• No results found

The long-run performance of Initial Public Offerings, and their relationship with IPO characteristics

N/A
N/A
Protected

Academic year: 2021

Share "The long-run performance of Initial Public Offerings, and their relationship with IPO characteristics"

Copied!
54
0
0

Bezig met laden.... (Bekijk nu de volledige tekst)

Hele tekst

(1)

The long-run performance of Initial

Public Offerings, and their

relationship with IPO characteristics

Institution: University of Amsterdam, Amsterdam Business School Course: MSc Business Economics, Finance track

Type: Master Thesis

Title: The long-run performance of Initial Public Offerings, and their relationship with IPO characteristics

Author: Ritesh Bisseswar (5744423) Date: 07-07-2015

(2)

Statement of Originality

This document is written by Student Ritesh Bisseswar 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.

(3)

Abstract

This study investigates the long-run performance of initial public offerings (IPOs), and the determinants of long-run returns of IPOs. Using a sample of 1925 IPOs on NASDAQ between 1980 and 2010, this paper investigates whether IPOs show abnormal performance in the long-run compared to various benchmarks over three years following the offering. Furthermore, this paper examines the effect of firm maturity (age) at time of the IPO and several other IPO characteristics on long-run performance.

To measure long-run performance, cumulative abnormal returns (CAR), buy-and-hold abnormal returns (BHAR) and wealth relatives (WR) are calculated against various market indices. When measuring long-run performance using CARs, IPOs do not underperform over three years, in fact the IPOs in the sample show slight overperformance. When measuring long-run performance using BHARs and WRs, IPOs appear to underperform the benchmarks three years after the offering. Furthermore, it appears that younger IPOs show larger underperformance after three years, while the IPOs in more recent years over performed on average. These results suggest that one must be careful in concluding whether IPOs underperform or over perform in the long-run. It appears that the findings of long-run abnormal IPO performance is sensitive to the sample, sample period, benchmarks used, and the measurement techniques used to evaluate long-run performance.

To identify the effects of firm age and IPO characteristics on long-run returns, a multiple regression model is used with the three year raw IPO returns as the dependant variable. The results of the regression analysis surprisingly suggest that firm maturity has a negative effect on long-run IPO returns, indicating that younger companies should perform better in the long-run. However, as the regression coefficient of firm age is not statistically significant, firm age therefore is not an appropriate indicator of long-run IPO performance. The regression models furthermore show that initial returns and venture backing might be better and more significant indicators of long-run performance.

In conclusion, this study evaluates the previous findings of long-run IPO performance. The initial evidence shows that IPOs appear to underperform in the long-run, and that younger companies underperform more on average. Furthermore, IPOs in more recent years did not underperform, indicating that the findings of underperformance in prior literature are not persistent when including more recent years in a sample. However, the significance of these results really depends on the considerations made in the research design; therefore this study fails to convincingly show that IPOs underperform in the long-run.

(4)

Table of Contents

Table of Contents ... 4

1 Introduction ... 5

2 Literature review and hypothesis development ... 8

2.1 The market for initial public offerings ... 8

2.2 Underpricing ... 9

2.2.1 Evidence for underpricing ... 9

2.2.2 Theories on underpricing ... 9

2.3 Long-run performance of IPOs ... 12

2.3.1 Evidence on long-run performance of IPO ... 12

2.3.2 Measurement issues in long-run performance ... 14

2.3.3 Factors Affecting long-run performance ... 15

2.4 Conclusion & Hypotheses development ... 22

3 Data ... 25

3.1 Data collection ... 25

3.2 Summary statistics... 25

3.2.1 IPO characteristics sort by year ... 25

3.2.2 IPO returns sort by year ... 27

3.2.3 IPO characteristics sort by age ... 29

3.3 Conclusion ... 30

4 Analysis of long-run IPO performance ... 31

4.1 Measures of long-run performance ... 31

4.2 Results ... 32

4.2.1 Monthly abnormal returns ... 32

4.2.2 3-year abnormal returns per age group ... 34

4.2.3 3-year abnormal returns per year ... 35

4.3 Conclusion ... 37

5 Determinants of IPO long-run performance ... 38

5.1 Regression design ... 38

5.2 Regression results ... 39

6 Summary and conclusions ... 41

Appendix ... 43

(5)

1 Introduction

An initial public offering (IPO) is the sale of a firm’s equity shares to investors on a public stock exchange for the first time. There are several reasons why companies decide to go public. One of the most important reasons of going public is to acquire capital. In addition to raising funds, IPOs create a market in which the founders and existing shareholders can sell their shares. A non-financial reason to go public is to increase publicity (Brau & Fawcett 2006). There are, however some disadvantages in conducting an IPO. Major disadvantages are the dilution of ownership and control, and high costs associated with going public (Ritter & Welch 2002).

The market performances of firms conducting an IPO have received much attention in prior literature. This performance is usually measured over two horizons, the short-run and long-run performance. The short-run performance is usually measured over a period of a few days to a month after an IPO, while the long-run performance is investigated looking at periods longer than a year (Ritter 1991). One phenomenon that is widely documented in the short-run performance of IPOs is that IPOs are underpriced on average. To analyse short-run market performance, usually the first-day average returns are used. The first-day return is the difference between the offer price and the first day closing price. A positive average return of the first trading day is identified as underpricing, and was first documented by Logue (1973) and Ibbotson (1975).

A pattern that is observed in the long-run analysis of IPOs is that IPOs appear to underperform (Ritter 1991, Loughran & Ritter 1995, Ang, Gu, & Hochberg 2007). However, evidence for this phenomenon is not as widespread as that of IPO underpricing. Long-run underperformance generally means that the subsequent share price of IPOs are usually lower than the first day trading prices, which results in negative abnormal returns in the long run for investors holding IPO stocks. Long-run IPO performance is still a debatable issue among researchers as shown by the conflicting findings that have been produced. Some studies show that IPOs underperform marginally or have no abnormal performance in the long run (Schultz 2003). With these findings, some have shown that underperformance disappears when different measures of performance or methodology are used (Gompers & Lerner 2003). Another explanation is the sample period considered (Hoechle & Schmid 2009), as the magnitude and direction of abnormal performance varies over the years.

Additionally to the debate on whether IPOs underperform in the long-run, a growing strand of literature focuses on the determinants of IPO long-run returns. Some argue that pre IPO factors such market timing (Yi 2001) or accounting accruals (Cotton 2008) can explain IPO performance.

Other studies focus on offer characteristics at time of the IPO and show that, amongst others, ownership characteristics (Gao & Jain 2011), underwriter reputation (Carter, Dark, & Singh 1998) and venture backing (Brav & Gompers 1997), all play a role in long-run returns of IPOs. In addition, some studies suggest that characteristics in the aftermarket of the IPO, such as analyst following (Rajan & Servaes 1997), initial returns (Santos 2010), or volatility (Gao, Mao, & Zhong 2006) are better predictors of long-run IPO performance.

The large body of literature devoted on IPOs and its subsequent performance, indicates that strong consensus is still missing on the direction and magnitude of long-run IPO performance. Prior literature also shows that multiple, and sometimes conflicting, factors exists that can influence long term IPO returns. This study aims to investigate the apparent patterns in IPO research. First, this paper investigates the long-run performance of IPOs against various benchmarks. Specific focus will be on the role of firm age or firm maturity in long-run abnormal performance of IPOs. Furthermore, this paper also investigates several the effects on several IPO characteristics, based on previous literature, on long-run IPO returns.

(6)

To identify long-run IPO performance and its determinants, a sample of 1925 IPOs on NASDAQ between 1980 and 2010 have been investigated. To determine whether IPOs show abnormal performance in the long-run, cumulative abnormal returns (CAR), buy-and-hold returns (BHAR) and wealth relatives (WR) are calculated, using the event-time approach, for the sample of IPOs against different market indices. These indices are the NASDAQ composite, the NYSE composite and the S&P500. This study further investigates whether the age of a firm when conducting its IPO plays a role in its long-run performance, therefore, the sample will be split into different age groups. To investigate how long-run abnormal returns vary over the years, abnormal returns are also calculated per year. Additionally, the effects of age and offer characteristics on long-run performance will be evaluated using a multiple regression model. The offer characteristics that are being evaluated are initial return, offer size, the percentage of secondary shares offered, and whether or not an IPO was venture backed.

The findings on long-run performance indicate that the results are sensitive to the measurement technique used. The CARs show that IPOs did not underperform after three years with values ranging from 2% to 11% across the different indices consistent with Gompers & Lerner (2003). The BHARs and WRs show the opposite, the average three year buy-and-hold returns for the IPOs in this sample range from -15% to -18%. The wealth relatives were all below one averaging around 0.80. These results on the BHARs and WRs show that IPOs underperformed against the benchmarks, consistent with Ritter (1991). The analysis when splitting the sample in age groups shows that the youngest age groups show the highest abnormal returns confirming Ritter & Welch (2002). When investigating long-run performance by year, results show that IPO underperformance was most persistent in the earlier years of the sample. IPOs in more recent years appear to over perform on average, which is in line with conclusions made by Hoechle & Schmid (2009).

The regressions analysis shows that firm age at time of the IPO negatively affects subsequent performance. However, as the results are not statistically significant, firm age is not an appropriate variable to determine long-run performance. The regression results furthermore show that of the explanatory variables, only initial returns and venture backing are appropriate predictors for long-run IPO returns.

The somewhat conflicting results in this study indicate that the research on long-run IPO performance highly depends on the methodology used. This is reflected by the opposite results of the CARs on one side, showing no underperformance, and the BHARs and WRs on the other hand indicating long-run IPO performance. Furthermore, it appears that IPOs in the earlier years of the sample underperformed on average, while IPOs in more recent years over performed. This might explain why a large part of the studies found long-run underperformance, as the majority of these studies use sample periods ending before the year 2000. From these observations and the regression results, it became apparent that considerations on measurement techniques, sample periods and benchmarks all influence the results in their own way. Therefore, one must be careful in making definitive conclusions on long-run IPO performance.

This study adds to the existing literature on IPOs by reexamining the general findings on IPO long-run performance, and aims to add an argument in the continuing debate on whether or not IPOs underperform in the long-run. First, this study uses a longer dataset with more recent data, and show that IPOs do not underperform in more recent years. Second, not many other studies specifically investigate the relationship of firm age and IPO long-run performance. Furthermore, this study shows that a combination of certain IPO variables can explain subsequent long-run IPO performance. The following chapter will discuss the main literature in IPO research, and the development of the hypothesis for this research. The third chapter describes the data and data collection process, as well as the results of the descriptive analysis of the data. Chapter four discusses the methods and results

(7)

of the long-run IPO analysis. The fifth chapter shows the regression design and results of the analysis on the determinants of long-run IPO performance. This paper will be concluded in chapter six, with a short summary of the chapters, as well as the major implications. The last sections of this paper consist of appendices and references.

(8)

2 Literature review and hypothesis development

An initial public offering, or IPO, is the sale of stocks by a company to the general public, on a securities exchange, for the first time. In this process, a private company becomes a public company. Private companies conduct an IPO to raise outside capital, after which the shares can be traded by the public. There are many reasons for a company to undertake an IPO. First, a company can conduct an IPO to raise money, especially when the cost of capital is low. Second, an IPO facilitates for early investors to cash out. Furthermore, an IPO can increase exposure, which allows for take-over opportunities. And finally, an IPO may serve as a strategic move, for example to increase reputation or for enhanced publicity (Brau & Fawcett 2006). Although IPO offers many advantages, there are also some disadvantages, the biggest amongst these are the significant direct and indirect costs associated with the IPO process. The direct cost includes the legal, auditing, and underwriting fees, furthermore as a public company, there are ongoing costs associated with the need to supply information on a regular basis to investors and regulators. The indirect costs are the management time and efforts spend in conducting the IPO and the dilution of shares and the loss of control for the original owners (Ibbotson & Ritter 1995).

2.1 The market for initial public offerings

Over the years, a lot has been written on Initial Public Offerings (IPOs). The market for IPOs shows several features that are not in line with general findings in market theories. One of the first major findings in IPO research is that IPOs showed systematic average positive first day returns (Ibbotson 1975). This implies that, on average, the shares in an IPO are underpriced. The first day closing price is generally used to assess what the true value of the shares according the market must be. If the closing price is higher than the price at which the shares are offered, it is typically assumed that the shares are underpriced. When IPOs are underpriced, it potentially means that the initial owners of the shares leave money on the table when they offer their shares to the market, as they could get a higher price, reflected by the closing price, for their initial offer.

Another pattern that is observed in the IPO market is that high underpricing is usually followed by a high volume of new IPOs (Ibbotson 1975). These fluctuation in the IPO market have been dubbed “hot” and “cold” markets, where hot markets sees high volumes of new issues, while in cold markets the volume of IPOs are generally lower than average. It is assumed that this fluctuation in issue volume is caused by companies that try to time their IPO. The owners of a company want the highest possible price for their shares. When, in a certain period of time, a lot of IPOs have been issued that were underpriced, it is assumed by the owners of the company that the market values shares higher on average. The owners want to take advantage of this positive valuation sentiment and offer their shares in periods where a lot of IPOs are underpriced on average. In cold markets, the volume of IPO issues is much lower (Ritter 1991).

A third pattern, which gets more attention in recent studies is that of long-term IPO performance. Starting with Ritter (1991), who was one of the first who documented evidence that IPOs on average underperform in the long-run, when compared to returns for similar companies or benchmark returns. This implies that investing in and holding IPO shares for a long period of time will, on average, result in lower returns than when one would invest in similar non IPO firms or in index stocks for the same holding period. Due to this observation, one can therefore predict that IPO shares will perform worse than comparable stocks.

The above mentioned anomalies inspired large bodies of literature trying to investigate and motivate why IPOs and the market for IPOs behaves as it does. Even though IPO underpricing is not the main

(9)

focus of this study, the following section will start with a short discussion on the evidence of IPO underpricing, after which the more popular theories on IPO underpricing will be discussed briefly. The remainder of this chapter will discuss the literature on IPO long-run performance in more detail, which is the main focus of this study. Initial evidence and the most important theories regarding long-run IPO performance will be reviewed. Due to space constraints, only literature focusing on IPOs in the United States will be discussed. This chapter will conclude with a section on hypotheses development and motivation.

2.2 Underpricing

IPO underpricing is usually defined as the percentage difference between the price at which the shares were offered to the market and the price at which the shares subsequently traded in the market. In addition, underpricing can also be measured in terms of “money left on the table” at the IPO. This is defined as the difference between the aftermarket trading price and the offer price, multiplied by the number of shares sold at the IPO. Over a long period of time, the average underpricing in the United States is between 10% and 20%, however, there are large fluctuations over time (Ljungqvist 2007).

2.2.1 Evidence for underpricing

IPO underpricing has interested financial researchers for decades. Loque (1973) and Ibbotson (1975) were one of the first to document that when firms go public, the shares they sold seems to be underpriced. That is, the share price jumped substantially on the first day of trading. This underpricing fluctuated substantially over the years, with an average underpricing of 21% during the 1960s, this dropped to 12 % in the 1970s, after which it started to rise again to 16% in the 1980s, 21% in the 1990s, and finally 40% from 2000 to 2005 (Ljungqvist (2007)). These initial findings on the irregularity of underpricing and the fluctuation over the years have inspired large bodies of literature, especially during the 1980s and 1990s.

In more recent studies, Ritter & Welch (2002) found that the average underpricing of IPOs was 18.8% in the period from 1980 to 2001 in the United States. Furthermore, Loughran & Ritter (2004) found that the average underpricing in the United States increased from 7% in the period between 1980 and 1989, to 15% from 1990 until 1998, after which it increased to 65% in 1999 and 2000. The authors argue that this large increase can be attributed to the change in composition of the firms going public, with a lot of small (technology) firms conducting an IPO during these years. These studies show that there is a consensus in the literature that IPOs are underpriced on average. Furthermore, it has been documented that this underpricing fluctuates over time with higher underpricing in more recent years.

2.2.2 Theories on underpricing

Many researchers have tried to explain the observed underpricing using several different theories. These theories of under pricing can be categorized into four broad groups. These are; theories based on asymmetric information, theories based on institutional explanations, theories based on ownership and control, and behavioural theories (Ljungqvist 2007). These theories will be discussed briefly below.

(10)

Asymmetric information theories:

The majority of explanations for IPO underpricing are based on asymmetric information. These theories assume that one of the key parties in an IPO process has more information than the others. The key parties in an IPO are usually the issuing firm, the underwriters handling the IPO, and the investors who buy the stock.

Asymmetric information between informed and uninformed investors

One of the most popular asymmetric information model is the winners curse model of Rock (1986). This model states that some investors are better informed about the true value of the share than others. This implies that informed investors only bid on IPOs that are attractive (underpriced), while uninformed investors cannot make this distinction and bid on all IPOs. This causes a “winners curse” for the uninformed investors as they receive all the shares in overpriced offerings, while in underpriced IPOs their demand is crowded out and rationed by the informed investors demand, thus leaving the uninformed investors with predominantly overpriced shares. To keep these uninformed investors in the market, shares therefore must be offered at a discount to offset the winners curse.

Asymmetric information between the issuer and the underwriter

These theories focus on the role of underwriters, book building methods and share allocation in IPO underpricing. Baron (1982) argues the underwriters have advanced information about the future demand of an IPO, compared to the issuers and the investors. When there is much uncertainty about the true value of an issuer, underwriters face a risk that not all shares are sold during an IPO. To minimize marketing effort and the sale risk of the IPO, underwriters intentionally underprice the IPO.

Asymmetric information between the issuer and potential investors

Under this assumption, "high quality" issuers are looking to signal their quality to the market by conducting an intentional underpricing. Better quality companies underprice their issues as they are confident that they can make up for the money left on the table. These companies use underpricing as a signal to distinguish themselves from lower quality issues (Welch 1989), but also to leave a positive note in investors’ minds for future issuance (Ibottson 1975).

Institutional explanations:

These theories can explain IPO underpricing in two major ways, first with legal liability and second with price support. The first is formulated by Ibotsson (1975), who argues that issuers deliberately underprice the IPO in order to avoid future lawsuits form investors who are disappointed with the post issue performance of the share. The second explanation is the price support approach, which states that underpricing reduces price drops, which stabilizes the price.

Underpricing to avoid legal liability

Tinic (1988) argues that underpricing is due to the legal liability for the underwriter as well as the issuer. Statements made in the prospectus vouch or guarantee for responsibility in the case of false, untrue or missing information in the issuing prospectus. Investors are less likely to litigate when initial returns are higher; therefore, underwriters intentionally underprice to protect their self against possible future claims.

Underpricing as a means for price stabilization

Ruud (1993) argues that underpricing can be traced back to the price support practice of the underwriters. That is, on average they know the fair value of the IPO and fix the price accordingly but underprice the issue in order to avoid negative performance. Furthermore, Asquith, Jones, & Kieschnick (1998) also argue that underpricing plays a role in price support activities, as is stabilizes against large drops in prices.

(11)

Ownership and control theories:

Ownership and control theories in IPO underpricing usually follow two opposing views. First, underpricing is used a mean to entrench managerial control, and to avoid monitoring by large outside shareholders. Second, underpricing is used to minimize agency cost by encouraging monitoring.

Underpricing to retain control

Brennan & Franks (1997) argue that underpricng gives managers the opportunity to protect their private benefits by allocating shares strategically when taking their company public. Underpricing makes that smaller investors can participate which disperse ownership, this reduces external monitoring. Furthermore, the more dispersed the ownership amongst smaller investors, the smaller the threat of a hostile takeover.

Underpricng to reduce agency costs

Stoughton & Zechner (1998) argue the opposite. They argue that issuers underprice their securities in order to attract institutional investors to buy large quantities of shares. This will lead to more monitoring buy large shareholders; this is usually beneficial for the company and smaller shareholders.

Behavioural theories:

Some authors try to explain IPO underpricing using behavioural theories. Behavioural theories assume the presence of “irrational investors” and biases that arise due to investor sentiments or behavioural biases associated with the issuer.

Investor sentiment

Irrational behaviour of the investors, over-optimism or market-modes can all be explanations as to why issuers and underwriters underprice IPOs. Welch (1992) argues that underpricing can cause a domino effect among investors that might raise demand for the issue. Furthermore, Demers & Lewellen (2003) argue that underpricing brings attention to the stock on the opening day, and Boehmer & Fishe (2001) document that underpricing increases the trading volume of the stock after the IPO.

Prospect theory

This theory states that the existing shareholders accept underpricing because they attracted more attention from investors and high underpricing suggests a relative high increase of the value of the remaining share. Loughran & Ritter (2002) furthermore argue that issuers are pleasantly surprised with their new found wealth (the amount they can raise in the IPO), that they are not significantly concerned with underpricing.

The above discussed theories show that the literature on IPO underpricing has fairly matured. The notion that IPOs are underpriced on average is widely accepted, as well as the fact that the extent of underpricing fluctuates over time. But even if short-run underpricing is documented as a persistent phenomenon in the IPO literature, the degree of underpricing and reasons for underpricing are not clear because of factors as sample size, sample period, and different measuring techniques and models. Therefore, there is no single dominant theoretical reason for underpricing. Nonetheless, there are several theories that try to explain why IPOs are underpriced, with most evidence favour explanations based on asymmetric information between the market participants in the IPO process. As there are multiple theories that can explain IPO underpricing, there is still debate as to which theory is superior.

(12)

2.3 Long-run performance of IPOs

The second major anomaly is the long-run performance of IPOs, which is the main focus of this study. The initial research on IPOs was predominately focused on IPO underpricing. However, more recently the long-run performance of IPOs received the majority of the attention. Research on the long-run performance of IPOs has documented some conflicting results regarding the behavior and determinants of long-term returns. The articles most relevant for this study will be discussed in more detail in this section.

2.3.1 Evidence on long-run performance of IPO

Ritter (1991) was one of the first who found and presented empirical evidence of long-run IPO underperformance. In his paper he documents the anomalous result that in the long-run, IPOs appear to be overpriced. He used a sample of 1526 IPOs between 1975 and 1984, and he calculated the three year buy-and-hold returns for the IPOs in the sample. He also calculated the three year buy-and-hold returns for a set of matching firms, matched by size and industry, listed on the American and New York stock exchanges. He finds that in the three years after going public, the IPOs significantly underperformed the set of comparable firms. The average holding period return for the sample of IPOs 34.47% in the three years after going public, while the average holding period return of the matching firms was 61.86% over the same period. Furthermore, he shows that younger companies and companies conducting their IPOs in high volume years perform worse than average. With this result, he concludes that in the long-run, IPOs underperform. He argues that his results are in line with the theory that many IPOs follow industry-specific fads and that IPO issuers are taking advantage of windows of opportunities when investors are irrationally over optimistic about the future potential.

The findings of Loughran & Ritter (1995) are in line with the initial findings of Ritter (1991). In their article, they show that IPOs from 1970 to 1990 significantly underperformed relative to non issuing firms, five years after the IPO date. They used a sample of 4753 IPOs and calculated the three and five year abnormal performance, adjusting for book-to-market effects, using annual holding-period returns for both IPOs and a set of matching firms matched on market capitalization. They furthermore ran cross-sectional regressions on monthly individual firm returns and 3-factor time series regressions of monthly returns for portfolios of IPOs and the matching firms. They found that the average return three years after the IPO was 5%, and after five years the average return was 11.6%. The average annual return over the whole five year period after the IPO was 5%, while the average annual return for the matched firms was 12% over the same period. With their analysis they reject their hypothesis of no underperformance with high degrees of statistical significance. They conclude that IPOs underperform in the long-run, and they also found that book-to-market effects can only explain a small part of the observed underperformance of IPOs. They conclude that their findings are in line with the theory that cycles in IPO volume are due to issuers taking advantage of windows of opportunity by issuing equity when, on average, they are substantially overvalued. They also argue that the rapid growth of many young companies made it easy to justify high valuations because investors are betting on long shots.

In a more recent study, Ritter & Welch (2002) also confirm, although with some caution, the initial evidence of long-term IPO underperformance. The authors examine IPO underperformance using a sample of 6249 IPOs in the United States over the period from 1980 to 2001. They find that the number of IPOs varies from year to year, and that their shares were underpriced by an average of 18.8%. Furthermore, by calculating buy-and-hold returns for IPOs and matching firms and using multifactor regressions with an equally weighted portfolio of IPOs, they find that over three years, the average IPO underperformed the CRSP value-weighted market index by 23.4% and

(13)

underperformed matched companies with the same market capitalization and book-to-market ratio by 5.1%. The authors note, however, that caution is advisable. They claim that the results are sensitive not only to methodology, but also to the time period chosen and how crisis periods are included. Furthermore, one must be careful when using Fama-French multifactor regressions, as these can produce biased results

There are also studies that do not find any significant abnormal performance. Gompers and Lerner (2003) for example, investigate the performance of IPOs by examining a more extended time period. Their sample consists of 3661 IPOs over the period from 1935 until 1972, and they measure their returns for up to five years after listing. The authors find some evidence for underperformance when performance is measured using value-weighted event-time buy-and-hold abnormal returns, but they conclude that these results are not consistently statistically significant. However, the underperformance disappears when either equal-weighted buy and-hold abnormal returns or cumulative abnormal returns are being used. They argue that the difference between the event-time and calendar-time results from the clustering of IPOs in periods immediately before poor IPO performance. They also find, with capital asset pricing model and Fama-French 3-factor regressions, that the intercepts are insignificantly different from zero, showing no abnormal returns. Finally, they conclude that the relative performance of an IPO sample depends on the method of examining performance.

Schultz (2003) also fails to find significant underperformance; however, his explanation differs from the above. In his paper he claims that the observed long-run underperformance of IPOs is more likely to appear in event time. He argues that the poor event-time performance of IPOs can be explained by a phenomenon he calls “pseudo market timing”. He explains pseudo market timing as more firms that go public when they can receive a higher price for their shares, and as a result, there are more offerings at peak valuations than at lower prices. This results in that the probability of observing long-run underperformance in event time is much higher. The author uses simulations with the distribution of historical market and IPO returns and the relation between the number of offerings and market levels between 1973 and 1997. His analysis reveals that underperformance of more than 25% in the five years following an offering is neither surprising nor unusual in event time, but that this is not an indication of any market inefficiency. He concludes that biases from pseudo market timing can be avoided by using calendar-time returns rather than event-time returns. Furthermore, he notes that if event-time returns must be used that the problem of pseudo market timing can be mitigated by using benchmarks that are as highly correlated with the firms being studied as possible. However, Ang, Gu, & Hochberg (2007) discuss the findings of Schultz (2003) and in lesser extent the findings of Gompers and Lerner (2003), and they reject their conclusions. Using a sample of 4843 IPOs between 1970 and 1996, they show that the sample of IPOs exhibits significant underperformance in both event time and calendar time. They use a self build Markov model that is able to capture small sample bias. They use this model to generate estimates of small sample distributions of IPO long-horizon abnormal returns. They compare the small sample distributions with the estimated point statistics of IPO long-horizon returns from actual data, and they find that the small sample effect cannot be used as an explanation for the IPO underperformance effect. They authors find that IPO underperformance remains robust both in event time and in calendar time. They explain that the different results found by Schultz (2003) are because he is only considering a short holding period horizon, which does not capture the IPO underperformance. Moreover, they argue that that IPO underperformance is highly unlikely to be the result of a statistical fluke. They conclude that the earlier findings of underperformance, starting with Ritter (1991), remains robust in both event and calendar time.

More recently, Hoechle & Schmid (2009) investigated IPO underperformance with a sample of 7,378 firms going public, in the period between 1975 and 2005. They find that there is significant

(14)

underperformance of IPO firms over the first year after going public, while there is virtually no underperformance in the following years. They argue that the underperformance disappears, as over time the characteristics of IPO firms converge to those of the more seasoned companies. They furthermore find that the observed IPO underperformance is due to fundamental differences in firm characteristics, like market-to-book ratio, leverage, and R&D spending, between IPO and more seasoned non-issuing firms. They argue that the documented IPO underperformance is associated with overly optimistic growth prospects and correspondingly high valuation levels. IPOs going public during hot issue periods performed even worse on average. They also conclude that IPO underperformance is likely to be the consequence of imperfect matching procedures. Furthermore, they argue that in the context of the traditional calendar time portfolio approach the documented IPO underperformance is likely to be the result of the Fama-French factors not being able to fully explain variations in the cross-section of stock returns. With these results, the authors conclude that there is no convincing evidence that IPOs underperform in the long-run.

As one can observe from the articles above, strong consensus is still missing on whether IPOs show underperformance in the long run. It becomes clear however, that at least some of the conflicting results can be attributed to differences in sample period and differences in measurement techniques. These differences across studies result in different conclusions on how IPOs perform in the long-run.

2.3.2 Measurement issues in long-run performance

Abnormal stock returns are in the focus of many different research questions, however, as we can see from the literature above, there is a debate in the literature on how abnormal returns should be estimated. There are three main areas of critique. The first discussion focuses on abnormal return measures mostly measured by using cumulative returns and buy-and-hold returns. The second discussion focuses on whether long-run performance should be measured in event time or calendar time. Finally, discussion also focuses on the reference portfolios used to determine abnormal performance. This section briefly discusses this debate.

Buy and hold returns versus Cumulative average returns

There are two main metrics used in the majority of studies to evaluate long-run performance. The first is buy-and-hold abnormal returns (BHARs), and the second is cumulative abnormal returns (CARs). There is still no consensus in the literature about the right metric for a long-term performance study, as both measures have their strengths and weaknesses. Authors in favor of cumulative abnormal returns, Fama (1998), Gompers & Lerner (2003), argue that these models are better and less biased performance measures. They argue that the empirical tests of asset pricing models typically use monthly returns. Therefore, cumulative average return models are simpler than buy and hold return models. Furthermore, methods based on average monthly returns reduce the problem of cross-correlation of returns across events, while buy-and-hold abnormal returns are not based on monthly average returns.

Proponents of buy and hold return models, like Lyon, Barber, & Tsai (1999) argue that buy-and-hold abnormal returns accurately mimic investors’ experience. Cumulative abnormal returns do not reflect the abnormal returns for an investor buying the event firms and shorting the benchmark over the full horizon. Furthermore, they argue that buy-and-hold returns are better than cumulative abnormal returns due to monthly portfolio rebalancing assumption associated with cumulative abnormal returns, which can create a downward bias in long-term CARs. Barber & Lyon (1997) also argue that when calculated using a reference portfolio such as a market index, CARs are seriously affected by a new listings bias, and as a result, significance levels of CARs are overstated.

(15)

Event time versus Calendar time

There are two ways of measuring long-run market performance, the first is the mostly used event-time approach and the second is the calendar-event-time approach. Some use the calendar-event-time approach instead of the event-time approach to evaluate the long-run performance of IPOs. They have argued that the event-time returns overstate the statistical significance of mean excess abnormal returns because of the cross-sectional dependence of observations. Some authors like Brav & Gompers (1997) favor the calendar time approach, as event-time abnormal returns calculated using reference portfolio can result in miss-specified test statistics. This will have different impact on buy-and-hold abnormal returns and cumulative abnormal returns. As a result, cumulative abnormal returns yield positively biased test statistics, while buy-and-hold abnormal returns and the associated test statistics are generally negatively biased. The calendar time approach avoids cross-sectional dependence problem because the returns on sample firms are aggregated into a single portfolio. Nevertheless, the most common approach used to measure the long-run performance of IPOs is the event-time approach. First, because according to Barber and Lyon (1997), calendar-time returns do not measure investor experience. Second, Lyon, Barber & Tsai (1999) argue that calendar-time returns are generally miss-specified in random samples. Third, Loughran & Ritter (2000) argue that using calendar time returns has lower power to identify time-varying misevaluations.

Reference portfolios

When measuring long-run performance, several benchmarks can be used to evaluate returns. The most used benchmarks are the markets return, as measured by market indexes, and returns on reference portfolios of similar companies with respect to size, book-to market or industry. When using the market return as a benchmark, Loughran & Ritter (2002) argue that long-run returns can be biased towards no abnormal returns because some of the IPOs firms are included in the benchmarks. When a benchmark is contaminated with many of the firms that are the subject of the test, than the test can be biased towards high explanatory power and no abnormal returns. Schultz (2003) argues however, that when event-time returns are used that the problem of pseudo market timing can be mitigated by using benchmarks that are as highly correlated with the firms being studied as possible. In reference portfolios, firms are usually matched by size, book-to-market, or industry. Loughran & Ritter (1995) argue that it is not advised to match IPOs by industry. First, companies may time their offers to take advantage of industry wide misvaluation and thus controlling for industry effects will reduce the ability to identify abnormal performance. Barber & Lyon (1997) furthermore argue that the use of control firm approach benchmark eliminate rebalancing and skewness bias. However, the use of size and book to market buy-and-hold reference portfolios does not eliminate this problem, resulting in the negatively biased test statistics.

The above show that there are different ways in documenting long-run performance. All methods for measuring and evaluating abnormal returns have advantages and disadvantages. The challenge is to find the method that is least likely to overstate the results or to introduce bias. Considerations on sample size, sample period and reference portfolio all influence the decision for the most appropriate measurement techniques. The different measurement techniques in combination with the considerations on the research design can heavily affect the eventual results.

2.3.3 Factors Affecting long-run performance

This section will discuss the theoretical and empirical explanations of long-run IPO performance and the factors that can influence IPO abnormal performance in the long-run. The literature on IPO aftermarket performance proposes several different theories about the determinants and factors in long-run performance of IPOs. These theories can be categorized in three broad groups, the first

(16)

group of theories discussing factor prior to an IPO that can influence subsequent performance. The second group consists of firm and IPO characteristics at time of the offering. The final group consists of aftermarket factors that can influence long-run abnormal performance. Theories and literature for each of these groups will be discussed below.

Pre IPO characteristics:

Pre-IPO characteristics are factors that occur before a firm is conducting an IPO. These factors are mostly based on behavioural theories; a change in the behaviour of firms is noticeable, prior to a potential IPO.

Window dressing

Window dressing theories state that firms try to make the firm look better prior to an IPO, to achieve greater valuations. Jain & Kini (1994) proposed this theory in their study. They investigated the change in operating performance of firms conducting an IPO. They found a significant decline in operating performance in firms prior to their IPO, over a six-year period extending from the year before the IPO until five years after the offering. They authors argue that IPOs start out with high market-to-book, and high price-earnings ratios relative to their industry counterparts but experience a decline in these measures after the IPO. The authors give several reasons for the decline in operating performance. One explanation is related to the potential for increased agency costs when a private firm becomes a public firm after their IPO. A second reason, according to the authors, could be that managers try to “window-dress” their accounting numbers, to make the firm look better prior to going public. A third explanation for the decline in operating performance is that entrepreneurs time their IPOs, and issue in periods of unusually good performance levels, which they know cannot be sustained in the future. Their results suggests that IPOs are unable to sustain their pre-issue performance level, and that investors appear to value IPOs based on the expectation that earnings growth will continue, while in reality the pre-IPO profit margins, on which the expectations are based, are not sustained in the aftermarket.

Similarly, Teoh, Welch, & Wong (1998), examine in their paper the relation between long-run post-IPO underperformance and the post-IPOs earnings management. They show that discretionary current accruals are higher for IPOs relative to similar seasoned firms. They also find that issuers with higher discretionary accruals have poorer stock return performance in the subsequent three years. With IPOs that are ranked in the highest quartile, of the IPO-year discretionary current accruals, show 15% to 30% worse three year performance on average compared to firms in the most conservative quartile. Additional tests suggest that the ability of accruals to predict IPO underperformance is derived from a general ability of accruals to predict returns in all firms. Moreover, the former general predictive ability of accruals is more significant for IPOs because the discretionary current accruals of IPOs are much larger than those of average non-IPO firms. This result can lead to IPOs that report high earnings by adopting discretionary accounting accrual adjustments that raise reported earnings relative to actual cash flows. The above results causes over optimism in investors, because they are guided by the high earnings, but are unaware that the earnings are inflated. The authors conclude that, keeping other things equal, the greater the earnings management at the time of the IPO, the larger the post IPO underperformance.

More recently, Cotten (2008) examined the earnings management explanation for IPO underperformance. More specifically, he compared the performance matched discretionary accruals and abnormal performance of IPO firms only issuing primary shares, with IPOs that issue both primary and secondary shares. Furthermore, regression analysis is used to examine if secondary shares can explain the level of discretionary accruals, and the relationship of these variables with long-run performance. The author uses a sample of 6652 IPOs between 1988 and 2002, and his initial results show that IPOs that only issue primary shares manage earnings upwards, IPOs that issue both

(17)

primary and secondary shares do not manage earnings upwards, while IPOs that only issue secondary shares manage earnings downwards. He furthermore finds that discretionary accruals are negatively related with performance over 24 and 36 months. He concludes that earnings management contributes to the long-run underperformance of IPOs, and the offering of secondary shares influences subsequent performance.

Market timing

There are also behavior theories that state that owners try to time their IPO to achieve greater valuations. Yi (2001) examines this in his article. More specifically, he examines whether or not the long-run performance of IPOs is related to the sign of earnings of the IPO firms at the time of issue. He motivates that many firms go public while having negative earnings per share, and that these firms typically project very high growth prospects, which can lead to an overoptimistic valuation of the IPO. He finds that IPOs underperformed the market index and control firms over a 3 year period after going public. And even though the long-run median returns indicate that all IPOs underperformed the NASDAQ index and the set of control firms matched by industry and size, only the firms going public with negative earnings had statistically and economically significant negative abnormal mean returns. The author concludes that his finding that the IPO firms with negative earnings significantly underperformed both the market index and more publicly established firms suggests that investors may have been too optimistic about future prospects of the IPO firms, especially those that had negative earnings.

Similarly, Chou, Gombola, & Liu (2009) examined the relation between growth opportunities and long-run IPO returns and operating performance of 371 IPOs between 1980 and 2000. They use Tobin’s q as a measure of growth opportunities where IPOs with high Tobin’s Q reflect high growth prospects. They show that the post-offering performance of IPOs is related to their growth opportunities. They find significant long-run underperformance in IPOs only for firms with high Tobin’s q. Furthermore, firms with higher Tobin’s Q experience not only poor stock price performance but also poor operating performance. The authors hypothesize that the finding of significant pre-offering run-up and positive earnings forecast revisions in IPOs with high Tobin’s Q, are both indicators of investor optimism for high growth firms. Their results are consistent with the view that investors are overly optimistic about the prospects of high growth firms.

Firm and IPO characteristics at time of the offering:

The following section will discuss firm and IPO variables that are observable at time of the IPO, and their relationship with long-run performance.

Ownership

Several researchers have tried to link the ownership structure to long-run performance. Howton, Howton, & Olson (2001) examined the role of the board of directors in IPO anomalies. They investigate the relation between IPO anomalies and the level of share ownership by members of the board of directors, and several other variables that measures board characteristics. In a sample of 412 IPOs between 1986 and 1994, they investigated the initial day, the one year, and the three year market adjusted returns. They found that the ownership position of board members, especially inside members of the board, plays a significant role in each of the three return measures. The initial day returns are directly related to the percentage of shares outstanding owned by both insiders and independent outside members of the board. While the one and three year market adjusted returns are directly related to insider ownership and not strongly related to outside ownership. The authors conclude that the results do not support the notion that higher board ownership IPOs experience less underpricing in the short-run, they do however support the idea that high board ownership IPOs perform better in the long run. Their explanation is that having insiders on the board that are also

(18)

shareholders leads to better alignment between manager and shareholder interests, which can lead to the observed superior long-run performance for these IPOs.

Similarly, Pukthuanthong, Roll, & Walker (2007) investigate whether the form of managerial compensation affects a firm's long-term performance. They document the method of equity compensation and operating performance for 5 years after an IPO, in a sample of 897 IPOs conducted between 1997 and 1999. They found that stock prices of IPOs in the sample declined in the post-IPO period, even after controlling for the market, comparable firms, and standard factor models. They argue that the stock price performance is less bad for firms managed by executives who are receiving a balanced form of compensation, in the form of both equity and options. They show that a particular compensation method is associated with better performance for at least 3 years after the IPO. They show that IPOs performed better when managers received a balanced combination of stock option grants and equity ownership. On the other hand, IPOs with unbalanced compensation arrangements; high equity ownership and few option grants, or many option grants and low equity ownership, do not perform as well. They conclude that this empirical finding is consistent with theoretical explanations based on managerial risk aversion and the alignment of managerial and owner incentives.

More recently, Gao & Jain (2011) analyze the long-run performance of founder and non founder CEO led IPOs. They investigate whether the post-IPO performance of founder CEO led firms is superior to that of similar non founder CEO led firms. Additionally, they investigate whether these results still hold in the context of high technology IPO firms. To test their hypothesis they document the five year post-IPO investment performance of founder and non-founder CEO led firms in both high and low technology environments, relative to several benchmarks using buy-and-hold abnormal returns. They found that while founder CEO led IPO firms outperform non-founder CEO led IPO firms, the significance of the results depends on choice of benchmark, portfolio weighting method, and factor regression model used to estimate abnormal returns. As a result, they do not find strong or consistent evidence of superior long-run performance of founder CEO led IPOs relative to similar non-founder CEO led IPOs. They do however find consistent evidence that founder CEO that led high technology IPOs showed significantly higher long-run returns relative to non-founder CEO led high technology IPOs. They conclude that that the unique nature of founder CEO leadership is particularly beneficial to IPO firms in high technology environments.

Underwriter

The type of underwriter and their reputation also might influence subsequent IPO performance. Carter, Dark, & Singh (1998) study the role of underwriters in abnormal IPO performance. The authors examine several measures of underwriter prestige and their association with initial and long-run IPO returns. They found that each of the reputation measure for underwriter prestige examined in isolation, to be significantly related to the initial return. Furthermore, when using long-run holding period returns, they find that the underperformance of IPO stocks relative to the market over a three year holding period is less severe for IPOs handled by more prestigious underwriters. They argue that more reputable underwriters are associated with less short-run IPO under pricing and less long-run IPO underperformance. Furthermore, they argue that of the measures used to evaluate underwriter reputation, the Carter-Manaster measure is the best measure to control for underwriter prestige in IPO long-run performance studies.

More recently, Chang, Chung, & Lin (2010) investigated the nature of earnings management in IPOs, by looking at the role of underwriter reputation. They use discretionary current accruals as a measure of earnings management, while controlling for changes in discretionary accruals which can be attributable to changes in performance. They find that underwriter reputation has a significantly negative correlation with earnings management. Furthermore, they find a negative relationship between earnings management and the post-offer performance of an IPO firm’s stocks, but only for

(19)

those firms associated with less prestigious underwriters. The authors argue that prestigious underwriters will protect their reputation by carefully monitoring and certifying financial information on IPO firms, which reduces any potential earnings manipulation. Those IPO firms that are associated with more prestigious underwriters are likely to exhibit substantially less aggressive earnings management.

Similarly, Dong, Michel, & Pandes (2011) analyze the relationship between the number of managing underwriters, underwriter reputation, and information production on the long-run performance of IPOs. They study a sample of 7407 IPOs issued between 1980 and 2006, and they test whether the quality of underwriters influences the three year long-run performance of IPOs, while control for size and book-to-market effects. Furthermore, they test whether underwriter quality has a greater effect on long-run performance when uncertainty about IPOs is higher. They find that, when excluding the internet bubble period of 1999 and 2000, the number of managing underwriters and underwriter reputation positively predicts long-run IPO performance, especially among firms with high uncertainty. The authors argue that these findings are consistent with the marketing and certification and screening roles of investment banks, but show little support for the information production role of underwriters. They conclude that if investors underestimate the importance of certification, they may overpay in IPOs with poor underwriter quality, which leads to the underperformance of these IPOs.

Venture backing

There is some evidence that venture backing in IPOs can influence their performance. Brav & Gompers (1997), who investigate the long-run underperformance of IPOs from 1972 until 1992, show that this might be the case. They use a sample of 934 venture backed IPOs and 3407 non venture backed IPOs, and used several measures to investigate the long-run performance in their sample. First, they show that over a five year period, venture backed IPOs outperformed non venture backed IPOs, when returns were weighted equally. Value weighing the returns significantly reduces the underperformance. They furthermore find that the underperformance in the non venture backed sample is primarily caused by small IPOs with market capitalizations less than $50 million. When using Fama-French 3-factor regressions, they find that venture-backed companies did not significantly underperform, while only the smallest non venture-backed firms did. They conclude that this underperformance is not specifically an IPO effect; because similar size and book-to-market matched firms that not issued equity perform just as poorly as IPOs. They argue that the underperformance is a characteristic of small, low book-to-market firms regardless of whether they conduct an IPO or not.

More recently, Yip, Su, & Ang (2009) analyze whether the choice of underwriters, venture capital support, industry and their interactions play a role in the long-term performance of IPOs. The sample of their study consists of 1772 IPOs that issued between 1996 and 2000, and the main focus is on the effect of choice of underwriters on performance of IPOs. They calculate abnormal monthly returns for the whole sample in the first 12 months; they furthermore use regression analyses on both abnormal returns and cumulative abnormal returns to investigate the effect of underwriter choice, venture capital support as well as industry and their interactions on the long-term performance of IPOs. They hypothesize that if potential investors demand that the issuing firm hires a reputable underwriter for certification of information, the aftermarket performances is expected to be different for IPOs underwritten by highly reputable investment banks and those underwritten by less active investment banks. They show that in their analysis, only significant underwriter and venture capital effects, and that industry effects disappears after controlling for the other effects. They conclude that even though, by the end of the first year, IPOs, on average, underperform the market, investors can still benefit by investing in IPOs underwritten by highly reputable investment bank, and the returns will be even higher if these IPOs are financed by venture capitalists.

(20)

IPO prospectus

Some argue that information adopted in the IPO prospectus can reveal factors that can influence long-run performance. Bhabra & Pettway (2003) study whether the information contained in the IPO prospectus is related to the subsequent long-run performance. They use a random sample of 242 IPOs in the United States between 1987 and 1991. The authors show that financial and operating characteristics as well as offering characteristics have a limited relation with the one year stock returns. Pre-IPO profitability, spending on R&D, relative offer size, firm size, and number of risk factors listed in the prospectus all have some relation with the one-year stock returns. However, they find no evidence that prospectus information is related to long-term performance over three years. They further find that firms that subsequently reissue equity or merge outperform their matched-firm benchmarks over three years. Underperformance is most severe for the smaller and younger firms. The authors conclude that prospectus information is useful in the aftermarket over a short window of a year and to some extent in predicting subsequent survival or failure, although the value of this information declines rapidly with time.

R&D spending

There is also research focused on the R&D spending of firms, as this might influence long-run IPO returns. Guo, Lev, & Shi (2006) examine the relationship between R&D activities of issuers and the subsequent performance. They used a sample of 2696 IPOs between 1980 and 1995, with R&D activities as a source of information asymmetry and valuation uncertainty. They find that R&D activities significantly affect both the initial underpricing of IPOs and their long-term performance. Furthermore, they find evidence that R&D intensity affects analysts’ forecasts of IPO long-term earnings. They conclude that the R&D activities of issuers are an important factor in IPO long-run performance.

After market characteristics:

This section will discuss factors that influence IPO performance, that occur or that are observable, after a company conducted their IPO.

Analyst following

Analyst following of newly conducted IPO might be an indicator for long-run performance. Rajan & Servaes (1997) examine the relationship between analyst following and IPO performance. Using a sample of 2725 IPOs between 1975 and 1987, they document how analyst coverage can influence IPO performance over three years. They find that higher underpricing leads to increased analyst following, analysts are overoptimistic about the earnings and growth performance of IPOs and more firms complete IPOs when analysts are particularly optimistic about the growth prospects of recent IPOs. They furthermore found that, in the long-run, IPOs perform poorly when analysts are more optimistic about their long-run growth projections. They conclude that the anomalies associated with IPOs may be partially driven by over optimism.

More recently, Das, Guo, & Zhang (2006), investigate the ability of analysts to forecast future firm performance, based on the selective coverage of IPOs. The authors hypothesize that the decision to provide coverage contains information about an analyst's underlying expectation of a firm's future prospects, which in turn can predict performance. They examine this using a sample of 4082 IPOs over the period from 1986 until 2000. The authors find that that in the three subsequent years, IPOs with high residual coverage have significantly better returns and operating performance than those with low residual coverage. More specifically, they find that the difference between high and low residual coverage portfolios of IPOs results in an annualized buy-and hold returns of 8.71% higher compared to a portfolio of firms matched by size and book-to-market ratio When comparing the results to a portfolio of matching firms matched by industry, this return is 7.17% higher. The authors

(21)

conclude that this indicates that analysts have superior predictive abilities and selectively provide coverage for firms where true expectations are favorable.

Initial returns

Initial returns, or initial under of overpricing, in IPOs might also influence long-run returns. In their paper, Affleck-Graves, Hegde, & Miller (1996) examine the conditional aftermarket performance of IPOs between 1975 and 1991. They use a sample of 2096 IPOs conducted on NASDAQ. They find that IPOs that were initially underpriced earn more compared to size matched firms in the first three months of trading, while overpriced IPOs underperformed in the same period. Additionally they find that both underpriced and overpriced IPOs exhibit significant underperformance in the subsequent 6 to 24 months, compared to the matched firms. They conclude that there is a strong relationship between market conditions and the number of overpriced IPOs, and that initial returns might influence subsequent performance.

Similarly, Purnanandam & Swaminathan (2002) study the valuation and its effect on subsequent performance of IPOs between 1980 and 1997, using comparable firm multiples. They find that overvalued IPOs have higher first day returns, but over the next five years they underperformed by about 20% to 40%, compared with undervalued IPOs. They furthermore show that the underperformance of overvalued IPOs is robust to various benchmarks and return measurement methodologies including size and book to market controls and the Fama-French 3-factor model. They conclude that their findings are in line with the theories based on investor confidence.

More recently, Santos (2010) examines the connection between IPO underpricing and long-term underperformance. With an initial sample of 6256 IPOs in the United States between 1973 and 2008, he shows that the average underpricing can determine the degree of subsequent long-run performance. He furthermore documents that the long-term underperformance of IPOs is driven by firms that go public in periods of high IPO underpricing. He concludes that investor sentiment is stronger in high underpricing periods and that the presence of overly optimistic investors in the market drives prices up. Low quality firms and underwriters then try to exploit these optimistic investors by issuing equity in periods of high underpricing.

Opening day characteristics

Some claim that opening day characteristics can be an indication of subsequent long-run performance. Houge, Loughran, Suchanek, & Yan (2001), investigate the relationship of early market indicators, divergence of opinion and the performance of IPOs. The authors use three opening day proxies associated with IPOs, to test their role in IPO underperformance. These proxies are: the percentage opening bid-ask spread, the time of the first trade, and the flipping ratio. The authors argue that these variables describe the uncertainty faced by different the IPO participants. In their sample of 2025 IPOs in the United States between 1993 and 1996 they find that IPOs with a wide initial spread, a late opening trade, or a high proportion of institutional flipping exhibit poor long-term returns. Their results are robust to indicators of IPO quality such as market capitalization, offer price, venture capital financing, underwriter prestige, and partial adjustment. They further find that flipping and time of first trade are more informative measures of long-run performance for large firms, while the opening spread is a stronger indicator for small firms. The authors conclude that the flipping ratio, opening spread and time of first trade each captures unique components in the uncertainty or divergence of opinion regarding IPOs.

Volatility

Gao, Mao, & Zhong (2006) investigate whether divergence of opinion plays a role in IPO long-term performance. In a sample of 4057 IPOs between 1980 an 2000 in the United States, they use early-market (the first 25 trading days after issuance) return volatility as their proxy of divergence of

Referenties

GERELATEERDE DOCUMENTEN

De Europese wetgever gaat verder dan de TRIPs-overeenkomst in die zin dat op grond van artikel 3 lid 2 Handhavingsrichtlijn de door lidstaten ingestelde maatregelen, procedures en

De Monitor doet namelijk niet alleen onderzoek wat door het publiek wordt aangedragen, maar wil het publiek ook echt iets bieden.. Ze willen een onderwerp niet

Figure 4-3: Influence of trans-membrane pressure time permeate pressure on molar flux through PAN-supported Teflon® AF2400 membrane at different feed side pressures..

The participants were asked to place four different names for each product category in one of the cells of the table for sound and semantics fit and misfit (for an overview of

The impact of venture capital reputation on the long run performance of Asian venture- backed initial public offerings.... Venture-backed initial public offerings in China, Japan

Another purpose of the thesis is to provide the readers with the information if the underperformance after the first day of the listing on a stock exchange can

The annual BHARs are corrected for the returns of the benchmark portfolio using size (expressed in market value of equity) and the market-to-book ratio. For BHAR1, BHAR2 and BHAR3

Barber and Odean (2008) posit the attention induced price pressure hypothesis, which states that attention grabbing stocks (e.g. stocks that are in the news, that have