• No results found

Master Thesis Underpricing and long-term performance of foreign and domestic IPO

N/A
N/A
Protected

Academic year: 2021

Share "Master Thesis Underpricing and long-term performance of foreign and domestic IPO"

Copied!
45
0
0

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

Hele tekst

(1)

Master Thesis

Underpricing and long-term performance of foreign

and domestic IPO

Student name: Yanwei Wang Student number: S3201503 Study Programme: MSc IFM Supervisor: Dr. Adri de Ridder

Abstract: This thesis examines the underpricing and long-term performance of IPO and compare the difference between foreign IPO and domestic IPO. 2896 IPO deals

from 20 countries from 2000 to 2016 are used in this thesis and both univariate

analysis and regression model are used. Supportive evidence is found, showing that

the level of underpricing is higher in foreign IPO deals than that of domestic IPO

deals, while the level of long-term performance is lower in foreign IPO deals than that

of domestic IPO deals. Furthermore, foreign IPO firms in high-tech industry have

lower level of underpricing and higher level of long-term performance than foreign

IPO firms in traditional industry. The prestigious underwriter is positively associated

with the level of underpricing of foreign IPO firms and it is negatively associated with

the level of long-term performance. Overall, my results provide a deeper

understanding of foreign IPO based on the theory of information asymmetry.

(2)

Table of content

1. Introduction ... 5

2. Previous Studies ... 8

2.1. IPO and underpricing ... 8

2.2. Literature related to foreign IPO ... 9

2.3. IPO and stock performance ... 10

2.4. Industry impact ... 12

2.5. Underwriter impact ... 12

3. Hypotheses, Data and Methodology ... 14

3.1. Hypotheses development ... 14 3.2. Data ... 16 3.2.1 Sample collection ... 16 3.2.2. Sample distribution ... 17 3.2. Methodology ... 20 3.3.1. Major variables ... 20 3.3.2. Control variables ... 22 3.3.3. Regression models... 23 4. Results ... 26 4.1. Univariate analysis ... 26

4.2. Long-term performance of IPO ... 34

(3)

5

1.

Introduction

Initial public offering, IPO for short, refers to a company selling its shares to the

public on the open market for the first time. Undertaking IPO means shares of a firm

can trade freely between investors on stock market and it makes a privately-held

company transform into a publicly-traded company. After IPO, firms can raise capital

on stock exchange, compensate existing shareholder and executives, and increase its

public awareness. For the importance of IPO, Fama and French (2004) point out that “It is the point of entry that gives firms expanded access to equity capital, allowing them to emerge and grow”. Historically, most of firms sell their shares only in the exchange market of their home country. As a result of the globalization and

integration of international capital markets, firms have various sources of their equity

capital. The most common practice is cross-listing, which means firms list on

different exchanges instead of only on their home exchange. Besides that, there are

another interesting phenomenon should be taken seriously, which is foreign IPO.

Foreign IPO refers to a firm that forego listing on its domestic stock market entirely

and make their IPO directly on a foreign exchange (Moore et al., 2012). According to

the data of Ritter (2018), the proportion of foreign IPOs, whose offer price is higher

than $5, in US market is around 25% in recent ten years. The main concern of firms

undertaking foreign IPO is to get a larger base of investors, access to foreign market

and achieve long-term growth (Daily et al., 2005). The main objective of this thesis is

to do a comparative analysis between foreign IPO and domestic IPO and to

investigate factors that influence firms choose to make their IPO on a foreign

exchange rather than on the home market.

At first, we should define a clear concept of foreign IPO. In this thesis, foreign IPO

refers to a company forgo undertaking IPO on its home exchange but doing IPO on a

non-home country exchange (Ding et al., 2010; Francis Hasan and Li, 2001). So, the

(4)

distinguishing foreign IPO from cross-listing, I use IPO sample that goes public at

first time and there is no repetitive firm in the sample. According to previous IPO

studies, underpricing and long-term underperformance are two main concepts and the

information asymmetry is the mostly used theory in this field (Rock, 1986; Carter and

Manaster, 1990; Ritter, 1991; Loughran and Ritter, 1995; Fama and French, 2004).

For underpricing, Rock (1986) finds that information of firms is unevenly distributed

among investor and the discounted offer price in IPO can be seen as the compensation

for uninformed investors. Beatty and Ritter (1986) suggest that the greater the degree

of uncertainty for characterizing an IPO, the more it is underpriced. For underpricing

of foreign IPO, the level of information asymmetry is larger than that on local

exchange, and foreign IPO will be more underpriced. I develop first hypothesis based

on the theory of information asymmetry and test whether the level of underpricing in

foreign IPO is higher than that in domestic IPO.

In long-term, some scholars find that IPO firms’ market performance is lower than the

market or their matching firms (Loughran and Ritter, 1995; Aggarwal and

Rivoli,1990; Affleck-Graves et al., 1996; Carter et al. ,1998). Bessler and Thies

(2007) suggest that the underperformance of IPO in long-term illustrates that

investors are too optimistic about IPO firm’s long-run prospect in earlier stage and

then investors get more realistic through time. However, some other studies find IPO

firms overperform in long-term (Chalk and Peavy, 1987; Brav et al., 2000; Eckbo and

Norli, 2005). The second hypothesis aims to test whether the level of

underperformance in foreign IPO is higher than that in domestic IPO.

Besides underpricing and underperformance, the industry and the reputation of

underwriter are other two important factors in researching IPO. Previous studies find

high-tech firms are more willing to do IPO on foreign exchange (Hursti and Maula,

2007; Blass and Yafeh, 2000; Ding et al., 2010). As a result of listing on foreign

exchange, high-tech firms can lower its cost of information transfer and access to

(5)

7 investigate the impact of industry, I analyze whether foreign IPO firms in high-tech

industry have higher level of underpricing and underperformance.

The underwriter is another important factor that influence the success of IPO and

extent studies argue prestigious underwriter is positively associated with IPO success

(Carter and Manaster, 1990; Carter et al., 1998; Loughran and Ritter, 2004). The

highly reputation of underwriter can help IPO firms reduce the influence of

information asymmetry (Stiglitz, 1985; Pollock et al., 2004; O'Brien and Bhushan,

1990). However, the interest conflict between underwriter and IPO issuer can make

underwriter decrease the offer price for achieving more profit in reselling activity

(Kojima, 2007; Ellis et al., 2002; Biais et al., 2002). I analyze whether foreign IPO

firms with prestigious underwriter have higher level of underpricing and

underperformance.

This thesis contains 2968 observations from 2000 to 2016 in 20 countries and uses

methods of univariate analysis and regression model. At first, by comparing the mean

statistic of foreign IPO firms and domestic IPO firms, we can see that underpricing

(IR) of foreign IPO firms is higher than that of domestic IPO firms, while the level of

long-term performance (BHAR and CAR) is lower in foreign IPO firms. Secondly, I

examine the impact of high-tech industry and prestigious underwriter on underpricing

and long-term performance. The result of mean statistic in industry impact show that

foreign IPO firms in high-tech industry have lower level of underpricing than firms in

traditional industry, and they have higher level of long-term performance than firms

in traditional industry. However, domestic IPO firms in traditional industry have

lower level of underpricing and achieve more profit in long-term period than that in

high-tech industry, which can be caused by a larger information gap in high-tech

industry (Aaji and Brounen, 2002). The mean statistic in underwriter impact find

domestic IPO firms using prestigious underwriter can reduce their level of

underpricing and increase their long-term performance. Thirdly, results of regression

(6)

follows results of univariate analysis and hypothesis. Moreover, it shows that total

IPO firms in high-tech industry underperform in 24 months, and foreign IPO firms in

high-tech industry overperform in long-term. Total IPO firms with prestigious

underwriter can lower their level of underpricing and overperform in long-term.

While, foreign IPO firms with prestigious underwriter suffer a higher level of

underpricing and underperform in long-term.

The rest of the thesis is constructed as follows. In Section 2, the previous literature on

IPO and its underpricing and performance are introduced, as well as the impact of

industry and the prestige of underwriter. In Section 3, I develop the hypotheses of this

paper and discuss the data collection, variable construction and methodology. Results

are discussed in Section 4, and Section 5 concludes.

2.

Previous Studies

2.1. IPO and underpricing

The underpricing is one of the most important concepts when investigating IPOs. “Underpricing is an opportunity cost to a firm going public” (Ritter, 2011). It is a well-recognized anomaly that the offer price of IPOs on average are undervalued and,

hence, associated with a positive first day return. There are a range of theories that

have been developed to explain the phenomenon of underpricing, and many of which

are based on asymmetric information. The asymmetric information theory was first

introduced by Akerlof (1970), and he develops this theory by investigating the case of

automobile market. Rock (1986) combines the IPO with the asymmetric information

theory, and he divides investors into informed investors and uninformed investors.

Informed investors are only attempting to buy underpriced shares, as they have more

(7)

9 for investors. Carter and Manaster (1990) develop a model based on Rock’s and show

that IPO firm must compensate investors for costs incurred during the information

gathering process. Beatty and Ritter (1986) find that the degree of uncertainty in

characterizing the IPO is positively related to the level of IPO underpricing.

Moreover, some studies focus on the signalling theory, they argue that managers of

IPO firms can be seen as the informed party, and the behaviour of underpricing IPO reflect firms’ quality (Allen and Faulhaber, 1989; Welch, 1989). These findings provide the theoretical proof that IPO on a foreign exchange will be more underpriced

than on the local exchange, based on the theory of asymmetric information (Francis et

al., 2010). For the firms of developing countries who want to list in a major market,

mainly Chinese companies, will obtain higher IPO costs, more strict accounting

requirements and regulations. So, foreign IPO will bring less short-term financial

benefits than domestic IPO (Ding et al., 2010). Francis, Hasan and Li (2001) find

empirical evidence that for a matched sample, foreign IPOs are significantly more

underpriced. In the research of US market, the information asymmetry, such as the

cultural differences and less analyst following, and the higher home country risk are

the two main reasons that caused the price discrimination of foreign issuers in

developed markets (Bruner et al., 2006).

2.2. Literature related to foreign IPO

Because of the globalization and integration of capital markets, more and more firms

decide to undertake an IPO in foreign capital markets (Bell et al., 2014). As Daily et

al. (2005) conclude, the goal of foreign IPO is to achieve the further growth rather

than obtain short-term benefits. According to previous literature, incentives of foreign

IPO include developed market environment and legal systems, a broader base of

investors, professional agencies and expertise, and so on (Francis, Hasan and Li,

(8)

among countries that have larger degree of integration between home and foreign

capital. Pagano et al. (1999) find that listing on foreign exchange reduces the financial

barrier to foreign investors and increases firm’s business reputation. Other studies also

point out that listing on the major exchange instead of local exchange is a signal of firm’s high quality and it represents the firm’s prospects for the investors (Cheung and Lee,1995; Fuerst, 1998). Ding et al. (2010) reviewed the theory of entrepreneurial

signalling and examine the differences between foreign IPO and domestic based on

the data of Chinese exchange and Hong Kong exchange. The results show that the P/E

ratio in the HK exchange is lower than that in the Chinese exchange and the issuing

costs are also higher in Hong Kong, which means the short-term benefits is not the

main purpose of a foreign IPO. They find that undertaking foreign IPO is the signal of

the company who aiming at a long-run growth strategy.

However, for the firm adopting IPO in a foreign capital market, there are a few

difficulties which should be solved. Moore et al. (2012) believe that the liability of

foreignness is the greatest challenge for foreign IPO. The higher level of underpricing

has reflected the weakness of foreign IPO. Besides the liability of foreignness, there

are also intangible costs associated with foreign IPO. Stulz (2009) suggests that the

complex agency and information asymmetry problems are negatively associated with

the stock market performance. Other studies point out that foreign IPO firms are

harder to value (Caglio et al., 2011) and that leads their valuation worse than their

peers (Sarkissian and Schill, 2012). Colak et al. (2014) suggest in order to determine

the location of IPO, one needs to consider all of these intangible costs and benefits, in

addition to the proceeds raised in IPO.

2.3. IPO and stock performance

The long-run IPO underperformance is another major puzzle of IPO study. A range of

empirical studies focus on investigating the performance of IPOs based on different

(9)

11 firms and finds that the average holding period return (BHAR) of IPO firms is 34.47%, which is lower than the matching firms’ performance of 61.86%. Besides Ritter (1991), other studies also provide the evidence of the long-term

underperformance in IPO. Aggarwal and Rivoli (1990), Affleck-Graves et al. (1996)

and Carter et al. (1998) adopted the measurement of cumulative abnormal returns

(CAR) and their results show that IPO firms’ long-term performance are lower than

the market benchmarks. Loughran (1993) and Loughran and Ritter (1995) adopted the

measurement of wealth relatives and get the same conclusion. The information

asymmetry (Myers and Majluf, 1984) or the agency theory are used to explain the reason of IPO’s long-run underperformance (Jensen, 1986). The negative CAR or BHAR reflect that IPO firms underperform in long-term period. The possible

explanation of underperformance can be optimistic prospect from investors, which

means that investors are too optimistic about the long-term prospect during initial

period and are getting more realistic through time. Another possible reason is the

short-term overvaluation, which means IPO firms at first trading day are overpriced

instead of underpriced and the underperformance in long-term is a correction of the

short-term overvaluation.

However, some studies didn’t support the result of underperformance among the

studies of IPO performance. Chalk and Peavy (1987) find IPO firms outperform the

market benchmark, which is around 18%. Brav et al. (2000) find a positive

buy-and-hold abnormal return (BHAR) of 6.60% and a positive value-weighted

BHAR of 1.40%. Eckbo and Norli (2005) use both unconditional and conditional

factor models to estimate expected returns and find no evidence of long-run

(10)

2.4. Industry impact

Besides the underpricing and post-IPO performance, there are several factors that also affect the decision of foreign IPOs. The thesis will focus on the aspects of firms’ industry and the reputation of the underwriter.

According to Hursti and Maula (2007), most foreign IPOs among European firms are

in high-tech industry. Blass and Yafeh (2000) use the data of Israeli IPOs, which are

issued on US exchange and Israeli Exchange (Tel Aviv). Their study shows that the

young, innovative and high-tech Israeli companies are more likely to choose IPO on

an US exchange, and they find the purposes of these high-tech innovative firms are to

reveal their value and distinguish themselves from competitors in home market.

Pagano, et al. (2002) suggest that high-tech and export-oriented European firms are

attracted by U.S. exchanges with rapid expansionary plans. Further, Ding et al. (2010)

examine foreign IPOs based on the Chinese data. They defined the high-tech firms as

the firms with greater growth potential and they also find that the high-tech firms are

more likely to do IPO in foreign market. Based on the theory of information

asymmetry, Subrahmanyam and Titman (1999) argue listing on a market where many

similar companies are already listed can help firm lower its cost of information

transfer. For the high-tech firm, opting for listing in foreign exchange can get a larger

base of both individual and institutional investors who understand their business

(Hursti and Maula, 2007). Moreover, the firms with high growth opportunity focus

more on the long-term performance rather than the short-term benefits and listing on a

major market can provide firm more reliable capital and larger market (Ding et al.,

2010).

2.5. Underwriter impact

(11)

13 influence the success of IPOs. Some papers have already proved that the prestigious

underwriters are positively associated with IPO success (Carter and Manaster, 1990;

Carter et al., 1998; Loughran and Ritter, 2004). For the success of foreign IPO, the

impact of highly reputational underwriter is from three aspects. First, prestigious

underwriters have a strict criterion of their client and IPO firms meet the strict

requirement generally represents highly quality (Stiglitz, 1985). Second, the

monitoring system of highly prestigious underwriters can reduce the problem of moral

hazard (Pollock et al., 2004). Finally, prestigious underwriters have more experience

of foreign IPO and it creates a cascading effect of monitoring. Furthermore,

prestigious underwriters have a greater resource of additional analysts and

institutional investors who can execute useful valuation and monitoring of the firm

(O'Brien and Bhushan, 1990). All these factors reduce the level of information

asymmetric and can improve IPO performance.

However, the studies also find a fact that there is a conflict of interest for underwriter

in IPO issuing process (Kojima, 2007; Ellis et al., 2002; Biais et al., 2002). There are

two methods that an underwriter can make profit in the process of IPO. The first

source of profit is the commission as a percentage of the issue price that payed by the

issuer. In this situation, the underwriter’s profit maximization is consistent with the issuer’s profit maximization, which means high issue price is profitable for both issuer and underwriter (Loughran and Ritter, 2002; Kojima, 2007). The other source

is that the underwriter buying issued shares and re-selling them in the post-issue

market. The profit is from the gap between issue price and post-issue market price. At

this situation, the underwriter has the incentive to reduce the issue price and it comes

with an interest conflict between underwriter and issuer (Kojima, 2007; Ellis et al.,

2002; Biais et al., 2002). Kojima (2007) finds the evidence that the underwriters

generally make profits from the stock price gap between former decrease of

(12)

3.

Hypotheses, Data and Methodology

3.1. Hypotheses development

The first hypothesis of this study is related to the difference of underpricing level

between foreign IPO and domestic IPO. According to the information asymmetry

theory (Akerlof, 1970), discussed above, the underpricing is a well-recognized

anomaly that the offer price of IPOs on average are undervalued and, hence,

associated with a positive first day return. According to Rock’s (1986) analysis, when a firm’s information is more unevenly distributed among investors, the IPO of the firm is more likely to be underpriced. Beatty and Ritter (1986) find that the greater the

degree of uncertainty for IPO firms, the more it is underpriced. These findings

provide the theoretical foundation that the firms IPO on a foreign exchange will suffer

more underpriced than on the local exchange. Furthermore, for the firms of

developing countries who opt for foreign IPO will get higher IPO costs, more strict

accounting requirements and regulations. Foreign IPO brings less short-term financial

benefits than domestic IPO (Ding et al., 2010). Francis, Hasan and Li (2001) find the

empirical evidence that for a matched sample, foreign IPOs are significantly more

underpriced. However, the data of these articles are ten or twenty years ago, whereas

this thesis use recent 16 years’ data to examine whether the foreign IPOs are more

underpriced than the domestic IPOs. I use the initial return (IR) to examine

underpricing of foreign IPO and the first hypothesis is as follow:

Hypothesis 1: The initial return is higher for foreign IPO firms than for domestic IPO

(13)

15 If the foreign IPOs are more underpriced, the next question is why these firms choose

to suffer these costs? Focusing on the long-term growth is the most reasonable

explanation for firms choosing foreign IPO. Daily et al. (2005) find foreign IPO firms

aims to achieve the further growth rather than obtain short-term benefits. Ding et al.

(2010) also find undertaking foreign IPO is the signal of the company who aiming at a

long-run growth strategy. In this paper, I use cumulative abnormal returns (CAR) and

buy-and-hold abnormal returns (BHAR) to investigate whether these foreign IPOs

achieve the long-term growth and compare them with the domestic IPOs. So, the

second hypothesis of the paper is:

Hypothesis 2: Long-term performance of IPOs is higher for foreign IPOs than for

domestic IPOs.

The studies show that the high-tech and innovative oriented firms are more willing to

suffer the cost of foreign IPO for achieving long-term performance and getting

understand from the investors (Hursti and Maula, 2007; Blass and Yafeh, 2000; Ding

et al., 2010). Issuing on a major exchange can help firms in high-tech industry get

more potential growth, while for the traditional industry, it is not attractive. I believe

the foreign IPO decision is different among the industries. So, my third hypothesis is

as follow:

Hypothesis 3: Foreign IPO firms in high-tech industry have lower level of

underpricing than firms in traditional industry.

Hypothesis 4: Foreign IPO firms in high-tech industry have higher level of long-term

performance than firms in traditional industry.

The high reputation of underwriter can deliver a signal of low risk to the investor, and

both Carter and Manaster (1990) and Carter et al. (1998) find that underwriter

(14)

Harris (1999) point out that for investors, a prestigious underwriter means having

strict selections of IPO companies, less risky and high growth potential. Caglio et al.

(2016) find the global underwriters are mainly the prestigious underwriters in the

ranking list and they replace prestigious underwriter by global underwriter in their

research. The result shows that the choice of using a global underwriter is associated

with the decision of foreign listing. I posit that firms prefer to employ a highly

prestigious underwriter for achieving the lower underpricing and higher performance

of its foreign IPO. In the thesis, I will use the reputation of underwriter as an

explanatory variable to test the hypothesis:

Hypothesis 5: Foreign IPO firms using a prestigious underwriter have lower level of

underpricing than firms using a non-prestigious underwriter.

Hypothesis 6: Foreign IPO firms using a prestigious underwriter have higher level of long-term performance than firms using a non-prestigious underwriter.

3.2. Data

3.2.1 Sample collection

The IPO sample of this paper is from Zephyr Database. The database provides the

information of each IPO case, which includes the issuer, the issue date, the offer price,

the closing price of first trade day, the number and type of shares offered, the IPO

exchange and so on. My sample meet the following criteria: (1) the deal type is initial

public offering, (2) the deal status is confirmed completed, (3) the deal offer price is

more than 1 USD, (4) the security type are ordinary or common shares, and (5) the

time period is from year 2000 to year 2016. Besides IPO deals from Zephyr

Databased, I also collect the IPO deals of Hong Kong exchange from a Chinses

financial database, Wind Datafeed, and add 146 observations into the sample. The

(15)

17 feature of IPO, some financial data is also used in this paper, which includes the firms’ total assets, total liabilities, total equity, and stock price. Some of data can download within Zephyr Database together with IPO deals, some of the other are

collected from DataStream. Finally, I exclude observations that are lack of available

information and my final sample comprise 2968 IPOs from 20 countries.

3.2.2. Sample distribution

Table 1 shows the sample distribution and there are 3 panels which are based on the

listing exchanges’ country, year and industry. Panel A reports the sample distribution

by the listing exchanges’ country. It shows the number of domestic, foreign and total

IPO from 20 countries. We can see that IPO deals in USA exchange is around 28

percentage and it includes 207 foreign IPO deals. The IPO deals in Chinese exchange

also has a high proportion of 22%, while there is no foreign company adopting IPO in

Chinese exchange. Hong Kong exchange also attract a large number of foreign IPO

deals in our sample, and the companies from mainland China are a big slice of these

foreign IPO. Furthermore, there are some foreign IPOs in Australia, Germany, UK

and other 9 countries. The distribution of foreign IPOs can indicate that US exchange

is the most attractive IPO location for foreign firms and Hong Kong is another global

financial centre of the Far East. Panel B reports the sample distribution by year from

2000 to 2016 and it shows an upward trend year by year. Panel C is the sample

distribution according to the industry. The high-tech industry firms are around 16% of

total observations, while the high-tech industry firms in foreign IPO deals is about

30%, which is higher than the percentage of domestic IPO (13%). It indicates that the

firms of high-tech industry are more likely to undertake foreign IPO and it is accord

with previous studies (Hursti and Maula, 2007; Ding et al., 2010; Blass and Yafeh,

(16)

Table 1 Sample distribution

Panel A Distribution of Countries (the listing country of each firm)

Countries Domestic

IPO Foreign IPO Total IPO

(17)

19 Panel B Distribution of Years

Year Domestic

IPO Foreign IPO Total IPO

% of total observations 2000 10 8 18 0.61% 2001 10 3 13 0.44% 2002 12 2 14 0.47% 2003 7 4 11 0.37% 2004 20 10 30 1.01% 2005 26 7 33 1.11% 2006 33 13 46 1.55% 2007 40 19 59 1.99% 2008 22 7 29 0.98% 2009 13 14 27 0.91% 2010 33 38 71 2.39% 2011 271 29 300 10.11% 2012 324 32 356 11.99% 2013 290 71 362 12.20% 2014 492 97 589 19.85% 2015 518 81 599 20.18% 2016 371 41 411 13.85% Total 2492 476 2968 100.00%

Panel C Distribution of Industries

Domestic

IPO Foreign IPO Total IPO

% of total observations

High-tech industries 360 126 486 16.37%

Traditional industries 2132 350 2482 83.63%

Total 2492 476 2968 100.00%

(18)

3.2. Methodology

3.3.1. Major variables

Measuring the choice of foreign IPO

Foreign IPO (FI): The difference of foreign IPO and domestic IPO is the main topic

of this paper. I create a dummy variable of FI in the regression models. The dummy

variable, FI, is equal to 1 if the firm do its IPO on a foreign exchange and it is equal to

0 if the firm do its IPO on domestic exchange.

Measuring underpricing of IPO

Initial Return (IR): The measurement of underpricing in the thesis will be the initial

return and it is calculated as the difference of the first closing price and the offer price

divided by the offer price. i.e. IR =𝑃1−𝑃0

𝑃0 , where P1 is closing price of first trading

day and P0 is the offer price.

Measuring the long-run performance of IPO in US

To evaluate the long-run performance of IPOs, there are two measurements are used

in previous studies. One measurement is using cumulative abnormal returns (CAR),

which usually using monthly cumulative abnormal returns of IPO firm and then

adjusted by the benchmark. This method is recommended by Aggarwal and Rivoli

(1990), Affleck-Graves et al. (1996), Carter et al. (1998) and Chalk and Peavy (1987).

Another measurement is using buy-and-hold abnormal returns (BHAR) that also using

monthly data and adjusted by the benchmark, which outcome is equal to the measure

of wealth relatives. This method is recommended by Ritter (1991), Loughran (1993),

and Loughran and Ritter (1995). This paper chooses both CAR and BHAR as the

measure of the long-run performance, and that can help us to check the robustness of

results. We can see from equation (1) and equation (2) that both measurements are

using abnormal returns and benchmark, and the difference between CAR and BHAR

(19)

21 The calculation of CAR following Ritter (1991). In equation (1), 𝑅𝑖,𝑡 is the raw

return of an IPO firm i in month t and 𝑅𝑚,𝑡 is the benchmark, which is the market

index at same period. CAR is computed as the difference between the raw return and

the benchmark return across T months:

𝐶𝐴𝑅𝑖,𝑡 = ∑𝑇𝑡=1[𝑅𝑖,𝑡 − 𝑅𝑚,𝑡] (1)

The time periods for calculating CAR are 12 months after IPO date and 24 months

after IPO date. The benchmark for adjusting the CAR is the market index of the

listing exchange for each firm.

BHAR is my alternative to the use of CAR, we also compute the buy-and-hold market adjusted return for firm i, defined as in equation (2):

𝐵𝐻𝐴𝑅𝑖,𝑡 = ∏𝑇𝑡=1(1 + 𝑅𝑖,𝑡)− ∏𝑇𝑡=1(1 + 𝑅𝑚,𝑡) (2)

where 𝑅𝑖,𝑡 is the raw return of an IPO firm i in month t and 𝑅𝑚,𝑡 is the benchmark, which is the market index at same period. The time periods for calculating BHAR are

12 months after IPO date and 24 months after IPO date. The benchmark for adjusting

is the market index of the listing exchange for each firm.

Measuring the impact of new industry

High-tech industry (High-Tec): This variable identified whether a firm is in the

high-tech industry and has a greater opportunity of growth potential, which introduced

by Ding et al. (2010). The NAICS 2017 industry code and description are used to

define the main business operation of each company. The measurement of high-tech

industry is a dummy variable, which equals to 1 if the main operation of the company

focuses on information technology, biotechnology, pharmaceuticals, new energy and

(20)

Measuring the impact of prestigious underwriter

Underwriter ranking (UW_rank): Among the previous research, there are two

methods of measuring the reputation of underwriter. Caglio et al. (2016) use the

global underwriter as the measurement. This measurement defines the global

underwriter as an underwriter takes at least two IPOs from different countries in

previous year. Otherwise, the underwriter only take part in domestic business cannot

be defined as global. However, the ranking of underwriter is adopted in more studies

(Carter and Manaster, 1990; Carter et al., 1998; Loughran and Ritter, 2004) and

comparing to the measurement of global underwriter, the ranking of underwriter is

more convenient for the research. The paper uses IPO Underwriter Reputation

Rankings (1980 – 2015) from Ritter’s (2018) website, and this ranking has combined

the results from Carter and Manaster (1990), Carter et al. (1998), and Loughran and

Ritter (2004). The underwriter prestige rankings are on a 0 to 9 scale, and in this

paper, I divided IPO deals into two groups. The UW_rank is a dummy variable and it

equals 1 if the ranking is more than 7, otherwise equals 0. For the unrecorded

underwriter in IPO cases, I rank it as below the average underwriter ranking, which

means UW_rank of these observations is 0.

3.3.2. Control variables

For the control variables, the thesis use leverage and age of the firm, which are used

in the most previous studies (Ding et al., 2010; Moore et al., 2012; Song et al., 2014).

The leverage (LEV) is ratio of total liabilities to total equity. The age of the firm

(AGE) is the number of operating years before firm goes public. Besides these three

common control variables, the IPO proceeds is a typical control variable in many IPO

studies, which is the natural log of offer size (Proceeds). Carter et al. (1998) argued

that the larger IPOs are generally from the more established and financially stronger

firms. Furthermore, the fraction of shares offered in IPO (Fraction) is another typical

control variable in most IPO-related studies (Francis et al., 2001). The control

(21)

23 3.3.3. Regression models

In my study, both univariate analysis and regression models are adopted. Following

the method of Francis et al. (2001), I use the univariate analysis to compare the mean

of IR from foreign and domestic IPOs, as well as the mean of CAR and BHAR from

both foreign and domestic IPOs. These mean differences can test the hypothesis 1 and

2. Moreover, the thesis also creates two sub-sample groups to test the data in different

dimensions. The first sub-sample group consists of IPOs in high-tech industry and

that in traditional industry. The second sub-sample group of consists of IPOs with

prestigious and that with non-prestigious underwriters. Comparing the mean

differences of IR, CAR and BHAR in these sub-sample groups aims to test the

hypothesis 3 and 4.

For the regression models, I use IR as the dependent variable in equation (3). In

equation (3), 𝛽1 is predicted to be positive, which means the level of underpricing on

foreign IPO will be higher than that on domestic IPO. 𝛽2 and 𝛽3 are used for

testing for the industry impact on underpricing level of total IPO and foreign IPO,

separately, and they are predicted to be negative. 𝛽4 and 𝛽5 is for testing the

underwriter impact, and they are predicted to be negative.

𝐼𝑅𝑖 = 𝛽0+ 𝛽1𝐹𝐼𝑖+ 𝛽2𝐻𝑖𝑔ℎ_𝑡𝑒𝑐𝑖 + 𝛽3𝐻𝑖𝑔ℎ_𝑡𝑒𝑐𝑖 ∗ 𝐹𝐼𝑖 + 𝛽4𝑈𝑊_𝑟𝑎𝑛𝑘𝑖

+𝛽5𝑈𝑊_𝑟𝑎𝑛𝑘𝑖 ∗ 𝐹𝐼𝑖+𝛽6𝐿𝐸𝑉𝑖 + 𝛽7𝐴𝑔𝑒𝑖+ 𝛽8𝐹𝑟𝑎𝑐𝑡𝑖𝑜𝑛𝑖 + 𝛽9𝑃𝑟𝑜𝑐𝑒𝑒𝑑𝑠𝑖+ 𝜀𝑖 (3)

The dependent variable of equation (4) and equation (5) are CAR (1,12) and CAR (1,

24) and they are used for testing the long-run performance. 𝛽1 is predicted to be

positive, which means the level of performance on foreign IPO will be higher than

that on domestic IPO. 𝛽2 and 𝛽3 are used for testing for the industry impact on long-run performance of total IPO and foreign IPO, separately, and they are predicted

(22)

predicted to be positive. 𝐶𝐴𝑅(1,12)𝑖 = 𝛽0+ 𝛽1𝐹𝐼𝑖 + 𝛽2𝐻𝑖𝑔ℎ_𝑡𝑒𝑐𝑖 + 𝛽3𝐻𝑖𝑔ℎ_𝑡𝑒𝑐𝑖∗ 𝐹𝐼𝑖 + 𝛽4𝑈𝑊_𝑟𝑎𝑛𝑘𝑖 +𝛽6𝑈𝑊_𝑟𝑎𝑛𝑘𝑖 ∗ 𝐹𝐼𝑖 + 𝛽7𝐿𝐸𝑉𝑖 + 𝛽8𝐴𝑔𝑒𝑖+ 𝛽9𝐹𝑟𝑎𝑐𝑡𝑖𝑜𝑛𝑖 +𝛽10𝑃𝑟𝑜𝑐𝑒𝑒𝑑𝑠𝑖 + 𝜀𝑖 (4) 𝐶𝐴𝑅(1,24)𝑖 = 𝛽0+ 𝛽1𝐹𝐼𝑖 + 𝛽2𝐻𝑖𝑔ℎ_𝑡𝑒𝑐𝑖 + 𝛽3𝐻𝑖𝑔ℎ_𝑡𝑒𝑐𝑖∗ 𝐹𝐼𝑖 + 𝛽4𝑈𝑊_𝑟𝑎𝑛𝑘𝑖 +𝛽6𝑈𝑊_𝑟𝑎𝑛𝑘𝑖 ∗ 𝐹𝐼𝑖 + 𝛽7𝐿𝐸𝑉𝑖 + 𝛽8𝐴𝑔𝑒𝑖+ 𝛽9𝐹𝑟𝑎𝑐𝑡𝑖𝑜𝑛𝑖 +𝛽10𝑃𝑟𝑜𝑐𝑒𝑒𝑑𝑠𝑖 + 𝜀𝑖 (5)

In equation (6) and equation (7), BHAR (1,12) and BHAR (1,24) are used also for

testing the long-run performance as the same as equation (4) and equation (5). 𝛽1, 𝛽2, 𝛽3, 𝛽4 and 𝛽5 are predicted to be positive

(23)

25 Table 2 Hypotheses and explanatory variables

Panel A hypotheses and expected sign of coefficients

Hypotheses Variables (coefficients)

1. The initial return (IR) is higher for foreign IPO firms than for domestic IPO firms.

FI (+)

2. Long-term performance of IPOs is higher for foreign IPOs than for domestic IPOs.

FI (+)

3. Foreign IPO firms in high-tech industry have lower level of underpricing than firms in traditional industry.

High-tec*FI (-)

4. Foreign IPO firms in high-tech industry have higher level of long-term performance than firms in traditional industry.

High-tec*FI (+)

5. Foreign IPO firms using a prestigious underwriter have lower level of underpricing than firms using a non-prestigious underwriter.

UW_rank*FI (-)

6. Foreign IPO firms using a prestigious underwriter have higher level of long-term performance than firms using a non-prestigious underwriter.

UW_rank*FI (+)

Panel B list of variables

Variables Symbol Description

Foreign IPO FI 1, if the company adopts foreign IPO; 0, if the company adopts domestic IPO

Initial Return IR The difference of the first closing price and the offer price divided by the offer price. i.e. IR =𝑃1−𝑃0

𝑃0 ,

Long-term performance

CAR The cumulative abnormal returns (CAR), which is adjusted by market index.

Long-term performance (alternative)

BHAR The buy-and-hold abnormal returns (BHAR), which is adjusted by market index.

Industry High-tec 1, if the company belongs to Hi-tech, Biotech, Pharmaceutical industry and so on; 0, if the company belongs to traditional industry

Underwriter UW_rank 1, if the company use a prestigious underwriter; 0, otherwise Leverage LEV Total debt divided by total assets in the year of IPO

Firm’s age AGE Age of the firm in number of years before IPO IPO proceeds Proceeds The natural log of offer size

Fraction of shared offers

(24)

4. Results

4.1. Univariate analysis

Table 3 reports the summary statistics of the variables used in this thesis. Panel A

presents descriptive statistics for the full sample of 2968 observations, which contains

2492 domestic IPO deals and 476 foreign IPO deals. Panel B presents descriptive

statistics for domestic IPOs of 2492 observations, and Panel C presents descriptive

statistics for foreign IPOs of 476 observations. In panel A, the mean of IR in total

IPOs is 0.030 and it indicates that IPO deals are underpriced by 3% on average. For

the mean of CAR in total IPOs, it is 0.096 for 12 months and 0.131 for 24 months.

The positive mean of CAR means that IPO firms, on average, perform better than the

market in long-run. And the mean of BHAR also indicates this phenomenon.

Table 3 Summary statistics

Panel A Descriptive statistic of full sample, 2968 observations

Variable Mean Median Max. Min. Std. Dev.

FI 0.160 0.000 1.000 0.000 0.367 IR 0.030 0.003 26.300 -1.000 1.140 CAR (1,12) 0.096 0.003 7.708 -2.341 0.649 CAR (1,24) 0.131 0.017 14.288 -3.504 0.790 BHAR (1,12) 0.064 -0.063 7.270 -1.286 0.724 BHAR (1,24) 0.114 -0.087 32.652 -1.689 1.205 High-Tec 0.164 0.000 1.000 0.000 0.370 UW_rank 0.645 1.000 1.000 0.000 0.479 LEV 1.512 0.506 249.567 0.000 6.693 Age 12.224 9.000 296.000 0.000 17.326 Proceeds 4.434 4.454 7.141 0.301 0.763 Fraction 0.453 0.307 1.000 0.003 0.282

Panel B Descriptive statistic of domestic IPO deals, 2492 observations

Variable Mean Median Max. Min. Std. Dev.

FI 0.000 0.000 0.000 0.000 0.000

IR 0.022 0.000 26.300 -0.985 1.232

(25)

27 CAR (1,24) 0.172 0.037 14.288 -3.504 0.827 BHAR (1,12) 0.079 -0.058 7.270 -1.286 0.737 BHAR (1,24) 0.134 -0.051 12.754 -1.659 1.047 High-Tec 0.144 0.000 1.000 0.000 0.352 UW_rank 0.650 1.000 1.000 0.000 0.477 LEV 1.453 0.520 249.567 0.000 6.406 Age 12.987 9.000 296.000 0.000 17.901 Proceeds 4.394 4.450 7.141 0.602 0.731 Fraction 0.431 0.286 1.000 0.003 0.269

Panel C Descriptive statistic of foreign IPO deals, 476 observations

Variable Mean Median Max. Min. Std. Dev.

FI 1.000 1.000 1.000 1.000 0.000 IR 0.076 0.030 7.158 -1.000 0.406 CAR (1,12) -0.045 -0.015 1.539 -2.001 0.374 CAR (1,24) -0.079 -0.039 2.221 -2.332 0.501 BHAR (1,12) -0.018 -0.119 6.434 -1.236 0.650 BHAR (1,24) 0.009 -0.252 32.651 -1.418 1.820 New_in 0.265 0.000 1.000 0.000 0.442 UW_rank 0.613 1.000 1.000 0.000 0.487 LEV 1.821 0.436 153.257 0.000 8.033 Age 8.231 4.000 134.000 0.000 13.254 Proceeds 4.646 4.478 6.640 0.301 0.886 Fraction 0.566 0.620 1.000 0.003 0.315

Note: Panel A presents descriptive statistics for the full sample of 2968 observations, Panel B presents descriptive statistics for domestic IPOs of 2492 observations, and Panel C presents descriptive statistics for foreign IPOs of 476 observations. FI is a dummy variable, and it is equal to 1 if the firm do its IPO on a foreign exchange and it is equal to 0 if the firm do its IPO on domestic exchange. IR is the initial return that is calculated as the difference of the first closing day price and the offer price divided by the offer price. CAR is the cumulative adjusted abnormal. BHAR is the adjusted buy-and-hold abnormal returns and the benchmark is the market index. High-Tec is a dummy variable that defines whether the firm is in the high-tech industry and UW_rank is another dummy variable that defines whether the underwriter of IPO is prestigious. LEV is ratio of total liabilities to total equity. AGE is the number of operating years before firm goes public. Proceeds is the natural log of offer size.

(26)

Based on the goal of investigating differences between domestic IPO and foreign IPO,

Table 4 lists the means of variables for both foreign and domestic IPO. By calculating

the mean difference between foreign IPOs and domestic IPOs, we can test our

hypotheses. The mean of IR in domestic IPOs is 0.022, while that is 0.076 in foreign

IPOs. The higher level of IR in foreign IPOs means foreign IPOs are more

underpriced by 0.055 than domestic IPOs, and the result is statistically significant at

10% significance level by using Mann–Whitney test. It illustrates that foreign IPOs

are more underpriced than domestic IPOs, which is in consistent with the hypothesis

1. The mean of CAR in domestic IPOs is positive and it indicates domestic IPOs

outperform in long-run. However, the mean of CAR in foreign IPOs is negative and it

indicates foreign IPOs underperform than the market in long-run. The mean differences of 12 months’ and 24 months’ CAR between domestic IPOs and foreign IPOs are 0.169 and 0.251, and this result is statistically significant. We can conclude

that domestic IPO firms outperform than foreign IPO firms in long-run. The

hypothesis 2 is rejected and firms that decide to undertake a foreign IPO cannot obtain

the better performance on stock market than domestic IPO firms. The mean

differences between BHAR in 12 month and in month are also positive, which

indicate that the performance of domestic IPO firms is better than foreign IPO firms.

The result is in consistent with the results of CAR differences. Further, we can see

that the mean difference of High-Tec is -0.120 at 1% significance level. And it

indicates that firms in high-tech industry are more likely to opt for a foreign IPO. Moreover, the mean of foreign IPO firms’ age is lower than that of domestic IPO firms, while the means of foreign IPOs’ size and fraction are higher than that of

domestic IPOs. Appendix B is the results of skewness and kurtosis of variables, and

we can see that variables are not normal distribution. The Mann–Whitney test is

(27)

29 Table 4 Univariate analysis

Variable Domestic IPO Foreign IPO Mean

Difference Mann–Whitney test p-value No. of observations 2492 476 IR 0.022 0.076 -0.055 0.000*** CAR (1,12) 0.124 -0.045 0.169 0.000*** CAR (1,24) 0.172 -0.079 0.251 0.000*** BHAR (1,12) 0.079 -0.018 0.097 0.003*** BHAR (1,24) 0.134 0.009 0.125 0.000*** High-Tec 0.144 0.265 -0.120 0.000*** UW_rank 0.650 0.613 0.037 0.122 LEV 1.453 1.821 -0.368 0.090* AGE 12.987 8.231 4.756 0.000*** Proceeds 4.395 4.646 -0.251 0.000*** Fraction 0.431 0.566 -0.135 0.000***

Note: Table 4 presents means of domestic IPOs and foreign IPOs. IR is the initial return that is calculated as the difference of the first closing day price and the offer price divided by the offer price. CAR is the cumulative adjusted abnormal returns. BHAR is the adjusted buy-and-hold abnormal returns and the benchmark is the market index. High-Tec is a dummy variable that defines whether the firm is in the high-tech industry and UW_rank is another dummy variable that defines whether the underwriter of IPO is prestigious. LEV is ratio of total liabilities to total equity. AGE is the number of operating years before firm goes public.

Proceeds is the natural log of offer size. Fraction is the percentage of the shares that the

company offered in its IPO. The null hypothesis of Mann–Whitney test is that the mean of testing variables between foreign and domestic IPOs are not different from each other. ***, **, and * indicate significance at the 1%, 5%, and 10% levels.

The thesis also creates two sub groups, which are classified by industry and

underwriter, to investigate the impact of high-tech industry and reputational

underwriter. Appendix B is the results of skewness and kurtosis of variables, and we

can see that variables are not normal distribution. The Mann–Whitney test is adopted

(28)

Table 5 is the results of industry impact by using comparative mean statistics. For

total IPO deals and domestic IPO deals, the mean of IR in high-tech industry is higher

than that in traditional industry, 0.040 (1% significant level) and 0.047 (1%

significant level); the mean of CAR and BHAR in high-tech industry is lower than

that in traditional industry. This indicates that domestic IPO firms in traditional

industry reduce level of underpricing and increase performance in long-term. Kim et

al. (2008) find the same results in their research and they explain this phenomenon in

the view of financial distress. They find high-tech firms’ cash flows tend to be more

volatile and they typically have few tangible assets, which will lead to a higher cost of

financial distress, and high-tech firms generally create little or no profit in early years.

Aaji and Brounen (2002) also find underpricing and underperformance of high-tech

IPO firms and they argue the scale of that is larger than traditional industry. Based the

information asymmetry, investors have more information of traditional industry in

comparison of high-tech industry and a larger information gap in high-tech industry

causes a larger-than-traditional level of underpricing and underperformance.

However, for foreign IPO firms, the mean of IR in high-tech industry is lower than

that in traditional industry (-0.011, at 1% significant level), and the mean of CAR and

BHAR in high-tech industry is higher than that in traditional industry. This result is in

consistent with the hypothesis that foreign IPO firms in high-tech industry have lower

level of underpricing than firms in traditional industry, and they have higher level of

(29)

31 Table 5 Industry impact

High-tech industry Traditional industry Difference Total IPO observations 486 2482 Mean of IR 0.064 0.024 0.040*** CAR (1,12) 0.052 0.105 0.053 CAR (1,24) 0.018 0.154 -0.136*** BHAR (1,12) 0.049 0.066 -0.017** BHAR (1,24) 0.113 0.114 -0.001*** Domestic IPO observations 360 2132 Mean of IR 0.062 0.015 0.047*** CAR (1,12) 0.069 0.133 -0.064 CAR (1,24) 0.058 0.191 -0.132** BHAR (1,12) 0.032 0.087 -0.056** BHAR (1,24) 0.065 0.146 -0.081*** Foreign IPO observations 126 350 Mean of IR 0.068 0.079 -0.011* CAR (1,12) 0.002 -0.062 0.064* CAR (1,24) -0.098 -0.073 -0.025 BHAR (1,12) 0.099 -0.061 0.160* BHAR (1,24) 0.252 -0.078 0.330**

(30)

Table 6 is the results of underwriter impact by using comparative mean statistics. For

total IPO deals and domestic IPO deals, the mean of IR in prestigious underwriter

group is lower than that in non-prestigious underwriter group, -0.241 (1% significant

level) and -0.282 (1% significant level); the mean of CAR and BHAR in prestigious

underwriter group is higher than that in non-prestigious underwriter group. This

indicates that domestic IPO firms using prestigious underwriter can lower their level

of underpricing and achieve more profit in long-term period. Previous studies have

proved this result that prestigious underwriters are positively associated with IPO

success (Carter and Manaster, 1990; Carter et al., 1998; Loughran and Ritter, 2004). Contrary to the result of domestic IPO firms, the mean of foreign IPO firms’ BHAR in prestigious underwriter group is lower than that in non-prestigious underwriter

group (-0.139 at 1% significant level, -0.084 at 5% significant level). It seems the

prestigious underwriter is a liability for foreign IPO. In the sample of foreign IPO

deals, prestigious underwriters are usually famous large investment banks while

non-prestigious underwriters are local financial institutions. In my opinion, when

facing the interest conflict between underwriter and issuer (Kojima, 2007; Ellis et al.,

2002; Biais et al., 2002), prestigious underwriters, large international investment

banks, will sacrifice the interest of foreign IPO firms. On the contrary,

non-prestigious underwriters as home country financial institutions will protect the

(31)

33 Table 6 Underwriter impact

Prestigious underwriter Non-prestigious Underwriter Difference Total IPO observations 1913 1055 Mean of IR -0.055 0.186 -0.241*** CAR (1,12) 0.160 -0.020 0.180*** CAR (1,24) 0.235 -0.056 0.291*** BHAR (1,12) 0.085 0.024 0.061*** BHAR (1,24) 0.149 0.050 0.098*** Domestic IPO observations 1621 871 Mean of IR -0.077 0.205 -0.282*** CAR (1,12) 0.196 -0.012 0.208*** CAR (1,24) 0.284 -0.037 0.321*** BHAR (1,12) 0.114 0.015 0.098*** BHAR (1,24) 0.180 0.048 0.131 Foreign IPO observations 292 184 Mean of IR 0.064 0.095 -0.031 CAR (1,12) -0.038 -0.057 0.019 CAR (1,24) -0.037 -0.146 0.109 BHAR (1,12) -0.072 0.066 -0.139*** BHAR (1,24) -0.023 0.061 -0.084**

(32)

4.2. Long-term performance of IPO

Figure 1, 2 and 3 include CAR, BHAR and cumulative raw returns for total IPO deals,

domestic IPO deals, and foreign IPO deals, respectively. We can see from each figure

that there is a big gap between domestic IPO performance and foreign IPO

performance. This gap indicates domestic IPO firms outperform than foreign IPO

firms in stock market. In figure 1, performance of domestic IPO firms is always

higher than the benchmark while that of foreign IPO firms is lower than the

benchmark in five months after IPO date. The figure 2 also shows the same results.

Patterns of foreign CAR and foreign BHAR illustrate the underperformance of IPO,

while patterns of total IPO and domestic IPO show a result of outperformance, which

coincide with studies of Chalk and Peavy (1987) and Brav et al. (2000). The possible

reason is that the information asymmetry as a factor of underperformance is

significant in case of foreign IPO deals (Myers and Majluf, 1984) and insignificant in

domestic IPO deals.

Figure 1 Cumulative average adjusted returns

Note: Figure 1 is the patterns of CAR for total IPO deals, domestic IPO deals, and foreign IPO deals,

respectively, with monthly rebalancing. The horizontal axis is month from 1st month after IPO date to

24th month. The vertical axis is CAR.

-15.00 -10.00 -5.00 0.00 5.00 10.00 15.00 20.00 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 CA R ( 1, t), % Month

(33)

35 Figure 2 Buy-and-Hold abnormal returns

Note: Figure 2 is the patterns of BHAR for total IPO deals, domestic IPO deals, and foreign IPO deals,

respectively, with monthly rebalancing. The horizontal axis is month from 1st month after IPO date to

24th month. The vertical axis is BHAR.

Figure 3 the long-term performance of IPO deals

Note: Figure 3 is patterns of raw returns (no adjusted) for total IPO deals, domestic IPO deals, and

foreign IPO deals, respectively. The horizontal axis is month from 1st month after IPO date to 24th

month. The vertical axis is cumulative return rate at t month.

-10.00 -5.00 0.00 5.00 10.00 15.00 20.00 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 B HA R ( 1,t), % Month

Total BHAR Dome BHAR Foregin BHAR

-5.00 0.00 5.00 10.00 15.00 20.00 25.00 30.00 35.00 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24

(34)

4.3. Regression models

Besides the univariate analysis, I also use regression models to investigate IPOs.

Table 6 shows the results of regression models. The variable FI is a dummy that to

test underpricing and long-term performance of foreign IPO. The coefficient of FI is

negative and it is statistically significant in model (3), which means foreign IPO firms

underperform in 24 months after IPO date.

The variable High-Tec is to test how industry influence the level of underpricing and

long-term performance. The coefficient is negative, and it is statistically significant in

model (3), which means IPO firms in high-tech industry underperform in 24 months

after IPO date. It is in consistent with the result in univariate analysis and it caused by

lager level of information asymmetry in high-tech industry (Aaji and Brounen, 2002).

The variable High-Tec*FI aims to test the industry impact on foreign IPO. We can see

coefficients are positive and they are statistically significant in model (2), (4) and (5),

which indicates foreign IPO firms in high-tech industry overperform in long-term.

The variable UW_rank is to test how reputation of underwriter influence the level of

underpricing and long-term performance. The coefficient is negative, and it is

statistically significant in model (1), which means IPO firms with prestigious

underwriter can lower their level of underpricing. Coefficients in model (2), (3), (4)

and (5) are positive and statistically significant, which indicates IPO firms with

prestigious overperform in long-term. The variable UW_rank*FI tests the underwriter

impact on foreign IPO. The coefficient in model (1) is positive at 10% significant

level, and it indicates foreign IPO firms with prestigious underwriter suffer a higher

level of underpricing. Coefficients in model (2), (3), (4), and (5) are negative and they

are statistically significant, which indicates foreign IPO firms with prestigious

underwriter underperform in long-term. The opposite result between variable

UW_rank and UW_rank*FI might due to the interest conflict between underwriter

(35)

37 Table 7 Results of regression models

Variables Model (1) [IR] Model (2) [CAR (1,12)] Model (3) [CAR (1,24)] Model (4) [BHAR (1,12)] Model (4) [BHAR (1,24)] Constant 0.812*** 0.005 -0.046 -0.126 -0.327** [0.135] [0.076] [0.092] [0.086] [0.143] FI -0.048 -0.069 -0.124* -0.006 -0.100 [0.096] [0.054] [0.065] [0.061] [0.102] High-Tec -0.002 -0.038 -0.089* -0.023 -0.029 [0.068] [0.038] [0.046] [0.043] [0.072] High-Tec*FI -0.103 0.138* 0.114 0.243*** 0.443*** [0.135] [0.076] [0.091] [0.086] [0.143] UW_rank -0.259*** 0.196*** 0.303*** 0.087*** 0.114** [0.048] [0.027] [0.033] [0.031] [0.051] UW_rank*FI 0.221* -0.171*** -0.171** -0.236*** -0.221* [0.117] [0.066] [0.080] [0.075] [0.125] LEV 0.002 -0.001 -0.002 0.001 0.000 [0.003] [0.002] [0.002] [0.002] [0.003] Age -0.002 0.000 0.000 0.000 0.002 [0.001] [0.001] [0.001] [0.001] [0.001] Proceeds -0.146*** 0.019 0.038** 0.004** 0.095*** [0.028] [0.016] [0.019] [0.018] [0.030] Fraction 0.090 -0.186*** -0.292*** -0.110** -0.114 [0.079] [0.044] [0.053] [0.050] [0.084] Observations 2968 2968 2968 2968 2968 R2 0.022 0.037 0.061 0.014 0.012

(36)

5. Conclusion

In this paper, I analyse the underpricing and the long-term performance of IPO deals,

by splitting total IPO deals into domestic IPO and foreign IPO. The thesis is based on

the theory of information asymmetry, and total collected sample consists of 2968

observations from 2000 to 2016 in 20 countries. The main objective in this thesis is to

examine the difference of underpricing and long-term performance between foreign

IPO firms and domestic IPO firms. It also investigates how the industry affect

underpricing and long-term for IPO firms. The reputation of underwriter is examined

as well.

This thesis adopts both univariate analysis and regression models and it produced

several findings. At first, by comparing the mean statistic of foreign IPO firms and

domestic IPO firms, we can see that the level of underpricing (IR) in foreign IPO

firms is higher than that of domestic IPO firms, while the level of long-term

performance is lower in foreign IPO firms. This finding is consistent with prior

studies (Ding et al., 2010; Francis, Hasan and Li, 2001). Secondly, I examine the

impact of high-tech industry and prestigious underwriter on underpricing and

long-term performance. I create two sub group, which are classified by the industry

and reputation of underwriter. The result of mean statistic in industry impact show

that foreign IPO firms in high-tech industry have lower level of underpricing than

firms in traditional industry, and they have higher level of long-term performance

than firms in traditional industry, which is in consistent with hypothesis. However,

domestic IPO firms in traditional industry have lower level of underpricing and

achieve more profit in long-term period than that in high-tech industry, which can be

caused by a larger information gap in high-tech industry (Aaji and Brounen, 2002).

The mean statistic in underwriter impact find domestic IPO firms using prestigious

underwriter can reduce their level of underpricing and increase their long-term

(37)

39 long-term performance than that using non-prestigious underwriter. My interpretation

of this findings is due to the interest conflict between underwriter and issuer (Kojima,

2007; Ellis et al., 2002; Biais et al., 2002). Thirdly, results of regression models show

foreign IPO firms underperform in 24 months after IPO date, which follows results of

univariate analysis and hypothesis. Moreover, it can be seen that total IPO firms in

high-tech industry underperform in 24 months after IPO date foreign IPO firms in

high-tech industry overperform in long-term. And total IPO firms with prestigious

underwriter can lower their level of underpricing and overperform in long-term.

While, foreign IPO firms with prestigious underwriter suffer a higher level of

underpricing and underperform in long-term.

The thesis makes several contributions to the literature. Firstly, it proved that

information asymmetric theory is useful to explain IPO research. Secondly, I extend

the extant IPO literature. Most of previous studies did not distinguish foreign IPO

from domestic IPO. I collect IPO data from 20 countries and compare the difference

between foreign IPO and domestic IPO. Finally, the extent IPO studies have different

views about the industry impact and the underwriter impact. My thesis provides new

evidence to this area.

There are also some limitations in this paper. Firstly, there are only 476 foreign IPO

deals used in this thesis. Further research can be improved by collecting more foreign

IPO observations. Secondly, only two explanatory variables, high-tech industry and

prestigious underwriter, are used in regression models. For foreign IPO study, there

should contain more explanatory variables, such as home country, foreign sales and so

on. Finally, some results in this paper are nor proved by previous studies and lack of

theoretical proof. For example, foreign IPO firms with prestigious underwriter are

underpriced more than that with non-prestigious underwriter and have lower

long-term performance. It requires future research to focus more on the theory and to

(38)

Appendix A

Table 8, Correlation matrix

Correlation FI IR CAR (1,12) CAR (1,24) BHAR (1,12) BHAR

(1,24) High-Tec UW_rank LEV AGE Proceeds Fraction

FI 1 IR 0.018 1 CAR (1,12) -0.096 -0.104 1 CAR (1,24) -0.117 -0.145 0.796 1 BHAR (1,12) -0.049 -0.089 0.733 0.560 1 BHAR (1,24) -0.038 -0.093 0.400 0.530 0.641 1 High-Tec 0.119 0.013 -0.031 -0.064 -0.009 -0.000 1 UW_rank -0.028 -0.101 0.133 0.176 0.040 0.039 0.074 1 LEV 0.020 0.006 -0.013 -0.021 0.008 -0.001 -0.042 0.014 1 AGE -0.101 -0.033 0.016 0.024 0.018 0.031 -0.074 0.037 0.028 1 Proceeds 0.121 -0.098 0.021 0.041 0.041 0.052 -0.177 0.043 0.041 -0.021 1 Fraction 0.176 0.042 -0.111 -0.148 -0.058 -0.040 0.283 -0.079 0.023 -0.108 -0.044 1

(39)

41

Appendix B

Table 9 Skewness and Kurtosis of variables

FI IR CAR_12 CAR_24 BHAR_12 BHAR_24 HIGH_TEC UW_RANK AGE LEV PROCEEDS FRACTION

Skewness 1.851 14.286 2.655 2.791 3.414 8.935 1.817 -0.604 5.096 24.005 -0.207 0.552 Kurtosis 4.426 257.650 22.254 42.118 24.788 194.786 4.303 1.365 47.169 761.999 4.366 1.869 Jarque-Bera 1946.46 0 8120319.0 00 49331.650 193089.700 64472.020 4588158.0 00 1843.688 511.121 254112.2 00 71526875.0 00 252.016 308.975 Probability 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 Observations 2968 2968 2968 2968 2968 2968 2968 2968 2968 2968 2968 2968

Referenties

GERELATEERDE DOCUMENTEN

One of the equilibria born in the saddle node bifurcation turns stable and an unstable limit cycle emerges through a subcritical Hopf bifurcation.. When we enter region (4), we are

Since the event violates the expected sequence of events, the reactor regexp publishes the reactor event violated, which is be bound to the event violated specified in the

Ook situationele kenmerken kunnen hier aan bijdragen, zoals een lage financiële behoefte (iemand gaat bijna met pensioen, kan op zijn partner bouwen of heeft geen kinderen) of de

This study is designed to address these questions by evaluating the extent to which preschool children (between three and five years old) can learn (foreign) literacy skills

In the user evaluation we wanted to examine whether the text representation form (full-text, key sentences, key phrases) had an influence on the correctness of the labels assigned to

A simple reading of the Gospels and the Pauline corpus shows that the concept of purity and impurity is used in different contexts by Jesus and Paul, a fact that is due, I suppose,

In this study, we present and evaluate a robotically actuated delivery sheath (RADS) capable of autonomously and accurately compensating for beating heart motions by using

As shown in the previous section, Plant Simulation provides a set of basic objects, grouped in different folders in the Class Library.. We now present the most commonly used