Msc Thesis International Financial Management
University of Groningen, Faculty of Economics and Business
The Mystery of Underpricing: The Influence of a Country ’s Institutional Environment and Legal Framework on the Level of Underpricing in Western Europe
Abstract
This paper examines the underpricing phenomenon by addressing the ex-ante uncertainty of the value of the firm and information asymmetry involved with an initial public offering (IPO) process in Western Europe in the post-financial crisis era spanning from 2008 to 2015. The research is executed through investor protection, the legal framework and institutional environmental in a Hierarchical Linear Model (HLM) in order to address the hierarchical structure of the data. A significant negative impact of multiple institutional environment indicators on the level of underpricing is found on country-level alongside firm-level significant negative relationships of venture capitalist backed firms and firms performing an equity carve- out on underpricing. The results depict that firm management should take country-level governance into consideration in order to limit ex-ante uncertainty and reduce information asymmetry, leading to a reduced level of underpricing and reduction in cost of capital.
Name: Robert Polet
Student number: s2003600 1
stsupervisor: S. Ursu 2
ndsupervisor: P. Smid
Field Key Words: Initial public offerings, underpricing, information asymmetry, ex-ante
uncertainty, institutional environment, legal framework, investor protection
The Mystery of Underpricing: The Influence of a Country’s Institutional Environment and Legal Framework on the Level of Underpricing in Western Europe
TABLE OF CONTENTS
1. INTRODUCTION 3
2. LITERATURE REVIEW 5
2.1 Underpricing 5
2.2 Asymmetric information 6
2.3 The institutional environment and underpricing 7
2.3.1 Governmental legislation and enforcement 8
2.3.2 Legal framework, investor protection and underpricing 9
2.4 Firm specific variables and underpricing 10
2.6 Hypotheses development 12
3. DATA & METHODOLOGY 14
3.1 Data 14
3.2 Variables 14
3.2.1 Measuring Underpricing 15
3.2.2 Minority Investor Protection 15
3.2.3 Worldwide Governance Indicators 15
3.2.4 Firm variables 17
3.2 Methodology 19
3.2.1 Hierarchical Linear Modelling (HLM) 19
4. EMPIRICAL RESULTS 20
4.1 Descriptive statistics 20
4.2 HLM null model 25
4.3 HLM Full Model 26
4.4 Tests for Robustness 29
5. DISCUSSION 31
6. CONCLUSIONS 34
7. REFERENCES 37
The Mystery of Underpricing: The Influence of a Country’s Institutional Environment and Legal Framework on the Level of Underpricing in Western Europe
1. INTRODUCTION
Initial public offerings (IPOs) have been a subject of extensive research over the past decades.
This seems logical because IPOs have a great impact on the firm making the transition from private to public ownership. Going public generates capital for a firm that can be used to pay current obligations or to finance future growth through e.g. capital intensive projects. Although firms generate a significant amount of capital through going public, firms tend to knowingly underprice their IPO relative to their actual value, resulting in firms missing out on capital at their first offering. Underpricing is commonly understood as the positive first day return of a stock from its offering price to its first day close and the potential first-day return for investors taking part in an IPO.
Many different theories have risen to find a reason behind this phenomenon of underpricing. Ljungqvist (2007) summarizes these researches into four categories namely, institutional reasons, ownership and control reasons, behavioral reasons and information asymmetry reasons. The focus in this paper will be on the information asymmetry reasons in order to further explain the level of underpricing apparent in Western Europe. Most studies have focused on firm-level reasons, finding significant contributors to the level of underpricing like, the firm ’s size, its age, the quality of the underwriter, whether the IPO market is ‘’hot’’ and so on and so forth. Little research, though, has been performed on country-level variables that might impact the level of underpricing. Therefore, in this research I will examine which governance indicators on country-level have an impact on the level of underpricing when addressing information asymmetry and ex-ante uncertainty of the value of the firm and its offering.
Using a dataset of 826 IPOs over a period from 2008 to 2015 in Western Europe I contribute to the current field of finance literature by including all the Worldwide Governance Indicators (WGI) in the research, giving management easy accessible measures to examine when considering an IPO in the Western European market. I include the Political Stability and
Absence of Violence/Terrorism and the Voice and Accountability variables, where prior research
has used the other variables of the WGI only partially and sporadically (Engelen & van Essen,
2010; Hopp & Dreher, 2013). I further distinguish myself from prior research by examining the
post-financial crisis era with which I aim to test whether prior found evidence on ex-ante
uncertainty and information asymmetry involved with it remains significant after such a
The Mystery of Underpricing: The Influence of a Country’s Institutional Environment and Legal Framework on the Level of Underpricing in Western Europe
distressing financial period. The purpose of the study is to guide managers seeking to reduce their cost of capital involved with an IPO while maintaining the investor satisfaction and ensuring the successfulness of the particular IPO through finding the perfect price level for the offer. The research question of this research is therefore the following: Do the institutional environment and the legal framework of a country have an impact on the level of underpricing in Western Europe?
I operationalize this question by using Hierarchical Linear Modelling, which enables me to examine firm-level factors that influence the level of underpricing and while on a second level I can examine whether country-level factors further explain the variance in underpricing. This method has not been frequently used in prior research concerning IPOs but should enable a more precise explanation of the variance in the level of underpricing among countries. Also, this research identifies three different institutional environment levels priory identified by Kaufmann, Kraay and Mastruzzi (2011), which are examined separately in order to identify the focus points for management regarding the institutional environment.
Through applying the winner ’s curse model as introduced by Beatty & Ritter (1986) I aim to explain underpricing by ex-ante uncertainty about the value of the firm and information asymmetry. I find significant results concerning the country-level factors concerning the legal framework, the enforcement of it and the quality of the legal system as a whole based on the helping hand approach by (Barth, Caprio Jr., & Levine, 2004).
This study is build up as follows. In the next section, the literature review will underpin
the theoretical background of the aforementioned aspects, after which the hypotheses are
developed. The third section introduces the methodology of the used HLM analysis, description
of the explanatory variables, the data description and the descriptive statistics. Then, the fourth
section will elaborate on the data description and the descriptive statistics. Fifth, the empirical
results are presented and interpreted. In the sixth and seventh section I will and discuss the
results and conclude the study.
The Mystery of Underpricing: The Influence of a Country’s Institutional Environment and Legal Framework on the Level of Underpricing in Western Europe
2. LITERATURE REVIEW
2.1 Underpricing
The research on underpricing has been initiated by Logue (1973) and Ibbotson (1975). These studies encountered significant positive first day returns in the U.S. IPO market over their
respective research years. After these initial researches a multitude of researchers have examined this phenomenon and found positive initial returns, and therefore underpricing, present in
practically every country around the world irrespective of time period.
Henceforth, underpricing is established as an economically significant and, most of all,
permanent phenomenon in the finance literature. Underpricing is also found to vary significantly over time, where higher underpricing occurs in periods of so-called hot IPO markets. An
example for this is the research executed by Loughran & Ritter (2002) whom research the IPO market in the U.S. in 1999 and 2000. In this period, new issues could see an average rise of its stock in the first trading day of 65%. They relate this irregularly high underpricing to the hot IPO market in that time period, where especially internet-based firms were deemed attractive due to the high potential growth opportunities of these firms. This increased percentage of underpricing led to a significant and somewhat disturbing amount of money left on the table by the issuing firms.
Underpricing seems to be a phenomenon that cannot be attributed to one simple factor of misevaluation or a predetermined risk premium for an IPO. Ritter and Welch (2002) are among the researchers that underpin this finding. They state that the substantial risk premium required by first-day investors is not rational, especially because first-day investors are deemed to be diversified and would not bear such an increased risk while the second-day investors, the
investors buying the shares from the first-day investors, would not bear this risk. Therefore, they assume that this substantially high risk premium spans over a longer time then one day, implying that multiple other factors influence the underpricing of an IPO.
Ljungqvist (2005) has summarized underpricing theories into several categories, which clarify
the findings found in prior research. The first classification depicted is based on the strategy used
by IPO-share buyers. These buyers are classified in rational and irrational buyers, which are
further subdivided into strategies based on symmetric and asymmetric information. Then, the
The Mystery of Underpricing: The Influence of a Country’s Institutional Environment and Legal Framework on the Level of Underpricing in Western Europe
former can be subdivided into institutional aspects and based on ownership and control
considerations, while the latter is further categorized in who holds the superior information in the IPO process (issuer, underwriter or investor).
This research focuses on the asymmetric information aspect depicted by Ljungqvist (2005) between the issuer and investor. This asymmetry leads to ex-ante uncertainty about the value of the firm and its IPO, which leads to a value assessing problem for the investors in the IPO. One of the main reasons this problem occurs is that there is no share value history of the firm pre- IPO, which makes valuation significantly more complicated than with an already established public firm. The following section will discuss the asymmetric information theory in depth in order to establish a common understanding of this phenomenon.
2.2 Asymmetric information
Multiple researchers have tried to further explain why underpricing occurs and which factors have an influence on underpricing. One strand of research is the research concerning the involvement of asymmetric information. (Rock, 1986) builds a theory surrounding information asymmetry named the winner ’s curse model. This theory depicts that in the IPO process a group of investors exist whom have attained superior information of the particular firm next to possible extra information on all the other involved investors. This beneficial position is crucial in the IPO process in the following way: When new shares would be offered at their expected value, the informed investors would crowd the market and buy all the shares when an issue is
qualitatively just, leaving no room for the uninformed investor to invest in the issue. On the other
hand, they would withdraw entirely when less attractive issues are offered at their expected
value, ending up with too little interest in the issue. Therefore, it makes sense for an offering firm
to put a discount on their shares, this in order to guarantee that less-informed investors invest in
the particular issue. Basically, Rock (1986) theorizes that uninformed investors buy shares of
every new IPO while informed investors only buy shares of attractive IPOs. Following this line
of reasoning, uninformed investors receive the entire supply of mediocre IPOs while they have
limited access to attractive IPOs due to the fact that when an IPO occurs, in general, the number
of shares offered are limited so, with an attractive IPO informed and uninformed investors
subscribe, often creating significant oversubscription. Henceforth, the expected return of an
uninformed investor would be below the average underpricing or even negative would a firm
The Mystery of Underpricing: The Influence of a Country’s Institutional Environment and Legal Framework on the Level of Underpricing in Western Europe
offer its shares at expected value (Ritter & Welch, 2002). Furthermore, Rock (1986) argues that the market of IPOs needs uninformed investors in order to exist, as the demand from informed investors is not sufficient for the existence of it. So, underpricing is needed for uninformed investors to gain positive returns from an IPO because without the expectation of positive returns they will not invest (Ljungqvist, 2005).
Building on the winner ’s curse model, Beatty & Ritter (1986) show that the degree of underpricing increases when ex-ante uncertainty about the value of the firm increases. Firms that, example given, have higher uncertainty about growth opportunities or autonomous managerial decisions post-IPO tend to have a higher degree of underpricing than others. After these seminal papers, it has now been widely agreed upon that prior-to-IPO uncertainty about the value of the firm is one of the most influential factors in the IPO process, with higher uncertainty leading to higher underpricing (Ljungqvist, 2005). Also, Beatty and Ritter (1986) prove that underpricing is a vehicle for compensation of costs induced by investors to become informed.
With an increase in the uncertainty of the true value of the firm, the costs for gaining the needed information tend to increase significantly too, therefore, underpricing should be higher in such a situation.
Alternatively, Principal/agent IPO models foster similar conclusions while focusing on information asymmetry between the issuer and its underwriter. Higher uncertainty about the value of the firm through, example given, uncertainty about expropriation risk leads to more asymmetric information between the issuer and its underwriters inducing the necessity to underprice further (Arthurs, Hoskisson, Busenitz, & Johnson, 2008; Baron, 1982).
2.3 The institutional environment and underpricing
Extensive research has been done on the firm- and issue-specific characteristics that can proxy for ex-ante uncertainty and information asymmetry relating to underpricing and firm value, but these phenomena are observable on multiple measuring levels and therefore country-specific characteristics should be accounted for too to further assess the variation in underpricing. This field of research is only limited hence, I deem it important to further examine this relationship (Banerjee, Dai, & Shrestha, 2011; Engelen & van Essen, 2010; Hopp & Dreher, 2013).
Therefore, this section will focus on the country-level institutional environment and legal
framework impacting IPO underpricing and firm valuation.
The Mystery of Underpricing: The Influence of a Country’s Institutional Environment and Legal Framework on the Level of Underpricing in Western Europe
Institutional theory depicts that an institutional environment concerns formal as well as informal institutions namely, regulative, normative and cultural-cognitive elements (DiMaggio, 1988). These elements are the rules of the game in society formed by the humans active in this society. Through the level of constraining the acceptable behaviors of society, institutions foster an environment of economic prosperity (Peng, 2013). Basically, institutions are created by society to decrease the uncertainty involved by a society without rules and regulations in order to establish a stable environment where the participants of the society can conveniently interact with one another. Formal institutions have proven to be of importance for corporate finance literature in general as well as for the underpricing phenomenon itself (Engelen & van Essen, 2010; Hopp & Dreher, 2013;La Porta, Lopez-de-Silanes, Shleifer, & Vishny, 1997) and are often divided in several segments. The segments used in this paper are discussed below.
2.3.1 Governmental legislation and enforcement
In their paper, Barth et al. (2004) research which banking regulation and supervision mechanism as established by the government is most effective for creation of a stable environment. Here, they identify two views on the role of the government in regulating the banking system, namely the helping-hand approach and the grabbing-hand approach. First of all, the helping-hand approach suggests that governments build regulations in order to insure and correct for market failures like, example given, externalities, monopoly power and information asymmetries.
Regulations which are exemplary for the helping-hand approach are, for example, deposit insurance schemes or official bank supervision frameworks. These regulations are essential in the case of underpricing because the investment banks (underwriters) involved with the IPOs have an incentive to underprice the IPO more than desired by the issuing firm itself. This is because investment banks tend to focus on winning and maintaining future business. An example for an IPO mechanism in this case is the offering technique of book building. With book
building, investment banks assess an indication of interest among investors in an IPO, which can
be seen through the non-binding offers at different prices given by the respective investors. Here,
the investment banks can decide to allocate more shares to the investors interested at a lower
price, guaranteeing the investors higher first day returns and therefore securing future business
for the investment banks themselves. So, in this approach, regulations would help establish a fair
offer price and therefore a reduced level of underpricing by reducing the information asymmetry
The Mystery of Underpricing: The Influence of a Country’s Institutional Environment and Legal Framework on the Level of Underpricing in Western Europe
between the investor and the issuer through enforcing regulations on underwriters. These
regulations also reduce the ex-ante uncertainty for a particular issue, which in turn should lead to a lower level of underpricing.
On the other hand, the grabbing-hand approach depicts that government failure is a greater threat than market failure and that government legislation is used to support political constituencies. This approach argues that a country ’s strict government legislation correlates with higher levels of corruption which in turn would lead to more ex-ante uncertainty and information asymmetries, leading to a negative impact on underpricing.
Also, in their paper, Hopp and Dreher (2013) argue that an increase in the quality of the institutional environment induces a decrease in the perceived risk of investing, which
simultaneously reduces asymmetric information issues leading to a reduction in the level of underpricing. They further argue that when a government forces a firm to disclose more precise information about the value of the firm, information asymmetry is reduced, reducing uncertainty about the value of the firm and inducing a separation between good and bad governance of a firm, leading to a reduction of ex-ante uncertainty. The results attained by Hopp and Dreher (2013) confirm their argumentation about information asymmetry, ex-ante uncertainty and the level of underpricing. This argumentation implicitly confirms the helping hand approach as depicted by Barth et. al. (2004).
In this research, I follow the helping-hand approach, hereby stating that government regulation and enforcement can effectively lower the level of information asymmetry and ex-ante uncertainty leading to lower levels of underpricing.
2.3.2 Legal framework, investor protection and underpricing
In their study on country-level institutional characteristics impacting the underpricing of an IPO, Engelen and Van Essen (2010) found that the quality of the country ’s legal framework, measured by the level of investor protection, the level of legal enforcement and the quality of the legal system as a whole, has a significant impact on the level of underpricing. They argue that the ex- ante uncertainty of a firm ’s value is influenced by the country's investor protection in two ways namely, on the one hand, weaker shareholder protection induces a higher uncertainty of
management decisions that might have a negative impact on the firm ’s value. On the other hand,
a conflict of interest between controlling shareholders and outside shareholders is more likely to
The Mystery of Underpricing: The Influence of a Country’s Institutional Environment and Legal Framework on the Level of Underpricing in Western Europe
occur with weaker shareholder protection. This increases the chance of expropriation by controlling shareholders and makes outside investors more reluctant to invest as the risk of investment is higher. So, both with the risk of ex-post management decisions as well as the expropriation risk, ex-ante uncertainty about the actual value of the firm increases. Therefore, the management of a firm filing for an IPO has to further underprice their offer in countries with weaker shareholder protection in order to attract outside investors to subscribe for the particular IPO. Also, they find that the quality of the legal background institutions reduces the ex-ante uncertainty and therefore, a lower level of underpricing.
2.4 Firm specific variables and underpricing
This section summarizes the theories behind the firm- specific risk factors that are reported in prior empirical IPO literature and used in this research to explain the varying levels of
underpricing. Since ex-ante uncertainty and information asymmetry itself are not observable variables, I will use several proxies in order to be able to explain these phenomena. These proxies are generally available pre-IPO and are therefore proper indicators for ex-ante uncertainty, they can be grouped into multiple categories namely, firm characteristics, issue characteristics, prospectus disclosure and certification (Jenkinson & Ljungqvist, 2001). In this research, I apply multiple firm- and issue characteristic proxies. Next to the firm-level proxies, the year of IPO is included in the analysis. The impact on the level of underpricing is further elaborated on below concerning the aforementioned proxies.
First of all, whether a venture capitalist is involved in a firm pre-IPO should have a significant impact on the level of underpricing, impacting ex-ante uncertainty on the value of the firm. This is due to the fact that venture capitalist tends to bring in both knowledge and
particular resources into the firm they get involved with. Typically, they also perform an
extensive due diligence on the firm prior to investing and after the investment decision they tend to occupy a seat in the board of directors as well as assist management in daily decision making.
In general, researchers expect a negative relationship between early investment of venture
capitalists and the level of underpricing. Although this expectation seems logical as the reduction in ex-ante uncertainty would lead to a lower level of underpricing, the results have been mixed in recent research (Dolvin & Jordan, 2008; Engelen & van Essen, 2010; Guo, Lev, & Shi, 2006).
Second, prior research has indicated that being a high-tech firm is another proxy for the
The Mystery of Underpricing: The Influence of a Country’s Institutional Environment and Legal Framework on the Level of Underpricing in Western Europe
level of ex-ante uncertainty in IPO firms. This is due to the fact that these firms are often viewed by investors as high growth potential firms because the products, being high tech, have a
significant future development potential for the firm. While on the other side, these firms have a relatively high risk because they often report either negative or low earnings before their IPO, which should indicate some form of information asymmetry on valuing the IPO. Also,
technology firms are, on average, more likely to be subject to agency problems (Amit, Brander,
& Zott, 1998). The higher agency risks for technology firms compared to other industries is due to the fact that information asymmetry is prevalent in these firms as the founders often have highly specialized knowledge, often not managerial, and prospects of future market in emerging industries are uncertain while they internationalize early (Ozdemir & Upneja, 2016). Loughran &
Ritter (2004) find that high-tech and internet IPOs in the late nineties, so during the internet bubble, attained a significantly higher initial return compared to other sectors, mainly due to the riskier nature of the firms involved. They argue that this is due to the fact that the products of these firms are difficult to value, which makes the valuation of the firm difficult. Therefore, in this research I expect to find a positive relationship between high-tech firms and IPO
underpricing.
Third, this research will examine the behavior of equity carve-out issues with respect to the level of underpricing. An equity carve-out is not an IPO in the traditional sense, instead a portion of a wholly owned subsidiary is offered to the market with the initial owner maintaining substantial ownership. This technique is most often used by owners to divest in their respective subsidiary. A multitude of explanations for using this technique have been discussed in finance and economic literature. One line of reasoning is that through using this technique, managers can focus on their core business lines, leading to a performance improvement (John & Ofek, 1995).
Also, (Gilson, Healy, Noe, & Palepu, 1998) argue that carve-outs like spin-offs or targeted stock offerings are used to improve the information disclosure towards investors. This reduction in information asymmetry will induce a lower level of underpricing of the respective issue.
Fourth, prior studies have empirically shown that the particular year of an IPO has an impact on the level of underpricing due to societal and economic pressures. Therefore, a dummy variable is included for the year of a firm ’s introduction. Research has been done on the so-called
‘’hot’’ and ‘’cold’’ IPO markets, which are deemed to be an explanation for the yearly
fluctuations in the level of underpricing (Engelen & van Essen, 2010; Ibbotson, Sindelar, &
The Mystery of Underpricing: The Influence of a Country’s Institutional Environment and Legal Framework on the Level of Underpricing in Western Europe
Ritter, 1988). A hot market is featured by an increasing number of firms going public and an increase in the average level of underpricing. After a hot market, the number of firms going public stays consistent while the average level of underpricing decreases again. As a reaction to this, less firms tend to go public and significantly lower levels of underpricing are found or even occasionally overpricing. An example for this pattern, as depicted before, is the period
surrounding the internet bubble, where a hot IPO market in the late nineties was preceded by a cold IPO market (Loughran & Ritter, 2004).
2.6 Hypotheses development
In this section I will briefly describe the hypotheses that are build up upon the literature review in the prior section for the country-level as well as for the firm-level variables that are examined in this research in order to answer our main research question: Do the institutional environment and legal framework of a country have an impact on the level of underpricing in Western Europe? Also, this research will distinguish between three different institutional environment levels, which are examined separately in order to identify the focus points for management regarding the institutional environment.
First, based on the helping-hand approach discussed by Barth et. al. (2004) I hypothesize that the level of government regulation and the enforcement of it have a negative relationship with the level of underpricing meaning that a higher level of government regulation induces a lower level of underpricing in that respective country because it reduces information asymmetry and especially ex-ante uncertainty about the value of the respective firm. The related hypothesis is constituted as follows:
Hypothesis 1: IPOs in countries with a higher (lower) governance level have a lower (higher) level of underpricing
Next, prior theory depicts that a country ’s level of investor protection has a negative relationship
with the level of underpricing due to the degree of expropriation risk and the risk of negative
managerial decisions ex-post the IPO. When these risks are high, the ex-ante uncertainty of the
value of the shares on offering increases significantly which will force the management of the
firm to further underprice the IPO. Also, when the investor protection level is lower, information
The Mystery of Underpricing: The Influence of a Country’s Institutional Environment and Legal Framework on the Level of Underpricing in Western Europe
asymmetry is one of the drivers of the aforementioned risks. For this argumentation, the hypothesis is the following:
Hypothesis 2: IPOs in countries with a higher (lower) level of minority investor protection have a lower (higher) level of underpricing
Then, for the firm specific variables, based on the theory discussed in the prior section multiple hypotheses have also been formulated. Prior research has shown, although not unanimously, that venture capitalist involvement pre-IPO is positive for the management of the firm as, example given, more expertise drives a reduction in a reduction in information asymmetry and uncertainty about the value of the firm, leading to a lower level of underpricing. Therefore, the hypothesis is formulated as:
Hypothesis 3: Venture capital ownership negatively relates to the level of underpricing
The fourth hypothesis is based on the growth of ex-ante uncertainty when a firm is operating in the high-tech sector. This growth in uncertainty about the value of the shares offered is due to multiple factors, one of them being the fact that high-tech firms often report negative or exceptionally low earnings pre-IPO. The hypothesis that conceptualizes the mentioned argumentation is:
Hypothesis 4: High-tech firms have a positive relationship to the level of underpricing
Finally, equity carve-out theory is focused on the fact that information asymmetry is reduced with an equity carve-out or related technique, leading to a lower level of underpricing. Therefore, the related hypothesis is the following:
Hypothesis 5: An equity carve-out has a negative relationship with the level of underpricing
The Mystery of Underpricing: The Influence of a Country’s Institutional Environment and Legal Framework on the Level of Underpricing in Western Europe
3. DATA & METHODOLOGY
3.1 Data
For this research, the SDC Platinum database from Thomson Reuters, accessed through
Thomson One, is used for the retrieval of IPO data. The sample period ranges from 01-01-2008 to 31-12-2015 and is based on first days of trading of the relevant issue. The initial sample of Western European IPOs in this time span consist of 1,423 IPOs. Then, firms in the financial sector (SIC 6000-6999) are excluded, erasing 436 firms. Following, firms with a missing offer price or closing price on first day of trade up to one week of trading are excluded (157 firms). In the sample, 62 firms did not have a first day closing price, in this case I replaced it with a closing price within one week of the first trading day. Lastly, following Banerjee, Dai, & Shrestha (2011), IPOs with abnormal high and low values, i.e. higher than 2000% and lower than -70%
are deleted from the sample (4 firms). The country-specific variables regarding the minority investor protection and the institutional environment are derived from the Worldbank and are based on research of Djankov, La Porta, Lopez-de-Silanes, & Shleifer (2008) and Kaufmann et al. (2011) respectively.
As depicted below, three firm specific variables are used as well as a dummy variable for IPO years. Initially, I have derived data for multiple other variables to control for like, firm age, firm size, profitability (return on assets and price earnings ratio) and offering technique, whom were all omitted for the final research. In the case of the first four variables, to little data could be retrieved from ThomsonOne or Datastream. For example, firm age was only retrievable for 426 firms of the total final sample of 826. In the case of the offering technique variable, the finance literature has shown that only book building significantly influences the variance in underpricing as depicted in the literature review. Next to that, my final sample consists of only four firms that makes use of book building in their IPO process and therefore this was not controlled for in the final research either.
3.2 Variables
In this section I will explicate the operationalization of the main variables namely, the dependent
variable and the firm- and country-level explanatory variables involved in this research.
The Mystery of Underpricing: The Influence of a Country’s Institutional Environment and Legal Framework on the Level of Underpricing in Western Europe
3.2.1 Measuring Underpricing
Underpricing can be measured in several forms of methodology, example given the standard underpricing with correction for market movements or the inclusion of the market beta. In this research, I will make use of the standard methodology of measuring underpricing as prior research has shown that the initial return of an IPO is, on average, significantly greater than the market movement as a whole of a single day, making the market movement only marginally impacting and therefore unneglectable. Therefore, underpricing in this research is measured by calculating the percentage change between the offer price and the closing price of the first day secondary market trading. As depicted before, when the first day closing price is missing, the first week closing price is used. This operationalization leads me to the following equation:
Underpricing = ((closing price − offer price)/offer price) * 100%. (1)
3.2.2 Minority Investor Protection
As first explanatory variable, I make use of the investor protection index (IPI), in particular the strength of the minority investor protection, in order to measure the level of investor protection active in the respective country. The index is built up upon averages of the extent of conflict of interest regulation and the extent of shareholder governance index (World Bank, 2016). The former is comprised of an extent of disclosure-, extent of director liability-, and ease of shareholder suits index. The latter encompasses the extent of shareholder rights-, extent of ownership and control- and extent of corporate transparency index. All indices have a score ranging from 0-10, the simple average of these indices are taken, resulting in the minority investor protection index. A negative relationship between the level of investor protection and the level of underpricing is expected.
3.2.3 Worldwide Governance Indicators
To further investigate the relationship between the quality of the institutional environment and
the underpricing phenomenon this research applies the Worldwide Governance Indicators (WGI)
as created by Kaufmann, Kraay, & Mastruzzi (2011). The indicators are derived from the World
Bank and are used in this research to further deepen prior research on this relationship. As
depicted above, prior papers have focused their research on certain parts of the institutional
The Mystery of Underpricing: The Influence of a Country’s Institutional Environment and Legal Framework on the Level of Underpricing in Western Europe
environment like legal, regulation and accounting indicators (Hopp & Dreher, 2013) or the quality of the legal system and judicial efficiency (Engelen & van Essen, 2010). Using all WGI variables, I distinguish myself from prior research by including measures for political stability and voice and accountability, which may further explain the quality of an institutional
environment in a country. Also, I will discuss each significant variable separately and identify theoretical- and managerial implications for the found results in order to deepen the
understanding of indicators for the level of underpricing.
The WGI is comprised of six indicators that keep track of a multitude of variables. All WGI are a perception of the relative indicator, they do not give actual objective values of any variables or indicators. All indicators mentioned below are held constant over the sample period with the 2015 score as variable value. The WGI scores range from -2.5 to 2.5 and are defined by Kaufman, Kraay and Mastruzzi (2011) as follows, divided in three subgroups:
The respect of citizens and the state for the institutions that govern economic and social interactions among them:
First, the Control of Corruption Index (CC) is defined as capturing the perceptions of the extent to which public power is exercised for private gain in smaller quantities as well as with significant corruption schemes. Also, it indicates the degree to which the state is captured by elites and private interests. A negative relationship is expected between the level of underpricing and the degree of control of corruption (Engelen & van Essen, 2010).
Next, the Rule of Law (RL) indicator captures the perception of agents in the country. It measures whether they have confidence in and abide by the rules like e.g. contract enforcement, police, property rights and courts. Also, it measures the confidence in the likelihood of crime and violence. Engelen and Van Essen (2010) argue and find that a higher quality legal framework in the form of abidance by and respect for the governing body and its regulations as depicted here reduces the ex-ante uncertainty, leading to a lower level of underpricing. The same is expected in this research.
The capacity of the government to effectively formulate and implement sound policies:
Third, Government Effectiveness (GE) captures perceptions of the quality of public
services alongside the quality of the civil service and whether these variables are independent
The Mystery of Underpricing: The Influence of a Country’s Institutional Environment and Legal Framework on the Level of Underpricing in Western Europe
from political pressures and the credibility of the government ’s commitment to such policies.
Expected is that a more effective government corrects market failure such as information asymmetry, which should negatively affect the level of underpricing.
Then, Regulatory Quality (RQ) captures the perceptions of the extent to which the ability of the government suffices in formulating and implementing sound policies and regulations promoting and permitting the development of the private sector fostering economic prosperity. It is expected that a higher regulatory quality leads to a reduction in information asymmetry and in turn a reduction in the level of underpricing.
The process by which governments are selected, monitored and replaced:
Fifth, Political Stability and Absence of Violence/Terrorism (PS) is used as measure to indicate the likelihood that the government will be destabilized through violent or unlawful means. Engelen & Essen (2007) use PS as proxy for the effectiveness and quality of legal enforcement and they find a negative relationship with underpricing. So, this research also expects a negative relationship between PS and underpricing.
Lastly, Voice and Accountability (VA) measures whether citizens in a country are part of the selection process of selecting their government next to whether they have freedom of speech and association and whether the media has these privileges too. Buckley, Raboy, Mendel, Duer, Price, & Siochr (2008) argue that media whom is independent and is free of voice significantly contributes to proper governance and accountability through delivering quality information to the public, especially investigative journalism is vital for a better governance. Therefore, I expect a negative relationship between Voice and Accountability and the level of underpricing.
3.2.4 Firm variables
As depicted in the theory section, firm- and issue-specific variables contribute to the model as level-1 predictors for the level of underpricing followed by the priory discussed country-level variables as level-2 predictors. The operationalization of the firm-level variables is the following:
Venture capital dummy
Following multiple studies on the involvement of venture capitalists in firms before IPO a
dummy variable is created with the value 1 if a firm filing for an IPO has venture capitalists
The Mystery of Underpricing: The Influence of a Country’s Institutional Environment and Legal Framework on the Level of Underpricing in Western Europe
whom attained a share in the company and 0 if not (Engelen & van Essen, 2010; Suchard, 2009).
As hypothesized above, I expect a negative relationship between venture capitalist backing and the level of underpricing.
High-tech dummy
Loughran and Ritter (2004), found significantly higher levels of underpricing for high-tech firms filing for an IPO. Like Roosenboom and Schramade (2006), I make use of a dummy variable in order to proxy for high-tech firms. The dummy will be valued 1 when a high-tech firm
undertakes an IPO and 0 when the firm participates in any other industry. A firm is deemed a high-tech firm when its primary SIC code is one of the following as used by Roosenboom and Schramaade (2006): 283, 357, 365, 366, 367, 376, 382, 384, 48, 737, and 8731. As argued in the literature review, I expect a positive relationship between the high-tech dummy and the level of underpricing.
Equity Carve Out dummy
In their research, Prezas, Tarimcilar, & Vasudevan (2000) find a significant relationship between the level of underpricing of an IPO and whether that firm performs an equity carve-out. They match equity carve-out IPOs with original IPOs by size and equity book-to-market ratio. The reason for this found relationship can be due to the fact that the original company has proven itself on the stock exchange due to its size, results and stock history, reducing the ex-ante uncertainty. For that reason, I expect a negative relationship between equity carve-out and an IPO. I operationalize this through making a dummy variable with the value 1 for an equity carve- out and a 0 for an original IPO.
IPO Year
Following the methodology of Engelen & van Essen (2010), this paper makes use of a dummy
variable for each particular IPO year. When an IPO is being undertaken in a certain year, it will
receive a value of 1 and it will receive a value of 0 when it did not take place in that particular
year. These variables will be created for 2008 until 2015.
The Mystery of Underpricing: The Influence of a Country’s Institutional Environment and Legal Framework on the Level of Underpricing in Western Europe
3.2 Methodology
3.2.1 Hierarchical Linear Modelling (HLM)
The data is analyzed through the use of Hierarchical Linear Modeling (HLM) (Raudenbush &
Bryk, 2002). HLM is used because the data in this research is structured hierarchically meaning that multiple variables are embedded in different levels of data. In this research, IPOs are nested within countries and may be more alike within these countries than between the countries. This is due to country-level factors that influence the IPO, for instance whether a government control for corruption or the quality of the law and the effectiveness of the government is high. This makes the IPOs in one country significantly different from IPOs in another country. Statistically, the basic assumption is that there is independence of observations, but with hierarchical data this assumption is frequently violated, leading an OLS regression obtaining too small standard errors to interpret the parameters estimates (Hox, Moerbeek, & van de Schoot, 2010). Alternatively, HLM does recognize these discrepancies and correct for it in the data. Also, at the first level, within the countries, the slopes and intercepts are allowed to differ between the respective countries, which should enhance the accuracy and reliability of the research as this is not possible with an OLS regression.
Earlier IPO studies have most commonly made use of OLS regressions (Banerjee, Dai, &
Shrestha, 2011; Hopp & Dreher, 2013), a scarce group of researchers have been focused on using HLM (Engelen & van Essen, 2010) in order to explain the institutional phenomena at country- level surrounding the variance in underpricing for the same reasons as mentioned earlier in this section. Like Engelen & van Essen (2010), I expect that the results of the HLM model will create a more reliable and accurate outcome than with an OLS regression because of the actual
hierarchical structure of the data, which OLS cannot handle properly.
The data in this research has a hierarchical structure which is two-leveled, the level 1 model predictors are variables on the firm-level concerning e.g. the size of the firm or the size of the offer. On the second level the predictors are the country specific variables and are example given the minority shareholder protection.
Following, the equation for the level 1 model (2) and the level 2 model (3 and 4) are described:
The Mystery of Underpricing: The Influence of a Country’s Institutional Environment and Legal Framework on the Level of Underpricing in Western Europe
= + + + 𝑒 (2)
= + + 𝜇 (3)
= + + 𝜇 (4)
where Yij is level of underpricing, X are firm-specific variables and Z are the country-specific variables.
HLM gives me the opportunity to run an One-Way ANOVA with random effects in order to examine how much of the variance of underpricing is explained by the difference among countries. The results from this test will clarify whether the use of a multi-level model is justified instead of a regular OLS regression. The analyses are executed in STATA 14.
4. EMPIRICAL RESULTS
In this section I present and discuss the descriptive statistics through a univariate analysis
followed by the HLM analysis. Before testing the hypotheses with the HLM analysis, I execute a one-way ANOVA model (HLM null model) with random effects in order to see whether there is a statistically significant difference in the variance of independent groups. Next to that, this test is done to calculate the overall mean underpricing adjusted for weight per country and to
determine the intra-class correlation and examine whether it is justified to use HLM in this case.
After justification, I execute the full HLM model to test the relevant hypotheses where the robust standard errors are applied.
4.1 Descriptive statistics
In this section I will discuss the descriptive statistics of the established sample. I will create an
overview of notable values found in the sample by performing a univariate analysis on the
different aspects of the sample depicted below.
The Mystery of Underpricing: The Influence of a Country’s Institutional Environment and Legal Framework on the Level of Underpricing in Western Europe
Table 1 shows us the descriptive statistics of the variables used in this research. The mean underpricing is 9.71% while the median is 3.64% indicating a limited amount of extreme values in underpricing which increase the average underpricing also visible in the maximum value of underpricing (600%). Furthermore, 14% of the sample is backed by a venture capitalist pre-IPO and 36% of the sample is a technology based firm. Also, 18% of the sample is a form of an equity carve-out. Concerning these three variables, the percentages show that there should be sufficient observations per variable in order to infer the influence of these characteristics on the level of underpricing.
Table 1. Descriptive statistics (N=826)
Variable Mean Median Std. Dev. Min Max
Underpricing 9.71 3.64 34.31 -69.72 600.00
Venture Capital 0.14 0.00 0.35 0.00 1.00
Technology 0.36 1.00 0.48 0.00 1.00
Equity Carve-Out 0.18 0.00 0.38 0.00 1.00
IPI 7.03 7.20 0.76 5.00 7.80
CC 1.59 1.71 0.49 0.00 2.42
GE 1.54 1.48 0.35 0.40 2.12
RQ 1.50 1.71 0.30 0.67 1.82
RL 1.58 1.71 0.37 0.38 1.97
PS 0.60 0.48 0.31 -0.08 1.38
VA 1.34 1.33 0.16 0.81 1.66
Underpricing is the initial return with a positive value being underpricing and a negative value being overpricing. Relevant variables:
Underpricing, Venture capital ownership dummy, High-tech dummy (Technology), Equity carve-out dummy, Investor protection index (IPI), Control of corruption index (CC), Government effectiveness (GE), Regulatory quality (RQ), Rule of law (RL), Political stability (PS), Voice and accountability (VA). IPI is measured on a scale 0-10, the World Governance Indicators (WGI) -2.5–2.5.
The country-level variables are characterized by a relatively high average (IPI=7.03 on 0
to 10 scale and e.g. CC 1.59 on -2.5 to 2.5 scale) as expected in this Western Europe sample. The
Political Stability and Absence of Violence variable (PS) is unexpectedly low for the Western
Europe sample. When looking at table 2 we can infer that Greece (0.07), Spain (-0.08), Italy
(0.46), France (0.48) and the United Kingdom (0.41) have a low value, relative to the other
sample countries, for PS while on the other hand Austria (1.21), Finland (1.38), Norway (1.24)
and Switzerland (1.29) score relatively high on PS. This may be due to social pressures like the
refugee crisis, euro crisis and sovereign debt crisis in Italy, Spain and Greece, while in France
and the United Kingdom, the terrorist threat and actual attacks can be explanatory for a lower
score.
The Mystery of Underpricing: The Influence of a Country’s Institutional Environment and Legal Framework on the Level of Underpricing in Western Europe
Further examining table 2 shows that Greece and Italy score relatively low on all variables in this research, while Austria, Denmark, Finland, Netherlands, Norway, Switzerland and Sweden score relatively high. I can identify a pattern that Scandinavian countries score high in general, so based on the build-up theory, underpricing should be relatively low in these countries and high in, for instance, Greece and Italy. As Greece and Italy consistently score low on the variables Control of Corruption and Rule of Law, I can infer that the respect of citizens and the state for the institutions governing social and economic interactions is low in these countries, while this respect is highest in the Scandinavian countries (Sweden, Denmark, Finland, Norway). According to Engelen & van Essen (2010), countries scoring high on these indicators tend to have a reduced level of ex-ante uncertainty about the value of the firm.
Next to that, in Greece and Italy, the government seems the least capable to effectively implement sound policies while the Scandinavian countries score the highest on Government Effectiveness and Regulatory Quality. Higher scores on these variables indicate an increase in the availability and quality of information alongside the creation of a stable environment with effective legal enforcement (Buckley et al., 2008; Engelen & Essen, 2007).
Third, the process by which transparent democracy is shown, monitoring is relatively well structured and replacement of the government is relatively easy is best shown, again, in the Scandinavian countries, whereas in this case Spain, Greece and Italy are the worst scoring across the variables (political stability and voice and accountability).
Lastly, minority investor protection is best organized in the United Kingdom, Ireland and
Norway whereas the worst minority investor protection is found in Switzerland, The Netherlands
and Finland. This could be partially explained by the differences in legal systems and its origin
as these differ between the UK/Ireland (common law) and mainland Europe (code law).
The Mystery of Underpricing: The Influence of a Country’s Institutional Environment and Legal Framework on the Level of Underpricing in Western Europe
Table 2. Institutional environment, legal framework and investor protection indicators
Country IPI CC GE RQ RL PS VA
Austria 6.50 1.69 1.68 1.53 1.87 1.21 1.42
Belgium 5.80 1.46 1.58 1.29 1.36 0.79 1.37
Denmark 7.20 2.42 2.10 1.82 1.96 1.00 1.62
Finland 5.70 2.31 2.12 1.77 1.97 1.38 1.56
France 6.50 1.39 1.48 1.20 1.44 0.48 1.25
Germany 6.00 1.77 1.60 1.57 1.70 0.87 1.39
Greece 6.30 0.00 0.51 0.67 0.59 0.07 0.81
Ireland 7.30 1.64 1.51 1.71 1.73 1.04 1.39
Italy 6.30 0.13 0.40 0.84 0.38 0.46 0.99
Netherlands 5.70 2.09 1.79 1.75 1.82 0.96 1.58
Norway 7.50 2.12 1.88 1.50 1.95 1.24 1.66
Portugal 5.70 0.98 1.05 0.93 1.06 0.82 1.15
Spain 6.50 0.95 1.08 1.07 1.08 -0.08 1.08
Sweden 7.20 2.23 1.93 1.71 1.93 1.14 1.60
Switzerland 5.00 2.13 1.96 1.64 1.84 1.29 1.60 United Kingdom 7.80 1.71 1.62 1.74 1.71 0.41 1.33
Taking this into account and looking at table 3, the average underpricing is surprisingly low for Greece and surprisingly high for Sweden while the averages of example given, Italy or
Switzerland are as expected based on the theory, being above and below the Western European
average respectively. For Greece, this may be due to the limited amount of observations over the
sample period. As visible in table 3, five countries have four observations or less, which is not
uncommon in prior research but it does make the inferences made upon these countries less
reliable, like Greece. Also, 40% of the sample is constituted out of IPOs in the United Kingdom
followed by 24.12% of the sample being French IPOs.
The Mystery of Underpricing: The Influence of a Country’s Institutional Environment and Legal Framework on the Level of Underpricing in Western Europe
Table 3 . Average underpricing per country over the period 2008 to 2015 Country Average underpricing Std dev. Obs.
Austria 7.31 12.54 3
Belgium 6.63 9.44 14
Denmark 8.93 11.60 9
Finland 1.44 10.92 14
France 5.08 21.48 200
Germany 0.77 11.29 62
Greece 2.47 31.60 3
Ireland 3.42 5.38 3
Italy 20.14 67.85 53
Netherlands 9.90 20.13 8
Norway 0.18 14.67 47
Portugal -0.76 1.40 4
Spain 11.27 23.00 13
Sweden 15.27 32.83 52
Switzerland 3.19 5.31 10
United Kingdom 13.90 39.80 331
Total 9.71 34.31 826
Table 4 shows us the average underpricing per year where the IPO has taken place. We can
derive from this univariate analysis that underpricing was lower than average in the years 2008,
2009, 2011 and 2014. Especially 2009 and 2014 are standing out in this analysis. When looking
at the fact whether the IPO market was hot or cold, I can derive that the IPO market in 2009
especially was cold with only 31 IPOs and an average underpricing of 5.11%. In this table, 2013
shows an average underpricing of 14.58% and 103 observations, indicating the start of a hot IPO
market again. This is contradicted though due to 2014, showing an average underpricing of
6.67% with a low standard deviation while the number of observations did grow and were more
than average over the years (average observations per year: 103). 2015 does confirm the hot IPO
market with an average underpricing of 12.69% and 150 observations. I further examined the
data and no explanation can be found in the distribution among countries in the years 2013, 2014
and 2015 as these are approximately the same.
The Mystery of Underpricing: The Influence of a Country’s Institutional Environment and Legal Framework on the Level of Underpricing in Western Europe
Table 4. Descriptive statistics IPOs per year from 2008 to 2015 IPO year Average underpricing Std. dev. Obs.
2008 8.107116 24.200949 82
2009 5.1092357 9.1265614 31
2010 10.019827 49.768135 108
2011 7.9584445 23.048825 99
2012 9.6448946 23.082134 79
2013 14.583022 28.298613 103
2014 6.6639364 16.716899 174
2015 12.69118 53.466703 150
Total 9.7099237 34.309314 826
Table 5 shows the correlation matrix. Since all WGI variables are highly correlated among each other, multicollinearity problems occur when examining these variables in one model combined.
When further examining the correlations between the country-level variables and firm-level variables, no alarmingly high correlation is found that can cause multicollinearity. In order to run the regressions without multicollinearity, I make use of six different models with each country- level explanatory variable tested separately. This is visible in table 7, where the HLM full model is depicted.
Table 5. Correlation matrix of variables
UP VC Tech ECO IPI CC GE RQ RL PS VA
UP 1
VC -0.0644 1
Tech -0.0029 0.3207* 1
ECO -0.0541 -0.1219* -0.0925* 1
IPI 0.0914* -0.1265* -0.0969* -0.0303 1
CC -0.0501 -0.0345 -0.0135 0.0174 0.3211* 1
GE -0.0619 0.0067 0.0159 -0.0129 0.2842* 0.9847* 1
RQ 0.0323 -0.1400* -0.0994* 0.0086 0.6383* 0.8099* 0.7556* 1
RL -0.0483 -0.0284 -0.0203 -0.0045 0.4180* 0.9820* 0.9775* 0.8415* 1
PS -0.0776* -0.0299 -0.0101 0.0711* -0.3112* 0.5498* 0.4950* 0.1779* 0.4243* 1
VA -0.0566 -0.0511 -0.0181 0.0474 0.1774* 0.9240* 0.8824* 0.6458* 0.8581* 0.7909* 1
*, Significant at a 5% significance level. Relevant variables: Underpricing (UP), Venture capital ownership dummy (VC), High-tech dummy (Tech), Equity carve-out dummy (ECO), Investor protection index (IPI), Control of corruption index (CC), Government effectiveness (GE), Regulatory quality (RQ), Rule of law (RL), Political stability (PS), Voice and accountability (VA).
4.2 HLM null model
First, I test if there exists significant variance between countries in the level of underpricing
because, if this is not the case there is no sound reason to make use of the HLM regression
The Mystery of Underpricing: The Influence of a Country’s Institutional Environment and Legal Framework on the Level of Underpricing in Western Europe
instead of a regular OLS regression as the assumption of independence of observations would not be violated and the OLS regression would suffice to explain the sought for relationships. As the ANOVA is significant at a 10% level, it supports our choice to execute HLM. Also, what can be inferred from the intraclass correlation is that 1.55% of the variance in underpricing is
established between the independent groups, the countries, indicating that variance exists and can be explained. I intend to this this with the HLM full model depicted below.
Table 6. One-Way ANOVA Model with random effects
Fixed effect Coefficient s.e.
Average underpricing 11.56 36.99
Random effect Variance Component df P value
Level 2 effect 1931.76 15 0.054
Level 1 effect 1163.16
Intraclass correlation 1.55%
4.3 HLM Full Model
In this section, the results on the HLM full model are presented. First, the results on the firm and issue specific variables are presented followed by the results of the minority investor protection and the institutional environment relationship with the level of underpricing.
Firm- and issue specific variables
The first model in table 7 presents the HLM analyses with the firm- and issue specific variables
exclusively. First of all, the results show a significant negative impact of the IPO year 2008,
2009, 2012 and 2014 at a 5%, 1%,10% and 5% significance level respectively. The significant
negative outcome of 2012 is remarkable due to the fact that the average underpricing in that
particular year (9.66%) is close to the sample ’s overall average (9.71%). Next to that, the results
depict a significant negative relationship between venture capitalist backing pre-IPO and the
level of underpricing at the 1% significance level, as expected. Also, IPOs that are some form of
an equity carve-out have a lower level of underpricing at a 1% significance level. High-tech
firms do have an expected positive relationship with the level of underpricing, however this
relationship does not bear enough significance to interpret the effect on the level of underpricing.
The Mystery of Underpricing: The Influence of a Country’s Institutional Environment and Legal Framework on the Level of Underpricing in Western Europe
Table 7 . Results of HLM analyses with firm- and country-effects on the level of underpricing
Variables Models
1 2 3 4 5 6 7 8
Fixed effect Level 1 predictors
IPO2008 -7.974** -7.616** -8.307** -8.292** -8.122** -8.358** -8.207** -8.270**
(3.62) (3.54) (3.57) (3.60) (3.60) (3.58) (3.39) (3.51)
IPO2009 -9.583*** -9.315*** -10.26*** -10.27*** -9.820*** -10.26*** -10.33*** -10.34***
(3.31) (3.15) (3.17) (3.21) (3.12) (3.20) (2.99) (3.09)
IPO2010 -5.564 -5.526 -5.432 -5.47 -5.534 -5.427 -5.462 -5.462
(5.08) (4.89) (5.24) (5.23) (5.18) (5.29) (4.97) (5.14)
IPO2011 -7.15 -7.172 -7.071 -7.121 -7.122 -7.04 -7.132 -7.117
(5.09) (5.03) (5.09) (5.11) (5.10) (5.13) (4.92) (5.02)
IPO2012 -4.186* -3.979* -4.219* -4.265** -4.156* -4.203* -4.395** -4.327**
(2.20) (2.19) (2.18) (2.17) (2.25) (2.20) (2.06) (2.08)
IPO2014 -7.782** -7.547** -7.736** -7.781** -7.752** -7.761** -7.592** -7.710**
(3.63) (3.67) (3.66) (3.63) (3.68) (3.65) (3.66) (3.65)
IPO2015 -1.286 -0.686 -1.365 -1.439 -1.336 -1.499 -1.1 -1.31
(4.19) (4.11) (4.25) (4.25) (4.26) (4.27) (4.11) (4.21)
VC -6.804*** -6.872*** -6.995*** -6.934*** -6.863*** -6.939*** -7.187*** -7.075***
(1.42) (1.39) (1.47) (1.49) (1.39) (1.49) (1.43) (1.47)
Technology 1.012 1.073 1.064 1.104 1.005 1.051 0.949 1.036
(1.38) (1.42) (1.27) (1.27) (1.35) (1.28) (1.31) (1.28)
ECO -6.812*** -6.803*** -6.828*** -6.910*** -6.846*** -6.867*** -6.824*** -6.808***
(1.94) (1.86) (1.98) (1.96) (1.96) (2.00) (1.87) (1.94)
Level 2 predictors
IPI 2.963
(2.06)
CC -4.516*
(2.69)
GE -6.918**
(3.04)
RQ -3.302
(7.47)
RL -6.424*
(3.29)
PS -7.776*
(4.46)
VA -13.29
(9.11) CONSTANT 15.39*** -4.74 22.75*** 26.23*** 20.20* 25.52*** 21.17*** 33.68**
(5.29) (14.80) (6.05) (5.92) (11.64) (5.96) (5.89) (13.41)
Random effect
Level 2 18.40*** 10.25 13.18*** 12.11*** 18.95*** 13.43*** 10.67*** 12.50***
(6.78) (20.72) (5.23) (5.57) (4.10) (5.42) (9.78) (5.83)
Level 1 1139.3*** 1140.7*** 1138.3*** 1138.0*** 1138.7*** 1137.9*** 1139.1*** 1138.8***
(346.70) (348.80) (345.60) (345.30) (346.30) (345.30) (347.80) (346.20) Chi2 58.54*** 62.25*** 145.43*** 199.42*** 88.21*** 222.59*** 177.97*** 133.69***
N 826 826 826 826 826 826 826 826
HLM analysis, with data from ThomsonOne and Worldbank between 2008 and 2015. Dependent variable is underpricing. ***, **, * Significant at the 1%, 5% and 10% level. The standard errors are given in parenthesis below the coefficient. Standard errors are adjusted for
heteroscedasticity by running a robust analysis. Chi-squared is included to show the significance of the model. Dependent variable:
underpricing between 2008 and 2015. Explanatory variables: Venture capital ownership dummy (VC), High-tech dummy (Technology), Equity carve-out dummy (ECO), Investor protection index (IPI), Control of corruption index (CC), Government effectiveness (GE), Regulatory quality (RQ), Rule of law (RL), Political stability (PS), Voice and accountability (VA). The IPO year 2013 has been omitted due to collinearity.
