Print Reference This Reddit This

Determinants Of Ipo Underpricing Empirical Evidence In China Finance Essay

This study examines the explanations of IPO underpricing in China by using cross sectional analysis and tests the information asymmetry hypothesis, institutional explanations and ownership and control theory. The sample used in this study consists of 50 firm-commitment IPOs of A-share which issued and listed in Shanghai Stock Exchange from July 2006 to March 2010. The study finds out that the market-adjusted initial return of IPOs is 58.882% in this investigation and the empirical evidence confirms that China enjoy a especially high underpricing level among the worldwide markets but the initial return is not abnormally high compared with evidence before 2005. The empirical evidence shows that the variables of offering size, firm’s age and money on the table are tested to be significantly correlated with the market-adjusted initial return, which suggests that the information asymmetry hypothesis is supported by empirical evidence to a large degree.

Chapter 1

Introduction

1.1. Context

An initial public offering (IPO) occurs when a company issues common stocks to the general public initially in the purpose to expand capital or to reform the state-owned enterprises. IPOs in China can be regarded as a privatization process of the state-owned companies, which provides a different perspective towards the IPO case study (Chi and Padgett, 2002).

IPO underpricing is defined as a high initial return of the new offerings for a short-run period. IPO underpricing is a “recurring phenomenon” across the world markets first documented by Reilly and Hatfield (1963), which is regarded as “one of the 10 puzzles in the financial researches” (Brealey and Myers, 1991). A common sense towards the IPO underpricing suggests that the IPO underpricing is a challenge to the efficient market hypothesis and may adversely affect the new company’s capital raising (Loughran, 1994). The IPO underpricing phenomenon has been documented in plenty of existing research papers. In the US stock market, the short-run performance of IPO shows a first day return of 11.4% (Ibbotson, 1975). Paudyal (1996) documents that UK IPOs gain an initial return of 38.7%. The level of underpricing is particularly high in the emerging markets. The initial returns for IPO in Malaysia, Korea, and Brazil are around 80% (Loughran, Ritter, and Rydqvist, 1994). The empirical evidence shows that China enjoys the highest initial return in the world markets. Plenty of researches figure out the initial return in China ranges from 200% to 900% before 2005.

Early empirical studies on the IPO underpricing phenomenon focus on the information asymmetry theory. Rock (1986)’s winner’s curse hypothesis is one of the important rationale for IPO underpricing. Besides, signal hypothesis supported by Ibbotson (1975) and principal-agent theory hypothesized supported by Baron (1982) explains the underpricing from different point of view. More recently, the explanations for IPO underpricing phenomenon can be categorized into four major groups: information asymmetry hypothesis, Institutional explanations, ownership and control explanations and behavioral approaches, which systematically summarized by Ljungqvist (2005).

1.2. Limitations of recent research

The IPO studies in China are developed from 1990s and majority of empirical evidences focus on the IPO market during 1995 to 2005. Su and Fleisher (1999) suggest that separating equilibrium under the information asymmetry hypothesis can explain the cross sectional pattern of short-run underpricing. Chi and Padgett (2002) test the information asymmetry hypothesis by cross-section regression analysis on the both Shanghai and Shenzhen Stock Exchange. Chen, Firth and Kim (2003)’s study mainly concentrates on the government ownership and investment risk. At present, all theories can be justified from different aspects, but without a general support of particular empirical evidence. The empirical study of IPOs underpricing on Chinese stock market have three main drawbacks. First of all, each IPO underpricing hypothesis is only valid for a specific time period and area. Secondly, whether foreign theory can explain the IPO underpricing in China is still inconclusive. Some scholars suggest that the Chinese stock market is inefficient and the high level of IPO underpricing can not be explained by information asymmetry hypothesis. Other financial researches suggest that the Chinese equity market is semi-efficiency. Finally, empirical studies on the IPO underpricing after the reform of non-tradable shares are rare. Therefore the previous empirical results can not explain the underpricing phenomenon after the reform.

1.3. Aims and objectives

Based on the existing empirical researches, the aim of this paper is to test whether the information asymmetry hypothesis can explain the IPO underpricing in China. Additionally it also takes the institutional explanation, ownership and control, investors’ behavior into consideration, which makes the results more sensible and comprehensive. The research also investigates the impact of special ownership structure in China. In this paper, the features of China primary market are discussed at the beginning. Then, the short-run underpricing of IPO on Chinese stock market is examined through Aggarwal’s methodology by using a sample of 50 IPO firms listed in Shanghai Stock Exchange form 2006 to 2010. The findings suggest that the market-adjusted initial return is 58.882%, 58.102% and 57.427% on the first, second and third day respectively, which show a significant underpricing in China. Finally, a cross sectional analysis is applied to investigate the determinants of short-run underpricing in China. The empirical results suggest that the winner’s curse hypothesis and underwriter price support hypothesis can well explain the underpricing. The explanation power of signaling hypothesis is not strong, while the results of ownership control and investor’s behavior hypothesis are contrary to the previous conclusions.

The main research questions are as follows:

Table 1.1: Research questions

To discuss the particular features of China primary market.

To examine the short-run underpricing of IPO on China stock market through Aggarwal’s methodology by using a sample of 50 IPO firms listed in Shanghai Stock Exchange form 2006 to 2010.

To investigate the determinants of short-run underpricing in China by applying the cross-section regression model.

1.4. Structure

The rest of this paper is organized as follows. Section 2 gives a brief description of China IPO market. Section 3 reviews the existing literatures that focus on the IPO underpricing theories. The empirical evidence of short-run underpricing in China and the cross sectional regression model which is used to analyze the determinants of underpricing are presented in section 4. Finally, the conclusion is presented in section 5.

Chapter 2

Background

2.1. An overview of China stock market

The basic role of stock market is to raise capital for firms, and the IPO provides an access to the public capital at a relative low costs. The establishment of Chinese stock markets is later than the developed countries. Shanghai Stock Exchange and Shenzhen Stock Exchange were found in 1990 and 1991 respectively. The China Security Regulation Commission (CSRC) was established in 1992 as the principal regulatory authority in China’s stock market. The share ownership structure in China has particular characteristics. Before 2005, the shares in China are separated into tradable and non-tradable shares. Non-tradable shares include state shares, legal person shares and employee shares. The tradable shares are ‘A’ shares, ‘B’ shares, ‘H’ shares and ‘N’ shares which can be freely traded on the secondary market. This empirical study is based on the ‘A’ shares listed in the Shanghai Stock Exchange.

2.2. IPO methods in China

From a global point of view, there are three major IPO methods (i.e. fixed price, auction and book building) have been widely. While making investigation on Chinese IPO underpricing, it is necessary to describe the evolution of IPO methods in China. Before 1994, fixed price is the primary IPO method, because the scale of IPO market is small and the offer price is determined by Commission. From 1994 to 1999, fixed price is replaced by public offer. The IPO price-earning ratio is strictly limited by administrative pricing. From 1999 to 2005, the offer price is determined by issuers and underwriters together, which represents the marketization of Chinese IPO. During this time period, two IPO methods, internet auction and quota system, coexisted at the same time. In 2005, “Several Issues Concerning the Trial Implementation of Price Inquiry System for Initial Public Offering of Shares Circular” is released by CSRC. Preliminary inquiry and quasi-book building are the major IPO methods up to now. Table 2.1 presents the evolution of Chinese IPO methods.

Table 2.1: Evolution of Chinese IPO methods

Time period IPO methods

1990-1994 Fixed price

1994-1999 Public offer which base on the fixed price

1999-2005 Internet auction and quota system

2005- Preliminary inquiry and quasi-book building

2.3. Chinese IPO market

The IPO market in China during the research period from 2006 and 2010 has its distinct characteristics attributed to both inside and outside factors. On one hand, the reform of non-tradable share in 2005 significantly influences the China primary market. The statistical data about the ownership structure of listed companies in Chinese stock market show that the only 30% shares can be traded on the secondary market before 2005, which adversely affects the development of stock market and the initial public offerings. In order to eliminate the negative effect of the existence of large proportion of non-tradable shares, the Chinese authority reforms the non-tradable shares. Ownership structural reform improves the Chinese stock market fundamentals and enhances the market liquidity.

On the other hand, the global credit crisis in 2008 affects the IPO market in China severely. Figure 2.1 shows that the Shanghai A share price index was dropping sharply from 6000 to 2000 during 2008. According to the financial crisis, the investors’ confidence is affected heavily and the investment risk increases which impose a negative impact on the IPO market.

Figure 2.1: Performance of the Shanghai A share price index

The Figure 2.2 below displays an overview of total offer numbers and capitalization in Shanghai Stock Exchange from 2006 to 2010. The volume of the IPO ranges from 5216561275 to 27786775587. The capitalization reaches its minimum figure of 6729.82 million Yuan form 2008 to 2009 due to the financial crisis in 2008.

Figure 2.2: Offer number and capitalization by year (2006 - 2010)

Chapter 3

Literature review

This section aims to provide an overview of the international empirical evidence on IPO short-run underpricing, and to review and summarize the hypotheses of short-run underpricing documented in the previous academic and working papers.

3.1. International empirical evidence on IPO underpricing

Numerous empirical and theoretical literatures on initial public offerings figure out various degree of short-run underpricing. At early stage, the investigation made in the US stock market about the short run performance of IPO shows an initial abnormal return of 11.4% (Ibbotson, 1975). From a global perspective, IPO underpricing is a common phenomenon. Loughran, Ritter, and Rydqvist (1994) study the abnormal returns on IPOs in 25 major markets, and the initial returns reported range from 4.2% to 80%. Consistent with previous evidence, Purnanandam and Swaminathan (2004)’s result, which is more recently, shows that the initial returns ranges form 14.5% to 50%. However, the international findings about the IPO underpricing phenomenon are mixture. The level of underpricing varies across countries.

There are four main explanations focusing on the difference in the level of underpricing: (1) The degree of economic development; (2) Sell mechanism; (3) Particular industries and issuing periods; (4) Regulatory.

The level of underpricing in the emerging markets is significantly higher than in the mature markets. On the US market, Loughran and Ritter (1995) report a mean initial return of 10%, whose data obtained from a survey paper. On the German market, the degree of underpricing is identical with US market. IPO underpricing on UK market is higher than other developed countries. Paudyal (1996) figures out that UK IPOs gain an initial return of 38.7%. The level of underpricing is much lower on the France market, which is only 4.2%, according to Loughran, Ritter, and Rydqvist(1994)’s investigation.

On the contrary, the emerging markets enjoy a higher amount of initial return. For instance, the initial returns for IPO in Malaysia, Korea, and Brazil are around 80% (Loughran, Ritter, and Rydqvist, 1994).

Majority researches on China IPO market show an incredibly high initial return. Su and Fleisher (1999) study a sample of 308 ‘A’ share offerings from 1987 to 1995, which states that the average initial return is 948.6%. More recently, Tian and Megginson (2007) find out that the average initial return on China stock market is 247%.

The differences in selling mechanism can lead to the IPO underpricing varies across countries. The contractual mechanism may affect the degree of IPO underpricing. Chowdhry and Sherman (1996)’s empirical work indicates that the “contractual arrangement” applied by UK and major Asian countries probability generates a higher degree of underpricing than the method applied by firm-commitment offerings in the US.

IPO underpricing phenomenon is changing over time and industries. Loughran and Ritter (2004)’s empirical work shows that the mean initial return of IPOs in the 1980s was 7%, and during 1980 to 1990, the initial return doubled to 15%. The figure jumped to 73% during the Dot-com Bubble period (Ljungqvist and Wilhelm, 2003). Ljungqvist and Wilhelm (2003) explain the abnormally high level of underpricing in the hot-issue market by “realignment of incentives hypothesis”. The empirical results provided by Loughran and Ritter are not support the “realignment of incentives hypothesis”. On the other hand, Ritter (1984) points out that the underpricing phenomenon is significant in particular industries, such as “oil and gas”.

Fiscal policy and financial regulation also affect the IPO initial return to some extent. According to Loughran, Ritter, and Rydqvist (1994), large amount of first-day return on the Sweden market can be explained by the motivation of tax avoidance. Tian and Megginson (2007) presented that the main cause of China’s high level of IPO underpricing is “government regulation”.

3.2. Explanations of IPO underpricing

After Ibbotson (1975) studying both short-run and long-term performance of initial public offerings during the 1960s, a large number of investigations shed lights on the underpricing mystery and attempted to solve the IPO underpricing puzzle. Previously, scholars attempt to test the interpretations based on their own hypothesis. Reilly (1977), Logue (1973), McDonald and Fisher (1972), and Reilly and Hatfield (1969) have documented their original findings about IPO short-run underpricing puzzle in their contributing empirical works.

Based on the previous literatures, most theoretical and empirical explanations of IPO underpricing phenomenon can be classified into four categories: (1) information asymmetry; (2) institutional explanations; (3) ownership and control; (4) behavioral approaches (Ljungqvist, 2005). The first three hypotheses are supported by the most existing literatures. Table 3.1 below summarizes the hypotheses of IPO underpricing which will be discussed later.

Table 3.1: Summary of underpricing hypotheses

Information asymmetry hypothesis

Winner’s curse Rock (1986) Uninformed investors only take part in the market whose shares are underpriced, which force other firms to discount their new issues for the compensation of averse-selection.

Principal-agent theory Baron (1982) the information asymmetry among issuers, underwriters, and outside investors creates agent costs, which results in an underpricing phenomenon.

Signaling hypothesis Ibbotson (1975) Good firms discount their offer price to signal investors their quality and “superior prospects”.

Institutional explanations

Underwriter price support Ruud (1993) The high abnormal initial return may be explained by “frequent market practice of underwriter price support or stabilization.”

Law avoidance hypothesis Tinic (1988) Since the Securities Act of 1933 makes all participants in the offer who sign the prospectus liable for any material omissions, one way of reducing the frequency and severity of future lawsuits is to underprice.

Ownership and control explanations

Retaining corporate control Ritter (1998) “Issuing firms may intentionally underprice their shares in order to generate excess demand and so be able to have a large number of small shareholders.”

Reducing the agency costs Stoughton and Zechner (1998) The asymmetric information between issuers, underwriters and outsiders may create the agency cost which is one of the determinants of excessive initial return of a new offering.

3.2.1. Information asymmetry

Early empirical studies on the IPO underpricing phenomenon focus on the information asymmetry theory. The basic idea behind information asymmetry theory depends on the existence of information asymmetry among issuers, underwriters, and outsider investors (Rock, 1986). Models of information asymmetry represent that “one of the involved parties during the process of taking a firm public is more informed than others thereby causing the underpricing of shares (Hopp and Dreher, 2007).”

Later studies are more systematic. Information asymmetry hypothesis has been developed into four aspects: signal hypothesis, adverse-selection model, principal-agent model and information revelation theories (Ljungqvist, 2005). Most literatures focus on the first three theories.

The signaling hypothesis suggests that good firms discount their offer price to signal investors their quality and “superior prospects”. Ibbotson (1975) assumes that IPOs are underpriced in order to “leave a good taste in the investors’ mouths” so that “future underwritings could be sold at attractive price.” Allen and Faulhaber (1988), Grinblatt and Hwang (1989) and Welch (1989) establish a class of signaling models to investigate the IPO underpricing phenomenon. However, the result of empirical test on signaling models seems not support the signaling hypothesis. Michaely and Shaw (1994) systematically test the IPO signaling theory and their empirical results contrast with the prediction. Table 3.2 shows a compare and contrast of empirical evidence of signaling hypothesis.

Table 3.2: Compare and contrast of empirical evidence of signaling hypothesis

Compare

Ibbotson (1975)

IPOs are underpriced in order to “leave a good taste in the investors’ mouths” so that “future underwritings could be sold at attractive price.”

Contrast

Allen and Faulhaber (1988)

Establish a class of signaling models to investigate the IPO underpricing phenomenon, and the result of empirical test on signaling models not support the signaling hypothesis.

Michaely and Shaw (1994)

Systematically test the IPO signaling theory and their empirical results contrast with the prediction.

One of the most important rationales behind the IPO underpricing phenomenon is adverse-selection theory. Adverse-selection theory suggests that uninformed investors will be worse off on the primary market and they would like to buy shares that are underpriced. In that case, they may face a “winner’s curse” problem that “if they get all of the shares which they ask for, it is because the informed investors don't want the shares.”(Ritter, 1998)

Rock (1986) first model the “winner’s curse” hypothesis. Rock’s model suggests that uninformed investors only take part in the market if the shares are underpriced, which force other firms to discount their new issues for the compensation of averse-selection. Most empirical results support the “winner’s curse” hypothesis. Depending on the Rock’s model, Carter and Manaster (1990) take underwriters’ reputation into consideration, which suggests that there is a positive relation between underwriter’s reputation and underpricing. To test the basic idea behind “winner’s curse”, Michaely and Shaw (1994) classify their data sample into two sets: a sample of “IPO master limited partnerships (MLPs)” and a sample of "regular" IPOs. The empirical results represent that the initial return of "regular" IPOs is significant but the mean initial return of MLPs is insignificant, which coincide with Rock’s winner’s curse hypothesis towards IPO underpricing. However, the explanation for IPO underpricing by winner’s curse hypothesis is not perfect. And the conflict with early studies has yet not been settled. Ritter (1984)’s study shows that “the hot-issue markets only occur in particular periods” which do not support Rock’s conclusion (Allen and Faulhaber, 1988). Table 3.3 shows a compare and contrast of empirical evidence of winner’s curse hypothesis.

Table 3.3: Compare and contrast of empirical evidence of winner’s curse hypothesis

Compare

Rock (1986)

First model the “winner’s curse” hypothesis.

Carter and Manaster (1990)

Depending on the Rock’s model, take underwriters’ reputation into consideration.

Michaely and Shaw (1994)

Classify their data sample into two sets: a sample of “IPO master limited partnerships (MLPs)” and a sample of "regular" IPOs. The empirical results represent that the initial return of "regular" IPOs is significant but the mean initial return of MLPs is insignificant, which coincide with Rock’s winner’s curse hypothesis towards IPO underpricing.

Contrast

Ritter (1984)

“The hot-issue markets only occur in particular periods” which do not support Rock’s conclusion.

The basic rationale of principal-agent theory is that the information asymmetry among issuers, underwriters, and outside investors could creates agency costs, which results in an underpricing phenomenon that the initial return is significant different from zero (Filatotchev and Bishop, 2002). The tests of single explanation of agency theory arrive at controversial outcomes.

Baron (1982) firstly presents a model for the underpricing of initial public offerings, whose argument based on the information asymmetry between issuers and underwriters. The Baron’s model predicts a positive relationship between underpricing and information asymmetry. His investigation is essential but not unchallenged. Muscarella and Vetsuypens (1989) test the Baron’s model and their findings contrast with Baron’s model, which suggest that “there is no statistically significant underpricing difference between self-marketed offerings and other IPOs.” However, Muscarella and Vetsuypens test Baron’s model by using a small sample of underwriters so that whether their empirical result is general should be taken into consideration.

Later research combines the agent theory with adverse-selection model and therefore achieves more meaningful empirical results. Habib and Ljungqvist(2001) test the relationship between IPO underpricing and optimal promotion strategy. Their model depends on the Baron’s signaling theory and Rock’s adverse-selection theory, and the empirical evidence supports the predictions, which indicates that the refutation of Baron’s model is premature.

In addition, principal-agent theory is the theoretical basic of ownership and control hypothesis. Habib and Ljungqvist(2001) state that “organizational theory has increasingly drawn on agency theory”(Beatty and Zajac, 1994; Brennan and Franks, 1997) and the IPO underpricing is explained from the perspective of corporate governance. The following table shows a compare and contrast of empirical evidence of principal-agent theory.

Table 3.4: Compare and contrast of empirical evidence of principal-agent theory

Compare

Baron (1982)

The Baron’s model predicts a positive relationship between underpricing and information asymmetry.

Habib and Ljungqvist(2001)

Their model depends on the Baron’s signaling theory and Rock’s adverse-selection theory, and the empirical evidence supports the predictions, which indicates that the refutation of Baron’s model is premature.

Contrast

Muscarella and Vetsuypens (1989)

Test the Baron’s model and their findings contrast with Baron’s model, which suggest that “self-marketed offerings are characterized by statistically significant underpricing compared with other IPOs.”

3.2.2. Institutional explanations

Besides the information asymmetry explanation, several scholars investigate the IPO underpricing phenomenon from a different point of view. Institutional explanations focus on two parts: underwriter price support and legal liability.

Considerable studies about the IPO underpricing phenomenon believe that the IPO underpricing is “taken deliberately” (Ruud, 1993). For instance, Baron’s information asymmetry theory, Rock’s winner’s curse hypothesis, Allen and Faulhaber(1988), Grinblatt and Hwang(1989) and Welch(1989)’s signaling hypothesis suggest that the issuers underpriced the offer price in order to sell their shares. On the contrary, Ruud(1993) find little evidence that support the theories hold by above scholars and he figure out that the motivation for underpricing to attract outside investors is insignificant. Ruud’s empirical findings indicate that the large amount of initial return could be explained by “frequent market practice of underwriter price support or stabilization.” The empirical studies on the underwriter price support theory and IPO underpricing not only assess the mean of initial return, but also take the shape of distribution into account. The investigation of distribution of IPO unperpricing suggests that the underwriter price support theory can explain the abnormal initial return to some extent. The shape of distribution of new issue’s initial returns indicates that “the original impact and gradual withdrawal of underwriter price support in a number of ways (Ruud, 1993).”

The underwriter price support theory for the interpretation of underpricing is supported by several empirical literatures. Both Ritter (1991)’s empirical result about the short run aftermarket performance and Muscarella and Vetsuypens’ (1989)’s evidence from the test of information asymmetry hypothesis by using the sample of self-marketed IPOs coincide with the underwriter price support hypothesis.

The law avoidance hypothesis suggests that “since the Securities Act of 1933 makes participants in the IPOs who sign the prospectus liable for any material omissions, the only method to reduce the probability of prospective lawsuits is to underprice (Ritter, 1998).” An early study on the “price performance of common stock” made by Ibbotson (1975) has given an original idea about the lawsuit avoidance hypothesis which documents that underpricing is a substitution of insurance which could reduce the odds of lawsuits. Tinic (1988) introduces the legal costs into Rock(1986)’s model and tests the implications of lawsuit avoidance hypothesis with two sets of samples before and after the “Securities Act 1933”, finding out that the Securities Act is correlated the underpricing phenomenon.

Hughes and Thakor(1992) develop the Tinic’s model by taking the impact of underwriter into consideration. Their empirical work regard the “litigation risk” as a determinant of excessive initial return, which indicates that the legal environment and underwriter reputation play an important part when issuers decide the degree of underpricing.

On the contrary, the lawsuits avoidance hypothesis does not supported by some empirical evidence. According to the lawsuits avoidance hypothesis, the litigation risk can be reduced by underpricing the new issues. However, empirical results provided by Darke and Vetsuypens (1993) are insufficient to support the insurance hypothesis which represent that underpricing is an inefficient method to avoid a legal suit.

3.2.3. Ownership and control explanations

Numerous empirical researches focusing on the IPO underpricing phenomenon from the perspective of corporate governance conclude that the underpricing of IPO may provides an efficient way of retaining corporate control and reducing the agency costs.

Ritter (1998) defines “ownership dispersion hypothesis” as “IPO firms may deliberately underprice the shares for the purpose of generating additional demand and as a result there will be able to attract a large number of smaller shareholders.” The ownership which contains a large proportion of small investors may improve the liquidity of the secondary market as well as reduce the challenge from outsider shareholders. One of the most importance purposes of a firm going public is to form a proper corporate governance system, which include the board structure, ownership dispersion and corporate control. Baker and Gompers (2003) investigate the board structure and the function of venture capital and outside shareholders by using a data set comprising 1,116 firms’ initial offerings through modeling the bargaining process between management team and the outside investors. Unlike other existing literature, this paper focuses on the function of venture capital, and investigates the proper corporate governance to retain the corporate control after IPO.

The extensive empirical evidence represents a positive relationship between ownership dispersion and underpricing. Brennan and Franks (1995) state that the ownership dispersion can be regarded as a major function of corporate control. Booth and Chua (1996) model and test the correlation between underpricing and ownership dispersion. Compared with previous studies, their sample of IPOs is divided into two sets: “firm-commitment offerings” and “best-efforts issues”. In their empirical framework, the results show a positive relation between underpricing and oversubscription and suggest information asymmetry hypothesis should be introduced into the model to make the empirical results more sensible. Brennan and Franks(1997)’s empirical study using a sample of UK IPOs examines “how corporate governance mechanism progresses as a result of an IPO and how the IPO underpricing can be used as an efficient way of maintaining corporate control.” The empirical framework tests the hypothesis that there is a negative correlation between degree of underpricing and the size of investors. And the evidence is consistent with the prediction.

Additionally, Fernando, Krishnamurthy and Spindt (1999) develop the ownership dispersion hypothesis by linking the level of underpricing with offer price and target ownership structure. They suggest that the high priced new issues are underpriced as “a compensation for the monitoring and information benefits provided by the institutional investors”. Their empirical framework takes the price level into account, and states that the underpricing decision made by issuers and underwriters may depends on the price level, and the argument that “the price level comprises information about firm’s ownership structure and prospective performance” has been supported by empirical evidence.

Filatotchev and Bishop (2002)’s empirical study shows that the level of underpricing is negative associated with “share ownership of executive directors”, “IPO's board diversity” and “share ownership of the IPO's non-executive directors”. Their empirical work introduces the ownership and control hypothesis to the signaling theory, and develops the previous investigations by analyzing the characteristics of not only financial organization but also partnerships in the enterprise.

IPO underpricing can be explained by agency cost hypothesis. The main idea behind the agency cost theory is that the asymmetric information between issuers, underwriters and outsiders might create the agency cost which is one of the determinants of excessive initial return of a new offering (Filatotchev and Bishop, 2002).

Loughran and Ritter (2004)’s investigation concentrates on the explanation of IPO underpricing changing over time, and at the same time, their empirical results represent that the changes of IPO underpricing, especially during the in during the doc-com bubble period, can be explained by the additional agency costs created by firm’s decision maker’s target changing. Loughran and Ritter (2002) combine the behavioral finance approaches and the agency costs theory, and apply the expectation hypothesis to explain the IPO underpricing phenomenon. Their model offers a explanation for IPO underpricing which suggests that high initial return is a form of compensation for underwriters, because “investors are willing to offer quid pro quos to underwriters to gain favorable allocations on hot deals (Loughran and Ritter, 2002)”.

The discussion and critical analysis of explanation of IPO underpricing phenomenon above imply that we can add testable hypotheses mentioned in section 3.2. According to the particular characteristics of Chinese IPO market and legal environment, the Investigation should concentrates on the aspects of large degree of information uncertainty, financial regulatory, and ownership structure.

Need help with your literature review?

Our qualified researchers are here to help. Click on the button below to find out more:

Literature Review Service

Related Content

In addition to the example literature review above we also have a range of free study materials to help you with your own dissertation: