Do Chinese underwriters grandstand to

attract more firms when they are ready

to go public?

Authors: Jian Jiao

Xuan Guo

Supervisor: Barbara Cornelius

Student
Umeå School of Business
Spring semester 2010
Master thesis, two-year, 30 hp
 Acknowledgments

At the end of our study, we feel that working and studying together for two years
without any doubts is our most wonderful experience. Our thesis could never been
completed without the combined effort of many people, who shaped our ideas and
thoughts, as well as those who provided us the motivation to carry forward. We would
like to thank everyone that has supported us along the way.

Firstly, we would like to show our gratitude to our supervisor: Professor Barbara
Cornelius who provided us with continuous and instructive feedback during the
progress of the study. She helped us to achieve everything with our maximum capacities.
She was always very patient, careful and helpful, and she spent a lot of time on our
thesis, even though helped us to correcting the grammars. Thank you very much for
your extraordinary support.

Second, we would like to thank Gisela Taube-Lyxzen, Program Coordinator，who made

Jian’s late application possible; as well as Monica Palmqvist, International Coordinator,
who helped us dealing with the problem of exchange study. We also would like to thank
the Umeå school of Business, Umeå University, which provides us with such great
facilitates and knowledgeable teachers.

Finally, but not least, we want to thank our respective parents who supported and
encouraged us all along our studies. Without their love we will not finish our study.

We love this small and peaceful city and the winter is so beautiful. We will always keep
this memory in hearts.

Jian Jiao
Xuan Guo
Umea, May 2010

i
 Abstract

The concept of grandstanding comes from Gompers (1996), in his article, he defined “to
grandstand” as “to act or conduct oneself with a view to impressing onlookers”. The
idea of grandstanding does not only apply solely to venture capital but also could apply
to underwriters of IPOs industry as well.

IPOs activities provide huge revenues for underwriters, so underwriters compete with
each other for IPO business. China’s stock market grows explosively after 2006, and it
has the highest underpricing, as well as more and more underwriters have emerged
recently, so our paper is constrained under Chinese stock market environment. We
empirically examine whether inexperienced underwriters grandstand when they conduct
IPOs in order to achieve more market shares, for example by deliberate underpricing or
charging lower fee rates.

This study is conducted from the underwriter’s perspective. We use two kinds of
reputation measurement methods to define “inexperienced” and “prestigious
underwriters” and employ a quantitative approach to analyze the data. Evidence from a
sample of 392 IPOs from June 19, 2006 to March 24, 2010 suggests that inexperienced
underwriters do not have incentives to grandstand. The number of IPOs that
underwriters have conducted and recent IPO performance do not always contribute to a
gain of market share directly. Therefore, inexperienced underwriters do not provide
more underpriced IPOs nor do they charge lower fee rates. Evidence also marginally
supports that underwriters do not intend to conduct small offer sized IPOs.

Key words: Initial Public Offering; underpricing; Grandstanding hypothesis; Chinese

stock market; underwriter

ii
 Table of Contents
1 Introduction ............................................................................................................. 1
1.1 Background .................................................................................................... 1
1.2 Main purpose of our study ............................................................................. 3
1.3 Research question .......................................................................................... 3
1.4 Delimitation .................................................................................................... 3
1.5 Disposition ...................................................................................................... 3
2 Theoretical Framework And Literature Review .................................................. 5
2.1 Initial public offering ..................................................................................... 5
2.1.1 IPO process ......................................................................................... 5
2.1.2 Book-building method ........................................................................ 7
2.1.3 Participants in IPO ............................................................................. 8
2.2 Information asymmetry................................................................................ 10
2.3 Underpricing phenomenon ...........................................................................11
2.4 Underwriting fee........................................................................................... 13
2.5 Financial intermediary and certification .................................................... 14
2.6 Reputation .................................................................................................... 15
2.7 Agency problem in IPO process................................................................... 16
2.8 Grandstanding hypothesis ........................................................................... 17
2.8.1 The origin of grandstanding hypothesis ......................................... 17
2.8.2 The implication on underwriter ...................................................... 17
3 Methodology ........................................................................................................... 19
3.1 Choice of subject and preconceptions ......................................................... 19
3.2 Theoretical perspective................................................................................. 19
3.3 Underlying philosophy ................................................................................. 20
3.4 Scientific approach....................................................................................... 21
3.5 Research strategy.......................................................................................... 21
3.6 Secondary sources and criticism.................................................................. 22
4 Empirical Study ..................................................................................................... 24
4.1 Research design ............................................................................................ 24
4.1.1 Comparison analysis......................................................................... 24
4.1.2 Correlation analysis .......................................................................... 24
4.1.3 Regression analysis ........................................................................... 24
4.2 Sample and data collection .......................................................................... 25
4.3 Variable measurement .................................................................................. 26
4.3.1 Measurement of underwriter’s reputation ..................................... 26
4.3.2 Variables used to examine the dependent variable-market share 27
4.3.3 Variables used to examine the IPOs related dependent variable . 28
4.4 Regression model.......................................................................................... 30
5 Empirical results and analysis .............................................................................. 31
5.1 Descriptive statistics ..................................................................................... 31
iii
 5.1.1 IPOs sample description .................................................................. 31
5.1.2 Comparison analysis of sample distribution .................................. 32
5.1.3 Comparison analysis of underpricing ............................................. 34
5.1.4 Market share presentation ............................................................... 38
5.1.5 Descriptive statistics on IPO related variables .............................. 40
5.1.6 Descriptive statistics on underwriters............................................. 43
5.2 Correlation analysis ..................................................................................... 44
5.3 Regression analysis ...................................................................................... 45
5.3.1 Market share ..................................................................................... 45
5.3.2 Underpricing ..................................................................................... 50
5.3.3 IPOs offer size ................................................................................... 54
5.3.4 Underwriting fee rate ....................................................................... 56
6 Conclusions............................................................................................................. 60
6.1 Summary....................................................................................................... 60
6.2 Limitation ..................................................................................................... 62
6.3 Further study ................................................................................................ 62
6.4 Credibility criteria ........................................................................................ 62
6.4.1 Reliability .......................................................................................... 62
6.4.2 Validity ............................................................................................... 63
Reference ....................................................................................................................... 64

iv
 List of Tables
Table 3.1: Research method summary .............................................................................................. 22

Table 4.1: Summary of variable measurement ................................................................................ 27

Table 4.2: Summary of variable measurement (continued) ........................................................... 28

Table 5.1: Summary of the sample (first reputation measurement method) ................................ 32

Table 5.2: Summary of the sample (second reputation measurement method) ........................... 33

Table 5.3: Summary of the whole sample group underpricing level ............................................. 36

Table 5.4: Descriptive statistics of variables (first reputation measurement method) ................ 41

Table 5.5: Descriptive statistics of variables (second reputation measurement method) ............ 42

Table 5.6: Descriptive statistics of underwriter (first reputation measurement method) ........... 44

Table 5.7: Descriptive statistics of underwriter (second reputation measurement method) ...... 44

Table 5.8: Pearson Correlations Matrix ........................................................................................... 46

Table 5.9: Regression Analysis of market share in the whole study period .................................. 47

Table 5.10: Regression Analysis of market share in year 2009 ...................................................... 48

Table 5.11: Regression Analysis of underpricing level (1) .............................................................. 51

Table 5.12: Regression Analysis of underpricing level (2) .............................................................. 52

Table 5.13: Regression Analysis of IPOs offer size .......................................................................... 55

Table 5.14: Regression Analysis of underwriting fee rate (1) ......................................................... 57

Table 5.15: Regression Analysis of underwriting fee rate (2) ......................................................... 58

v
 List of Figures
Figure 2.1: The IPO process .............................................................................................7

Figure 5.1: Summary of the underpricing of IPOs according to the issue size (1)........ 35

Figure 5.2: Summary of the underpricing of IPOs according to the issue size (2)........ 35

Figure 5.3: Summary of underpricing in different year (1) ........................................... 37

Figure 5.4: Summary of underpricing in different year (2) ........................................... 38

Figure 5.5: Shanghai stock exchange composition index from year 2006 to 2010 ........ 38

Figure 5.6: Total market shares (first reputation measurement method) ..................... 39

Figure 5.7: Total market shares (second reputation measurement method) ................. 39

Figure 5.8: Market share in year 2009 (first reputation measurement method)........... 40

Figure 5.9: Market share in year 2009 (second reputation measurement method) ...... 40

Figure 6.1: Top 10 underwriters in Wang (2010)’s report ............................................. 61

vi
1 Introduction

In this chapter we present the background and main purpose of our study first. Then
three research questions and the delimitation of our study are also presented.

1.1 Background

The concept of grandstanding comes from Gompers (1996). In his article he defined “to
grandstand” as “to act or conduct oneself with a view to impressing onlookers”, where
he reported that young venture capital firms are willing to bear the cost (e.g., greater
underpricing and hold smaller equity stakes) by taking firms public earlier than older
venture capital firms in order to establish reputation and raise new fund. The idea of
grandstanding does not only apply solely to venture capital but also could apply to
underwriters as well. So we apply the grandstanding hypothesis to underwriters in IPO
industry.

IPOs activities provide huge revenues for underwriters, so underwriters compete with
each other for IPO business. During the IPO process, underwriters act as financial
intermediaries and information providers. Their credibility and reputation is essential.
Brau and Fawcett (2006) found that issuing companies select underwriters mainly by
two criteria, namely, underwriter’s reputation and expertise. The prestige of
underwriters that have more successful IPO records is the evidence of expertise and
reputation, which helps the underwriters to attract potential clients who are willing to go
public. In order to reach this level of performance, inexperienced underwriters (having
less successful IPOs records) have the incentive to grandstand when they conduct IPOs
in order to signal their worth to the market that they also have the ability to fulfill the
client’s need.

The reason for us to conduct our paper on underwriter’s perspective is that we find a
gap in previous research. There is a tremendous amount of literature reporting studies in
the IPO underpricing area, most of researchers extend their research mainly from the
investors or issuing companies’ perspective (Logue, 1973; Beatty & Ritter, 1986;
Titman & Trueman, 1986). In our research, we draw from existing theories regarding
IPO underpricing issues, i.e. agency conflicts between underwriters and issuers, and
information asymmetry in order to examine whether the inexperienced underwriters
acting as an agent of issuing companies deliberately offer underpriced issues in order to
achieve their own benefits.

Furthermore, our paper is constrained under Chinese stock market environment. The
reason is that Chinese stock market is characterized with highest underpricing level (Su
& Fleisher, 1999; Dimovski & Brooks, 2004) and the stock market is quite young. The
most important fact is that more and more new underwriters emerged in recent years in

1
China which creates a clear boundary of inexperienced underwriters and existing
prestigious underwriters. Also Chinese stock market was applied the book-building
method to set up the final offer price in 2005 (Cheung, Ouyang & Tan, 2009; Gao,
2010). Only if there is competition, it makes the “grandstanding” concept applicable on
underwriting industry in China. By taking this advantage, we could empirically examine
do these inexperienced underwriters grandstand when they take issuing companies
public in order to attract more clients.

China’s stock market was established in 1990. There are two securities markets, namely,
Shanghai Stock Exchange (SHSE) and Shenzhen Stock Exchange (SZSE). These were
founded on December 19, 1990 and July 3, 1991 respectively. Only Chinese citizens
were allowed to trade A-shares in these two stock exchanges. While in 1992, foreign
investors were allowed to trade B-shares exclusively, or shares listed in Hong Kong
stock market. B-shares are identical to A-shares and have the same ownership and
voting rights, receive the same amount of dividends. The only difference is that
B-shares are traded with U.S dollar for Shanghai B-shares and Hong Kong dollar for
Shenzhen B-shares. Ordinary A-shares are traded in local currency which is yuan,
formally called Renminbi (RMB). Recently, domestic investors are allowed to trade
B-shares also. The listed companies’ number increased from 53 in 1992 to 1718 in 2009
(CSRC, 2010). For simplicity, in this paper, we only focus on A-share IPOs.

China’s stock market grew explosively after 2006 and it ranks the second largest market
by market value after the U.S (China daily, 2010). Because of the bull-run market
periods, more and more firms listed in public and raised huge amount of money. So we
have a large number of IPOs as raw data. Also China’s IPOs market has the highest
underpricing level in the world (Su & Fleisher, 1999; Dimovski & Brooks, 2004);
normally the first day return is exceeding 80%. Due to the changing degree of
government regulation, the underpricing level decreased simultaneously (Cheung et al,.
2009). Cheung et al. (2009) also mentioned that Chinese IPOs market transformed from
a regulatory controlled system into a more market-oriented system. At the beginning of
2005, the book-building mechanism (detailed information will be explained in literature
review chapter) was introduced into market which allows underwriters to set the final
offer price based on the information received from investors (Cheung et al., 2009; Gao,
2010). Because the market orientation was not adopted in that country until that point,
more meaningful results would be obtained if data collected were restricted to a time
after this event. So our research will pay attention to the period when the book-building
method applied in IPOs.

All in all, we are trying to fix the gap by examining the IPOs from underwriter’s
perspective. This could be the academic contribution of this study. Also the results will
be meaningful for investors, underwriters, issuing companies and regulatory officials in
Chinese stock market. This research provides evidence to issuing companies and
regulatory officials whether inexperienced underwriters sacrifice issuing companies’
interest in order to achieve underwriters’ own benefits. Do investors gain higher

2
abnormal return by participate the IPOs which are conducted by inexperienced
underwriters? This study would also be meaningful to underwriter, because we will
provide evidence to empirically examine this issue.

1.2 Main purpose of our study

The main purpose of this paper is that first, we empirically examine to what extent
underpricing exists. Then, we examine whether inexperienced underwriters grandstand
when they conduct IPOs, for example by deliberate underpricing or charging lower fee
rates etc. Third (if possible), if inexperienced underwriters do not grandstand, what
other factors could determine their underpricing, e.g. offer size and fee rate for IPOs?
We will empirically to find out answers to these questions.

1.3 Research question

 To what extent does the underpricing phenomenon exist in China?

 Do inexperienced underwriters grandstand when they bring firms public in order to

attract more IPO clients? (This is our main research question)

 What characteristics are associated with IPOs conducted by inexperienced

underwriters?

1.4 Delimitation

The study is limited to the issuer who launched their IPOs in Chinese stock market.
Sample selected from those IPOs which are after the book-building pricing system was
applied into Chinese stock market. It gives the underwriter and issuer rights to price
their IPOs and makes the grandstanding hypothesis applicable. If the price is regulated
by officials, it is impossible for underwriter to deliberately underprice the issue.

1.5 Disposition

Chapter 1 Introduction: In this chapter we present the background and main purpose of
our study first. Then three research questions and the delimitation of our study are also
presented.

Chapter 2 Theoretical framework and literature review: In this chapter we present

the theoretical framework for this study. The theoretical choices are based on a literature
review from which we have derived our four hypotheses.

3
Chapter 3 Methodology: In this chapter, we will present the methods which were used
in conducting our study. As well as the choice of subject, research perspective, research
philosophy, research approach, research strategy and information about secondary
sources used.

Chapter 4 Empirical study: In this chapter we will explain our research design and
variable measurement method. We also include how we collected the data, and present
our regression model at the end of this chapter.

Chapter 5 Empirical results and analysis: In order to examine whether the

inexperienced underwriters grandstand or not when they bring firms public, we firstly
find out what are the characteristics of these IPOs under different choices of
underwriters. The relevant findings will be presented in this chapter. Tables, figures and
graphs will be used in order to make the results clear. Moreover, the results within the
different choice of underwriters will be tested in statistical way to add to the validity
and reliability of the study.

Chapter 6 Conclusion: In this chapter, we summarize and conclude research findings

and limitation of this paper, also make the proposal for further research.

4
2 Theoretical Framework And Literature Review

In this chapter we present the theoretical framework for this study. The theoretical
choices are based on a literature review from which we have derived our four
hypotheses.

2.1 Initial public offering

Before we introduce the theoretical framework, we feel that it is necessary to provide a

whole picture of the initial public offering (IPO) which will help the reader to
understand the following section.

Initial public offering (IPO) is the process of selling shares to the public for the first
time. Going public offers the existing shareholders with flexible choice of exit and
greater liquidity. IPOs also enable them to diversify their wealth; they can easily cash
out later (Rock, 1986). In addition, Merton (2005) found that private companies that
have no access to public equity markets cannot easily raise new capital by issuing
shares. As the firm grows, they find that the shortage of capital becomes an obstacle
against chasing new growth opportunities. Public companies have better access to large
amounts of capital. This motivates companies to go public. The non-financial reason
motivating firms to go public include drawing attention from publicity, strengthening
the company image etcetera (Ritter & Welch, 2002).

2.1.1 IPO process

When companies decide to go public, the whole process can be divided into several
stages. At the beginning, issuing companies will choose a stock market in which to list.
An international capital flow gives firms the opportunity to choose which capital market
they are willing to go to. Different capital markets have various characteristics. Some
are labeled as developed markets, normally with less underpricing and more mature
investors. The rest are considered as developing markets with higher underpricing
(Loughran, Ritter & Rydqvist, 1994).

After determining the choice of stock market in which to list, issuing companies can
hire underwriters to help them in going public. Underwriters manage the offerings, for
example, they may have an arrangement at the first, by choosing one offer mechanism.
There are three offering mechanisms in total, firm-commitment, best-efforts and auction.
We will describe each of them separately.

First, referring to the work of Berk and Demarzo (2007), they describe that in the
best-efforts IPOs, underwriters do not guarantee that it will sell all of the stock but
instead they promise that they will try to sell the issues at the best price and raise

5
maximum amount of capital. Using this method, underwriters actually act as the
intermediary between the buyers and sellers, and underwriters get rid of the risk of a
failed IPO. More commonly, as Dunbar (1998) argued, unsuccessful issues are costly
for issuers both in time and money, so issuers are more likely to choose the underwriting
method which can provide them with greater ex ante probability of success. Clearly,
best-efforts mechanism does not guarantee the success of IPOs.

Second, in the firm-commitment IPOs, underwriters guarantee that they will purchase
all the shares from the issuing companies at a specific price and then re-sell these shares
to the public. The purchase price is lower than the public offer price. The difference
between the purchase price and offer price is called the spread. This spread is a discount
to compensate underwriters for taking the risk of unsuccessful offerings. In this
mechanism, the risk of failed IPOs is transferred from issuers to underwriters.

Another mechanism by which underwriters can determine a price for an IPO is the
auction, involving selling shares directly to the public through an online system, rather
than setting the price itself in the traditional way, such as firm-commitment and
bests-efforts. Auctions allow investors to bid the price as they want. After the bidding
period, the IPO final offer price is automatically determined. It is the lowest bid price
which the total bidding volume equals the number shares included in the offer. All the
bidders get a share allocation with this price even if they bid higher. But this mechanism
is not so popular recently (Sherman, 2005).

After deciding the offer mechanism, underwriters start doing the paper work for issuing
companies which is referred to as the prospectus. Prospectus is the most important file
to both regulatory officials and public investors (Daily, Certo & Dalton, 2005).
Normally, IPO’s prospectus must disclose the firm specific information like agreements
between issuer and underwriters, the issuer self-introduction, financial information, the
capital ownership structure and further investment opportunity of raised capital (Daily
et el., 2005). The prospectus serves as a communicating document, providing the
issuing companies’ internal information to the market. Jing, Bilson and Powell (2008)
found that those younger and less reputable companies do not disclose earning forecasts
in their IPO prospectus, which causes higher underpricing as compensation. As we see,
the more information the prospectus discloses, the less information asymmetry between
the market and the company.

In the next stage, underwriters are going to plan for the timing and marketing of the IPO.
There are two types of pricing methods: book-building and fixed price (Ma & Faff,
2007). The most commonly used pricing method is called book-building. Because all
IPOs in our sample used book-building method to set up the offer price, so we also
discuss this concept more in a later section. Once the price range is set out, underwriters
start to arrange the road show (where senior management and underwriters explain their
rationale to investors) to promote the IPO. In the book-building period, they gather the
purchase information from potential investors. Investors give their feedback as how

6
many shares they wish to bid and the price they would like to accept. And then
underwriters will actively adjust the final offer price to satisfy both issuer and investors
(Hanley, 1993). As Cornelli and Goldreich (2001) mentioned, book-building is the
process for adding the total demand of potential bid and adjust the price unless the IPO
is seems to success. The other pricing method is fixed price (Ma & Faff, 2007). The
offer price is fixed before the potential investors submit their bidding demand. Fix
pricing method is commonly used if the market is under regulatory control.

Finally if the IPO ends by successfully raise the money for issuing companies, issuing
companies will pay the commission fee for the services received (the fee varies with the
choice of offer mechanism). Now we have briefly described the whole IPO process
(Figure 2.1) and the three main participants (issuer, underwriter and investor) here. We
will explain and discuss the role and function of each participant in detailed in the
following section.

Figure 2.1 The IPO process

Modified from: Jenkinson, T. & Ljungqvist, A. (2001).

2.1.2 Book-building method

As we mentioned in introduction, we choose the data after year 2005 at which point the
book-building method was applied in Chinese stock market. So this pricing method is a
critical selection measurement of our data, we explain the detail of this method here and
gives the reader an overview.

Sherman (2005) mentioned that book-building is the most used and controversial
method for setting the issue price because it allows underwriters the opportunity to
preferentially allocate shares. The normal procedure is that: First, underwriter work
closely with the client to come up with a price range before they set up the offer price.
After that underwriter will arrange a road show (a marketing process which aims to

7
promote the issuing company to known investors, especially institutional investors) to
promote company and gather information mainly from these institutional investors
(Cornelli & Goldreich, 2001).

The potential investors will inform the underwriter about their interests and tell them
how many shares they would like to purchase. Cornelli, Goldreich (2003) and Berk,
Demarzo (2007) said although this bid is not an actual purchase but normally the total
demand from customers is accurate. So underwriters could adjust the offer price based
on the potential purchase information they have acquired. The process that underwriters
add up investors’ total demand and then actively adjust the offer price until the IPOs
seems not to fail is called book-building. Because underwriters can change the offer
price as they wish, book-building method may let underwriters underprice the IPO. The
privilege of underwriters is that underwriters allocate the shares depending on the
information they collected. If they believe one specific investor has contributed to them,
underwriters can allocate the shares to that investors as reward. Without that privilege,
investors do not have enough incentives to provide their information accurately
(Benveniste & Spindst, 1989). In this way, underwriter may sacrifice the issuer’s
interest to consort with the investors which cause potential agency conflicts. For
example, underwriters can underprice the IPOs and allocate to specific investors. We
will discuss this agency problem in the section 2.7.

2.1.3 Participants in IPO

Issuers, investors and underwriters are three most important parties in IPO process. And
the following discussions are closely tied with these three parties’ relationship, such as
the agency problem between underwriters and issuers, information asymmetry between
investors and issuers, etc. So it is essential to explain and discuss them one by one.

Issuer

Issuers refer to issuing companies who go public to raise equity capital funds for their
growth and provide a public market in which the existing shareholders can convert their
shares into cash (Ritter & Welch, 2002). They could sell their existing shares or issue
new shares to investors. They want to sell their shares at a reasonable price, or issue
new shares with higher price in case of the dilution of their existing share’s value thus
avoiding leaving too much money on the table. Issuing companies will hire underwriters
to help them selling the shares.

Researchers have demonstrated that the issuing company’s choice of underwriter is

based on their company‘s own specific criteria and then pick up the underwriters which
satisfy their needs (Titman & Trueman, 1986; Habib & Ljungqvist, 2001). In contrast,
Fernando, Gatchev and Spindt (2005) suggest that issuers and underwriters are mutually
selected. That is, the quality of the issuer and the ability of underwriter interact in the
selection process. The potential motivation for issuers to choose to hire underwriters

8
rather than sell the shares directly is that issuing company expects underwriters to fulfill
their requirements and maximize their interests. For example, underwriters could help
them to attract more investors, add to the credibility of their IPOs etc.

Investor

From the investor’s perspective, they want to make quick profits through their
participation in IPOs or invest on those companies that have a prosperous future
(Cornelli, Goldreich & Ljungqvist, 2006). So how much information they have about
the issuing company will affect their choice of investment. For example, the price they
would like to bid and the volume they tend to hold are affected by their valuation of the
IPO. If the offer price is higher than their valuation, they will bid for less shares or
nothing at all.

According to investors’ fund size and expertise, investors are divided into two
categories, namely, institutional investors (also called regular investors) and retail
investors (Rock, 1986; Hanley & Wilhelm, 1995; Aggarwal & Prabhala, 2002).
Institutional investors have a higher ability to detect the intrinsic value of IPO than
retail investors (Rock, 1986). Institutional investors share their information with
underwriters and issuers at the book-building stage where underwriters query them
regarding their interest in purchasing shares of the IPO including how many shares
investor desire and the bid price they will accept (Certo, 2003). “The offering price is
based on results from a book-building process directed at institutional investors
(Cheung et al., 2009, p. 696)”.

Commonly, institutional investors have an important status in the IPO process due to
their supportive role and large bidding volume. More specifically, investors provide
their private information (i.e. their bidding price reflects their valuation about this IPO)
in the book-building stage and support the IPO, the more shares institutional investors
bid, and the more shares underwriters can count into their pre-selling estimation. The
more pre-selling they get, the less failure possibility of the IPO. Benveniste, Spindt
(1989) and Cornelli, Goldreich (2001) found that the underwriter awards more shares to
bidders who provide information in the book-building stage and the regular investors
receive favorable allocations.

If investors do not submit enough bidding, underwriters have to adjust the offer price to
attract more investors’ attention. Rocholl (2009) found that institutional investors
support the underwriter when the issue is weak, and demand the favorable share
allocation in those underpriced IPOs as return for proceeds. He named this relationship
as “a friend in need is a friend indeed” (Rocholl, 2009, p. 284). Underwriters have
incentives to build relationships with institutional investors rather than issuer. In that
situation, underwriters could underprice the IPO to consort with their investors.

9
Underwriter

Underwriters who help the issuing companies to sell their shares to investors, they have
a clear contract which includes the terms and commitments with issuer. The job of
underwriters is that they use their professional knowledge to help issuing companies
pricing their issue accurately, add credibility and certify that the offer price represents
the true value of the issuing companies. In Booth and Smith's (1986) model,
underwriters could certify that the issue is not overpriced by using their reputation
because overpricing issues would harm their reputation. Also Beatty and Ritter (1986)
argue that “the first-day return is costly since future issuers would avoid underwriters
that leave too much money on the table”. James (1992) found out the pricing errors at
the time of IPO could lead to the decision of changing lead underwriter. So underwriters
cannot only consort with their investors, but also have to balance their relationship with
issuer. Otherwise, they will lose their client and potential issuer.

The prior researchers divided underwriters into prestigious underwriters and

inexperienced underwriters with different methods (Johnson & Miller, 1988; Megginson
& Weiss, 1991; Carter & Manaster, 1990). The importance of reputation of seller has
already been recognized on the spread of corporate bond (Takaoka & Mckenzie, 2006).
So in IPOs market, there is expectation that underwriter’s reputation also has a great
influence on the underwriting outcomes (Mckenzie & Takaoka, 2008). We will test
whether the prestigious underwriters have larger and more positive effects on IPOs than
inexperienced underwriters in the empirical results section.

2.2 Information asymmetry

We present this section here, because information asymmetry is factor said to lead to
IPO underpricing (Ritter & Welch, 2002) which is a key issue of our study. As we
mentioned before, that asymmetric information exists between each of the three
participants in IPOs, namely, underwriter, issuer and investor. We found several studies
related to information asymmetry in IPOs.

First, underwriters as the financial intermediary between issuer and investor have the
opportunity to obtain privileged information through pre-selling activities. The
asymmetric information between underwriters and the issuing company may cause the
underwriters to set a lower offer price by using their superior market information and
knowledge, reduce their marketing efforts and making their offer more attractive to
please their bidders (Baron & Holmstrom, 1980; Baron, 1982). Consistent with Ritter’s
(1984) work, underpriced offers harm small, less well known issuing companies.
Underwriters can build and maintain long-term relationships with investors through
deliberately under pricing the offer to favor selected investors. For example, if investors
have their personal information regarding the valuation of the issue, underwriters would
compensate sophisticated investors with underpriced offers in exchange for their

10
accurate bidding and pricing information at the book-building stage (Benveniste &
Spindt, 1989; Benveniste & Wilhelm, 1990; Spatt & Srivastava, 1991).

Second, Allen, Faulhaber (1989) and Chemmanur (1993) argued that issuing companies
have the superior information compared with outsiders. Insiders (issuers) could benefit
from the situation when outsiders (investors) overestimate the issue value (Myers &
Majluf, 1984). If the asymmetric information between issuing companies and potential
investors exist, outsiders will consider the issuing company has higher ex ante
uncertainty. Beatty and Ritter (1986) found that there is a positive relationship between
ex ante uncertainty of the issuing company and underpricing. The large sized firms are
considered as well-known and less uncertain. Also a firm with a long history is
considered with less uncertainty and the issue size has a negative relationship with IPO
underpricing. So issuing company surrounded with higher uncertainty normally has a
higher level of underpricing.

Third, the asymmetric information exists between informed investors (normally refer to
institutional investors) and uninformed investors (refer to retail investors). Rock (1986)
found that because the informed investors (those who have privileged information) have
stronger ability to know the intrinsic value of IPOs. As Rock developed the winner’s
curse model, explained that uninformed investors will receive a higher proportionate
allocation of IPOs with lower initial returns, and lower proportionate allocation of IPOs
with higher initial returns because of the increased demand from informed investors.
Informed investors will only bid the good issue with a discount and withdraw from the
market when bad issue is offered. In contrast, those uninformed investors will bid all the
issues. So the adverse selection which refers to winner curse: uninformed investors get
all shares as the demand from others is low and the IPO is likely to perform poorly
(Rock, 1986; Levis, 1990). In this case, the offer must issue at a discount in order to
encourage those uninformed investors to anticipate the IPO.

2.3 Underpricing phenomenon

Many researchers argued that underpricing is partly due to the asymmetric information,
ex ante uncertainty between issuers and investors (Baron & Holmstrom, 1980; Beatty &
Ritter, 1986; Allen & Faulhaber, 1989; Grinblatt & Hwang, 1989; Welch, 1989;
Chemmanur, 1993). We present this section in order to provide reader the background
regarding this issue. Underpricing is the most important measurement factor in our
study, used to examine whether the grandstanding phenomenon exists.

Underpricing is also called “the money left on the table” and is defined as the difference
between offer price and first trading day closing price multiplied by the total amount of
shares sold (Loughran & Ritter, 2002). Underpricing is the phenomenon that the first
trading day return is positive, which is found all around the world in all financial
markets (Loughran, Ritter & Rudqvist, 1994; Loughran & Ritter, 2002).

11
Empirical documentation of underpricing has been done by several researchers.
Ibbotson (1975) was the first researcher who found that IPOs have a positive initial
return. Ritter and Welch (2002) reported that from 1980 to 2001, the average
underpricing in U.S is 18.8%. In China, Gannon and Zhou (2008, p. 492) reported that
the underpricing declined through time, which supported by prior researches: “Su and
Fleisher (1999) 949% during 1987-1995, Mok and Hui (1998) 289% during 1990-1993,
Yang (2003) 285% between 1991-2000, Gu (2003) 217% during 1994, Chi and Padgett
(2005) 129% between 1996 and 2000”. Su and Fleisher (1999) and Deng and
Dorfleitner (2008) found that during year 2002 to 2004, the underpricing level in
Chinese stock market is 88.67%. The adequate researches supported Loughran et al
(1994) who demonstrated that a developing market has higher underpricing than a
developed market. In our study, we use the newest data which is between 2005 and 2010
to examine whether this high level of underpricing still exists. If so, we will test whether
this high level of underpricing is the results of grandstanding phenomenon.

Underwriters play a key role in determining the offer price and influence the level of
IPO underpricing (Flagg & Margetis, 2008). Examples can be found in the data on
venture capital backed IPOs where significant differences in underpricing occurred
between VC backed IPOs and non-VC backed IPOs (Megginson & Weiss, 1991; Lee &
Wahal, 2004). By contrast, Bradley and Jordan (2002) found that after controlling for
underwriter quality, there were no significant differences in underpricing between VC
backed and non-VC backed IPOs. One conclusion that could be reached is that the
underwriters selected by the venture capitalists had a big influence on the IPO. Thus
underwriters directly impact on the degree of underpricing.

Other researchers, outside the area of venture capital, concur. Beatty and Ritter (1986),
Carter and Manaster (1990) found a negative relationship between underwriter prestige
and underpricing in 1990s. This means that the issuing firms who choose prestigious
underwriters would have higher proceeds from IPOs and have a lower level of
underpricing. But this relationship between underwriters’ prestige and underpricing has
reversed in the 1990s (Beatty & Welch, 1996; Cooney, Singh, Carter & Dark, 2001).
The reversed relationship motivates us to find out whether underwriters are determining
the level of underpricing.

From this, we develop our hypothesis as follows:

H1: IPOs provided by inexperienced underwriters have higher underpricing levels than
those provided by prestigious underwriters.

12
2.4 Underwriting fee

The issuing companies pay for underwriters’ services according to the underwriting
agreements. One typical example from Chen and Ritter (2000) shows that underwriting
fee are composed of gross spread, management fee and other underwriting fee. In a
firm-commitment offer, gross spread is the discount of the offer price which is
calculated as the difference between the offer price and underwriter’s purchase price.
Management fee and underwriting fee normally is the fix cost of issuing company.
Underwriting business is very profitable which Chen and Ritter (2000) reported that U.S
investment banker admitted that the IPO business is very profitable and apparently lack
of price competition. They do not want to compete on fees, over 90 percentage of the
medium size IPOs’ gross spread is seven percent (Chen & Ritter, 2000). Hansen (2001)
also reported that seven percent spread is optimal in competition for IPO contract which
is coherent with the results from Chen and Ritter (2000), and underwriters compete for
business in other ways such as reputation and the degree of underpricing of IPOs. In our
research we will test whether it have the same result that underwriters do not compete
on fee in Chinese market.

Furthermore, Chen and Ritter (2000) argued that the fee charged signals the
underwriter’s quality. For example if Goldman Sach cut its spread, issuers are
suspicious that Goldman Sach had become a low-quality underwriter. This assertion is
consistent with the empirical research done by Krigma, Wayne and Kent (2001). They
supported that underwriters do not compete on the basis of IPO fees. The issuing
companies pay more attention to underwriters’ quality, because underwriting fee is not
the most important factor to the issuing companies.

We argue that prestigious underwriters can receive higher fees, a factor which relies on
their reputation. In contrast, inexperienced underwriters do not have the reputation to
invest in IPO issue. The issuing company does not want to pay the same level of fee to
those inexperienced underwriters due to the fact that they lack sufficient reputation. In
that situation, inexperienced underwriters have to reduce the fee level as compensation.
Consistent with Torstila’s research (2001) that in Europe, IPOs backed by prestigious
underwriters have higher spread than those IPOs backed by inexperienced underwriters.
And Ljungqvist, Jenkinson and Wilhelm (2003) mentioned that European issuing
companies were likely to pay a premium, if top prestigious underwriters (e.g., Golden
Sach, etc.) were hired. So we will test whether this is consistent with the way it works in
Chinese stock market.

Except the reputation factors, underwriters’ fees also vary with the issuing companies
specific characteristics. Pichler and Wilhelm (2001) showed that the spread declines
with the increasing of offer size. Ahn, Kim and Son (2007) found that there is no
clustering pattern in underwriting fee in South Korea, but two separate clustering ranges
for small and large IPOs are found. This research supports the assertion that when the

13
IPO has a larger offer size, the underwriter will charge a smaller fee rate.

Bartling and Park (2009) argued that issuers would like to enjoy the proceeds of higher
a issue price, but underwriters only have incentives to set high offer pricse with higher
fee rates. This is because higher issue prices will increase the risk taken by the
underwriter. Most investors are aware of this process (Bartling & Park, 2009). Based on
our hypothesis 1, if IPOs backed by inexperienced underwriters have higher
underpricing level, inexperienced underwriters would compensate the issuer with lower
fee rates. Since institutional (or regular) investors have influence on underwriters, they
will force underwriters to offer more underpriced issues with lower fee levels as they
can benefit from these issues. Underwriters are more likely to serve institutional
investors than issuers (Hoberg, 2007). It motivates underwriters to choose to offer more
underpriced issues and charge lower fee rates.

In addition, Carter and Manaster (1990) reported that prestigious underwriter charge
higher fees to the issuer than inexperienced underwriters. This is because inexperienced
underwriters cannot provide as much reputation as prestigious underwriters for
attracting investors and certifying the IPO’s quality. We will extend the underwriter’s
certification function in the next section.

From the discussion above, we develop our second hypothesis as follows:

H2: Inexperienced underwriters charge lower fee rates than prestigious underwriters.

2.5 Financial intermediary and certification

Financial intermediary is the middleman between the insider and outsider. Klein and
Leffler (1981, cited in Booth & Smith 1986, p. 261) demonstrated that “The
intermediary reputation capital is an effective bond to guarantee the quality”. Some
researchers apply the intermediary reputation capital theory into financial market later.
For example, DeAngelo (1981) mentioned that auditor as an intermediary between
public investor and corporation, invest their reputation as a kind of guarantee of
financial statement accuracy. Wakeman (1981) discussed the similar issue that rating
agency as an intermediary between investor and bond issuer certifies the corporate bond
grade (e.g. AAA bond).

We apply the financial intermediary theory to our study. Underwriter is the financial
intermediary between issuer and investor. On one hand, issuers need the intermediary
because they are not the professionals. They need to hire specialists to help them to
conduct the IPO. Underwriters repeatedly doing business in equity market, pricing the
IPO and communicating information to investors, that makes the idea of the underwriter
as a financial intermediary possible (Chemmanur & Fulghieri, 1994). As a specialist,
underwriters also could help issuing companies to attract more investors, and certify the

14
offering price is accurate and reliable. On the other hand, investors also need
intermediary, because they believe that asymmetric information occurs when the seller
has superior information compared with the buyer. It is like the “lemons issue” which
discussed by Akerlf (1970). Investors do not know the intrinsic value of the issue firm
and their inside information. So investors are afraid of overpaying for the IPO, and they
need intermediary (underwriter) to disclose the true information of issuer.

Underwriters fulfill the needs of both parties. So underwriters, as a financial

intermediary, reduce the information asymmetry, and also reduce the cost for the issuer.
By adding credibility, the reputation of the underwriter is important. Booth and Smith
(1986) derived a certification hypothesis that issuing companies could benefit from
underwriters who use their reputation as an effective bond to certify the issue.
Megginson and Weiss (1991) in an empirical study proved that prestigious underwriters
certified issue had a lower level of underpricing. Prestigious underwriters more
effectively reduce the information asymmetry, increase the transparency and
creditability of the information disclosure which can reduce the underpricing level.

Both investors and issuers valued underwriters’ reputations by their past performance
(Chemmanur & Fulghieri, 1994). And their reputation is gained from repeatedly doing
business in the market. So it’s important for underwriters to set up strong and reliable
reputations to convince their clients and principals, make their certification more
powerful. We will discuss the reputation issue in the next section.

2.6 Reputation

Reputation is not only an invisible way used for self-control, but it also like a kind of
guarantee that is used for promoting a business to potential clients and earning
economics benefits. Diamond (1989) proved that reputation is important for having
access to debt and equity markets. Less reputable companies find it harder to raise funds
from the market. Fang (2005) found that economic rents are earned by reputation, which
motivates underwriters to maintain their reputation.

In Booth and Smith’s (1986) work, they suggest that underwriters can bond with
investors by putting their reputation capital into the issue to certify the IPO issue, at that
time investors would regard the certification as effective and reliable. If the increased
value (additional money raised) from hiring prestigious underwriters is higher than the
cost of certification (extra payment for hiring prestigious underwriter), it means that
certification will increase the issuing companies’ offer value. Issuing companies would
benefit from using prestigious underwriters. In order to attract potential clients,
prestigious underwriters need to maintain their reputation. In contrast, inexperienced
underwriters do not have reputations yet. To build a reputation involves a deliberate
sacrifice of short-term profits, like underpricing the issue, and receiving fewer fees in
order to get more successful records and signal their ability to bring firms public. These

15
issues are what we are going to test.

Fernando et el. (2005) suggest that issuers and underwriters are mutually selected. From
the issuer’s view, issuer would like to choose prestigious underwriter in order to attract
more investors, have larger analyst coverage, less underpricing and better long-term
performance (Michaely & Shaw, 1994; Carter et al., 1998). From the underwriter’s view,
Zhang, Bartol, Pfarrer and Khanin, (2008) mentioned that “individuals are loss-averse
and tend to forgo the possibility of a gain when it involves the potential for loss relative
to one’s current position” (p. 242). In the IPO case, underwriters are loss-averse, and
they do not want to risk their reputation to gain small underwriting fees. Underwriters
are likely to underwrite those IPOs with large offering size (Carter & Manaster, 1990).
The smaller size issuers then have to work with inexperienced underwriters.

From discussion above, we derive our next hypothesis as follow:

H3: Inexperienced underwriters work for smaller offer sized IPOs than prestigious
underwriters.

2.7 Agency problem in IPO process

In the IPO process, the issuing company hires an underwriter to give them advice, to
help them to determine the offer price, to promote and distribute the issues to potential
investors (Chen, Fok & Wang, 2006). By applying agency theory in the IPO case,
issuers become principals and underwriters act as an agent. They both have their own
interests and goals. For the issuing company, they want to get a higher offer price in
order to raise more capital and maximize the benefits of the IPO. Underwriters only
have incentives to set up higher offer price with appropriate rewards. In contrast, the
underwriter has incentives to care more about interests of their own such as their
relationship with their investors and, as a consequence, they demand more underpricing
to ensure the issue is attractive in the market. Loughran and Ritter (2004), Reuter (2006)
found some evidence that underwriters do provide favorable share allocation to
institutional investors. Binay, Gacthev and Pirinsky (2007) found that regular
institutional investors are likely to participate in IPOs offered by the same underwriters.
Moreover, more underpricing may please the investor who has strong ties with the
underwriter. Thus if issuer and underwriter’s interests are not aligned well, underwriters
will violate the issuer’s interest in order to please institutional investors thus supporting
their interests in the future.

According to agency theory, there are two ways to ease the conflicts between agent and
principal, which are effective monitoring and appropriate incentives (Agrawal &
Knoeber, 1996). First, principal can monitor the agent’s actions by themselves. The
second way is to use performance based instruments. Bonding is the more popular
mechanism (Eisenhardt, 1988).

16
To solve the agency problems in the IPO case, issuers and underwriters interact to reach
an underpricing equilibrium. Issuers cannot tolerate underpricing if it is too high. Also
underwriters will not set the offer price with too low a level of underpricing without
incentives. Researchers have provided evidence that “underwriters who cheat on the
underpricing equilibrium will lose either potential investors if they do not underprice
enough or issuers if they underprice too much” (Beatty & Ritter, 1986). Underwriting
fees can be treated as a bonding method. If issuers choose to use a floating underwriting
fee, that means the total payment of the issue varies with the proceeds of IPO. The
floating underwriting fee has an impact on underwriter’s behavior and it aligns the
issuer with underwriter directly. If issuers provide higher fee rates for the IPO, it is a
strong incentive for underwriters.

2.8 Grandstanding hypothesis

In this section, we explain the concept of the grandstanding hypothesis and how we
apply this concept to underwriters. In addition, our research models are based on this
hypothesis, so it is essential to detail this hypothesis for better understanding of the
following empirical study
.
2.8.1 The origin of grandstanding hypothesis

Gompers (1993) developed the model of grandstanding and applies this model in a
venture capital market environment. Then Gompers (1996) provided evidence that
young venture capital firms take companies public earlier and more underpriced in order
to establish their own reputation and raise new funds. They also have been on board a
shorter period, and incur the cost of holding smaller equity stakes. Before the young
venture firms have their first IPO, they can hardly raise enough money for their second
fund. But after their successful records of IPO, they raise enough money in a very short
period (Gompers, 1996).

Gompers (1996) designed a way to derive the benefits of additional IPO. He argued that
if the investors will increase their assessment of venture capitalists ability, venture
capitalists gained the benefits from additional IPOs. And the amount of next fund a
venture capital can raise is related to their perceived ability. The cost of younger venture
capitalist bring firms public earlier is not trivial. Younger venture capitalists will benefit
from the next capital raising. Old venture capital firms with good reputations do not
need to incur this cost to signal their ability.

2.8.2 The implication on underwriter

In this paper, we apply this grandstanding hypothesis to underwriters in the Chinese

IPO market environment. Inexperienced underwriters have incentives to grandstand if it

17
helps them to accumulate reputation and they can benefits from it. By deliberately
underpricing the offer or charging lower fees rate in order to have a more successful
IPO record, they can achieve market recognition, attract more potential clients and
obtain more market share.

Similarly, issuing companies believe that bringing firms public is a strong signal that
underwriters are skilled at underwriting business. We develop a suggestion that
inexperienced underwriters deliberately underpriced the IPO more, as well as charge
lower fee rate than prestigious underwriters. Inexperienced underwriters conduct IPOs
with smaller offer size than prestigious underwriters. We assert that after a successful
IPO, the issuer will increase have a more positive attitude toward to inexperienced
underwriters. Because issuer already knows more about the prestigious underwriters, an
additional IPO would not affect their judgment on prestigious underwriter’s ability as
much as on inexperienced underwriters. If underwriter’s market share is an increased
function of their competence, then these costs (i.e. extra underpricing, low fee rate) is
not trivial.

From this, we develop our fourth hypothesis as follows:

H4: After each additional IPO, inexperienced underwriters’ market share has more
positive effects than prestigious underwriters.

18
3 Methodology

In this chapter, we will present the methods which were used in conducting our study. As
well as the choice of subject, research perspective, research philosophy, research
approach, research strategy and information about secondary sources used.

3.1 Choice of subject and preconceptions

Both of the two authors have studied several financial courses in Umeå School of
Business. Without study into this area, we cannot pick this topic and conduct this study. In
financial market and corporate finance course, we heard that China have the highest
level of underpricing in initial public offering all around world. So it inspires us and we
hope to figure out whether it is the truth and why this phenomenon exists. It is the initial
idea for our choice of subject. Since the Chinese stock market is an emerging market
and regulation is not as mature in the US or European market, we have the opportunity
to explore some interesting results. Meanwhile, we are both very interested in the stock
market. Especially one of the authors has been an active investor in Chinese stock
market for several years. Moreover, he also has the experience of bidding initial public
offering. He has a feeling that sometimes the IPOs have huge underpricing, but
sometimes the first day initial return even is negative. Because he does not pay attention
on the IPOs’ underwriter, so in this research he will go to examine the impact of
underwriters on underpricing. For all the reason above, we decide to study the subject of
underpricing phenomenon in initial public offering and find deeper related factor around
these.

Our experiences and theoretical knowledge lead us to this topic. However, these kinds
of experience and training created our understanding of the surroundings by
subconscious, which is recognized as preconceptions (Bryman & Bell 2007). Based on
the knowledge we got from Master’s program in accounting, assisting with the
statistical tools and prior researches help us to treat the real figure more objectively.
Because we are aware of the importance to maintain objectivity, we have collected data
from reliable resources and from official agency. To avoid inserting our own bias we
have statistics report of data by using standard scientific principles.

3.2 Theoretical perspective

As we mention in last chapter, there are three participants in IPO process, who are issuer,
investor and underwriter. The investor and issuer perspective have been adopted in the
majority of prior empirical studies. In this study, we set out to answer the research
question from underwriter perspective, which guide us to gather and analysis the data.
Underwriters play two important roles in IPO process: the financial intermediary

19
between issuers and investors, also the agent for issuers. This double fold perspective
brings us deeper understanding of the study.

In this paper, we will test whether the Chinese underwriter grandstand when they bring
firm public. So grandstanding hypothesis and the implication on underwriter were the
most important to be described. Except that, there are a large number of theories related
to our study. We choose to provide our reader with the most relevant theories in the
literature review chapter. First, we introduce the background of IPOs process and its
participants, which helps our reader to gain a clear overview of our study. Then we
discussed about the information asymmetry theory which is one main reason of
underpricing. After that, we explain the underpricing phenomenon and how it was
measured by prior researchers. In order to provide a clear explanation regarding the
underwriter’s two functions which are financial intermediary and agent of issuer, we
present the financial intermediary theory and agency theory respectively. At the end, in
order to justify the rationale of dividing underwriter as prestigious underwriter and
inexperienced underwriter, we developed the reputation section.

3.3 Underlying philosophy

Research philosophy contains assumption about how we view the knowledge. So

understanding philosophical issues well can help us to recognize which research
strategy and research method will work for a given circumstance, how to use these
strategies and methods outside our previously experience, and how to avoid no use
work through clarifying the chosen methods (Smith & Milina, 1999). Ontology and
epistemology view usually been considered as this point.

Ontology refers to what the real world is and what the nature of reality, which of
independent particulars (Bryman & Bell, 2007). Objectivism and constructionism are
two branches of ontology. According to Bryman and Bell (2007), objectivism view is
that social reality is external to the researchers’ mind and the reality is objective
remained independent of research. While constructionism is more subjective, which
means reality in the social world does not wholly exist out there but is constructed.
“Social phenomena and their meanings are continually being accomplished by social
actors” (Bryman & Bell, 2007).

Epistemology position describes how we study the real world and explains human
behaviors (Bryman & Bell, 2007). Two branches are positivism and interpretivism,
respectively, which guide research to examine the truth from a given ontological
orientation. Positivism holds a natural science view. Researchers can measure the reality
by some scientific method, in order to test hypotheses and then get knowledge about it,
which is objective study process. On the other hand, interpretivism points that social
science is different from the natural sciences, researchers require a new way to interpret
the reality. Social action is a subjective position and hardly be measured like natural

20
science directly, so researchers need understanding what is happened in order to
develop a new idea regarding that.

Considering our undertaken research, the reality is that underwriters helped issue
company went to public. We just observed this objective process, which happened
whatever we observe it or not. Thus, our ontological consideration is objectivism. Based
on this ontological position, our epistemological consideration is positivism. We
formulated the hypotheses by some measurements, like underpricing level, offer size,
fee rate, and then using liner regression tools to test the result. It is the objective
analyzing way for data.

3.4 Scientific approach

The purpose of business research is to “speculate, discover, and document, as well as to

provisionally order, explain, and predict, (presumably) observable social processes and
structures that characterize behavior in and of organizations” (Maanen, Sørensen &
Mitchell, 2007). Further, theory is "an ordered set of assertions about a generic behavior
or structure assumed to hold throughout a significantly broad range of specific
instances" (Sutherland, 1975, p. 9). Scientific approach can help us build the
relationship between research and theory. Deductive and inductive are two basic
approach for scientific research (Bryman & Bell, 2007).

Deductive approach means that based on the specific knowledge and existing theory,
researchers create some hypotheses and then collect data to test these hypotheses
(Bryman & Bell, 2007). At the end of this, researchers usually reject or confirm the
hypotheses. It is a process from theory to finding. In contrast, inductive approach works
from finding to theory. Through the data collection methods like interview or
observation, researchers study the unknown phenomena in order to develop new
theories (Bryman & Bell, 2007).

Regarding to our study, the purpose is not to develop new theory, but to test the existing
theory under different environments. More specifically, we applied the grandstand
concept which comes from Gompers (1996) on underwriters in IPO process under
Chinese stock market. We generated hypotheses and analyzed the data using statistical
methods to test the theory. For these stages, we believe the deductive approach is
appropriate.

3.5 Research strategy

According to Bryman (2007), there are two strategy systems for business research,
quantitative method and qualitative method, which will affect the way of data collection.
Quantitative method emphasizes quantification in the collection and analysis of data,

21
more about the numeric data (numbers). On the other hand, qualitative method focuses
on non-numeric data (words) rather than quantification in the collection and analysis of
data (Bryman & Bell, 2007). Qualitative usually helps to deeply understand or evaluate
a specific case.

Except for data, the other essential distinction is the assumption of research. A
quantitative method is often associated with deductive approach to test the theories by
develop hypotheses, and shows a positivism epistemological and objectivism
ontological position. In contrast, a qualitative method always uses inductive research to
generate theories, and shows an interpretivist epistemological and constructionist
ontological orientation (Bryman & Bell, 2007). In order to present these relationship
clearly, we will sum them by the table below:

Table 3.1 Research method summary

Strategy Quantitative Qualitative
Approach deductive inductive
Epistemology positivism interpretivism
Ontology objectivism constructionism
Original source: Bryman & Bell. (2007), p. 28.

The choice of research strategy depends upon the aim of research. For our research
question, we tend to exam whether the underwriters grandstand when they take the
company go to public. We need to gather IPO related data from company’s prospectus
which listed on China Securities Regulatory Commission (CSRC), also the first trading
day stock price information from Tonghuashun stock information software (2010). All
of these deal with number, which gives us appropriate measurement of each IPO’s
character, like underpricing level, size and so on. The data provides us evidence to
confirm the theories our research based on. So the quantitative method is suitable for
our study.

In summary, the combination of objectivism ontology and positivistic epistemology

with deductive approach using quantitative method is the most appropriate for our
research.

3.6 Secondary sources and criticism

Searching secondary sources is an essential part of research, which provides researchers

background information and valuable knowledge around the research area. “While the
researchers also should treated secondary sources with the same caution as any primary
data and the secondary sources should be able to answer the research question and meet
the research purpose” (Saunders, Lewis & Thornhill, 2007, p. 263).

In our research, the objective is to discuss about the IPO underpricing issue, in order to

22
examine whether the inexperienced underwriters acting as an agent of issuing
companies deliberately offer underpriced issues or not. For this, we searched the
literature and articles through the databases Business Source Premier and Emerald Full
text at the library of Umeå University, also the book from ALBUM library book
searching system. Firstly, by using the key words “IPO” and “underwriter”, we gather
the whole picture of IPO process from the previous research. Secondly, key word
“underpricing” gives us another theoretical support for our study. Next, we search
“IPO” together with “financial intermediary”, “information asymmetry”, and “agency
problem” respectively, which helped us to understand the underwriter’s role in IPO the
process and inspired us in the following study. Finally, “reputation” helps us divide the
underwriters into two groups: experienced and inexperienced underwriters. Except these
accesses, we also search the sources from internet, for example “Google Scholar”.

However, when we use these previous studies, it is important to evaluate the

independence of these materials independence, whether they are free from false views,
how recent it is and how relevant (Bryman & Bell, 2007). Because the inaccurate
materials or unreliable sources will lead us to the wrong way of research, so we should
treat these supplementary materials cautiously and critically. For instance, internet
source as one popular tool should be treated carefully, in determining whether it is
reliable or could be considered scientific. Meanwhile, we try to be objective to find
credible and relevant sources. Many of these articles have been examined, reviewed and
published on the high quality scientific journals.

23
4 Empirical Study

In this chapter we will explain our research design and variable measurement method.
We also include how we collected the data, and present our regression model at the end
of this chapter.

4.1 Research design

In this section we will explain the methods that were applied in the analysis and how
this research design helps us to derive answers regarding our research questions. To find
out the potential benefits of IPO for underwriter, and what kinds of action underwriters
take in order to reap those benefits. Three types of analysis were used to prove the
results both descriptively and statistically.

4.1.1 Comparison analysis

In our paper, we conduct this part in order to provide readers with a general picture
about whether there is a difference between the IPOs depending upon the choice of
underwriters. Specifically, are there direct underpricing differences resulting from the
use of prestigious underwriters. As our IPO samples are distributed in different years
and stock exchanges, they also have different age and size scales. All of these results are
reported in a comparison analysis. It is important to examine the underpricing level if
we want test the grandstanding hypothesis. So we fully describe it in detail, tables and
graphs are used where necessary. At the end of this part, we present the two kinds of
underwriting business market share, namely, the total market share segmentation and
the latest market share in year 2009 segmentation.

4.1.2 Correlation analysis

In order to find out the relationship between each variable, we conducted a correlation
analysis to deal with our data. For example, we input our data into SPSS, and conduct
correlation analysis. We present the Pearson correlation matrix in Table 5.8 that
includes the correlation of each variable. Using the Pearson correlation matrix allowed
us to avoid the multi-collinearity problem which would like skewed our results.

4.1.3 Regression analysis

After the comparison analysis and correlation analysis, we use regression analysis to
examine the statistical relationship of between variables and report which factors have
statistics significance in our models. We developed four models in order to examine
different dependent variables. In this section we only provide the reason why we build
these models. In addition, the detailed models are reported in section 4.4. The first

24
model is used to check does additional IPOs number and recent IPO performance could
provide benefits for underwriter. Hypothesis 4 could be tested in that model. Second
model is used to show whether the underpricing level is associated with each variable, if
the dummy variable (underwriter reputation) shows significance, we could find out if
underpricing varies with the choice of underwriter. So hypothesis 1 could be tested in
this model. Similar method in model 3 and 4 does the dummy variable show significant
difference on this dependent variable is the main purpose. Hypothesis 2 and 3 are tested
in each model. In the mean time, we also could take a look at which variables have the
critical impacts on each model’s dependent variables. We will explain the specific
variables of each model in the 4.3 section.

4.2 Sample and data collection

In our research, data has been collected from Chinese stock market and all the IPOs in
our sample are issued under stand-by underwriting method. This method is similar to
firm-commitment method except that firm-commitment method requires purchasing the
issue upfront (How & Yeo, 2000). Since China adopted the book-building mechanism in
IPOs from January 1, 2005, and because the market orientation was not adopted in that
country until that point, more meaningful results would be obtained if data collected
were restricted to a time after this event (Gao, 2010). So our research sample data (to
make it clear, our sample is time sample, but it’s the whole population of the research
period) will be collected in this period when the book-building method applied in IPOs.
Another fact is that the book-building method was adapted in 2005, but the first
book-building case emerged in June 19, 2006. The reason for that is the stock market
was not fully functioning, also the market performance was poor and was illiquid.
Regulators stopped IPO issues from 2005 until the first case emerged in June 19, 2006
(Zhongguancun, 2010). So in fact we collected our data starting from June 19, 2006
until March 24, 2010. This included all 393 IPOs from Shanghai, Shenzhen A share
stock exchange market. One of these IPOs called “Lili electronic” failed and was unable
to be listed in SZSE, so we deducted it from our study and our final sample includes
392 IPOs.

All IPOs prospectus were listed on China Securities Regulatory Commission (CSRC)
official information disclosure website (Juchao Information, 2010), which is publicly
available information. So we checked all the prospectus one by one and acquired all the
variable data we needed, which included：issuing firm name，the age of the issuing
firm，company industry, issue volume，offer size，underwriting fee，underwriter name，
total number of shares outstanding after issuance，book value per share before issuance
and the firm's past performance data like ROA. We use ROA as our firm performance
indicator because it is the compulsory information disclosure requirement both in
issuing companies’ financial reports and prospectus (Ping & Li, 2003). However, some
variables are not shown in IPOs prospectus, for example, the listing date，issuing date，
offer price, first day market close price and offer price P/E. We obtained this

25
information from Tonghuashun stock information software, which is a leading stock
software including all Chinese history stock price charts and each company’s financial
performance data. Then we manually entered this data into Excel and double-checked
the data.

4.3 Variable measurement

4.3.1 Measurement of underwriter’s reputation

In this section, first we reviewed the prior researches regarding how to measure the
underwriter’s reputation. Then we introduce the reputation measurement of our study
and justify why we use this kind of measurement to define the prestigious underwriter
and inexperienced underwriter.

The prior researchers (Johnson & Miller, 1988; Megginson & Weiss, 1991; Carter &
Manaster, 1990) developed a method of measurement on underwriter’s reputation. First,
Johnson and Miller (1988) developed a four-tier reputation measurement method. A
rank of three is termed for the most prestigious underwriters. A two is assigned to junior
underwriters who are less reputable than the last group. One is assigned to normal
underwriters. Zero is used for the least prestigious underwriters (Carter, Dark & Singh,
1998). Second, Megginson and Weiss (1991) use relative market share to measure the
reputation rank for each underwriters. Carter. el. (1998, p. 289) described that “Lead
underwriter for each IPO is given the full credit for the total amount underwritten, and
the relative market share for each underwriter is determined by dividing the
underwriter’s total credits by the industry’ total”. Third, Carter and Manaster (1990)
built a ten-tier (From 0 to 9) reputation measurement ranking scale established by
comparing tombstone announcements. For example, Cater and Manaster (1990)
assigned a rank for each underwriter in the tombstone announcement according to its
position. The underwriter whose name showed at the top of the announcement was
assigned the rank of nine. Underwriters who stayed in the later position were assigned a
rank of eight, etc.

In our paper, we use two types of measurement to define the prestigious underwriter.
The reason for that is the prior methods were developed under U.S stock market
environment. Many of the IPOs are underwritten by an underwriter syndicate in U.S. In
addition, in China, in most cases of IPOs are underwritten by a single underwriter. The
stock market is quite young, Chinese underwriters do not have a clear level which could
separate them. So Johnson and Miller (1988) and Carter and Manaster (1990)
measurement methods were not applicable in Chinese stock market. The applicable
method is Megginson and Weiss (1991) which used relative market share as the
measurement. Since most of Chinese IPOs were backed by a single underwriter, we
modified the Megginson and Weiss (1991) method and applied it to Chinese stock
market.

26
We set up two indicators to reflect the underwriter’s market position. First, it is based on
the total issue size. Second, it is based on the number of issues. Since our data sample is
starting at year 2006. We gather the data of underwriter in year 2006 from Securities
Association of China (SAC, 2010), which reported the underwriters of top 10 number of
issues and top 10 issue size. So our sample was divided into two groups based on the
two measurement methods. First method is that underwriters who rank in top 10 issue
size are defined as prestigious underwriters. The other underwriters are defined as
inexperienced underwriters. Second, underwriters who rank in top 10 of the number of
issues are treated as prestigious underwriters. The left underwriters who do not rank in
any top 10 are coded as inexperienced underwriters.

4.3.2 Variables used to examine the dependent variable-market share

In the first regression model, the dependent variable is the underwriter’s market share.
The first independent variable is a dummy variable used to control for underwriter
reputation. Then we use the total number of IPOs for each underwriter as the second
variable. This variable is similar with Gompers (1996) study where he used the total
number of IPOs as one variable to examine whether there is a positive effect on the
dependent variable. The next variable is the recent IPOs performance. We have used the
latest IPOs offer size as a proxy for this variable. Specifically, we choose two types of
market share to examine the relationship. First, the market share of the whole sample
period helps us to observe the whole picture of market. Second, the latest market share
of year 2009 helps us to capture the recent movements of market. This two dimensional
analysis help us to avoid bias and increases the accuracy and creditability of the results.
The following variables are defined and all the variables are summarized in the table:

Table 4.1 Summary of variable measurement

Variables Definition
Underwriter reputation - measured by issue size 1-Inexperienced;
(Reputation 1) 0-Prestigious
Underwriter reputation - measured by number of 1-Inexperienced;
issues (Reputation 2) 0-Prestigious
Total number of IPOs (Number) Total number of IPOs
Numbers of IPOs * Reputation 1 Number * dummy variable 1
Numbers of IPOs * Reputation 2 Number * dummy variable 2
Most recent offer size (Recent Size) Recent offer size amount
Most recent offer size * Reputation 1 Size * dummy variable 1
Most recent offer size * Reputation 2 Size * dummy variable 2

Underwriter reputation (measured by issue size): use underwriter data of year 2006
from Securities Association of China. Underwriters who rank in the offer size top 10 are
treated as prestigious underwriters and code as 0. The others defined as inexperienced
underwriters and code as 1.

27
Underwriter reputation (measured by number of issues): use underwriter data of
year 2006 from Securities Association of China. Underwriters who rank in top 10 of the
number of issues are treated as prestigious underwriters and code as 0. The others
defined as inexperienced underwriters and code as 1.

Total number of IPOs: The total number of IPOs which had been conducted in the
whole sample periods (from year 2006 to 2010).

Numbers of IPOs * Reputation: The total number of IPOs multiplies the dummy
variable of underwriter reputation.

4.3.3 Variables used to examine the IPOs related dependent variable

The rest of the regression models are conducted to check different factors of IPOs (ie.
underpricing, offer size and fee rate). Dummy variable of underwriter reputation is put
into each model in order to examine the relationship of choice of underwriter and
dependent variable, namely, difference in underpricing, offer size and fee rate with the
choice of underwriter.

Table 4.2 Summary of variable measurement (continued)

Variables Definition
Offer size (Size) Offer price multiply the total offer shares
Firm age (Age) Days between offer date and firm
establishment date
Net asset before IPOs (NTA) Total net asset before IPOs
Return on net asset (ROA) Net income divided by total net asset
Offering P/E ratio (P/E) Offer price divided by EPS
Audit firm choice (Audit) 1-Non Big 4 audit firm; 0-Big 4 firm
Earnings per share (EPS) Net income divided by total shares in the
last reported year
Turnover (TO) First day trade volume divided by total
offer share
Lottery rate (LR) Offer size divided by total subscription
Stock exchange (SE) 1 -Shenzhen stock exchange;
0 –Shanghai stock exchange
Underwriting fee rate (Fee) Total underwriting fee divided by total
offer size
Time lag between issue and listed date Days between the issue date and listed
(Lag) date
Underpricing (UP) (First trading day close price - offer price)
/ Offer price

We also use several variables to control for other effects. Like the previous studies had

28
been done before on IPOs, for instance Gannon and Zhou (2008) use offer size and
ROE as the indicator for asymmetric information, market P/E, offer P/E, initial turnover
as proxies for bandwagon effects in their model. Welch (1992, cite in Yong, 2007, p.
257) explained bandwagon effects as “potential investors pay attention not only to their
own information about a new issue, but also to whether other investors are purchasing.”
Initial turnover in Chinese stock market is much higher than developed market. Zhu and
Tian (2002) find that initial turnover in Chinese stock market is 57.91% compared with
8.2% in developed market. We extend it by adding the firm age as another proxy.
Aggarwal (2000) found cold issues are over allocated, and hot issues are normally
flipped which proved market demand of issue is also need to be considered. We use
lottery rate as control for market demand. Some financial and market effects variables
are defined and summarized in the table as follow:

Return on net asset (ROA): The net income divided by the total net asset before IPOs.

Offering P/E ratio (P/E): The offer price divided by the earning per share (last reported
year).

Underpricing: First trading day close price minus the offer price first, and then divided
by the offer price.

Lottery rate: The total offer size divided by total effective subscription.

Underpricing:

The first day initial return is usually as the measurement of IPO underpricing. First day
initial return is defined as the differences between the offer price and first day closing
price. It is calculated as follows:

P1
Undepricing (First day initial return) = −1
P0

P1 denotes the first-day closing price; P0 denotes the offer price (Chan, Wang & Wei,
2004; Chi & Padgett, 2005; Deng & Dorfleitner, 2008; Guo & Brooks, 2008).

Some researchers were considering the market effects between the issuing day and
listing day, so they use the market adjusted initial return as the measurement of IPO
underpricing. The market adjusted initial return is calculated as follow:

P1 Pm1
Market adjusted undepricing (Market adjusted initial return) = −
P0 Pm0

Pm1 is the closing market index of the corresponding issue’s first trading day.
Pm0 is the closing market index of the corresponding issue’s offer day (Chan et al.,

29
2004; Chen et al., 2004; Deng & Dorfleitner, 2008; Guo & Brooks, 2008).

Underwriting fee:

We want to measure the underwriting fee rate, and the total commission fees as well as
the total IPO offer size. Some prior research split the fee into several parts (i.e. gross
spread, management fee and underwriting fee) (Chen & Ritter, 2000). In our study, we
only need to count the total payment of issuer as their expense for hiring the
underwriters. So we denote total amount of money (F0) charged by underwriters as their
fee. The offer price (P0) is the final offer price for the IPO issue. The total amount of the
share denotes as V0. The offer size (P0 × V0) represents the total proceeds from the
IPO (before deduct the fee). Then we calculate the underwriting fee rate as follow:

F0
Underwriting Fee = .
P0×V0

4.4 Regression model

As we have introduced the variables used in each model, we construct the regression
models as follows:

Model 1:
Market share = β0 + β1Reputation + β2Number + β3 Reputation*Number + β4Recent
Size + β 5Recent size*Reputation+ ε

Model 2:
Underpricing = β0 + β1Reputation + β2Size + β 3Age + β 4ROA+ β 5P/E + β6Audit +
β7TO + β 8LR + β 9Fee + β 10Lag + β 11SE + ε

Model 3:
Offer size = β 0 + β1Reputation + β2NTA + β 3Age + β 4ROA + β 5EPS + β6Audit + β7TO
+ β 8LR + β 9SE + ε

Model 4:
Fee rate = β 0 + β1Reputation + β2Size + β3UP + β 4Age + β 5ROA + β 6P/E + β 7Audit +
β8TO + β9LR + β 10SE +ε

30
5 Empirical results and analysis

In order to examine whether the inexperienced underwriters grandstand or not when

they bring firms public, we firstly find out what are the characteristics of these IPOs
under different choices of underwriters. The relevant findings will be presented in this
chapter. Tables, figures and graphs will be used in order to make the results clear.
Moreover, the results within the different choice of underwriters will be tested
statistically.

5.1 Descriptive statistics

5.1.1 IPOs sample description

In our study, all the 392 companies hired underwriters when they went public. Fifty
seven different underwriters served all these IPOs. “Guoxin Secuirities” underwriter
was most frequently used, which was 46 times out of the 392 IPO issues. The following
most active underwriter was “Ping’an Securities” underwriter with 41 issues.

We also categorized the whole sample depending on different size scale, namely, large
sized IPO and small sized IPO. With the measurement that if the total offer’s size is
higher than 500,000,000 RMB, the IPO was categorized as a large sized IPO.

The IPOs could be listed on two stock exchanges, namely, SHSE and SZSE. SHSE is
usually to accommodate the large offer volume IPOs, called “blue chips”. SZSE is
suited for small and medium sized enterprise (SME).

Another type of grouping method is to separate samples by the issuer firm age. Because
in our sample, we have the date of the issuer’s establishment and the offer date it was
possible for the issuer’s firm age to be computed by Excel formula. It makes the number
quite big, the median is around 3000, so we set the cut-off point at 3000 days. When the
issuer’s firm age is bigger than 3000 it is termed as an older firm, otherwise it is a
young firm.

Our IPO samples were conducted by different types of underwriters, i.e. prestigious
underwriters and inexperienced underwriters respectively. In our research, we have two
types of measurement to identify prestigious underwriters. First, reputation is measured
by issue size in the starting year, 2006. If an underwriter is ranked in top 10 issue size of
year 2006, then we define the underwriter as a prestigious underwriter. Second,
reputation is measured by the number of issue in 2006, if an underwriter is ranked top
10 in the number of issue in 2006, then we define the underwriter as a prestigious
underwriter. These two types of measurement may give us the opportunity to derive
different results with different measurement methods. We could make a comparison and

31
discussion based on two potential distinct results. If we get the similar results with
different measurements, then the results are robust and consistent with the research done
by Yu and Tse (2006), Yu and Tse (2006) used four different proxies to measure one
variable and proved their results were robust in a similar fashion.

Our IPO samples are distributed over 5 years, the first sample taken at June 5, 2006.
The last one ends at March 24, 2010. We also categorized sample by year of 2006-2010.
In a nutshell, we use five grouping methods to take a first glance on our sample data,
namely, by offer size, stock exchange, firm age, issue year of 2006-2010 and
underwriter reputation.

5.1.2 Comparison analysis of sample distribution

Refer to Table 5.1, under the first underwriter measurement method, we find that SHSE
group has 44 IPOs conduct by prestigious underwriters and only 15 IPOs bring by
inexperienced underwriters. Prestigious underwriters not only have the high proportion
of in SHSE, but also in large sized firms, with 102 IPOs (55.4% of their total IPO
issues). In contrast, SZSE are the stock exchange which suits for small and medium
sized firms listing (Deng & Dorfleitner, 2008). SZSE group has more IPOs taken by
inexperienced underwriters with the number of 206 compare to 127 (38.1% of total
number in SZSE) taken by prestigious underwriters.

Table 5.1 Summary of the sample (first reputation measurement method)

Prestigious (by issue size) Inexperienced
Number Underpricing Number Underpricing
Year 2006 29 75.2% 40 92.08%
Year 2007 54 182.9% 65 204.26%
Year 2008 36 105.48% 40 125.13%
Year 2009 30 62.62% 33 64.43%
Year 2010 22 35.21% 43 50.37%
SHSE 44 (74.6%) 59.51% 15 90.99%
SZSE 127 (38.1%) 125.11% 206 120.84%
Large sized issue 102 (55.4%) 64.8% 82 82.86%
Small sized issue 69 (33.2%) 172.44% 139 140.14%
Young issuer 86 (41.5%) 132.42% 121 132.3%
Old issuer 85 (45.9%) 83.76% 100 102.48%
Total 171 108.23% 221 118.81%
[72.18%] [87.58%]
Number in the small bracket is the percentage of that group. For example, SHSE: 44
(74.6%) out of 59 IPOs in SHSE were in prestigious group.
Number in the mid bracket is the median number. For example, [72.18%] is the median
underpricing of prestigious group.

32
In addition, inexperienced underwriters are more likely to bring small sized firms public,
which proved by the number that 139 out of 221 in total. The possible reasons for that
are issuing companies which belong to “blue chips” types are more likely to be listed in
SHSE. Since prestigious are more likely to conduct IPOs with large volume because
large offer size means higher commissions and the issuing companies have less
uncertainty and the higher quality. Issuing companies also prefer to hiring prestigious
underwriters to signal their quality. The mutual selection procedure causes prestigious
underwriters take more IPOs (102 account for 55.4% of total number) which have large
offer size.

We find that prestigious underwriters have the similar proportion of IPOs both in
younger and older issuers, 41.5% and 45.9% respectively. So firm age is not a
significant variable to judge issuers what type of underwriters they would like to choose.
The possible explanation is that younger firms prefer to hire a prestigious underwriter to
add the credibility and signal their quality (Ljungqvist et al., 2003). Consistent with the
prior research, use prestigious underwriter signals the lower risk of the IPO and certifies
the price is true (Rock, 1986).

Table 5.2 Summary of the sample (second reputation measurement method)

Prestigious (number of issues) Inexperienced
Number Underpricing Number Underpricing
Year 2006 37 76.73% 32 94.54%
Year 2007 68 182.83% 51 210.21%
Year 2008 47 107.59% 29 129.17%
Year 2009 39 63.18% 24 64.19%
Year 2010 29 42.92% 36 47.1%
SHSE 42 (71.2%) 66.15% 17 70.88%
SZSE 178 (53.5%) 119.43% 155 125.96%
Large sized issue 117 (63.6%) 71.3% 67, 75.31%
Small sized issue 103 (49.5%) 152.38% 105 149.36%
Young issuer 111 (53.6%) 130.45% 96 134.55%
Old issuer 109 (58.9%) 87.67% 76 102.79%
Total 220 109.26% 172 120.51%
[79.08%] [86.32%]

As we refer to the Table 5.2, by using the second underwriter measurement method, we
find that prestigious still have the majority position in SHSE, they take 42 firms public
out of the total 59 firms, which consistent with the results above. They take 117 large
sized IPOs public compared with 67 IPOs are taken by inexperienced underwriters. In
contrast, inexperienced underwriters are no longer dominant in SZSE anymore. They
only take 155 (46.5%) firms going to public, their market share decrease by 15.4% as
we compare with the analysis above. The proportion of small sized IPOs is similar to
both types of underwriters, 103 versus 105 in number. The possible explanation is that if

33
the reputation measured by number of issues, due to the large numbers of IPOs provided
in SZSE, it is easier for underwriters to accumulate reputation by issuing more IPOs.
The difference of the proportion for prestigious underwriters in young firm group and
old firm group is still limited, 53.6% versus 58.9% which consistent with the first
measurement method. Moreover, the proportions for inexperienced underwriters still
fall behind the prestigious underwriters in large sized IPO even if we change another
reputation measurement method.

5.1.3 Comparison analysis of underpricing

In this section, we will analyze the underpricing phenomenon with the two different
measurements, namely, by issue size and the number of issues.

Issue size, our first measurement method, we found similar results to those mentioned in
the last section. That is, there it was suggested that prestigious underwriters are more
likely to play a key role in large sized IPOs. We found less underpricing with large sized
IPOs although the degree of underpricing is quite different and varies with the
measurement system used. In Table 5.1 under the first measurement method (by issue
size), the large sized IPOs brought to the market by prestigious underwriters have
significantly lower average underpricing levels (64.8% versus 82.86% by inexperienced
underwriters). The mean of underpricing in Table 5.1 and 5.2 supports the
grandstanding hypothesis that inexperienced underwriters bring firms public with higher
underpricing. But we needed to further test its significance in the later section. In Table
5.1, the average [median] underpricing is 108.23% [72.18%] for all the IPOs in the
sample period brought to the market by prestigious underwriters. Average [median]
underpricing is 118.81% [87.58%] for all the IPOs brought by inexperienced
underwriters. In Table 5.2, number of issues, our second measurement method, the
comparable average [median] underpricing is 109.26% [79.08%] and 120.51% [86.32%]
for prestigious and inexperienced underwriters respectively. We further examined this
relationship using regression analysis to determine the finding was supported later.

We find another interesting phenomenon in Table 5.1 where small sized IPOs
underwritten by prestigious underwriters have much higher underpricing compared with
less experienced underwriters (172.44% versus 140.14%). We made a bar chart to show
this result in Figure 5.1. The possible explanation is not from the underwriters’ side.
Since the IPOs were supplied in the primary market and after that were listed and traded
in the secondary market. We assume that this is caused by second market overreaction
to prestigious underwriters’ additional creditability. Because investors believe that
hiring prestigious underwriters is a strong signal revealing issuing companies’ quality,
investors in the secondary market are fascinated with those IPOs who hire prestigious
underwriters (Zhu & Tian, 2002; Gannon & Zhou, 2008). This overreaction is
consistent with the analysis before that younger firms hire prestigious underwriters to
signal their quality (Rock, 1986).

34
In contrast, under the second measurement method (by number of issues), the
comparable underpricing level in large sized IPOs is 71.3% for the prestigious
underwriter group versus 75.31% for the inexperienced group. The underpricing levels
are, however, close to each other (See Figure 5.2). It shows that the prestigious
underwriters do not provide extra certification effects to reduce underpricing, or
alternatively we could assume that inexperienced underwriters do not have a negative
effect on underpricing level based on this measurement. It also could reflect that
measured by number of issues no distinctive difference separates the underwriters. The
same results occur for small sized IPOs, no big differences on underpricing levels
(152.38% versus 149.36%). We will further examine both measurement effects in the
regression analysis section later.

Until now, we have made a brief comparison analysis of underpricing and have draw
two pictures to indicate that. But we have not tested the significance of difference yet.
In order to support the phenomenon and provide statistical support, we will conduct
t-test in the later section.

Figure 5.1 Summary of the underpricing of IPOs according to the issue size (1)

Figure 5.2 Summary of the underpricing of IPOs according to the issue size (2)

35
Description of underpricing in general

As we could see that the underpricing level varies with the year dramatically. In a bull
market, the underpricing is quite high and peak in year 2007 (194.57%). The maximum
of underpricing happened in year 2007 with 538.12%. This is quite high number which
means that the first day market closing price is over five times the offer price. In the
same year, even the lowest IPO has over 30% underpricing. It is strong evidence that
bull market boost the underpricing level. After year 2007, there is a significant decline
both in IPO numbers and underpricing levels. This is consistent with the great impact of
the financial crisis, equity market collapse and market stagnation. In year 2009 and
2010, the average underpricing is 63.56% and 45.24% respectively.

Table 5.3 Summary of the whole sample group underpricing level

Number of IPOs Mean Max Min
Year 2006 69 84.99% 345.71% 0%
Year 2007 119 194.57% 538.12% 32.25%
Year 2008 76 115.82% 403.54% 7.66%
Year 2009 63 63.56% 206.93% 2.34%
Year 2010 65 45.24% 147.3% -4.12%
SHSE 59 67.51% 329.53% -4.12%
SZSE 333 122.47% 538.12% -3.32%
Large issue size 184 72.76% 428% -4.12%
Small issue size 208 150.85% 538.12% 6.42%
Prestigious (size) 171 108.23% 496.07% -3.32%
Inexperienced 221 118.81% 538.12% -4.12%
Prestigious (number) 220 109.26% 496.07% -3.32%
Inexperienced 172 120.51% 538.12% -4.12%
Total sample group 392 114.2% 538.12% -4.12%

In general, large sized IPOs have lower underpricing (72.76%) when compared with
small sized IPOs (150.85%). Although we argued above that the second measurement
seems not to providing a distinction between underwriters. Both measurement methods
showed that prestigious underwriters bring firms public with less underpricing even if
the difference is small (108.23% versus 118.81% and 109.26% versus 120.51%
respectively). The mean of each group shows a difference but we need a statistical
examination to support this result. So we will put the all necessary statistics together in
the later section. All in all, the average underpricing of the whole sample is still in the
high level (114.2%). The results are consistent with prior research which shows that the
Chinese stock market has the highest underpricing level around the world (Su &
Fleisher, 1999; Dimovski & Brooks, 2004). Chi and Padgett (2005) empirically tested
that the average underpricing between 1996 and 2000 finding it was 129%.

36
Argument on market adjusted underpricing

Some prior researchers (Chan et al., 2004; Chen et al., 2004; Deng & Dorfleitner, 2008;
Guo & Brooks, 2008) adjusted the raw underpricing into market adjusted underpricing.
They suggest that the movements of market have an impact on the IPOs, so they
subtract these effects from the raw data. But in our study, we will not adjust the
underpricing into market adjusted underpricing. There are three reasons for that:

First, our study focuses on the comparison and analysis of underpricing variations with
the different choices of underwriters, not the underpricing change year by year. Even if
the market does have impact on underpricing, it affects both types of underwriters
simultaneously. So it is the one reason that we will not adjust the underpricing.

Second, after we summarize the data within different years, we observe from Figure 5.3
and 5.4 that the underpricing level peaks in 2007 and declines in the following years.
From Figure 5.5, we found the market index also peaks in year 2007 and had a huge
drop after that and rebounded in year 2009 and 2010. In our study we use Shanghai
Composition Index (SCI) as our market index (see Figure 5.5) because SCI is the most
influential market index in China (Tatro, 2010; Bloomberg, 2010). The underpricing of
each year has the similar trend with the market index which means the valuation of
IPOs is correlated with the market index. A conclusion can be reached that the
underpricing had already been adjusted the market itself. If we adjust the underpricing
by market index again, it will cause a double adjustment.

Third, the average of time lag between issue date and listing date is 12 to 13 trading
days in our data. And we compute that the average 12 days market movements in
Chinese stock market are less than 10% which it is too tiny, compared with the high
degree of underpricing (in our data, the average underpricing is 114.2%). So we can
neglect the market index effect and do not adjust the underpricing which maintained the
original characteristic of underpricing level.

Figure 5.3 Summary of underpricing in different year (1)

37
Figure 5.4 Summary of underpricing in different year (2)

Figure 5.5 Shanghai stock exchange composition index from year 2006 to 2010

Data source: Tonghuashun stock information software (2010)

5.1.4 Market share presentation

The purpose of this section is to provide an overall impression of the market share
segmentation. We examine whether the change of market share are caused by
grandstanding effect.

We provide four kinds of market share segmentation. The reason to do that is we

thought the underwriters may have incentives to grandstand if they can achieve a market
share increase. So it is meaningful to look into the pattern of both total market share and
the latest market share in year 2009. If we only take a look at the total market share, we
cannot observe the latest change in our sample. If one inexperienced underwriter
absorbs adequate expertise through these years, now he is capable of being the
prestigious underwriters and obtains large enough market shares already. Only using
total market shares may not help us to find out this. So use the latest market share
provides us an opportunity to explore some new results. In addition, regarding with the

38
two kinds of measurement on prestigious underwriters, we have drawn these four
market share pie chart below.

Market share in whole study period

In this study, for simplicity, we use the word “total market share” instead of “market
share in whole study period” in the following paragraph. As we could see, Figure 5.6
and 5.7 show similar results even if we use two types of measurement. It means that the
total market share is quite stable. Prestigious underwriters take up around 70% of total.

Figure 5.6 Total market shares (first reputation measurement method)

Figure 5.7 Total market shares (second reputation measurement method)

The latest market share in year 2009

We demonstrate that the latest market shares segmentation changes about 10% in Figure
5.8 and 5.9. In year 2009, the prestigious underwriters only take up around 60% of the

39
market share. In addition, the two figures also show a slight difference in market share
with different measurement (59.72% versus 64.3% in prestigious group).

Figure 5.8 Market share in year 2009 (first reputation measurement method)

Figure 5.9 Market share in year 2009 (second reputation measurement method)

Here we could see that the prestigious underwriter measurement method does not affect
the total market share too much, which is 70.18% and 69.08% respectively. In contrast,
the different prestigious underwriter measurement does have impact on the market share
of year 2009, which is 59.72% and 64.3% respectively. Both measurement methods
indicate that there is a change between the total market share and latest market share. It
motivates us to further examine whether this is caused by grandstanding effects. The
different market share will be used in a regression analysis in the later.

5.1.5 Descriptive statistics on IPO related variables

So far we have discussed some important variables (underpricing, market share),

especially the choice of IPO underwriters (Prestigious versus inexperienced
underwriters) in different circumstances. Overall, the IPOs offered in large size and
listed on SHSE are more likely to choose prestigious underwriters. Inexperienced

40
 underwriters are more likely to be selected in SZSE and for small sized IPOs. The
comparison analysis shows that firm age has less effect on the choice of underwriters.
The following sections of the descriptive statistics will focus on the variables that have
not been fully discussed yet. To test the grandstanding hypothesis, we divided the
sample into two groups, namely, prestigious and inexperienced underwriters as we
described in the research design chapter.

Table 5.4 Descriptive statistics of variables (first reputation measurement method)

Prestigious Inexperienced P-value test of
no difference
Average offering size in millions of RMB 4664 838 0.000
[640] [345] [0.000]
Average net asset of issuing companies before 19299.6 632.8 0.020
IPOs (in millions of RMB) [295.2] [209.0] [0.000]
Average underpricing level 108.23% 118.81% 0.298
[72.18%] [87.58%] [0.019]
Average age of issuing companies 3986 3239 0.023
[2970] [2893] [0.418]
Average ROA ratio 25.7% 24.5% 0.284
[24.2] [23.4] [0.438]
Average P/E ratio 35.25 33.74 0.332
[29.97] [29.96] [0.864]
Average turnover of the first trading day 69.7% 73.6% 0.003
[73.2%] [74.6%] [0.019]
Average lottery rate 0.532% 0.344% 0.035
[0.274%] [0.211%] [0.016]
Average underwriting fee rate 5.23% 6.55% 0.000
[4.85%] [6.43%] [0.000]
Average of time lag between issue data and 12.1 12.9 0.035
listing date [12] [13] [0.044]
* In the p-value column, p-value of t-test is reported and Mann-Whitney U test in the bracket.

Table 5.4 summarizes the relevant information for IPOs conducted by prestigious and
inexperienced underwriters. Prestigious underwriter backed IPOs are larger in size, have
an average [median] of 4664 [640] millions RMB of offer size for prestigious
underwriters and 838 [345] millions RMB for inexperienced underwriters. Similarly,
prestigious underwriter backed IPO firms have large firm net assets before going public
with an average (median) 19299.6 [295.2] millions versus 632.8 [209.0] millions for
inexperienced underwriters’ firms.

The summary statistics in Table 5.4 for underpricing level does not support the
grandstanding hypothesis because of the insignificant t-test results. Inexperienced
underwriters have a slightly higher underpricing level but the p-value (0.298) is
insignificant. Table 5.4 also suggests that issuers who choose prestigious underwriters
41
 are older, an average [median] of 3986 [2970] days since the firm was established
compared with 3239 [2893] days for inexperienced underwriters offering firms. The
t-test for the equality of mean is significant but Mann-Whitney U test is insignificant.
The insignificant results of Mann-Whitney U test supports that the variance is not
different which we interpret as no significant difference regarding the firm ages in the
two groups.

Table 5.5 Descriptive statistics of variables (second reputation measurement

method)
Prestigious Inexperienced P-value test of
no difference
Average offering size in millions of RMB 3564 1155 0.001
[561] [380] [0.001]
Average net asset of issuing companies before 14966.8 857.0 0.023
IPO (in millions of RMB) [257.9] [210.2] [0.001]
Average underpricing level 109.26% 120.51% 0.265
[79.08%] [86.32%] [0.168]
Average age of issuing companies 3750 3328 0.138
[2975] [2824] [0.395]
Average ROA ratio 25.1% 24.9% 0.843
[23.6%] [23.9%] [0.952]
Average P/E ratio 34.95 33.69 0.391
[29.96] [29.96] [0.81]
Average turnover of the first trading day 71.2% 72.7% 0.230
[73.8%] [73.6%] [0.332]
Average lottery rate 0.508% 0.322% 0.027
[0.246%] [0.221%] [0.303]
Average underwriting fee rate 5.74% 6.29% 0.036
[5.58%] [6.13%] [0.029]
Average of time lag between issue data and 12.3 12.9 0.082
listing date [12] [13] [0.03]
* In the p-value column, p-value of t-test is reported and Mann-Whitney U test in the bracket.

Table 5.4 also reveals that prestigious underwriter backed IPOs have a slightly higher
return on asset (ROA) ratio. The average [median] ROA before initial public offering is
25.7% [24.2%] for issuing companies brought public by prestigious underwriters versus
24.5% [23.4%] for the firms brought public by inexperienced underwriters.
Unfortunately, the t-test shows that the result is not significant. Similarly, Price earnings
(P/E) ratios were also insignificant. We conclude that the firm performance indicator
(ROA) and offer price indicator (P/E) are not significantly different between the two
groups. Both groups of underwriters deal with the same groups of firms of varying
quality.

The summary statistics in Table 5.4 indicate that inexperienced underwriter backed
42
IPOs have a significantly higher turnover rate in the first trading day. The average
[median] turnover rate is 73.6% [74.6%] with IPOs backed by inexperienced
underwriters compared to 69.7% [73.2%] for prestigious underwriter backed IPOs.
Consistent with the higher demand signaled in first trading day turnover, inexperienced
underwriter backed IPOs have a lower lottery rate (which means it is harder to receive
the desired share allocation than with prestigious underwriter backed IPOs). The
average [median] number is 0.344% [0.211%] of those IPOs brought public by
inexperienced underwriters while the lottery rate backed by prestigious underwriters is
0.532% [0.274%]. Those IPOs conducted by inexperienced underwriters have a longer
time lag between the issue date and listing date than do the other group. The average
[median] time lag for inexperienced underwriters is 12.9 [13] compared with 12.1 [12]
for prestigious underwriters.

Similarly, inexperienced underwriters charge higher rates or underwriting fees. The

average [median] fee rate is 6.55% [6.43%] charged by inexperienced underwriters. In
contrast, prestigious underwriters only charge 5.23% [4.85%] as their commission fee.
If consistent with the prior research, that fee charging decrease is associated with the
offer size (Pichler & Wilhelm, 2001). In other words, this summary does not support the
grandstanding hypothesis that inexperienced underwriters incur the cost of taking firms
public by charging a lower fee rate.

In Table 5.5, we divided the sample into two groups by using the second underwriter
measurement method (by number of issues). Here we only discuss some different
results when we compare Table 5.4 with Table 5.5. This shows that there is no
significant difference on the issuer’s average age. Also there is no difference for
underpricing levels as well. In addition, the turnover rate turns out to be insignificant
different under the second measurement method, where there was a significant
difference. Similar results for the lottery rate also exist. This shows that there is a
significant difference on the equality of means, but Mann-Whitney U test shows the
results are insignificant.

All in all, the significant results of independent sample t-tests do not suggest the
variance and the variable distribution is different between the two groups. In order to
test both of them, we conduct the Mann-Whitney U test and T-test simultaneously. Only if
both of the statistical results are significant could we say that there are differences on that
specific variable for the two sample groups.

5.1.6 Descriptive statistics on underwriters

We summarize the relevant information for each underwriter which we will use later. In
order to test the grandstanding hypothesis, we have to calculate that the total number of
IPOs for each underwriter had been done. Then we pick up the latest IPO offer size of each
underwriter in order to examine the effect of recent IPOs performance (use offer size as
proxy) on market share. The following two tables show the descriptive statistics within two

43
types of underwriter reputation measurements. P-value of t-test and results of
Mann-Whitney U test were also reported.

Table 5.6 Descriptive statistics of underwriter (first reputation measurement method)

Prestigious Inexperienced P-value test of no
difference
Total market share 7.02% 0.63% 0.053
[2.71%] [0.14%] [0.000]
Market share in 5.97% 1.12% 0.098
year 2009 [2.01%] [0.51%] [0.003]
Total IPOs 18.6 4.9 0.007
number [16.5] [3] [0.000]
Latest IPO issue 280418 157876 0.31
size [175101] [50896] [0.010]
* Mean is reported and median is in the bracket.
* In the p-value column, p-value of t-test is reported and Mann-Whitney U test in the bracket.

Table 5.7 Descriptive statistics of underwriter (second reputation measurement

method)
Prestigious Inexperienced P-value test of no
difference
Total market share 6.91% 0.66% 0.058
[2.71%] [0.14%] [0.000]
Market share in 6.43% 0.99% 0.061
year 2009 [2.51%] [0.51%] [0.000]
Total IPOs 23.4 3.9 0.001
number [21] [3] [0.000]
Latest IPO issue 323436 148724 0.274
size [175101] [50896] [0.001]
*Mean is reported and median is in the bracket.
* In the p-value column, p-value of t-test is reported and Mann-Whitney U test in the bracket.

In both Tables 5.6 and 5.7, the prestigious underwriter group has a significantly larger
market share. As well as total IPOs number. Although the recent IPO offer size in the
prestigious underwriter group is larger than for the inexperienced underwriter group.
The p-value of the t-test does not suggest that this difference is significant. In addition,
when we compared the latest market share with the total market share, we found that
inexperienced underwriters’ market share increased a little. We will further test whether
this is caused by grandstanding effects in the regression section.

5.2 Correlation analysis

Before we do the regression analysis, we have to test the correlation of each variable
44
first. Although Multi-collinearity does not bias the results, it produces large standard
errors in related independent variables (Brien, 2007). In that case, we will avoid putting
the independent variables together if their correlation is over 0.6. Stock exchange
variables are highly correlated (over 0.8) with issue size. So we delete this variable from
our model. We report the correlation of each variable in Table 5.8. Some variables are
standardized in order to get the normal distributional variable (e.g. firm age etc.). We
rewrite the variable name to reflect what we have done (e.g. firm age is called ln age).

5.3 Regression analysis

Some statistical techniques are used in this section. For example, natural logarithm of
offer size, age etc. in order to alleviate skewness. Underpricing was standardized to
reduce its range and still maintain the initial trend. We reported standardized coefficient
in the statistics results to support our analysis and examine our hypothesis.

5.3.1 Market share

Total market share as dependent variable

Results from the regression for the underwriter’s total market share are presented in
Table 5.9. The grandstanding hypothesis predicts that the underwriter’s market share
should be positively correlated with the number of IPOs they conduct. Moreover, IPOs
should be more sensitive to the dependent variable for inexperienced underwriters.

In Table 5.9, the first and second columns are the regression results achieved by
applying the first underwriter measurement method (by issue size). The regression in
the first column shows that the number of IPOs conducted by underwriters does not
provide additional benefits to the total market share because of the insignificant
coefficient. In the second column when the two interaction products (inexperienced
underwriters dummy variable with total IPOs number and the most recent IPOs offer
size) are added into the model, the coefficient (0.610) of total the number of IPOs
suggests that each additional IPO helps the underwriter to increase market share. But
the interaction products are only marginally significant, the negative coefficient (-0.048)
means that inexperienced underwriters total market share is negatively related to each
additional IPO. The Hypothesis 4: Inexperienced underwriters’ market share has more
positive effects towards each additional IPO than prestigious underwriters is not
supported. This result clearly shows that additional IPOs do not have positive impact on
inexperienced underwriter’s total market share. It contradicts grandstanding hypothesis
suggesting that inexperienced underwriters do not have incentives to grandstand.

We assumed IPO offer size as a proxy for the issuing company’s quality. We did so as

45
Table 5.8 Pearson Correlations Matrix
Prestigious Prestigious Ln issue Ln net Lottery Audit
(size) (number) size asset underpricing ROA P/E Turnover rate Fee rate Time lag firm Ln age
Prestigious Correlation 1
(size) Sig. (2-tailed) .
Prestigious Correlation .655(**) 1
(number) Sig. (2-tailed) .000 .
ln issue size Correlation -.315(**) -.137(**) 1
Sig. (2-tailed) .000 .007 .
Ln net asset Correlation -.287(**) -.143(**) .903(**) 1
Sig. (2-tailed) .000 .005 .000 .
underpricing Correlation .118(*) .069 -.452(**) -.336(**) 1
Sig. (2-tailed) .019 .175 .000 .000 .
ROA Correlation -.056 .012 .033 -.199(**) -.026 1
Sig. (2-tailed) .268 .819 .516 .000 .608 .
P/E Correlation -.051 -.052 .253(**) .060 -.345(**) -.120(*) 1
Sig. (2-tailed) .315 .308 .000 .236 .000 .017 .
Turnover Correlation .156(**) .047 -.399(**) -.378(**) .295(**) .133(**) -.096 1
Sig. (2-tailed) .002 .358 .000 .000 .000 .008 .058 .
Lottery rate Correlation -.103(*) -.092 .501(**) .485(**) -.349(**) -.045 .123(*) -.354(**) 1
Sig. (2-tailed) .041 .070 .000 .000 .000 .371 .015 .000 .
Fee rate Correlation .255(**) .082 -.763(**) -.659(**) .338(**) -.162(**) -.099 .312(**) -.342(**) 1
Sig. (2-tailed) .000 .105 .000 .000 .000 .001 .051 .000 .000 .
Time lag Correlation .105(*) .074 -.226(**) -.158(**) .175(**) -.072 -.208(**) -.045 -.009 .065 1
Sig. (2-tailed) .037 .142 .000 .002 .001 .153 .000 .371 .862 .197 .
Audit firm Correlation .194(**) .047 -.576(**) -.578(**) .160(**) .047 -.020 .274(**) -.215(**) .372(**) .060 1
Sig. (2-tailed) .000 .351 .000 .000 .001 .354 .692 .000 .000 .000 .238 .
Ln age Correlation -.053 -.048 .321(**) .299(**) -.195(**) -.098 .224(**) -.069 .096 -.176(**) -.142(**) -.241(**) 1
Sig. (2-tailed) .296 .342 .000 .000 .000 .052 .000 .172 .056 .000 .005 .000 .
** Correlation is significant at the 0.01 level (2-tailed).
* Correlation is significant at the 0.05 level (2-tailed).

46
Table 5.9 Regression Analysis of market share in the whole study period
Regression model: Market share = β0 + β1Reputation + β2Number + β3 Reputation*Number + β4Recent Size + β5Recent size*Reputation+ ε
Dependent variable
Independent variable Underwriters’ Market share (in the whole sample period)
Dummy variable 1 (reputation measured by issue size) -0.38*** 0.065
(0.006) (0.76)
Dummy variable 2 (reputation measured by number of issues) -0.466** 3.301**
(0.015) (0.018)
Total number of IPOs 0.106 0.610** -0.053 0.163
(0.412) (0.035) (0.769) (0.466)
Number of IPOs * Dummy variable 1 -0.048*
(0.098)
Number of IPOs * Dummy variable 2 0.033
(0.814)
Logarithm of the most recent IPO issue size 0.323*** 0.331*** 0.324*** 1.135***
(0.006) (0.004) (0.007) (0.000)
Logarithm of the most recent IPO issue size * Dummy variable 1 -0.200
(0.363)
Logarithm of the most recent IPO issue * Dummy variable 2 -3.346***
(0.005)
Adjusted R square 0.363 0.429 0.343 0.422
F-statistics 11.66 9.43 10.734 9.172
(0.000) (0.000) (0.000) (0.000)
Maximum of Variance inflation factors (VIF) 1.548 7.792 2.954 175.931
Number 57 57 57 57
*** Coefficient is significant at the 0.01 level (2-tailed).
** Coefficient is significant at the 0.05 level (2-tailed).
* Coefficient is significant at the 0.10 level (2-tailed).

47
Table 5.10 Regression Analysis of market share in year 2009
Regression model: Market share = β0 + β1Reputation + β2Number + β3 Reputation*Number + β4Recent Size + β5Recent size*Reputation+ ε
Dependent variable
Independent variable The latest market share in year 2009
Dummy variable 1 (reputation measured by issue size) -0.234 0.140
(0.142) (0.576)
Dummy variable 2 (reputation measured by number of issues) -0.266 3.407**
(0.234) (0.039)
Total number of IPOs 0.351** 1.020*** 0.266 0.497*
(0.025) (0.003) (0.216) (0.062)
Number of IPOs * Dummy variable 1 -0.643**
(0.043)
Number of IPOs * Dummy variable 2 -0.006
(0.973)
Logarithm of the most recent IPO issue size 0.139 0.124 0.140 0.896**
(0.302) (0.345) (0.312) (0.014)
Logarithm of the most recent IPO issue size * Dummy variable 1 -0.003
(0.992)
Logarithm of the most recent IPO issue * Dummy variable 2 -3.345**
(0.022)
Adjusted R square 0.271 0.343 0.258 0.321
F-statistics 6.573 5.69 6.21 5.249
(0.001) (0.000) (0.001) (0.001)
Maximum of Variance inflation factors (VIF) 1.503 7.277 2.938 169.454
Number 46 46 46 46
*** Coefficient is significant at the 0.01 level (2-tailed).
** Coefficient is significant at the 0.05 level (2-tailed).
* Coefficient is significant at the 0.10 level (2-tailed).

48
taking higher quality firm public is a strong signal showing an underwriter’s ability. It
also makes sense that if underwriters bring large offer sizes to the public, they will
obtain higher market share than small sized issuer. Both coefficients (0.323 and 0.331)
of recent IPO offer size with statistical significance can prove this relationship. But the
interaction product (reputation dummy variable * Logarithm of offer size) coefficient
(-0.200) is not significant. This means that the total market share is not sensitive to the
issuing company’s quality (size) for an inexperienced underwriter. Both the IPO number
and recently backed IPO quality do not provide additional benefits for an inexperienced
underwriters’ market share. Again, we cannot find evidence to support the
grandstanding hypothesis.

By using the number of issue, our second reputation measurement method, the third and
forth column in Table 5.9 shows that the total number of IPOs does not contribute to the
total market share of the underwriter given the insignificant results. Also the
insignificant coefficient (0.033) of the interaction product again does not support the
Hypothesis 4: Inexperienced underwriters’ market share has more positive effects
towards to additional IPOs than prestigious underwriters. The coefficients of both the
recent IPOs offer size (0.324 and 1.135 respectively) and interaction product (-3.346)
suggest that the issuer quality (size) has positive effects on total market share, but has
less contribution for inexperienced underwriters. Using another underwriter reputation
measurement gives us the same results here. These results do not support the
grandstanding hypothesis.

The different scale of variance inflation factors (VIF) show that when we add
interaction products into model, the multi-collinearity problem is severe. If VIF is
higher than 10, it is regarded by many practitioners as a sign of serious
multi-collinearity problems (Brien, 2007). The possibility of multi-collinearity is very
high especially when the regression models contain both indicator and interaction terms
(Chambers, 2010). Hair, Black, Babin, Anderson and Tatham (2005) argued that the key
point to test the interaction term effect is to examine the incremental effect (R2) not only
to examine whether the interaction term is significant or not. We also find that the
adjusted R2 is increased by around 8%. It supports the assertion that the interaction
effects do exist, although interaction terms have negative coefficients which do not
support the grandstanding hypothesis.

All in all, although the two measurements provided us with different results in each
model, under the first measurement method, the relation between total market share and
numbers of IPOs an underwriter has taken public is positive. The second method does
not indicate the same results. Both the interaction product with two measurement
methods do not support the grandstanding hypothesis, which additional IPOs and recent
IPOs offer quality (size) do not contribute additional total market share for
inexperienced underwriters.

49
The latest market share (2009)

In Table 5.10, we provide a similar presentation of regression analysis with the first and
second column providing results from the first underwriter reputation measurement
method. The third and forth columns are used to present results by using the second
method. We find that the variable for the total number of IPOs is significant and positive
(coefficient is 0.351 and 1.020) related to market share in the year 2009 in the first set
of regressions. In contrast it is only marginally supported (coefficient is 0.497) under
10% significance level in the forth column presenting the second set of regressions.

The recent IPOs offer size presented in the forth column shows a positive relation
(0.896) to the dependent variable. The other three coefficients (0.139, 0.124 and 0.140)
in each column are insignificant. Although the interaction product between the
‘inexperienced’ dummy variable and offer size is significant in the forth column, the
negative coefficient shows that inexperienced underwriters do not gain additional
benefits in terms of the number of IPOs and issuer quality (size). These findings
contradict the prediction that taking higher quality firms public is a strong signal to
show an underwriter’s ability. In contrast, inexperienced underwriters could not gain
more market share than prestigious underwriters by bringing higher quality issuers
public. When we used the latest year’s market share data, the results still did not support
the grandstanding hypothesis.

We had similar problems with multi-collinearity when we add the interaction product
into our regression model as mentioned above. Hair et al. (2006) suggested the key
point to test the interaction term effect is to examine the incremental effect (R2) not only
to examine whether the interaction term is significant or not. The increased adjusted R
square could show that the interaction effect exists and we could go into the analysis of
the coefficient of each term. The negative coefficients (-0.643 and -0.006) also do not
support the grandstanding hypothesis.

After refer to the two tables which use different market shares as dependent variables,
the results are robust and show that inexperienced underwriters do not gain additional
benefits from additional number of IPOs or recent IPO performance. The grandstanding
hypothesis is not supported by these two regression analyses. Using these variables at
least, does not support an assertion that inexperienced underwriters have incentives to
provide more underpriced IPOs in order to set up more successful records. We will
discuss the underpricing issue and examine the hypotheses further in the next section.

5.3.2 Underpricing

As the grandstanding hypothesis predicted, the sensitivity of an underwriter’s market

share to the number of IPOs and their recent IPO performance (in our study we use
recent IPO offer size as a proxy for performance) is stronger for inexperienced
underwriters. The grandstanding hypothesis suggests that inexperienced underwriters

50
Table 5.11 - Regression Analysis of underpricing level (1)
Underpricing = β0 + β1Reputation + β2Size + β 3Age + β 4ROA+ β 5P/E + β6Audit +
β7TO + β 8LR + β 9Fee + β 10Lag + β 11SE + ε
Dependent Variable
Independent Variable Standardized Underpricing Level
Dummy variable 1 (reputation measured by issue size) -0.018 -0.015 -0.018
(0.695) (0.743) (0.696)
Logarithm of IPO offering size -0.450*** -0.314*** -0.261***
(0.000) (0.000) (0.006)
Logarithm of firm age -0.028 -0.050 -0.048
(0.545) (0.281) (0.301)
ROA -0.038 -0.069 -0.061
(0.393) (0.118) (0.176)
Offering P/E ratio -0.232*** -0.235*** -0.229***
(0.000) (0.000) (0.000)
Audit Firm (1 For Non Big four) -0.106* -0.097* -0.088
(0.053) (0.071) (0.109)
Turnover 0.134*** 0.141***
(0.006) (0.004)
Lottery rate -0.136*** -0.146***
(0.008) (0.005)
Fee rate 0.036
(0.607)
Time between issue and listing day 0.067
(0.148)
Adjusted R square 0.260 0.290 0.290
F-statistics 23.846 20.936 16.973
Significance 0.000 0.000 0.000
Maximum of Variance inflation factors (VIF) 1.836 2.450 4.845
Number 392 392 392
*** Coefficient is significant at the 0.01 level (2-tailed).
** Coefficient is significant at the 0.05 level (2-tailed).
* Coefficient is significant at the 0.10 level (2-tailed).

51
Table 5.12 - Regression Analysis of underpricing level (2)
Underpricing = β0 + β1Reputation + β2Size + β 3Age + β 4ROA+ β 5P/E + β6Audit +
β7TO + β 8LR + β 9Fee + β 10Lag + β 11SE + ε
Dependent Variable
Independent Variable Standardized Underpricing Level
Dummy variable 2 (reputation measured by number of 0.000 -0.002 -0.004
issues) (0.995) (0.968) (0.931)
Logarithm of IPO offering size -0.444*** -0.309*** -0.256***
(0.000) (0.000) (0.006)
Logarithm of firm age -0.029 -0.051 -0.049
(0.533) (0.274) (0.292)
ROA -0.037 -0.069 -0.060
(0.402) (0.121) (0.181)
Offering P/E ratio -0.232*** -0.236*** -0.229***
(0.000) (0.000) (0.000)
Audit Firm (1 For Non Big four) -0.106* -0.098* -0.088
(0.053) (0.071) (0.107)
Turnover 0.133*** 0.140***
(0.006) (0.004)
Lottery rate -0.137*** -0.147***
(0.007) (0.004)
Fee rate 0.036
(0.611)
Time between issue and listing day 0.066
(0.151)
Adjusted R square 0.259 0.290 0.290
F-statistics 23.81 20.917 16.952
Significance 0.000 0.000 0.000
Maximum of Variance inflation factors (VIF) 1.749 2.338 4.761
Number 392 392 392
*** Coefficient is significant at the 0.01 level (2-tailed).
** Coefficient is significant at the 0.05 level (2-tailed).
* Coefficient is significant at the 0.10 level (2-tailed).

52
are motivated to conduct more IPOs. Although the hypothesis is not supported by the
regression analysis above, we still went on to examine the reputation effects on IPOs
underpricing issue. We present the regression results with different variable
combinations. We also used two reputation measurement methods to doubly confirm the
results as we have done before in Table 5.11 and 5.12.

Three models were tested and presented in Table 5.11 showing the results of an
examination of the relationship between underpricing and various sets of variables. F
value and adjusted R squares show that the model has a certain explanatory power. In
addition, the variance inflation factors are small suggesting that our models do not have
multi-collinearity problems. The stock exchange variable is deleted from the regression
model due to multi-collinearity problems as we mentioned in the correlation analysis
section. The insignificant results of underwriter reputation variables indicate that the
underwriter’s reputation does not have a significant impact on underpricing. This is
another factor negating the grandstanding hypothesis that inexperienced underwriter
backed IPOs have higher underpricing levels after controlling for other factors. So
Hypothesis 1: IPOs provide by inexperienced underwriters have higher underpricing
level than those provide by prestigious underwriters is not supported under the first
reputation measurement method. The results are consistent with Tian (2000). They find
that reputation of underwriter has no explanatory power on Chinese IPO underpricing.

When we examine the issuing company related variables in the first column, we find
that offer size, P/E ratio and audit firms (marginally) have significant effects on the
dependent variable. Firm age and ROA do not contribute to explaining the dependent
variable. The adjusted R square (0.26) shows that 26% of the dependent variable’s
variance was explained by firm related variables. When we add market variables
(turnover and lottery rate) into the model, we find that there is a small improvement on
R square (0.29). Both of the two variables have an impact on underpricing which is
significant at 1% level. The turnover on the first trading day has a positive relation
(0.134 and 0.141) with underpricing. In addition, the share allocation lottery rate has
negative effects (-0.136 and -0.146). Also the higher lottery rate means it is easier to
receive share allocations, which suggests that the issue has low competition. The high
turnover shows the great demand or high degree of speculation in the secondary market.
It is a clear hint that the underpricing anomalies are caused by secondary market higher
demand. We argue that we do not get the expected results which could support the
grandstanding hypothesis among underwriters and attribute the anomalies, at least
partially, to mispricing in the secondary market.

When we use the second underwriter reputation measurement method to find the results
presented in Table 5.12. These results are similar to previous ones demonstrating that
the results of underpricing regressions are robust. Hypothesis 1: IPOs provide by
inexperienced underwriters have higher underpricing level than those provide by
prestigious underwriters is not supported under either of the reputation measurement
methods. In conclusion, large offer size, higher P/E ratio at the issue stage and using Big

53
4 auditing firms can reduce underpricing, but the dominant influence is the first trading
day turnover. A small turnover could significantly result in lower underpricing.

5.3.3 IPOs offer size

In Table 5.13, we reports results for offer size regressions. Hypothesis 3 was used to
predict that inexperienced underwriters would bring firms public with smaller offer
sizes. The regression results show that inexperienced underwriters (measured by the
first method) backed IPOs have smaller offer sizes. The coefficient (-0.034 and -0.033,
significant at 10% level) in the first and second columns support that. So Hypothesis 3:
Inexperienced underwriters work for smaller offer sized IPOs than prestigious
underwriters is supported under first reputation measurement method. Issuing company
related variables like firm net asset, firm age, ROA and the choice of audit firm, all have
a significant relationship with issue size. The R square is 86.8% which suggest that the
models have good explanatory power on the dependent variable. The firm net asset
before IPO is consistent with the IPO offer size due to the positive coefficient (0.888,
significant at 1% level). This largest scale of the coefficient suggests that IPO offer size
is heavily dependent on the issuing company’s size. Since regression results prove that
inexperienced underwriters conduct smaller sized IPOs and firm net assets before IPOs
is correlated with offer size. So we could derive that inexperienced underwriters are
more likely to have clients who have small net assets.

The next important variable is ROA, its coefficients (0.219) suggests higher earning
companies are more likely to raise larger amounts of capital. Another positive
coefficient is firm age (0.063), indicating that older firms tend to raise higher amounts
of capital in their IPOs. This is consistent with the prior research suggesting that older
firm have less uncertainty because firms with a long history provide more information
to the market (Barry and Brown, 1985). The negative coefficients (-0.051) show that
issuing companies who choose non-Big 4 as their audit firms normally raise less capital.

When we put market related variables into the model, we find that both turnover rates
and lottery rates are significant at the 1% level. The coefficient of turnover (-0.066)
shows that the first trading day turnover is negative when related to offer size. If the
offer size is large, the secondary market seems to have less enthusiasm for that IPO. The
variable on lottery rate shows similar results, the coefficient (0.065) suggests that the
higher the offer size, the less competition on share allocation. Investors find it easier to
get a bid with higher lottery rate.

In the third and forth column in Table 5.13, we use the second underwriter reputation
measurement method, the regression results in the third and forth column in Table 5.13
show similar results except for the underwriter reputation dummy variable. This means
that prestigious underwriters, if measured by number of issues, are not significantly
different from less experienced underwriters on IPOs underpricing levels. So
Hypothesis 3: Inexperienced underwriters work for smaller offer sized IPO than

54
Table 5.13 - Regression Analysis of IPOs offer size
Offer size = β 0 + β1Reputation + β2NTA + β 3Age + β 4ROA + β 5EPS + β6Audit + β7TO
+ β 8LR + β 9SE + ε
Dependent Variable
Independent Variable Logarithm of offer Size
Dummy variable 1 (reputation measured by issue -0.034* -0.033*
size) (0.076) (0.085)
Dummy variable 2 (reputation measured by -0.006 -0.005
number of issues) (0.741) (0.777)
Logarithm of IPO firm’s net asset 0.888*** 0.829*** 0.898*** 0.838***
(0.000) (0.000) (0.000) (0.000)
Logarithm of firm age 0.063*** 0.068*** 0.062*** 0.067***
(0.001) (0.000) (0.002) (0.000)
ROA 0.219*** 0.217*** 0.223*** 0.221***
(0.000) (0.000) (0.000) (0.000)
EPS -0.004 0.001 -0.004 0.001
(0.847) (0.952) (0.850) (0.949)
Audit Firm (1 For Non Big four) -0.051** -0.053** -0.053** -0.054**
(0.028) (0.019) (0.024) (0.018)
Turnover -0.066*** -0.068***
(0.001) (0.001)
Lottery rate 0.065*** 0.063***
(0.002) (0.003)
Adjusted R square 0.868 0.876 0.867 0.875
F-statistics 429.238 345.745 425.364 342.781
Significance 0.000 0.000 0.000 0.000
Maximum of Variance inflation factors (VIF) 1.780 2.278 1.711 2.201
Number 392 392 392 392
*** Coefficient is significant at the 0.01 level (2-tailed).
** Coefficient is significant at the 0.05 level (2-tailed).
* Coefficient is significant at the 0.10 level (2-tailed).

55
prestigious underwriters is not supported under the second reputation measurement
method.

Here we argue that the second method indicates that reputation is accumulated by (the
number of issues) and not by issue size. There is a possibility that some of the
prestigious underwriters (by number of issues undertaken) conduct many smaller sized
IPOs. Thus, inexperienced underwriters and prestigious underwriters show no difference
on the offer size except in number. If that is the case, it could explain why inexperienced
underwriters do not show a significant negative relationship with offer size. In
conclusion, hypothesis 3 is weakly supported due to the different results in Table 5.13.
Also the variance inflation factors (VIF) are quite small indicating that we do not have a
multi-collinearity problem.

5.3.4 Underwriting fee rate

In Table 5.14, we report the regression results for IPO commission fee rates which could
be a direct cost of grandstanding. Fee rate are the percentage charged of the total capital
raised. As we mentioned in the literature review chapter, underwriters have incentives to
increase the offer price only if they charge higher fee rates. The grandstanding
hypothesis predicted that inexperienced underwriters would underprice more in order to
achieve more successful IPOs, thus they would charge lower fee rates. In addition, if
underwriters lower their charge for their IPOs, it makes them more attractive to some
issuers.

The regressions show that inexperienced underwriters (using the first measurement
method) do not charge a significantly smaller (coefficient is 0.013 and insignificant
result) fee rate as expected. So Hypothesis 2: Inexperienced underwriters charge lower
fee rate than prestigious underwriters is not supported under the first reputation
measurement method. There is no difference on the fee rate either group charges. This
insignificant underwriter reputation dummy variable does not support our grandstanding
hypothesis. The most important factor to explain the dependent variable is offer size of
each IPO. This variable has the highest coefficient (-0.785). It indicates that
underwriters will charge lower fee rates on larger offer sized IPOs. That the coefficient
on firm age (0.076) is significant at the 5% level means that underwriters charge higher
fee rates on older firms. In fact, when ROA, P/E ratio and choice of audit firm are
included, both the significance and coefficient of firm age are reduced. It indicates that
firm age is no longer the main effect when we consider other variables together.

Also issuing firm’s ROA, issue P/E ratio and the choice of audit firm has a significant
impact on fee rates charged. But their coefficients (-0.113, 0.094 and -0.104) reflect they
have smaller explanatory power on dependent variables than has offer size. When we
add the market effect variable (turnover and lottery rate), turnover is not significant and
lottery rate is only marginally significant (under 10% level). So the fee rate is mainly
affected by issuing company related variables.

56
Table 5.14 - Regression Analysis of underwriting fee rate (1)
Fee rate = β 0 + β1Reputation + β2Size + β3UP + β 4Age + β 5ROA + β 6P/E + β 7Audit +
β8TO + β9LR + β 10SE +ε
Dependent Variable
Independent Variable Fee rate
Dummy variable 1 (reputation measured by issue size) 0.013 0.007 0.002
(0.717) (0.828) (0.960)
Logarithm of IPO offer size -0.785*** -0.852*** -0.876***
(0.000) (0.000) (0.000)
Standardized Underpricing -0.004 0.006 0.010
(0.904) (0.877) (0.801)
Logarithm of firm age 0.076** 0.042 0.044
(0.028) (0.219) (0.204)
ROA -0.113*** -0.114***
(0.000) (0.001)
P/E ratio 0.094*** 0.096***
(0.007) (0.006)
Audit Firm (1 For Non Big four) -0.104*** -0.114***
(0.009) (0.004)
Turnover 0.043
(0.238)
Lottery rate 0.069*
(0.068)
Adjusted R square 0.583 0.609 0.612
F-statistics 137.763 88.173 69.42
Significance 0.000 0.000 0.000
Maximum of Variance inflation factors (VIF) 1.485 2.113 2.592
Number 392 392 392
*** Coefficient is significant at the 0.01 level (2-tailed).
** Coefficient is significant at the 0.05 level (2-tailed).
* Coefficient is significant at the 0.10 level (2-tailed).

57
Table 5.15 - Regression Analysis of underwriting fee rate (2)
Fee rate = β 0 + β1Reputation + β2Size + β3UP + β 4Age + β 5ROA + β 6P/E + β 7Audit +
β8TO + β9LR + β 10SE +ε
Dependent Variable
Independent Variable Fee rate
Dummy variable 2 (reputation measured by number of -0.023 -0.023 -0.021
issues) (0.494) (0.477) (0.509)
Logarithm of IPO offer size -0.793*** -0.858*** -0.880***
(0.000) (0.000) (0.000)
Standardized Underpricing -0.005 0.006 0.009
(0.901) (0.880) (0.802)
Logarithm of firm age 0.077** 0.042 0.044**
(0.026) (0.216) (0.203)
ROA -0.113*** -0.114***
(0.000) (0.001)
P/E ratio 0.094*** 0.096***
(0.007) (0.006)
Audit Firm (1 For Non Big four) -0.105*** -0.115***
(0.008) (0.004)
Turnover 0.042
(0.24)
Lottery rate 0.069*
(0.069)
Adjusted R square 0.584 0.610 0.612
F-statistics 137.967 88.344 69.547
Significance 0.000 0.000 0.000
Maximum of Variance inflation factors (VIF) 1.372 2.019 2.475
Number 392 392 392
*** Coefficient is significant at the 0.01 level (2-tailed).
** Coefficient is significant at the 0.05 level (2-tailed).
* Coefficient is significant at the 0.10 level (2-tailed).

58
The adjusted R square (0.583) in the first column shows the model already has good
explanatory power with the limited number of variables. When add additional variables
like ROA, P/E ratio, and choice of audit firm, only 2% of adjusted R square is increased.
Similar results arise when putting turnover and lottery rates into the model only
increasing an adjusted R square by 0.3%. This all confirms the other variables do not
have much effect on the fee rates. The results are consistent with the Pichler and
Wilhelm (2001) that the spread declines with large issue sizes. The higher the
profitability of the issuer, the larger the amount of capital they are willing to raise.
Results are also consistent with Bartling and Park (2009). Fee rate increases with the
issuer price. Big-4 audit firms provide higher quality audits, and partly reduce the
potential information asymmetry (Khurana & Raman, 2004). Underwriters charge fewer
fees on the issuer who chooses Big-4 audit firm. The small scale of VIF shows that we
do not have a multi-collinearity problem in these models.

In Table 5.15, we show similar results when we use the second reputation measurement
method to define prestigious and inexperienced underwriters. There are still no
significant differences on the fee rates charged. Our results are robust after comparing
them with Table 5.14. Hypothesis 2: Inexperienced underwriters charge lower fee rate
than prestigious underwriters is also not supported under the second reputation
measurement method.

59
6 Conclusions

In this chapter, we summarize and draw conclusions from our research findings and
present the limitations of this paper. We also make proposals for further research.

6.1 Summary
At the beginning we will review our research question:

 To what extent does the underpricing phenomenon exist in China?

 Do inexperienced underwriters grandstand when they bring firms public in order

to attract more IPO clients? (This is our main research question)

 What characteristics are associated with IPOs conducted by inexperienced

underwriters?

In our study, the grandstanding hypothesis provides us a great opportunity to apply an

existing theory to a new area. We have taken advantage of the explosive stock market
growth in an emerging country where lots of new market players have emerged. This
has provided us with a balanced and appropriate mixed data set which made our test
of grandstanding hypothesis possible. Also we examined the underwriters’
performance with their IPOs and find that the average underpricing in China is still at
a high level (114.2%, even if we use the recent market data from 2006 to 2010).

The latest market share (year 2009) of prestigious underwriter and inexperienced
underwriters has changed, when compared with the market share segmentation in the
whole study period (June 19, year 2006 to March 31, year 2010). Inexperienced
underwriters’ market share has increased although the grandstanding hypothesis is not
supported.

In the descriptive statistics section, there are significant differences in offer size, fees
or rates charged issuers when comparing the prestigious underwriters group with the
inexperienced underwriters group. But when we add other factors into the regression
model to examine this phenomenon, we obtained insignificant results on fees or rates
and marginally significant results on offer size. This means that the different choices
of underwriters do not have significant impact on either variable. Inexperienced
underwriters do not deliberately offer higher underpricing for IPOs nor do they charge
lower fee rates or choose to conduct smaller offerings. The grandstanding hypothesis
was not supported based on our empirical study.

In addition, after conducting a regression analysis on all the underwriters, we found

that the number of IPOs and the most recent IPO performance (using offer size as a

60
proxy for performance) do not always contribute to the market share of the
underwriter significantly. Especially the number of IPOs offered and most recent IPO
performance does not have more positive effects on inexperienced underwriters. We
believe this indicates that inexperienced underwriters do not have incentives to
grandstand. The conclusion is that the grandstanding hypothesis is not supported.

This examination of the grandstanding hypothesis and the empirical results provide
evidence for the Chinese IPOs industry. Grandstanding and signaling could create real
wealth losses for issuers. We show that underwriters do not have incentives to
grandstand, so issuers do not need to be concerned with these costs from excessive
underpricing. More and more issuers have gone into the Chinese stock market. The
listed companies’ numbers increased from 53 in 1992 to 1718 in 2009 (CSRC, 2010).
The underpricing level decreased from 949% during 1987–1995 (Su and Fleisher,
1999) to 114.2% of our study period. The tremendous entry of new issuers and
apparent fair dealing by underwriters (no grandstanding) potentially explain the
decline underpricing.

Chinese underwriter’s market share affects their ability to compete for large IPO
business. Wang (2010) reports that the top 10 underwriters (Figure 6.1) obtained 70%
of the underwriting business and the largest IPO offers are reserved for the leading
underwriters (e.g. China State Construction Engineering Corp, the largest IPOs of
offer size in 2009 was conducted by China International Capital Corp; China
Shipbuilding Industry Corp and China CNR Corp were the second the third largest
IPOs which were conducted by CITIC Securities). Even though Wang does not
mention his measurement method, we also found that prestigious underwriters have a
dominant market position. Less experienced underwriters need not grandstand as,
despite the role of prestigious underwriters, plenty of business is left for others. We
could see that in Figure 6.1, the top underwriter only conduct three IPOs but its
market share is over 40%. Ultimately, we suggest, underwriters should compete on
quality instead of quantity. They do not need to rush IPOs into the public market in
order to accumulate reputation.

Figure 6.1 Top 10 underwriters in Wang (2010)’s report

61
Data source: Bloomberg, cited in Chinese economics website.

6.2 Limitation

Our study is conducted in an emerging stock market. It is believed that many

investors in an emerging market are not as mature as the investors on developed stock
markets. We realize that the results may not be fully applicable on developed markets.
In addition, we constrained our study period to that after the year 2006. In that time
the Chinese stock market was applying the book-building method to set up the price
for IPOs offered in the market. This method is quite new in the Chinese stock market
and maybe not be fully functional. Furthermore, in our study period, we found that
before the year 2008, the market was full of energy (e.g. lots of firms going public,
enough liquidity). But there was a big financial turmoil after year 2008. We also
should take the financial crisis impact into consideration, because it may cause some
negative impacts on IPO markets (e.g. Freeze the new IPO offers, increase market
competition). Maybe we also need to include the cyclicality issue into our study.

6.3 Further study

How does the issuing company choosing their underwriters? We did not examine how
an issuer chooses their underwriter. In further studies, it may be useful to take a look
at the issuer perspective as well as to examine the long-run performance of the issuer.
It would provide us with a better understanding of the relationship between issuers
and underwriters. Other variables also could be considered in further studies. For
example, ownership structure, share lock up period, issuer area and industry
differences could be useful factors to consider.

6.4 Credibility criteria

According to Bryman (2007), reliability and validity are two of the most critical
criteria to measure the quality of research, especially for quantitative research.

6.4.1 Reliability

Reliability refers to whether the results of a research are repeatable and whether the
measurements of the study are consistent. (Bryman & Bell, 2007) It is means that,
under similar circumstances, using similar methods, the measures should be stable.

Regarding our study, it is important to acquire trustworthy sources. We gathered IPO

related data like issuer firm size, age, and underwriting fee etc. from issuing
company’s prospectus which listed on China Securities Regulatory Commission

62
(CSRC) official information disclosure website (Juchao Information, 2010) and had
been investigated by CSRC before the firm went public. So we believe this data is
accurate and provides us with a high level of reliability. Furthermore, we also
double-checked this data in order to be more accurate, and then we documented all
this information in Excel, which is a consistent data collection procedure.

Another fact makes us have confidence of our reliability is that we have done a total
time frame study of IPO process within the chosen period, not only a random sample
selection. Moreover, with using the statistics tool in SPSS, our model has also been
tested many times by previous researchers. All these approaches could increase our
degree of reliability degree.

6.4.2 Validity

Validity is the most important criteria of one study as Bryman (2007) mentioned. A
valid study can be said to have succeeded. There are four perspectives related to
validity: measurement validity means whether the indicators really measure the
concept they are intended to measure; internal validity is concerned with whether
there is a causal relationship between the variables; external validity relates to
whether the outcome can be generalized beyond the specific research and ecological
validity stands for social environment with our routine life (Bryman & Bell, 2007).

In order to build up the validity of our study, we used research models to connect the
research, the theory and the research question together. For theory part, the prestigious
journal articles and previous studies have been quoted in this study. Then the
empirical data was collected from an official statistical database. Next we used
statistical tools, such as Office Excel, and SPSS to deal with the comparison analysis,
correlation analysis and regression analysis between different variables. Since we use
several proxies to measure the variables, an argument about their validity could be
raised. For example, we use the recent IPO offer size as the proxy for underwriter’s
recent IPO performance. We realize that other measures might have been selected as a
proxy for performance but believe our choice, applied appropriately, was unbiased.
Regarding the internal validity of the study, we conducted correlation analysis and
avoided putting highly correlated variables into one regression model. At least in this
regard we have justified our claim to internal validity. Because of the particular
research environment (an emerging market), this study cannot easily be generalized
into more mature markets, although we have no reason to suppose that they are
radically different in regard to the grandstanding issue.

63
Reference

Journal

Ahn, O.H., Kim, J., & Son, P. (2007). The pricing of underwriting services in the
Korean IPO market: Characteristics and determinants. Asia-Pacific Journal of
Financial Studies, 36 (5), 731-764.

Allen, F., & Faulhaber, G. (1989). Signaling by underpricing in the IPO market.
Journal of Financial Economics, 23, 303-323.

Akerlof, G.A. (1970). The market for ‘lemons’: Quality uncertainty and the market
mechanism. Quarterly Journal of Economics, 84, 488-500.

Aggarwal, R. (2000). Stabilization activities by underwriters after initial public

offerings, Journal of Finance, 55, 1075-1103.

Aggarwal, R., & Prabhala, N.R. (2002). Institutional allocation in initial public
offerings: Empirical evidence. The Journal of Finance, 57 (3), 1421-1442.

Agrawal, A., & Knoeber, C. (1996). Firm performance and mechanisms to control
agency problems between managers and shareholders. Journal of Financial &
Quantitative Analysis, 31 (3), 377-397.

Baron, D.P., & Holmstrom, B. (1980). The investment banking contract for new
issues under asymmetric information: Delegation and the incentive problem. Journal
of Finance, 35, 1115-1138.

Baron, D.P. (1982). Amodel of the demand for investment banking advising and
distribution services for new issues. Journal of Finance, 37, 955-976.

Barry, C.B., & Brown, S.J. (1985). Differential information and security market
equilibrium. Journal of Financial and Quantitative Analysis, 20 (4), 407-422.

Bartling, B., & Park, A. (2009). What determines the level of IPO gross spreads?
Underwriter profits and the cost of going public. International Review of Economics
and Finance, 18 (1), 81-109.

Beatty, R.P., & Ritter, J.R. (1986). Investment banking, reputation, and the
underpricing of initial public offerings. Journal of Financial Economics, 15, 213-232.

Beatty, R., & Welch, I. (1996). Issue expenses and legal liability in initial public

64
offerings. Journal of Law and Economics, 39, 545-602.

Benveniste, L.M., & Spindt, P.A. (1989). How investment bankers determine the
offer price and allocation of new issues. Journal of Financial Economics, 24,
343-362.

Benveniste, L.M., & Wilhelm, W.J. (1990). A comparative analysis of IPO proceeds
under alternative regulatory environments. Journal of Financial Economics, 28,
173-208.

Binay, B.M., Gatchev, V.A., & Pirinsky, C.A. (2007). The role of
underwriter-investor relationships in the IPO process. Journal of Financial and
Quantitative Analysis, 42, 785-810.

Bradley, D., & Jordan, B.D. (2002). Partial adjustment to public information and IPO
underpricing. Journal of Financial and Quantitative Analysis, 37, 595-616.

Brien,O., & Robert, M. (2007). A caution regarding rules of thumb for variance
inflation factors. Quality and Quantity, 41 (5), 673-690.

Brau, J.C., & Fawcett, S.E. (2006). Initial public offerings: An analysis of theory and
practice. The Journal of Finance, 61, 399-436.

Booth, J.R., & Smith, R.L. (1986). Capital raising, underwriting and the certification
hypothesis. Journal of Financial Economics, 15, 261-281.

Carter, R., & Manaster, S.M. (1990). Initial public offerings and underwriter
reputation. Journal of Finance, 45, 1045-1067.

Carter, R.B., Dark, F.H., & Singh, A.K. (1998). Underwriter reputation, initial returns
and the long-run performance of IPO stocks. Journal of Finance, 53, 285-311.

Certo, S.T. (2003). Influencing initial public offering investors with prestige:
Signaling with board structures. The Academy of Management Review, 28 (3),
432-446.

Chan, K., Wang, J., & Wei, K.C.J. (2004). Underpricing and long-term performance
of IPOs in China. Journal of Corporate Finance, 10, 409-430.

Chen, H.C., & Ritter, J.R. (2000). The seven percent solution. The journal of finance,
55 (3), 1105-1131.

Chen, G., Firth, M., & Kim, J.B. (2004). IPO underpricing in China’s new stock
markets. Journal of Multinational Financial Management, 14, 283-302.

65
Chen, H.C., Fok, R.C.W., & Wang, Y.J. (2006). Why underwriters charge low
underwriting fees for initial public offerings in Taiwan? Journal of Business Finance
and Accounting, 33 (7-8), 979-1005.

Cheung, Y.L., Ouyang, Z., & Tan, W. (2009). How regulatory changes affect IPO
underpricing in China. China Economic Review, 20 (4), 692-702.

Chi, J., & Padgett, C. (2005). Short-run underpricing and its characteristics in Chinese
Initial Public offering (IPO) markets. Research in International Business and Finance,
19, 71-93.

Chemmanur, T. (1993). The pricing of initial public offerings: A dynamic model with
information production. Journal of Finance, 48, 285-304.

Chemmanur, T.J., & Fulghieri, P. (1994). Investment bank reputation, information

production, and financial intermediation. Journal of Finance, 49 (1), 57-59.

Cornelli, F., & Goldreich, D. (2001). Bookbuilding and strategic allocation. The
Journal of Finance, 56 (6), 2337-2369.

Cornelli, F., & Goldreich, D. (2003). Bookbuilding: How informative is the order
book? The Journal of Finance, 58 (4), 1415-1443.

Cornelli, F., Goldreich, D., & Ljungqvist, A. (2006). Investor sentiment and Pre-IPO
markets. The Journal of Finance, 61 (3), 1187-1216.

Daily, C.M., Certo, S.T., & Dalton, D.R. (2005). Investment bankers and IPO pricing:
Does prospectus information matter? Journal of Business Venturing, 20 (1), 93-111.

DeAngelo, L. (1981). Auditor independence, ‘low balling’, and disclosure regulation.

Journal of Accounting and Economics, 3, 113-127.

Deng, H.N., & Dorfleitner, G. (2008). Underpricing in Chinese IPOs-some recent

evidence. Applied Financial Economics, 18 (1), 9-22.

Diamond, D. (1989). Reputation acquisition in debt markets. Journal of Political

Economy, 97, 828-862.

Dimovski, W., & Brooks, R. (2004). Do you really want to ask an underwriter how
much money you should leave on the table? Journal of International Financial
Markets, Institutions and Money, 14, 267-280.

Doug, D. (1989). Reputation acquisition in debt markets. Journal of Political

Economy, 97, 828-862.

66
Dunbar, C.G. (1998). The choice between firm-commitment and best-efforts offering
methods in IPOs: The effect of unsuccessful offers. Journal of financial
intermediation, 7 (1), 60-90.

Eisenhardt, K.M. (1988). Agency and institutional explanations of compensation in

retail sales. Academy of Management Journal, 31, 488-511.

Fang, L.H. (2005). Investment bank reputation and the price and quality of
underwriting services. Journal of finance, 60 (6), 2729-2761.

Fernando, C.S., Gatchev, V.A., & Spindt, P.A. (2005). Wanna dance? How firms and
underwriters choose each other. Journal of Finance, 60 (5), 2437-2469.

Flagg, D., & Margetis, S. (2008). What characteristics of underwriters influence

underpricing? Journal of Private Equity, 11(2), 49-59.

Gannon, G., & Zhou, Y.W. (2008). Conflicts of interest and China's A-share
underpricing. International Review of Financial Analysis, 17, 491-506.

Gao, Y. (2010). What comprises IPO initial returns: Evidence from the Chinese
market. Pacific-Basin Finance Journal, 18, 77-89.

Gompers, P.A. (1996). Grandstanding in venture capital industry. Journal of

Financial Economics, 42 (1), 133-156.

Grinblatt, M., & Hwang, C.Y. (1989). Signalling and the pricing of new issues.
Journal of Finance, 44, 393-420.

Gu, Y. X. (2003). State ownership, firm size, and IPO performance: Evidence from
Chinese ‘A’ share issues. American Business Review, 21-101.

Guo, H.F., & Brooks, R. (2008). Underpricing of Chinese A-share IPOs and short-run
underperformance under the approval system from 2001 to 2005. International
Review of Financial Analysis, 17, 984-997.

Hanley, K.W. (1993). The underpricing of initial public offerings and the partial
adjustment phenomenon. Journal of Financial Economics, 34, 231-250.

Hanley, K.W., & Wilhelm, Jr.W.J. (1995). Evidence on the strategic allocation of
initial public offerings. Journal of Financial Economics, 37 (2), 239-257.

Habib, M.A., & Ljungqvist, A. (2001). Underpricing and entrepreneurial wealth

losses in IPOs: Theory and evidence. Review of Financial Studies, 14, 433-458.

67
Hansen, R. (2001). Do investment banks compete in IPOs? The advent of the “7%
plus contract”. Journal of Financial Economics, 59 (3), 313-346.

Hoberg, G. (2007). The underwriter persistence phenomenon. Journal of Finance, 62

(3), 1169-1206.

How, J.C.Y., & Yeo, J.J.L. (2000). The pricing of underwriting services in the
Austrialian capital market. Pacific Basin Finance Journal, 8 (3-4), 347-373.

Ibbotson, R.G. (1975). Price performance of common stock new issues. Journal of
Financial Economics, 2, 235-272.

James, C. (1992). Relationship-specific assets and the pricing of underwriter services.

Journal of Finance, 47, 1865-1885.

Jing, S., Bilson, C., & Powell, J. (2008). Valuation uncertainty risk compensation and
IPO prospectus earnings forecasts. Applied Economics Letters, 15 (5), 331-335.

Johnson, J., & Robert, M. (1988). Investment banker prestige and the underpricing of
initial public offerings. Financial Management, 17, 19-29.

Kahneman, D., & Tversky, A. (1979). Prospect theory: An analysis of decision

making under risk. Econometrica, 42, 262-291.

Khurana, I.K., & Raman, K.K. (2004). Litigation risk and the financial reporting
credibility of Big 4 versus Non-Big 4 audits: Evidence from Anglo-American
countries. The Accounting Review, 79 (2), 473-495.

Klein, B., & Leffler, K. (1981). The role of market forces in assuring contractual
performance. Journal of Political Economy, 89, 615-641.

Krigman, L., Wayne, H.S., & Kent, L.W. (2001). Why do firms switch underwriters?
Journal of Financial Economics, 60 (2-3), 245-284.

Lee, P., & Wahal, S. (2004). Grandstanding, certification and the underpricing of
venture capital backed IPOs. Journal of Financial Economics, 73, 375-407.

Levis, M. (1990). The winner’s curse problem, Interest costs and the underpricing of
initial public offerings. Economic Journal, 100 (399), 76-89.

Ljungqvist, A., Jenkinson, T., & Wilhelm, W. (2003). Global integration in primary
equity markets: The role of U.S. banks and U.S. investors. Review of Financial
Studies, 16 (1), 63-99.

68
Logue, D. (1973). On the pricing of unseasoned equity issues: 1965–1969. Journal of
Financial and Quantitative Analysis, 8, 91-103.

Loughran, T., Ritter, J., & Rydqvist, K. (1994). Initial public offerings: international
insights. Pacific-Basin Finance Journal, 2, 165-199.

Loughran, T., & Ritter, J.R. (2002). Why don’t issuer get upset about leaving money
on the table in IPOs? The Review of Financial Studies, 15 (2), 413-443.

Loughran, T, & Ritter, J.R. (2004). Why has IPO underpricing changed over time?
Financial Management, 33, 5-37.

Maanen, J.V., Sørensen, J.B., & Mitchell, T.R. (2007). The interplay between theory
and method. Academy of Management Review, 32 (4), 1145-1154.

Ma, F., & Faff, R. (2007). Market conditions and the optimal IPO allocation
mechanism in China. Pacific-Basin Finance Journal, 15, 121-139.

Mckenzie, C.R., & Takaoka, S. (2008). Underwriter reputation and switching.

Mathematics and Computers in Simulation, 78, 215-222.

Megginson, W., & Weiss, K. (1991). Venture capitalist certification in initial public
offerings. Journal of Finance, 46, 879-904.

Merton, R.C. (2005). You have more capital than you think. Harvard business
review, 83 (11), 84-94.

Michaely, R., & Wayne, H.S. (1994). The pricing of initial public offerings: Tests of
adverse selection and signaling theories. The Review of Financial Studies, 7, 279-319.

Mok, H. M. K., & Hui, Y. V. (1998). Underpricing and aftermarket performance of

IPOs in Shanghai, China. Pacific-Basin Finance Journal, 6, 453-474.

Myers, S., & MaJluf, N. (1984). Stock issues and investment policy when firms have
information investors do not have. Journal of Financial Economics, 13, 187-221.

Pichler, P., & Wilhelm, W. (2001). A theory of the syndicate: Form follows function.
Journal of Finance, 56 (6), 2237-2264.

Ritter, J.R. (1984). The ‘hot issue’ market of 1980. Journal of Business, 57, 215-240.

Ritter, J.R., & Welch, I. (2002). A review of IPO activity, pricing, and allocations.
Journal of Finance, 57 (4), 1795-1828.

69
Reuter, J. (2006). Are IPO allocations for sale? Evidence from the mutual fund
industry. Joumal of Finance, 61, 2289-2324.

Rock, K. (1986). Why new issues are underpriced. Journal of Financial Economics,
15, 187-212.

Rocholl, J. (2009). A friend in need is a friend indeed: Allocation and demand in IPO
bookbuilding. Journal of Financial Intermediation, 18, 284-310.

Sheridan, T., & Trueman, B. (1986). Information quality and the valuation of new
issues. Journal of Accounting and Economics, 8, 159-172.

Sherman, A.E. (2005). Global trends in IPO methods: Book building versus auctions
with endogenous entry. Journal of Financial Economics, 78 (3), 615-649.

Spatt, C.S., & Srivastava, S. (1991). Preplay communication, participation restrictions,

and efficiency in initial public offerings. Review of Financial Studies, 4, 709-726.

Su, D.W., & Fleisher, B.M. (1999). An empirical investigation of underpricing in

Chinese IPOs. Pacific-Basin Finance Journal, 7, 173-202.

Smith, M.E., & Milina, D. (1999). Cross cultural Collaborative research: Toward
Reflexivity. Academy of Management Journal, 42 (1), 76-86.

Tian, Z. (2000). The reputation of investment bank and IPO underpricing. The Journal
of the Graduate School of the Institute of Chinese Social Science, 4, 33-36.

Takaoka, S., & McKenzie, C.R. (2006). The impact of bank entry in the Japanese
corporate bond underwriting market. Journal of Bank and Finance, 30 (1), 59-83.

Titman, S., & Trueman, B. (1986). Information quality and the valuation of new
issues. Journal of Accounting and Economics, 8, 159-172.

Torstila, S. (2001). What determines IPO gross spreads in europe? European

Financial Management, 7 (4), 523-541.

Wakeman, L. (1981). The real function of bond rating agencies. Chase Financial
Quarterly, 1, 18-26.

Welch, I. (1989). Seasoned offerings, imitation costs and the underpricing of initial
public offerings. Journal of Finance, 44, 421-449.

Welch, I. (1992). Sequential sales, learning, and cascades. Journal of Finance, 47,
695-732.

70
Yang, D. (2003). Shell resource and IPO underpricing: An empirical study in China's
stock market. USA–China Business Review, 3 (15).

Yong, O. (2007). A review of IPO research in Asia: What's next?. Pacific-Basin

Finance Journal, 15, 253-275.

Yu, T., & TSE, Y.K. (2006). An empirical examination of IPO underpricing in the
Chinese A-share market. China Economic Review, 17, 363-382.

Zhu, J., & Tian, Y.H. (2002). The empirical study on the initial return of IPOs.
Decision Studies, 1, 27-31.

Zhang, X.M., Bartol, K.M., Pfarrer, M.D., & Khanin, D.M. (2008). Ceos on the edge:
Earnings manipulation and stock-based incentive misalignment. Academy of
Management Journal, 51 (2), 241-258.

Thesis

Cooney, J., Singh, A., Carter, R., & Dark, F. (2001). IPO initial returns and
underwriter reputation: Has the inverse relationship flipped in the 1990s? Working
paper. Lexington: University of Kentucky.

Gompers, P.A. (1993). The theory, structure, and performance of venture capital.
Doctoral dissertation. Cambridge, MA: Harvard University.

Book

Berk, J., & Demarzo, P. (2007). Corporate finance. International edition. Boston:
Pearson Addison Wesley.

Bryman, A., & Bell, E. (2007). Business research methods. 2nd edition. Oxford: OUP
Oxford.

Jenkinson, T. & Ljungqvist, A. (2001). Going public – The theory and evidence on
how companies raise equity finance. 2nd edition. New York: Oxford University Press.

Hair, J.F., Black, B., Babin, B., Anderson, R.E., & Tatham, R.L. (2005). Multivariate
Data Analysis. 6th Edition, London: Prentice Hall.

Saunders, M. Lewis, P., & Thornhill, A. (2007). Research Methods for Business
Students. 4th edition, Edinburgh: Prentice Hall

Sutherland, J.W. (1975). Systems: Analysis administration and architecture. New

York: Van Nostrand Reinhold.

71
Web Pages

Bloomberg. (2010). Shcomp:Ind Shanghai Se Composite. Bloomberg website. <

http://www.bloomberg.com/apps/quote?ticker=SHCOMP%3AIND> [Retrieved 201
0-05-12].

China Securities Regulatory Commission (CSRC). (2010, March 19). Market S-

tatistics data of Februrary 2010. China Securities Regulatory Commission Websi
te. <http://www.csrc.gov.cn/pub/zjhpublic/G00306204/zqscyb/201003/t20100319_17
8552.htm> [Retrieved 2010-04-15].

Chambers, M.S. (2010). Using Indicator and Interaction Variables. Matt’s home
page. <http://pages.towson.edu/mchamber/chapter7eco306.pdf> [Retrieved 2010-05
-08].

Juchao Information. (2010). Initial public offerings prospectus disclosure. Juchao

formation website. < http://www.cninfo.com.cn/default.htm > [Retrieved 2010-0
4-10].

Ping, X.Q., & Li, Z.R. (2003). How to determine the quality of listed compan
y raise capital and fraud information disclosure. China economics and educatio
n website. <http://cenet6.nsd.edu.cn/userfiles/2009-9-15/20090915231253664.pdf>
[Retrieved 2010-04-28].

Tatro, Q. (2010). Why the Shanghai Index Should Be in Your Sights. Minyanv
ille website. <http://www.minyanville.com/businessmarkets/articles/china-Shanghai-
Stock-Index-traders-short/3/4/2010/id/27133> [Retrieved 2010-05-12].

Tonghuashun stock information software. (2010). Software download center. To

nghuashun financial service website. http://www.10jqka.com.cn/download [Retrie
ved 2010-03-10]

Wang, H.M. (2010). China International Capital Corporation hits the top again-
in IPOs market, and the underwriting fee is over 1.23 billion RMB. JRJ web
site. <http://finance.jrj.com.cn/2010/01/0416536745799.shtml> [Retrieved 2010-05-
19>

Zhongguancun. (2006, July). Haidianyuan Zhonggongguoji listed on Shenzhen st

ock exchange. Wei-pu information website. < http://www.cqvip.com/qk/86959x/2
006007/22347887.html > [Retrieved 2010-05-10].

72
Umeå School of Business
Umeå University
SE-901 87 Umeå, Sweden
www.usbe.umu.se