IPOs Underpricing: Evidence from Chinese Companies
Jie Zhang 10618643
Supervisor: Razvan Vlahu
University of Amsterdam
Faculty of Economics and Business
Bachelor thesis Finance and Organization
Abstract
Over the past decades, a growing number of Chinese companies have been listed overseas. Chinese initial public offerings (IPOs) underpricing have drawn significant attention due to their magnitude. Existing literature has investigated A-share IPOs underpricing patterns. In this thesis, 899 A-share IPOs and 50 N-share IPOs are examined to investigate the driving forces behind IPOs underpricing. Ordinary least-squares regressions are employed to assess several potential explanatory variables. The results indicated that information asymmetry and elapsed time between the IPO announcement day and the IPO issuing day can partly explain the different degree of A-share IPOs and N-share IPOs underpricing and the influence of the overall financial environment on IPOs underpricing is significant for A-share IPOs but not for N-share IPOs.
Statement of Originality
This document is written by Zhang Jie who declares to take full responsibility for the contents of this document. I declare that the text and the work presented in this document are original and that no sources other than those mentioned in the text and its references have been used in creating it. The Faculty of Economics and Business is responsible solely for the supervision of completion of the work, not for the contents.
1. Introduction
When seeking for funds to support further growth, companies often choose to go public in order to generate capital. Initial public offerings (IPOs) are referred to the first time when private companies issue stocks to the public to raise funds. It has been reported by Ritter and Welch that more than one company was listed in U.S. everyday between 1980 and 2001 (2002). Moreover, based on the data obtained, from 2010 to 2014, nearly 112,361,578,153 dollars are generated from Chinese IPOs market.
Established in the early 1900s, Chinese IPOs market offers two kinds of shares: A-shares and B-shares. A-shares are defined as standard stocks issued by Chinese companies and sold in the domestic market exclusively, while B-shares are first offered to foreign investors only but later to domestic investors also (Wu, 2014). The Chinese IPOs market is highly regulated and was controlled by Chinese authorities until early 2000s, when the control over IPOs pricing and quotas were abolished.
As China’s rapid economic development attracted large amounts of foreign investors, an increasing amount of leading Mainland Chinese-based companies chose to go public in foreign stock exchanges to raise foreign capital. Based on the records of NYSE and NASDAQ, 515 Chinese companies listed in America referred to as Chinese concept stocks (N-shares) (2016). Chinese concepts stocks are defined as Chinese stocks, whose main financial transactions take place in Mainland China but are listed overseas (Luo, Fang, and Esqueda, 2012). According to Hsieh and Walkling (2006), concept stocks are referred to as stocks with extremely high price to sales ratios. Thus, China concepts stocks are considered an investment in the long-term economic growth of China and have a growing influence on the world stock market (Luo, Fang, and Esqueda, 2012).
When companies go public, abnormal returns of IPOs are often observed, which are considered as compensation for first-day investors (Ritter and Welch, 2002). In the past few years, Chinese companies’ IPOs have drawn significant attention due to their magnitude as the highest underpricing worldwide. From 1990 to 2007, the average abnormal first-day return of A-share IPOs exceeded 200% (Wu, 2013). Su and Fleisher argues that B-shares are priced at discount compared with A-shares to avoid currency controls and enhance companies’ reputations (1999). Same underpricing patterns can be applied to N-shares IPOs, since N-shares IPOs also target foreign investors.
The central focus of this study is to determine in which IPO are underpricing patterns more significant, A-share IPOs, which are targeted to domestic investors, or N-share IPOs, Chinese stocks listed in the United States and targeted towards international investors. If there is a difference between the two IPO underpricing patterns, thus arises the question, what are the driving forces behind these circumstances? Data of 949 sample companies that went public in China and U.S. between 2010 and 2014 are collected and analyzed through ordinary least-squares (OLS) regressions to investigate the determinants of different underpricing patterns.
The rest of this thesis is organized as follows: the theoretical framework is introduced in section 2 and 3, in which IPO markets background in China and U.S. and related explanatory theories of IPOs underpricing will be discussed. Subsequently, methodology and hypotheses are presented in section 4. Section 5 presents and discusses the results of the analysis. Lastly section 6 offers a conclusion and closing remarks.
2. IPO market background
From 1990 to 2001, an Administrative Review and Approval system (Quota system) was employed by the Chinese authorities (Cheng and Li, 2015). Under the Quota system, an annual quota was determined by the State Planning Commission in conjunction with the People’s Bank of China (the central bank) and the China Securities Regulatory Committee (CSRC) and allocated to each province regarding aggregate economic planning (Su and Fleisher, 1999). The quota system was highly regulated and controlled by authorities. As a result, development opportunities of companies are limited under this system. Thus, the quota system was abolished and replaced by the Channel and Approval system (Approval system),
which was more market-oriented, in March 2001 (Cheng and Li, 2015). Under the new system, there were no limitations on the shares issued each year, and underwriters take charge over part of the approval authority and make recommendations on issuers to the CSRC. The Approval system was determined to be more effective than the quota system (Guo and Brooks, 2009). Average IPO abnormal returns decreased from 948.6% between 1987 and 1995 reported by Su and Fleisher (1999) to 93.49% from 2001 to 2005 reported by Guo and Brooks (2009) after the adoption of the Approval system. Moreover, under the Quota system, the average time gap between the IPO announcement day and listing day was 260 days (Su and Fleisher, 1999), while Guo and Brooks (2009) reported that the elapsed time varies from 10 to 39 days, after adopting the Approval system. Although the waiting period has decreased dramatically, it is still greater than the U.S. exchanges, in which the typical elapsed time between the offering date and issuing day is 0-1 day.
Compared with the U.S. capital market where IPO pricing and share supply are derived by market forces (Chiou, Li, Cheng, and Chang, 2010). A-share IPOs pricing and shares supply are strictly controlled and must be proved by CSRC for a buffer period until 2004, when the Company and Securities Laws were amended (Chiou, Li, Cheng, and Chang, 2010). Three main pricing and allocation mechanisms were implemented by the CRSC: fixed price, auctions and bookbuilding (Wu, 2014). The fixed price system was the first allocation system adopted by the CRSC. A fixed price is set by the issuer and underwriter; then the shares are allocated to investors (Wu, 2014). Later, the auction system was employed for a short period; where the price was mainly set by investors’ online bidding, regarding a minimum accepted price set by issuers and underwriters. Later, the bookbuilding system with a price cap taking into account the P/E ratio was adopted in 2004. Under the bookbuilding system, a price range is set by issuers after investigating the interests of investors. The price cap was abolished in 2008 and IPO pricing is mainly derived by market force. IPO pricing and allocation in the Chinese capital market became more market-orientated afterwards. However, as a developing capital market, Chinese IPOs mechanisms were still considered inefficient. Thus, Chinese companies have incentives to go public overseas.
As a result of China’s recent openness and globalization, Chinese companies sought to go public in foreign exchanges. Since the early 1990s, growing numbers of Chinese companies,
attracted by better economic and financial fundamentals in the U.S., have become listed on NYSE and NASDAQ (Luo, Fang, and Esqueda, 2012). Better economic fundamentals in a country such as higher income and growth opportunities, are associated with more internationalization of their firms including listing, trading and capital raising in international exchanges (Claessens, Klingebiel and Schmukler, 2006). The liquidity theory also suggested that Chinese companies benefit from the U.S. stock market, a developed market with considerable market size and liquidity. Foreign investors began to show interest in Chinese issuers after the Asian financial crisis, due to the strong performance of Chinese stocks (Zhang and King, 2010). China’s rapid economic growth with a huge potential growth attracted a great amount of foreign investors (Zhang and King, 2010). Based on past records, the Industry and Commerce Bank of China (ICBC) and Alibaba Group Holding Limited are the two largest Chinese companies listed in America, collected $21.6 billion and 21.8 billion respectively.
Moreover, Chinese companies benefit from issuing costs when going public in the U.S. capital market. Abnormal initial return of IPOs is referred to as evidence of IPO underpricing, which is considered as part of the issuing costs of listing (Güçbilmez, 2014). The average underpricing of U.S. IPOs was 17.9% from 1980 to 2012 (Güçbilmez, 2014) while the average underpricing of Chinese IPOs was 111% from 2001 to 2009 (Lin and Tian, 2012). Thus, the issuing costs are relatively lower in the U.S. than in China.
3. Literature review
In recent decades, an increasing amount of literature has examined the abnormal returns of initial public offerings (IPOs) in various countries. Since the 20th century, Chinese IPO underpricing aroused widespread concern, as high IPOs with abnormal returns was observed. As reported by Su and Fleisher (1999), an average IPOs’ return of 948.6% between 1987 and 1995, was observed in a sample of 308 Chinese A-shares. Moreover, Zhou and Zhou (2010) examined a sample of 1380 A-share IPOs and reported an average initial underpricing of 238% from 1991 to 2005. Existing literature has investigated the IPO underpricing patterns of A-share common stocks and plentiful explanatory theories have been tested. However, there remains a lack of research examining the underpricing pattern of Chinese concept stocks.
The subsequent three factors can describe the difference in underpricing patterns in China and America:
a. Asymmetric information
Theoretical explanations of IPOs underpricing have been classified by Ritter and Welch, into two categories based on whether asymmetric information or symmetric information is assumed (2002). In this paper, two scenarios of asymmetric information, information asymmetry between investors and issuers and asymmetric information asymmetry among investors, will be discussed.
Under the scenario of asymmetric information between issuers and investors, issuers process private information and have a more accurate forecast of companies’ future cash flows. Asymmetric information can be an explanation of the underpricing pattern as a strategy to signal a company’s value (Su and Fleisher, 1999). With respect to the compensation theory, underpricing patterns can be utilized to compensate for uninformed investors. Allen and Faulhaber (1989) suggested a category of signaling models where IPO underpricing is the equilibrium in a situation where issuers process private companies’ information relative to investors. It is optimal for issuers to signal their quality by IPOs underpricing and investors will be informed about companies’ future prospects through underpriced initial shares (Allen and Faulhaber, 1989).
On the other hand, the greater the difficulty in acquiring information for rational investors, the more compensation on the price of the IPO required. Differences in IPO returns between N-shares and A-share can be referred to the compensation theory, since foreign investors have more limited access to information on Chinese than domestic investors. The winner’s course proposed by Rock (1986), implies that informed investors will buy in underpriced new shares to gain profits; while uninformed investors will leave the IPO market because of unprofitable investment due to lack of information in the long run. As a result, IPOs underpricing ensures uninformed investors’ participation in IPOs market. Moreover, Su and Fleisher (1999) argued that it is reasonable to assume that in international markets, information asymmetry poses as a larger obstacle to fund raising than in the domestic market so that companies with well-established domestic histories have competitive advantages when raising foreign capitals. In brief, foreign investors require more compensation on the price of IPOs.
b. Waiting period
The waiting period between offerings and listings is one of the significant differences in the issuing systems of IPOs in China and America. According to Mok and Hui’s (1998) study, the longtime gap between the IPO announcement date and listing date is observed in China and a positive relationship between the elapsed time and IPO underpricing is reported. Su and Fleisher described the same results in a sample of 308 A-share IPOs (1999). On the contrary, based on the data obtained, the waiting period is normally between 0 to 1 day in America. Thus, there is a significant difference in the waiting period in China and America.
Theoretical explanation has been documented by Chowdhry and Sherman (1996). When the elapsed time between IPO announcement day and issuing day, private price information can be revealed to investors before the bidding period. As a result, when abnormal return of IPO is realized ex ante, a large oversubscription is observed and when IPO price is realized to be “too high”, IPO fails (Chowdhry and Sherman, 1996). To avoid the failure of IPO, IPO is normally underpriced. Furthermore, the listing lag is considered as a risk associated to IPOs and uninformed investors need more compensation due to higher uncertainty of the investment (Chen, Firth and Kim, 2004). The longer the lagging period between offering and listing the risker the investment for two reasons that (1) asymmetric information is distributed among investors, and that (2) funds raised from IPOs are locked up during this period (Chan, Wang, and Wei, 2004).
c. Different financial environments
Different financial environments also lead to different levels of underpricing patterns since China is a typical developing capital market and the United States is a typical developed capital market. Dewenter and Malatesta (1997) argued that IPO returns are substantially greater in developing capital markets such as Malaysia than in highly developed capital markets such as Canada. Loughran, Ritter and Rydqvist also reported that the magnitude of underpricing varies gratefully across 25 countries and argued that the risker the companies went public the larger the degree of IPOs underpricing (1994). The annual IPO volume and the inflation level of the stock market are positively correlated (Loughran et al., 1994). Loughran et al. suggested that as investment opportunities are controlled by the authorities during the earlier 1990s, issuers benefited from time-variation in stock market valuations regarded real GDP
growth (1994). This phenomenon implies that the overall situation of the economy with respect to the overall investment risk is considered, when issuing IPOs. Since the 21st century, IPOs markets are more market-orientated. IPOs are more likely to be influenced by the fluctuation of the stock market.
4. Methodology and hypotheses
OLS regressions are employed to examine different magnitude of A-share IPOs and N-share IPOs underpricing patterns. Empirical models formulated in this paper are built with respect to the three explanatory theories presented in the literature review. The aim of these models is to examine the explanatory variables that drive different underpricing patterns of A-share IPOs in China and N-share IPOs in America.
4.1 Variables
Measurement of IPOs underpricing
The main concern of this thesis is the IPO abnormal return. According to Su and Fleisher (1999), the return on the first closing day is employed as a measurement of IPOs underpricing, defined as:
𝑅𝑒𝑡$ =𝑝'$− 𝑝)$ 𝑝)$
Where 𝑝)$ refers to the initial offering price and 𝑝'$ refers to the closing stock price of the first day. The higher the initial return the more serious the IPO underpricing.
Explanatory Variables
Table 1: Explanatory variables and explanation Explanatory Variable Explanation
Chinai A dummy variable described the listing location
China=1 A-share IPOs in China China=0 N-share IPOs in U.S.
Agei Established years of a listed company before its IPOs
Valuei Capital raised of a company from IPOs
ValueIPOsi=PriceIPOsi*SharesIPOsi
PriceIPOs: initial public offering price per share SharesIPOs: number of shares of IPOs issued
PEi Price to earnings ratio of a company 1 year after the listing
date of a company
Wi Elapsed days between the IPO issuing date and the listing
date of a company
SS300Indexi China Securities Index, an A-share stocks market index
derived from 300 identical listed companies on Shanghai Stock Exchange and Shenzhen Stock Exchange
S&P500Indexi The Standard & Poor’s 500 Index, an American stock market
index derived from 500 identical listed companies on the NYSE or NASDAQ
4.2 Data and descriptive statistics
Data of all A-shares listed in China and N-shares listed in America from January 2010 to December 2014 were examined. The sample period excluded the period of financial crisis, thus eliminating the influence of the turbulent capital market from 2007 to 2009. Based on financial records, 949 companies went public during this period, 899 A-share IPOs and 50 N-share IPOs. Two sub-samples were extracted for the purpose of analyzing the impact of different financial environments on IPO underpricing. One contained 899 A-share IPOs, and the other contained 50 N-share IPOs. All the data were collected from official stock exchange websites. Descriptive statistics are contained in Table 1 presented below.
Table 2: Descriptive statistics
a. Sub-sample of 899 A-share IPOs from January 2010 to December 2014
Variable Mean Standard deviation Minimum Maximum
Ret 0.33456 0.49678 -0.44135 8.93621 China 1 1 1 1 Age 11.13557 4.98269 1 52 Value 799904449.6 2364473854 44400000 59590587900 lnValue 19.93008 0.88270 17.60875 24.81076 PE 45.53577 21.54514 6.23 150.82 W 11.02225 2.93499 0 23
SS300Index 2840.36159 360.64155 2122.84 3533.71 lnSS300Index 7.94330 0.13102 7.66051 8.17010 b. Sub-sample of 50 N-share IPOs from January 2010 to December 2014
Variable Mean Standard
deviation Minimum Maximum Ret 0.64067 1.70473 -0.91538 8.85 China 0 0 0 0 Age 8.84 4.57304 2 26 Value 584615749.28 3067988545.41898 8905000 21767215616 lnValue 18.39212 1.27732 16.00212 23.80367 PE 48.47935 87.66496 1.32000 359.12000 W 0.44 0.70450 0 4 S&P500Index 1518.72776 369.80055 1022.58 2126.64 lnSP500Index 7.292828 0.23557 6.93008 7.66230
As presented in Table 2, two significant differences between A-share IPOs and N-share IPOs can be seen. The first-day average abnormal return of A-share IPOs were 33.456% while for N-share IPOs was 64.067%. The difference between the two shares’ IPOs was 30.611%. The average waiting period from the IPO announcement day to the IPO issuing day of A-share IPO was nearly 11 days, which is reduced sharply from an average elapsed time of 260 days reported by Su and Fleisher (1999). However, for N-share IPOs, the time elapsed between announcement day and listed day was less than 1 day.
4.3 Hypotheses
Hypothesis 1: N-share IPOs listed in America were more seriously underpriced than A-share
IPOs.
Model 1 is employed in order to examine the central question of this thesis: if there exists a difference in the underpricing patterns of A-share IPOs and N-share IPOs and to measure which underpricing pattern is more serious.
Model 1:
The dependent variable is the IPO abnormal return (Ret). The independent variable is the dummy variable, China. The difference in IPO underpricing patterns between A-share and N-share is examined by the dummy variable China.
Su and Fleisher (1999) reported different degree of underpricing of B-share IPOs targeted international investors and A-share IPOs targeted domestic investors. As observed, B-share IPOs offered the same rights and dividends as A-share IPOs but are priced lower than A-share IPOs. The average initial return is 178% for A-share IPOs and 11.6 for B-share IPOs from 1993 to 1998 (Chan et al., 2004). Issuers try to avoid currency controls and enhance their reputations when raise foreign capitals. Similar to B-share IPOs, N-share IPOs are targeted to international investors. Moreover, the longer lagging period between IPO offering and issuing also implies that A-share IPOs is more underpriced than N-share IPOs. However, according to the theory of asymmetric information, N-share IPOs is predicted to be more underpriced than A-share IPOs as a compensation to international investors. Issuers underprice initial shares to signal their quality of IPOs. Meanwhile, international issuers need more compensation to stay in the IPOs market than domestic investors because they have limited access to companies’ information. Hence, how N-share IPOs’ underpricing pattern is different from A-share IPOs’ is ambiguous.
Hypothesis 2: The influence of information asymmetry on N-share IPOs is more significant
than on A-share IPOs.
There sub-hypotheses are examined to investigate the magnitude of the influence of information asymmetry on A-share and N-share IPOs. Hypotheses 2b and 2c are derived from singling models according to Su and Fleisher (1999), and testable regarding the data set obtained.
Hypothesis 2a: IPO underpricing and the established years of companies before listing are
negatively correlated. Model 2a:
𝑅𝑒𝑡$ = 𝜅)+ 𝜅'𝐴𝑔𝑒$+ 𝜅5𝐶ℎ𝑖𝑛𝑎$∗ 𝐴𝑔𝑒$+ 𝜆$
The dependent variable is the IPO abnormal return (Ret). The explanatory variables are Age and China*Age. The relationship between the established year of companies before listing and IPOs underpricing is investigated by Model 2a. Age represented companies’
established years and China*Age assesses the different degree of influence of established years of companies before IPOs between A-shares IPOs and N-shares IPOs.
The influence of asymmetric information in international IPOs market is greater compared with domestic IPOs market so that companies with longer established histories are viewed as a strong evidence of future prospects (Su and Fleisher, 1999). Well-established companies have competitive advantages in international capital markets. A negative relationship between established year of a company and IPOs underpricing is assumed because the more years a company founded the more information revealed to investors so that less influence of asymmetric information on IPOs underpricing.
Hypothesis 2b: IPOs underpricing and IPOs price are negatively correlated.
Model 2b:
𝑅𝑒𝑡$ = 𝜌)+ 𝜌'𝐴𝑔𝑒$+ 𝜌5𝑙𝑛𝑉𝑎𝑙𝑢𝑒$+ 𝜌=𝐶ℎ𝑖𝑛𝑎$∗ 𝑙𝑛𝑉𝑎𝑙𝑢𝑒$+ 𝜏$
The dependent variable is the IPO abnormal return (Ret). According to Su and Fleisher (1999), IPO price is defined as the amount companies’ received from IPOs regarding the fraction of shares offered. The explanatory variables are lnValue, logarithm of the amount of capital a company raised from its IPO, and China*lnValue, holding other variables constant. The capital raised from IPO is in the natural logarithm form to avoid extreme values. The coefficient of lnValue is predicted to be negative. China*lnValue examines the different degree of the signaling of A-shares IPOs and N-shares IPOs. Investors infer IPOs value from IPOs price. Only high-quality companies will retain a large proportion of shares when issuing IPOs (Su and Fleisher, 1999). Thus a negative relationship between IPOs price and IPOs underpricing is assumed.
Hypothesis 2c: IPOs underpricing and companies’ initial value are negatively correlated.
Model 2c:
𝑅𝑒𝑡$ = 𝜋)+ 𝜋'𝐴𝑔𝑒$+ 𝜋5𝑙𝑛𝑉𝑎𝑙𝑢𝑒$+ 𝜋=𝑃𝐸$+ 𝜋B𝐶ℎ𝑖𝑛𝑎$∗ 𝑃𝐸$. +𝜇$
The dependent variable is the IPO abnormal return (Ret). The explanatory variables are PE and China*PE, holding other variables constant. The different magnitude of the signaling of A-share IPOs and N-share IPOs is judged by the variable China*PE. An optimal signaling schedule of companies’ IPOs is regarded with their initial value (Su and Fleisher, 1999). In this paper, the initial value of a company is represented by its price to earnings ratio one year
after the issuing referred to Su and Fleisher’s study and a negative relationship between PE ration and IPOs underpricing is predicted.
Hypothesis 3: The waiting period between the IPO announcement day and the listing day is
positively correlated to IPO underpricing. Model 3:
𝑅𝑒𝑡$ = 𝛾)+ 𝛾'𝐴𝑔𝑒$+ 𝛾5𝑙𝑛𝑉𝑎𝑙𝑢𝑒$+ 𝛾=𝑃𝐸$+ 𝛾B𝑊$+ 𝛾G𝐶ℎ𝑖𝑛𝑎$∗ 𝑊$+ 𝜉$
The dependent variable is the IPO abnormal return (Ret). Independent variables are W and China*W, holding other variables constant. The lagging period between IPO offering and issuing is presented by W and the difference in the degree of the influence of the lagging period on A-share and N-share IPOs underpricing is tested by China*W.
A positive correlation between the elapsed time of the IPO announcement day and issuing day is observed by Su and Fleisher (1999) and Chan et al. (2004). When the lagging period is long, private information can be revealed and the investment risk is increased. Thus, the longer the elapsed time, the larger the underpricing. The lagging period of A-share IPOs is generally longer than N-share IPOs based on the data obtained. Hence, it is reasonable to estimate that when holding constant control variables, the coefficients of W and China*W are positive.
Hypothesis 4: Overall financial environment has a significant impact on the IPOs
underpricing pattern.
Two Sub-samples are employed to run the following models separately, aimed to examine the influence of the overall financial environment on IPOs underpricing:
Model 4a for China=1:
𝑅𝑒𝑡$ = 𝜃)+ 𝜃'𝐴𝑔𝑒$+ 𝜃5𝑙𝑛𝑉𝑎𝑙𝑢𝑒$+ 𝜃=𝑃𝐸$+ 𝜃B𝑊$+ 𝜃G𝑙𝑛𝑆𝑆300𝐼𝑛𝑑𝑒𝑥$+ 𝜑$ Model 4b for China=0:
𝑅𝑒𝑡$ = 𝜗)+ 𝜗'𝐴𝑔𝑒$+ 𝜗5𝑙𝑛𝑉𝑎𝑙𝑢𝑒$+ 𝜗=𝑃𝐸$+ 𝜗B𝑊$+ 𝜗G𝑙𝑛𝑆𝑃500𝐼𝑛𝑑𝑒𝑥$+ 𝜔$ The dependent variable is the IPO abnormal return (Ret) for Chinai=1 and Chinai=0 respectively. lnSS300Indexi and lnS&P500Indexi are the explanatory variables in the two models, holding other variables constant. SS300 Index and S&P500Index are in natural logarithm forms and represent the overall financial environments of stock markets in China and America, respectively.
Loughran et al. reported a positive relationship between IPOs volume and inflation-adjusted level of the stock market in 14 countries out of 15 (1994). Gross national product (GNP) was employed as the measurement of the inflation level of the stock market since the authorities took in charge of the IPOs market regarded the aggregate economic situation in 1990s. The influence of overall financial environment on IPOs market is significant. Companies experienced success when the stock market is optimal (Loughran et al., 1994). As a developed stock market, the American IPOs market is market-orientated, and after the adoption of the Approval system, the Chinese IPOs market also tends to be determined by the market. Therefore, stocks market index is employed as the measurement of the fluctuation of the stock market in this thesis because it reflects the tendency of the stock market. Accordingly, the coefficients of both variables are predicted to be significant and negative.
5. Results
5.1 Correlation between IPO underpricing and listing countries Model 1: 𝑅𝑒𝑡$ = 𝛽)+ 𝛽'𝐶ℎ𝑖𝑛𝑎$+ 𝜀$
Table 3: OLS regression estimates of model 1
Ret Coefficient P-value
Intercept 0.640667 0.000
China -0.3061068 0.001
The theory of signaling discussed in section 3a implies a negative correlation between 𝑅𝑒𝑡$, representing the abnormal return of IPO, and 𝐶ℎ𝑖𝑛𝑎$, a dummy variable described the listing location of the IPO. Table 3 exhibits the empirical results of Model 1, indicating a negative correspondence between the dependent variable Ret and explanatory variable China corresponded with the signaling theory. The coefficient of China is -0.3061068 inferring that the underpricing is larger when Chinese companies are listed in America than in China and the difference on average is 30.61%. The result is also complied with the data description.
Denoted 10% level of significance, the difference in abnormal return between A-share IPOs and N-share IPOs is significant. Thus, hypothesis 1 is supported by the estimated result. Though N-share IPOs and B-share IPOs are both targeted international investors, N-share
IPOs are more underpriced than A-share IPOs while B-share IPOs are less underpriced than A-share IPOs. Further hypotheses needed to be examined to determine the driving forces behind this phenomenon.
5.2 The degree of influence of asymmetric information Model 2a: 𝑅𝑒𝑡$ = 𝜅)+ 𝜅'𝐴𝑔𝑒$+ 𝜅5𝐶ℎ𝑖𝑛𝑎$∗ 𝐴𝑔𝑒$+ 𝜆$ Table 4: OLS regression estimates of model 2a
Ret Coefficient P-value
Intercept 0.3185707 0.000
Age 0.0318127 0.001
China*Age -0.0301726 0.001
The regression results of model 2a is presented in Table 4. The coefficients of China*Age, -0.0318127, are negative as predicted, while the coefficient of Age, 0.0318127, is positive, opposite to the prediction. The estimated coefficients imply that: (1) the established year of a company before IPOs and the IPOs first-day abnormal return of that company are positively correlated which is opposite to the prediction, and that (2) with A-share IPOs, when companies’ established years increase by 1 year, IPOs abnormal return will increase by 0.16401%, and with N-share IPOs, when companies established years increases by 1 year, IPOs abnormal return will increase by 3.18127%, which is larger than A-share IPOs.
Denoted at the 10% level of significance, the results of model 2a is significant according to the P-values of Age, 0.001, and China*Age, 0.001. Though hypothesis 2a is not supported by the results, it shows that N-share IPOs underpricing suffers more from asymmetric information than A-share IPOs with respect to the established years of companies before listing.
Model 2b: 𝑅𝑒𝑡$ = 𝜌)+ 𝜌'𝐴𝑔𝑒$+ 𝜌5𝑙𝑛𝑉𝑎𝑙𝑢𝑒$+ 𝜌=𝐶ℎ𝑖𝑛𝑎$∗ 𝑙𝑛𝑉𝑎𝑙𝑢𝑒$+ 𝜏$ Table 5: OLS regression estimates of model 2b
Ret Coefficient P-value
Intercept 1.47701 0.001
Age 0.0020678 0.611
China*lnValue -0.0115075 0.027
As can be seen from Table 5, holding Age constant, the coefficients of lnValue and China*lnValue are both negative as predicted, -0.0469461 and -0.0115075. The results imply that: (1) the capital raised from IPOs and the IPO abnormal return are negatively correlated as predicted, and that (2) for A-share IPOs, when the capital raised from IPOs by 1%, the IPO abnormal return will decrease by 58.4536%, and for N-share IPOs when the capital raised from IPOs by 1%, the IPO abnormal return will decrease by 4.69462%.
At a denoted 10% level of significance, the regression estimate of lnValue is not significant while the estimated coefficient of China*lnValue is sufficient. Though hypothesis 2b is not fully supported by the results of model 2b, the influence of capital raised on IPO underpricing patterns of A-shares and N-shares are significantly different from each other. Thus, N-share IPOs underpricing suffers more than A-share IPOs underpricing from asymmetric information with regarded to the capital raised from IPOs.
Model 2c: 𝑅𝑒𝑡$ = 𝜋)+ 𝜋'𝐴𝑔𝑒$+ 𝜋5𝑙𝑛𝑉𝑎𝑙𝑢𝑒$+ 𝜋=𝑃𝐸$+ 𝜋B𝐶ℎ𝑖𝑛𝑎$∗ 𝑃𝐸$. +𝜇$ Table 6: OLS regression estimates of model 2c
Ret Coefficient P-value
Intercept 1.90128 0.000
Age 0.0001373 0.473
lnVlaue -0.0766206 0.000
PE -0.0014423 0.195
China*PE 0.0008027 0.526
The estimated results of Model 2c are shown in Table 6. Holding Age and lnValue constant, the estimated coefficients of PE and China*PE are -0.0039296 and 0.0042291 respectively. With respect to the significant level of 10%, the estimated coefficients of both explanatory variables are relatively small and not significantly different from 0. Therefore, there is no sufficient evidence to support hypothesis 2c. As presented in descriptive statistics, the range of PE ratio is large and its standard deviation is high. Thus, the insignificant result can be attributed to the measurement of the initial value of IPO, PE ratio.
Model 3: 𝑅𝑒𝑡$ = 𝛾)+ 𝛾'China$+ 𝛾5𝐴𝑔𝑒$+ 𝛾=𝑙𝑛𝑉𝑎𝑙𝑢𝑒$+ 𝛾B𝑃𝐸$+ 𝛾G𝑊$+ 𝛾Y𝐶ℎ𝑖𝑛𝑎$∗ 𝑊$+ 𝜉$
Table 7: OLS regression estimates of model 2c
Ret Coefficient P-value
Intercept 1.967533 0.000 Age -0.0000704 0.386 lnVlaue -0.0752103 0.001 PE -0.0008552 0.273 W -0.2476781 0.031 China*W 0.2401687 0.034
Holding Age, lnValue and PEconstant, it is shown in Table 7 that the coefficients of W and China*W are -0.247676781 and 0.2401687, respectively. The estimated coefficients imply that (1) the elapsed time between the IPOs announcement day and the IPOs issuing day is negatively correlated with IPOs abnormal return which is opposite to the prediction, and (2) for A-share IPOs, when the waiting period increases by 1 day the IPOs abnormal return will increase by 0.75004%, and for N-share IPOs, when the waiting period increases by 1 day the IPOs abnormal return will decrease by 24.76781% which is not complied with what is found by Su and Fleisher (1999).
Regarding the significant level of 10%, the estimated coefficients are significant. However, there is no sufficient evidence to support hypothesis 3. The results showed that the abnormal return of A-share IPO is positively correlated with the lagging period which is constant with the empirical results reported by Chan et al. (2004). The percentage increase of IPO abnormal returns is relatively small with 1-day increase in the lagging period. The result implies that the lagging period of A-share IPO is short enough so that investors do not require large underpricing in A-share IPO as compensations. In contrast to A-share IPO, the abnormal return of N-share IPO is negatively correlated with the time gap. One explanation can be that the lagging gap is relatively small in America and investors may require more time to process companies’ public information to make decision. It can be assumed that
investors need more compensation when they have to make decisions roughly in a short period.
5.4 Correlation between different financial environment and IPOs underpricing Model 4a: 𝑅𝑒𝑡$ = 𝜃)+ 𝜃'𝐴𝑔𝑒$+ 𝜃5𝑙𝑛𝑉𝑎𝑙𝑢𝑒$+ 𝜃=𝑃𝐸$+ 𝜃B𝑊$+ 𝜃G𝑙𝑛𝑆𝑆300𝐼𝑛𝑑𝑒𝑥$+ 𝜑$ Table 8: OLS regression estimates of model 4a
Ret Coefficient P-value
Intercept -0.59234 0.622 Age 0.00308 0.368 lnVlaue -0.05294 0.006 PE -0.00030 0.743 W 0.00619 0.278 lnSS300Index 0.23834 0.098
Table 8 presents the estimated results of model 4a. As showed in the table, holding other variables constant, the coefficient of the explanatory variable lnSS300Index, 0.0000928, is positive. The coefficient implies that the overall financial situation of the Chinese stock market is positively related with A-share IPOs underpricing. 1 percentage increase of the SS300 Index is accompanied by an increase of 23.834% in A-share IPOs.
Denoted at the 10% significant level, the estimated result is significant. Thus, there is sufficient evidence to support hypothesis 4. The influence of the overall financial environment on the Chinese stock market has significant influence on the A-share IPOs market. The result implies that after adopted the Approval system, Chinese IPOs market is more market-orientated. However, the coefficient of lnSS300Index is negative which is opposite to the prediction. A-share IPOs are more underpriced when Chinese stock market is in good condition and vice versa.
Model 4b: 𝑅𝑒𝑡$ = 𝜗)+ 𝜗'𝐴𝑔𝑒$+ 𝜗5𝑙𝑛𝑉𝑎𝑙𝑢𝑒$+ 𝜗=𝑃𝐸$+ 𝜗B𝑊$+ 𝜗G𝑙𝑛𝑆𝑃500𝐼𝑛𝑑𝑒𝑥$+ 𝜔$ Table 9: OLS regression estimates of model 4b
Ret Coefficient P-value
Intercept 16.19385 0.183
lnVlaue -0.11193 0.570
PE -0.002659 0.476
W -0.00872 0.987
lnSP500Index -1.86479 0.265
As presented in Table 9, the estimated coefficient of SP500Index, -0.0010971, is negative, holding other variables constant. At a significant level of 10%, the result is not significant. Therefore, there is no evidence to support that the overall financial situation of America has significant influence on N-share IPOs underpricing. One potential explanation can be that the American IPOs market is relatively mature and the stock market index is increasing steadily over the years so the stock market does not influence IPOs underpricing remarkably. The small sample size of N-share IPOs can also lead to insignificant result.
6. Conclusion and discussion
In this thesis, OLS regressions analysis is applied to investigate the abnormal return of A-share IPOs and N-share IPOs and the driving forces behind the different degree of underpricing patterns. As predicted, the magnitude of underpricing patterns of A-share IPOs and N-share IPOs are significantly different from each other. N-share IPOs are more underpriced than A-share IPOs. Several hypotheses on how N-share IPOs’ initial returns are different from A-share IPOs are examined. These hypotheses are constructed upon the following three potential explanatory factors of the different magnitude of underpricing patterns: asymmetric information, elapsed time between the IPOs announcement day and the IPO issuing day, and different financial environment in China and America. The summary of the empirical results and main findings are summarized as below:
Table 10 Summery of the empirical results
Variables Ret
Model 1 Model
2a
Model 2b
Model 2c Model 3 Model 4a Model 4b Intercept 0.64067 0.31857 1.47701 1.90128 1.96753 -0.59234 16.1938 5 China -0.30611 ** Age 0.03181 ** 0.00207 0.00014 -0.00007 0.00308 0.02598
lnValue -0.04695 * -0.07662 *** -0.07521 ** -0.05294 ** -0.1119 3 PE -0.00144 -0.00086 -0.00030 -0.0026 59 W -0.24768 ** 0.00619 -0.0087 2 China*Age -0.03017 ** China*lnV alue -0.01151 ** China*PE 0.0008 China*W 0.24017 ** lnSS300Ind ex 0.23834 ** lnSP500Ind ex -1.8647 9
Note: ***, **, * denotes significant level of 1%, 5% and 10%, respectively.
a. Asymmetric information partly explains the underpricing phenomenon. Evidence showed that N-share IPOs suffered more from asymmetric information than A-share IPOs.
b. The elapsed time between the announcement day and the issuing day of IPOs has significant influence on IPOs underpricing. The lagging period is positively correlated with A-share IPOs underpricing which is constant with the perdition made according to Su and Fleisher (1999) and Chan’s et al. (2004) study. On the contrary, the waiting period is negatively correlated with N-share IPOs underpricing opposite to the prediction. c. The overall Chinese financial environment has significant positive relationship with
A-share IPOs underpricing while there is no sufficient evidence to support that the overall American financial environment has significant influence on N-share IPOs underpricing. The driving forces behind the different degree of underpricing pattern of A-share IPOs and N-share IPOs remain to be ambiguous.
Certain limitations can be modified in further studies. The sample size of N-share IPOs is relatively small. Data of N-share IPOs only counts for 5.6% of the total sample. The small proportion of N-share IPOs in the data set obtained can be an explanation of insignificant result. A longer period can be investigated in future studies. Moreover, a testable variable, seasoned equity offerings (SEO), is not included in this study due to the short period.
Loughran et al. (1994), Su and Fleisher (1999) and Chowdhry and Sherman (1996) examined B-share IPOs as SEO of A-share IPOs. In another word, issuers offered B-share IPOs after their A-share IPOs. N-share IPOs can also be examined as SEO of A-share IPOs in future studies. The reason why N-share IPO is underpriced more than A-share IPO while B-share IPO is underpriced less than A-share IPO can be an interesting topic to investigate.
Reference
Allen, F., & Faulhaber, G. R. (1989). Signaling by underpricing in the IPO market. Journal of
financial Economics, 23(2), 303-323.
Chan, K., Wang, J., & Wei, K. J. (2004). Underpricing and long-term performance of IPOs in China. Journal of Corporate Finance, 10(3), 409-430.
Chen, G., Firth, M., & Kim, J. B. (2004). IPO underpricing in China’s new stock markets.
Journal of Multinational Financial Management, 14(3), 283-302.
Cheng, S. & Li, Z., (2015). The Chinese Stock Market Volume I. London: Palgrave Macmillan.
UK.
Chiou, J. R., Li, M. Y. L., Cheng, L., & Chang, S. Y. (2010). Pricing and allocation mechanisms in underpricing of Chinese IPOs. Chinese economy, 43(1), 93-108. Chowdhry, B., & Sherman, A. (1996). International differences in oversubscription and
underpricing of IPOs. Journal of Corporate Finance, 2(4), 359-381.
Claessens, S., Klingebiel, D., Schmukler, S.L., (2006). Stock market development and internationalization: do economic fundamentals spur both similarly? Journal of
Empirical Finance, 13, 316–350.
Dewenter, K. L., & Malatesta, P. H. (1997). Public offerings of state-‐owned and privately-‐ owned enterprises: an international comparison. The Journal of Finance, 52(4), 1659-1679.
Güçbilmez, U. (2014). Why do some Chinese technology firms avoid ChiNext and go public in the US? International Review of Financial Analysis, 36, 179-194.
Guo, H., & 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(5), 984-997.
Hsieh, J., & Walkling, R. A. (2006). The history and performance of concept stocks. Journal of
Banking & Finance, 30(9), 2433-2469.
Lin, Z. J., & Tian, Z. (2012). Accounting conservatism and IPO underpricing: China evidence.
Journal of International Accounting, Auditing and Taxation, 21(2), 127–144.
Loughran, T., Jay R. Ritter & Kristian, R., 1994, Initial public offerings: International insights,
Luo, Y., Fang, F., & Esqueda, O. A. (2012). The overseas listing puzzle: Post-IPO performance of Chinese stocks and ADRs in the US market. Journal of multinational
financial management, 22(5), 193-211.
Mok, H. M., & Hui, Y. V. (1998). Underpricing and aftermarket performance of IPOs in Shanghai, China. Pacific-Basin Finance Journal, 6(5), 453-474.
Ritter, J. R., & Welch, I. (2002). A review of IPO activity, pricing, and allocations. The
Journal of Finance, 57(4), 1795-1828.
Rock, K. (1986). Why new issues are underpriced. Journal of financial economics, 15(1), 187-212.
Su, D., & Fleisher, B. M. (1999). An empirical investigation of underpricing in Chinese IPOs. Pacific-Basin Finance Journal, 7(2), 173-202.
Wu, C. (2014). Underpricing of homecoming A-share IPOs by Chinese firms already listed abroad. Review of Quantitative Finance and Accounting, 43(3), 627-649.
Zhang, C. X., & King, T. H. D. (2010). The decision to list abroad: The case of ADRs and foreign IPOs by Chinese companies. Journal of multinational financial management, 20(1), 71-92.
Zhou, Z. & Zhou, J., (2010). Chinese IPO activity, pricing, and market cycles. Review of
