, Fraction Sold at the IPO is the proportion of the firm that is being sold at the IPO, calculated as number of shares sold at the IPO divided by the total number of shares outstanding following the IPO.
All IPOs (n = 389)
Variables Mean Std. Dev Min. Max.
Underpricing 12.5677 21.0083 -91.18 138.777
Aftermarket Returns 1 0.2283 13.3293 -48.4044 44.1413
Aftermarket Returns 2 0.7863 16.09 -57.0337 60.12831
Proceeds (million USD) 181.6422 221.8245 5.775 1875.5
Fraction Sold at the IPO 0.3057 0.1541 0.0761 1
Domestic IPOs (n = 351)
Variables Mean Std. Dev Min. Max.
Underpricing 12.6460 32.0301 -91.18 138.777
Aftermarket Returns 1 0.3580 13.0591 -48.4044 44.1413
Aftermarket Returns 2 0.5239 16.2303 -57.0337 60.1283
Proceeds (million USD) 182.4837 227.8508 5.775 1875.5
Fraction Sold at the IPO 0.3001 0.1426 0.0761 1
Foreign IPOs (n = 38)
Variables Mean Std. Dev Min. Max.
Underpricing 11.8449 21.0712 -7.17 88.93
Aftermarket Returns 1 -0.9703 15.7575 -30.6408 38.3646
Aftermarket Returns 2 3.2104 14.0667 -25.8457 48.1570
Proceeds (million USD) 173.8689 157.5861 6.667 671.698
16 (models 1-4), the subsample of domestic IPOs (models 5-8) and the subsample of
foreign IPOs (models 9 and 10).
5.1.1. Full Sample
Table 3 presents the results of the estimated logit model. Results for the full sample (domestic and foreign IPOs) are reported in column 1-4. Francis et al. (2010) state that only some firms value underpricing as a signaling device. Furthermore, they argue that firms that face higher information are more likely to use a signaling strategy. Since in the full sample no difference is made between firms that are likely to use a signaling strategy and firms that are not, a weak or nonexistent relation between underpricing and the probability of a SEO is expected.
In line with expectations based on the existing literature, the coefficient of
underpricing is insignificant in all four models for the full sample. There is no significant relation between the degree of underpricing and the probability of a SEO in all four models. The results for the full sample do not provide any evidence for the signaling hypothesis.
Instead, the returns of the stock in the period following the IPO are better predictors of SEO activities. The coefficient of Aftermarket returns 1 is significant (at a significance level of 10%) and positive for all four models. The coefficient of
Aftermarket Returns 2 is only significant (10% significance level) in models 1 and 3. After controlling for potential differences across industries and years (model 2 and 4), the coefficient of Aftermarket Returns 2 remains positive, but is insignificant. The results with respect to Aftermarket Returns 1 provide evidence in favor of the market-feedback hypothesis (Welch, Weinstein and Welch, 1993), which states the returns in the period following the IPO are better in predicting subsequent SEOs. In more detail, the higher the aftermarket returns, the higher the probability of a SEO within three years of the IPO. However, the relationship between Aftermarket Returns 1 and the probability of a SEO is only significant at a significance level of 10%. This indicates that the evidence provided in favor of the market-feedback hypothesis is weak for the full sample.
As expected, for all four models, the effect of the natural logarithm of the size of the IPO is significant and positive (1% significance level). In models 3 and 4, Fraction Sold at the IPO is included as a control variable. Based on existing literature, firms that
17
Table 3. Logit regression estimates of the probability of SEOs for different (sub)samples.
This table presents logit regression estimates of the relation between the probability of a seasoned equity issue within three years of the IPO (dependent variable) and stock returns at and around the IPO date (independent variables of main interest) in the 2005-2010 period. Estimates are given for the full sample (column 1-4), the subsample of domestic IPOs (column 5-8) and the subsample of foreign IPOs (column 9-10). The dependent variable is a dummy variable, taking a value of 0 if a firm does not issue seasoned equity within three years of the IPO, and a value of 1 if it issues seasoned equity within three years of the IPO. The independent variables of main interest are Underpricing, Aftermarket Return 1 and Aftermarket Return 2. Underpricing is defined as (P₁ - P₀)/P₀ x 100, Where P₁ is the first-day closing price and P₀ is the initial offering price of the IPO. Aftermarket return 1 is the
cumulative abnormal return for the first 20 days following the IPO. Similarly, Aftermarket return 2 is the cumulative abnormal return for days 21-40 following the IPO. LnIPOsize is the natural logarithm of the proceeds of the IPO. Fraction Sold at the IPO is the proportion of the firm that is being sold at the IPO, calculated as number of shares sold at the IPO divided by the total number of shares outstanding following the IPO. In some models, industry fixed effects and year fixed effects are included. ***, **, * indicate significance at 1%, 5% and 10%, respectively. p-values are reported in parentheses.
All IPOs Domestic IPOs Foreign IPOs
Variables (1) (2) (3) (4) (5) (6) (7) (8) (9) (10) Constant -2.299*** (0.000) -2.185*** (0.006) -2.239*** (0.000) -2.049** (0.011) -2.130*** (0.001) -1.706** (0.022) -2.014*** (0.002) -1.516** (0.046) -5.882* (0.076) -6.272* (0.065) Underpricing 0.001 (0.775) (0.975) 0.000 (0.875) 0.001 (0.873) -0.000 (0.460) 0.004 (0.605) 0.003 (0.578) 0.003 (0.752) 0.002 (0.144) -0.075 (0.205) -0.065 Aftermarket Returns 1 1.517* (0.071) (0.060) 1.619* (0.073) 1.504* (0.061) 1.609* 2.097** (0.022) 2.247** (0.017) 2.054** (0.024) 2.188** (0.020) (0.146) -6.459 (0.119) -7.326 Aftermarket Returns 2 1.245* (0.069) (0.144) 1.033 (0.068) 1.235* (0.155) 1.009 1.440** (0.044) (0.112) 1.175 1.432** (0.046) (0.123) 1.144 (0.737) 1.505 (0.590) 2.527 LnIPOsize 0.380*** (0.002) 0.368*** (0.004) 0.125*** (0.001) 0.388*** (0.003) 0.350*** (0.007) 0.316** (0.018) 0.385*** (0.004) 0.361*** (0.009) (0.112) 1.000 (0.145) 0.922
Fraction Sold at the IPO -0.465
(0.518) (0.413) -0.639 (0.269) -0.914 (0.168) -1.224 (0.398) 1.652
Year fixed effects No Yes No Yes No Yes No Yes No No
Industry fixed effects No Yes No Yes No Yes No Yes No No
n 389 389 389 389 351 351 351 351 38 38
19 only sell a small fraction of their shares at the IPO, may be more likely to issue SEO. However, Fraction Sold at the IPO is insignificant. This indicates that the Fraction Sold at the IPO is not an important determinant of SEO activities.
5.1.2. Domestic IPOs
Results for the subsample of domestic IPOs are reported in column 5-8. As stated before, the U.S. has a financially integrated market. Based on existing literature, the market-feedback hypothesis is expected to be a better predictor of subsequent SEOs for firms from financially integrated markets.
For all four models (5 to 8), the coefficient of underpricing is insignificant. The degree of underpricing does not have a significant effect on the probability of a SEO for domestic firms. Again, the results do not provide any evidence for the signaling
hypothesis.
However, there is a significant and positive relationship between the probability of a SEO and the Aftermarket Returns 1 and 2. For models 5 and 7, the coefficients of both Aftermarket Returns 1 and Aftermarket Returns 2 are positive and significant at the 5% level. After controlling for differences across industries and years, the coefficient of Aftermarket Returns 2 remains positive but is not significant anymore. However, the coefficient of Aftermarket Returns 1 remains significant at the 5% level in all models. The relationship between Aftermarket Returns 1 and the probability of a SEO is stronger for the subsample (significant at a 5% significance level) of domestic firms than for the full sample (significant at the 10% level). This indicates that the higher the returns are in the first period following the IPO, the higher the probability of a SEO is. The results show that for domestic firms, Aftermarket Returns are indeed better predictors of SEO activities than the degree of underpricing.
5.1.3. Foreign IPOs
Results for the subsample of foreign IPOs are reported in columns 9 and 10. Industry and year dummies are not included for this subsample. If included, some of the industry and year dummies predict the probability of a seasoned equity issue perfectly, which leads to biased results. As argued earlier, foreign firms are more likely to value
underpricing as a signaling device. This would result in a positive relationship between underpricing and the probability of a seasoned equity issue.
20 However, there is no significant positive relationship between underpricing and the probability of a seasoned equity issue in models 9 and 10. The coefficients
(p-values) are -0.075 (0.144) and -0.065 (0.205) in models 9 and 10, respectively. The relationship is negative but insignificant. These results show that foreign firms are not more likely to engage in a signaling strategy per se. Francis et al. (2010) find a positive and significant relationship between underpricing and the probability of a seasoned equity offering for foreign firms from segmented markets only. Their findings show that some firms actually value underpricing as a signaling device. However, this research shows that foreign firms in general not value underpricing as a signaling device.
In both models (9 and 10) the coefficients of Aftermarket Returns 1 and 2 are insignificant. This is in line with the findings of Francis et. Al (2010). They do not find a significant relationship between Aftermarket Returns 1 and 2 in the full sample and subsample of firms from segmented markets. For foreign firms, no evidence is found for the signaling hypothesis, neither for the market-feedback hypothesis.
5.2 Robustness check
As a robustness check, additional logit models are estimated to test whether the results hold under different specifications of the model. In contrast to the previous section, all models estimated in this section include all 389 IPOs. Instead of splitting the sample up into different subsamples, interaction effects are included to allow for differences between domestic IPOs and Foreign IPOs. Table 4 presents the results of the four estimated models.
The dummy variable foreign takes a value of 0 in case of a domestic IPO and a value of 1 in case of a foreign IPO. The coefficient of this variable indicates whether, all else being equal, the probability of a seasoned equity is higher (indicated by a positive coefficient) or lower (indicated by a negative coefficient) for foreign firms. The
variables Foreign * Underpricing, Foreign * Aftermarket Returns 1 and Foreign * Aftermarket Returns 2 are interaction terms. Their coefficients indicate whether there is a difference with respect to the effect of independent variables between domestic IPOs and foreign IPOs. In more detail, the coefficients of Underpricing, Aftermarket Returns 1 and Aftermarket are estimates for domestic IPOs. The coefficients of the interaction effects, in turn, indicate whether the effect of the independent variables Underpricing, Aftermarket Returns 1 and Aftermarket Returns 2 is different for foreign
21 firms. A positive coefficient of an interaction term indicates that the relationship
between the probability of a SEO and the independent variable of interest is more positive (or less negative) for foreign IPOs than for domestic IPOs.
In all four models, the coefficient of Underpricing is not significantly different from 0. The coefficient of Aftermarket Returns 1 is significant at the 5% level in all four models, whereas the coefficient of Aftermarket Returns 2 is significant at the 5% level in models 1 and 3. After controlling for industry and year fixed effects, the relationship
Table 4. Logit estimates of the probability of SEOs with interaction effects
This table presents logit regression estimates of the relation between the probability of a seasoned equity issue and stock returns at and around the IPO date in the 2005-2010 period. In all four models, the dependent variable is a dummy variable taking a value of 1 if a firm issues seasoned equity, and a value of 0 otherwise. Underpricing is defined as (P₁ - P₀)/P₀ x 100, where P₁ is the first-day closing price and P₀ is the initial offering price of the IPO.
Aftermarket return 1 is the cumulative abnormal return for the first 20 days following the IPO. Similarly, Aftermarket return 2 is the cumulative abnormal return for days 21-40 following the IPO. Fraction sold is the proportion of the firm that is being sold at the IPO. LnIPOsize is defined as the natural logarithm of the proceeds of the IPO. Foreign is a dummy variable taking a value of 1 in case of a foreign IPO and a value of 0 otherwise. Foreign * Underpricing, Foreign * Aftermarket Returns 1 and Foreign * Aftermarket Returns 2 are interactions between Foreign and Underpricing, Foreign and Aftermarket Returns 1 and Foreign and Aftermarket Returns 2, respectively. In model 2 and 4, industry and year fixed effects are included. ***. **, * indicate significance at 1%, 5% and 10%, respectively. P-values are reported in parentheses.
All IPOs Variables (1) (2) (3) (4) Constant -2.286*** (0.000) -2.124** (0.011) -2.227*** (0.000) 1.977** (0.019) Underpricing 0.004 (0.473) (0.675) 0.002 (0.536) 0.003 (0.753) 0.002 Aftermarket Returns 1 2.100** (0.022) (-0.017) 2.237** 2.077** (0.023) 2.213** (0.018) Aftermarket Returns 2 1.442** (0.044) (0.096) 1.235* 1.437** (0.045) (0.103) 1.214 LnIPOsize 0.382*** (0.002) 0.359*** (0.007) 0.401*** (0.002) 0.382*** (0.005) Foreign -0.414 (0.392) (0.292) -0.560 (0.473) -0.353 (0.361) -0.487 Foreign * Underpricing -0.070 (0.135) (0.125) -0.077 (0.124) -0.073 (0.107) -0.082 Foreign * Aftermarket Returns 1 -7.189*
(0.065) -8.253** (0.048) -6.995* (0.072) -8.060* (0.053) Foreign * Aftermarket Returns 2 -1.321
(0.721) (0.756) -1.187 (0.688) -1.502 (0.712) -1.433
Fraction Sold at the IPO -0.472
(0.524) (0.358) -0.734
Year fixed effects No Yes No Yes
Industry fixed effects No Yes No Yes
n 389 389 389 389
22 becomes insignificant in model 4 and only significant at the 10% level in model 2. These results are in line with the results of the subsample of domestic IPOs reported in table 3 and thus confirm that the higher the returns in the period following the IPO are, the higher the probability of a SEO is for domestic IPOs.
The coefficient of Foreign is insignificant in all four models. This indicates that, all else being equal, the probability of a SEO is not significantly different for foreign IPOs compared to domestic IPOs. The coefficients of the interaction terms Foreign *
Underpricing, Foreign * Aftermarket Returns 1 and Foreign * Aftermarket Returns 2 are negative in all four models. However, the coefficients of Foreign * Underpricing and Foreign * Aftermarket Returns 2 are insignificant.
The coefficient of Foreign * Aftermarket Returns 1 is negative and significant in all four models. This indicates that the relationship between Aftermarket Returns 1 and the probability of a SEO is less positive or even negative for foreign firms. This is in line with the results for foreign IPOs reported in table 3, where the coefficient of
Aftermarket returns 1 is insignificant.
The results of the robustness check confirm that returns in the period
immediately following the IPO are better in predicting SEO activities than the degree of underpricing is for domestic IPO firms only. For foreign firms, neither the degree of underpricing nor the returns in the period following the IPO is significantly related to the probability of a SEO.
6. Conclusion and discussion 6.1. Conclusion
Using a sample of domestic and foreign IPOs in the U.S., the relation between the probability of a SEO and underpricing is examined in order to test the signaling
hypothesis. In addition, the relation between aftermarket returns and the probability of a SEO is examined to test the alternative market-feedback hypothesis. The different hypotheses are tested for foreign firms and domestic firms separately.
Jegadeesh, Weinstein and Welch (1993) find only weak evidence for the signaling hypothesis. Instead, they formulate an alternative hypothesis; the market-feedback hypothesis. This hypothesis predicts that returns in the period immediately following the IPO are better predictors of seasoned equity offering activities. However, Francis et
23 al. (2010) find that some firms actually value underpricing as a signaling device, but that it is important to identify firms that are likely to do so first. They state that firms that experience more information asymmetry are more likely to engage in a signaling strategy.
In line with the findings of Jegadeesh, Weinstein and Welch (1993) and Francis et. al. (2010), aftermarket returns are found to be better in predicting subsequent equity offerings than underpricing is for domestic firms. The results for domestic firms provide evidence in favor of the market-feedback hypothesis. This indicates that managers of domestic firms do not deliberately leave money on the table at their IPO. Instead, they use the information provided by price changes in the post-IPO period to decide whether or not to increase the firm’s investment by raising additional capital through seasoned equity offerings.
For foreign firms, no evidence in favor of the signaling hypothesis is provided either. This indicates that foreign firms do not voluntarily underprice their IPO in order to raise additional under more favorable terms in the future. Therefore, it can be
concluded that foreign firms are not more likely to engage in a signaling strategy per se. This is in contrast with the findings of Francis et al (2010). They state that foreign firms are more likely to use underpricing as a signaling device. In addition, there is no positive relation between aftermarket returns and the probability of a seasoned equity issue. This indicates that foreign firms do not use the information provided by returns in the post-IPO to decide whether or not to issue seasoned equity. This is in line with the findings of Francis et al. (2010).
The results indicate that neither foreign firms nor domestic firms in general voluntarily leave money on the table to raise additional capital in the future under more favorable terms. However, domestic firms use the information provided by stock price changes in the period following the IPO to decide whether or not to raise additional capital through seasoned equity issues.
6.2. Discussion
The lack of evidence for the signaling hypothesis for foreign firms might be caused by the number and the type of firms that are included in the subsample of foreign IPOs. After excluding firms for which data was missing, the number of foreign IPOs decreased from 70 to 38. Based on the classification of the International Monetary Fund (2014),
24 only 14 (37%) out of the 38 firms included in the final subsample of foreign firms can be classified as coming from a country with a developing economy.
Firms from developing countries are likely to experience a high degree of information asymmetry. In addition, they are likely to face more difficulties in raising external capital. Therefore, firms from developing countries are expected to be more likely to engage in a signaling strategy. The overrepresentation of firms from developed
countries (63%) in the final subsample of foreign firms, after excluding firms for which data was missing, could therefore have a negative effect on the relationship between underpricing and the probability of a SEO for foreign firms.
Foreign firms are not more likely to engage in a signaling strategy than domestic firms per se. However, it can be still be the case that some firms engage in a signaling strategy. In order to examine this, firms have to be identified that possibly experience a high degree of information asymmetry. After identifying those firms, the signaling hypothesis can be tested again for those firms only.
References
Allen, F., & Faulhaber, G., R. (1989). Underpricing and firm prospects. Journal of
Financial Economics, 23, 303-323
Amihud, Y., Hauser, S., & Kirsch, A. (2003). Allocations, adverse selection, and cascades in IPOs: Evidence from the Tel Aviv Stock Exchange. Journal of Financial
Economics, 68(1), 137-158
Bekaert, G., & Harbey, C. R. (1995). Time-varying world market integration. Journal of
Finance, 50(2), 403-444
Bekeart, G., Harvey, C. R., & Lundblad, C. (2005). Does financial liberalization spur growth? Journal of Financial Economics, 77(1), 3-55
Beneveniste, L. M., & Spindt, P. A. (1989). How investment bankers determine the offer price and allocation of new issues. Journal of Financial Economics, 24(2), 343-361 Booth, J. R., & Chua, L. (1996). Ownership dispersion, costly information, and IPO
underpricing. Journal of Financial Economics, 41, 291-310
Boulton, T. J., Smart, S. B., & Zutter, C. J. (2011). Earnings quality and international IPO underpricing. The Accounting Review, 86(2), 484-505
Brennan, M. J., & Franks, J. (1997). Underpricing, ownership and control in initial public offerings of equity securities in the UK. Journal of Financial Economics, 45(3), 391-413
Cornelli, F., & Goldreich, D. (2003). Bookbuilding: How Informative Is the Order Book? Journal of Finance, 58(4), 1415-1443
Drake, P. D., & Vetsuypens, M. R. (1993). IPO underpricing and insurance against legal liability. Financial Management, 22(1), 64-73
Ellul, A., & Pagano, M. (2006). IPO underpricing and after-market liquidity. The Review of
25 Francis, B. B., Hasan, I., Lothian, J. R., & Sun, X. (2010). The signaling hypothesis
revisited: evidence from foreign IPOs. Journal of Financial and Quantitative
Analysis, 45(1), 81-106
Garfinkel, J. A. (1993). IPO underpricing, insider selling an subsequent equity offerings: Is underpricing a signal of quality? Financial Management, 22(1), 74-83
Hovakimian, A., & Hutton, I. (2010). Market feedback and equity issuance: evidence from repeat equity issues. Journal of Financial and Quantitative Analysis, 45(3), 739-762
Ibbotson, R. G. (1975). Price performance of common stock new issues. Journal of
Financial Economics, 2(3), 235-272
International Monetary Fund (2014). Country composition of WEO groups. Retrieved from http://www.imf.org/external/pubs/ft/weo/2014/02/weodata/groups. htm
Jegadeesh, N., Weinstein, M., & Welch, I. (1993). An empirical investigation of IPO returns and subsequent equity offerings. Journal of Financial Economics, 34, 153-175
Keloharju, M. (1993). The winner's curse, legal liability, and the long-run price
performance of initial public offerings in Finland. Journal of Financial Economics,
34(2), 251-277
Krigman, L., & Shaw, W. H., & Womack, K. L. (1999). The Persistence of IPO Mispricing and the Predictive Power of Flipping. Journal of Finance, 54(3), 1015-1044
Liu, X., & Ritter, J. R. (2011). Local underwriter oligopolies and IPO underpricing. Journal
of Financial Economics, 102(3), 579-601
Ljungqvist, A., & Wilhelm, W. J. (2005). Does prospect theory explain IPO market behaviour? Journal of Finance, 60(4), 1759-1790
Ritter, J. R., & Welch, I. (2002). A review of IPO activity, pricing and allocations. Journal
of Finance, 57(4), 1795-1828
Rock, K. (1986). Why new issues are underpriced. Journal of Financial Economics, 15, 187-212
Tinic, S. M. (1988). Anatomy of initial public offerings of common stock. Journal of
Finance, 43(4), 789-822
Welch, I. (1989). Seasoned offerings, imitation costs, and the underpricing of initial public offerings. Journal of Finance, 44(2), 421-449
Welch, I. (1992). Sequential sales, learning, and cascades. Journal of Finance, 47(2), 695-732
