Country of origin effects and foreign IPO success : investigating the relationship between investment freedom and foreign IPO underpricing

(1)

Country of Origin Effects and Foreign IPO Success:

Investigating the relationship between Investment Freedom

and Foreign IPO Underpricing

University of Amsterdam Supervisor: Dr. Ilir Haxhi

Second Reader: Msc. Erik Driksen Final: 4-08-2014

Student: Temur Tuzhba Student ID: 10605320

Master of Science Business Studies Track: International Management

(2)

Abstract

Building on previous literature on Initial Public Offerings (IPOs) and the role of legitimacy in capital markets, the current study explores the country of origin effect on Foreign IPO Underpricing. Existing literature reveals the relationship between home-country institutional development and Foreign IPO Underpricing. However, it does not investigate whether institutional development of a home-country in general, or the development of finance-connected institutions has more influence on Foreign IPO Underpricing. The current research addresses this issue by testing whether Investment Freedom (IF) is a better predictor of Foreign IPO Success than other measures used by our priors (i.e., Economic Freedom (EF)). Moreover, the current study examines whether companies from countries with low Investment Freedom can compensate for their Liability of Foreignness (LoF) by sending positive signals to investors, such as increasing the independence of Board of Directors (BD) and broadening Geographic Scope prior to IPO, and thus moderate the relationship between home-country IF and Foreign IPO Underpricing. A sample of 198 foreign (non-American) firms that had previously undergone an IPO on American Capital Market was selected and analyzed in order to determine whether the above-mentioned relationships were supported. The results indicated that, first, there was a negative relationship between home-country IF and Foreign IPO Underpricing; then, IF was found to be a better predictor of Foreign IPO Underpricing than home-country EF, and, finally, that moderating effects of both Board Independence and Geographic scope on the relationship between IF and Foreign IPO Underpricing were significant. These findings contribute to the existing literature by establishing a link between home-country IF and Foreign IPO Underpricing. From a managerial perspective, this research suggests that companies can partially compensate for their LoF in foreign capital market by increasing the number of independent directors in BD and broadening company’s geographic scope prior to the IPO.

(3)

2.2 Linking Investment Freedom, Economic Freedom, Board Independence, Geographic Scope, and IPO underpricing ... 18 3. Methodology ... 25 3.1 Data Sources ... 25 3.2 Sample ... 26 3.3 Dependent variable ... 26 3.4 Independent Variables ... 27 3.5 Moderating Variables ... 28 3.6 Control Variables ... 29 3.7 Data Analysis ... 30

4. Results and analysis ... 35

4.1 Descriptive Analysis ... 35

4.2 Variation Inflating Factors test ... 37

4.2 Correlations ... 41 4.3 Regression Analysis ... 46 5. Discussion ... 52 5.1 Findings ... 52 5.2 Managerial Implications ... 55 5.3 Limitations ... 56 5.4 Future Research ... 56 6. Conclusion ... 58 References ... 61 Appendices ... 71

(4)

Introduction

The integration of foreign capital markets provides new opportunities to firms from all over the world, making it easier for companies to gain equity capital through more developed foreign markets (Yoshikawa, Rasheed, Datta, & Rosenstein, 2006). However, the level of capital market development varies significantly between different countries, making it such that some capital markets can provide more opportunities than others.

The size of the foreign market is not the only reason for a firm to list abroad. As Hursti and Maula (2007) point out, other factors, such as type of industry, foreign experience of top management team, venture capital investors’ characteristics, etc., can also influence foreign Initial Public Offering (IPO) market choice. Thus, foreign equity capital market choice is an important decision, and one which is actively studied in the fields of financial and international management and international finance (Blass & Yafeh, 2001; Hursti & Maula, 2007; Kadiyala & Subrahmanyam, 2002; Moore, Bell, & Filatotchev, 2010; Moore, Bell, Filatotchev, & Rasheed, 2012). As a result of this trend, the number of foreign firms undergoing IPOs in the United States (US) has been increasing for almost two decades (Bruner, Chaplinsky, & Ramchand, 2004).

The IPO process in general is complicated, often associated with high levels of uncertainty, especially regarding the initial offering price (Lowry & Schwert, 2004). Existing literature emphasizes the crucial role of the information asymmetry that exists between company insiders and company outsiders (Bell, Moore, & Al-Shammari, 2008; Bell, Moore, & Filatotchev, 2012; Bruner, Chaplinsky, & Ramchand, 2006; Hursti & Maula, 2007; Moore et al., 2010). This discrepancy, and especially the lack of information provided to outsiders often results in the underpricing of company shares. As Bell et al. (2008) suggest, the lower

(5)

the underpricing, the better the IPO performance, as a low level of underpricing means that the company was able to capture more value.

When a firm decides to be listed on a foreign capital market, this process becomes even more complicated because the level of information asymmetry increases, and legitimacy problems arise (Bell et al., 2008; Bell, Moore, et al., 2012; Moore et al., 2012). Scholars are actively studying this problem both on country- and firm-levels. Specifically, existing research has provided support for the influence of host-country and home-country effects (Moore et al., 2010), country of origin (COO) effects (Bell et al., 2008), and corporate governance effects (Moore et al., 2012), on foreign IPO success.

A company can change itself relatively easily, but quite often a firm faces problems with Foreign IPO performance not because of firm-level factors, but because of country-level factors, which it does not have the power to change. Additionally, there is clear evidence of Zaheer’s (1995) Liability of Foreignness (LoF), which every company undergoing an IPO in a foreign country must deal with (Bell, Filatotchev, & Rasheed, 2012). However, as many previous studies have demonstrated, it is possible to compensate for LoF to some degree by implementing legitimate corporate governance practices intended to send positive signals to investors (Certo, Daily, & Dalton, 2001; Francis, Hasan, Lothian, & Sun, 2010; Moore et al., 2010). For a foreign IPO, investors and managers must consider not only firm factors and market factors, but also the company’s home-country macroeconomic factors (Francis et al., 2010; Moore et al., 2010). Unfortunately, many scholars still agree (Bell, Moore, et al., 2012; Moore et al., 2012), that, in spite of receiving more attention in recent years, this field of research remains understudied. Hence, there are still many gaps in the existing literature that should be covered.

Bell et al. (2008) unveil the relationship between home-country economic freedom and foreign IPO underpricing by using the Heritage Foundation Index of Economic Freedom

(6)

(Beach & Miles, 2004) as a proxy. This is a macroeconomic index that consists of ten different dimensions: Property Rights Freedom, Freedom from Corruption, Fiscal Freedom, Government Spending, Business Freedom, Labor Freedom, Monetary Freedom, Trade Freedom, Investment Freedom and Financial Freedom (Miller, Holmes R., & Kim B., 2014). This index is calculated by taking an average of all ten parameters, which implies that they are all equally important. However, this is clearly not the case for an investor who is considering buying shares of a foreign company. Investors might not look at the level of country economic freedom as a whole when they are considering whether to invest or not, but may instead pay more attention to certain areas, such as the level of Investment Freedom. This relationship between Investment Freedom and Foreign IPO Underpricing clearly warrants further consideration.

Thus, building on Bell et. al.’s (2008) work, the current study analyzes the effect of home-country institutional development on foreign IPO performance. First, following the role of legitimacy in capital markets rationale (Bell et al., 2008; Deeds, Mang, & Frandsen, 2004; Moore et al., 2012), it is assumed that, in terms of Investment Freedom, home-country institutional development levels influence Foreign IPO Underpricing. As Economic Freedom is known to have significant long-term effects on the development of the economy (de Haan & Sturm, 2000; J. D. Gwartney, Lawson, & Holcombe, 1999), and capital markets such as the American market are known to be associated with primarily short-term oriented behavior (Marginson & McAulay, 2008; Miles, 1993; Reilly, 1973), it is assumed that home-country Investment Freedom is a more appropriate predictor of Foreign IPO Underpricing than Economic Freedom. As there is support for the fact that companies from countries with weak investment institutions have to employ a large number of legitimate corporate governance practices in order to raise investors’ perceptions of the company’s value (Bell, Filatotchev, & Aguilera, 2013), it is hypothesized that the level of board independence moderates this

(7)

relationship. Furthermore, because it significantly contributes to the long-term success of the company (Bell et al., 2008), geographic scope is seen as a positive signal by investors (Goerzen & Beamish, 2003; Tallman & Li, 1996), Thus it is predicted that geographic scope moderates the relationship between Investment Freedom and Foreign IPO Underpricing.

Therefore, this research aims to answer the following questions:

1 - How does home-country level Investment Freedom affect Foreign IPO performance?

2 – Does Investment Freedom have more influence than Economic Freedom on Foreign IPO Underpricing?

3 – To what extent do firm’s characteristics, such as Board Independence and Geographic Scope, moderate the relationship between Investment Freedom and Foreign IPO Underpricing?

Data on 198 foreign firms that underwent an IPO on the US Capital Market (Both NYSE and NASDAQ) between 2000 and 2012 was gathered. After collecting the data, it was observed that the American capital market was especially popular among Chinese firms, as 96 companies (48% of the sample) originated from China. In order to ensure the robustness of the research, a second, non-Chinese sample was constructed, consisting of 102 companies. The choice of American equity market was not random, as it is one of the most popular destinations for foreign IPO (Blass & Yafeh, 2001; Bruner et al., 2004, 2006; Kadiyala & Subrahmanyam, 2002). In order to test the significance of the above-mentioned relationships and answer the research questions posed, the current study research employs a hierarchical regression.

This research contributes both to scientific and managerial fields. It elaborates on the relationship between Foreign IPO Success and COO effects, by establishing a link between home-country Investment Freedom and Foreign IPO Underpricing. In addition, it adds to the

(8)

existing legitimacy in capital markets literature by supporting the prediction that Investment Freedom is a better predictor of Foreign IPO Performance than Economic Freedom, thus revealing that capital market actors are more interested in certain institutional areas than in institutional development levels in a broader sense. Furthermore, this paper serves as a insight for managers of companies that are considering listing abroad, to compensate for LoF and minimize underpricing by implementing certain corporate governance practices or by increasing geographic scope of the company prior to IPO.

The remainder of this paper is organized as follows. The following section reviews existing literature on foreign IPO performance and other relevant theories. In section 3, the methodology and research design are discussed. Section 4 presents the results of the analysis. In section 5, the results are discussed and linked to existing theory. The limitations of the study are mentioned and areas of future research are proposed. The last section provides conclusion.

(9)

2. Literature Review

To fully understand the nature of the connection between Country of Origin (COO) factors and foreign IPO underpricing in terms of Economic Freedom and Investment Freedom, as well as the moderating effect of Geographic scope and Board Independence on the relationship between Investment Freedom and IPO underpricing, it is important to revise the relevant existing literature.

2.1 Relevant Theories

Initial Public Offering

It is quite common for an entrepreneur to start his company by getting capital from a limited number of investors (e.g., venture capitalists, angel investors etc.), providing them with a possibility of great returns in exchange for high degrees of risk and temporary lack of liquidity of its investments. After a period of considerable success, a company may decide to go public in order to reward their investors, provide them with an exit opportunity, and raise extra capital to continue its growth (Ibbotson & Ritter, 1995). Security offering is a significant event in a life cycle of a firm, and many factors contribute to its importance. For instance, after the offering ownership structure often changes and a large amount of capital flows into the issuing firm (Eckbo, 2008).

Brau and Fawcett (2006) distinguish four main reasons that firms go public. First, they note that cost of capital scholars (e.g., Modigliani & Miller, 1963; Scott, 1976) argue that firms undergo IPO when external equity gains are predicted to drive their cost of capital down. Second, as previously mentioned, an IPO allows insiders to sell their parts of the company and gain personal wealth (Ritter & Welch, 2002; Zingales, 1995). Third, an IPO may be taken as a step to make a takeover easier to implement, as shares may serve as a form of “currency” (Brau, Lambson, & McQueen, 2005; Zingales, 1995). Fourth, an IPO may be

(10)

to the firm (Certo et al., 2001). There is also some evidence that an IPO helps a firm to use a first-mover advantage (Maksimovic & Pichler, 2001). An IPO can also be seen as a reputation-enhancing event, which is especially important for high-tech companies (Brau & Fawcett, 2006).

Many scholars have noticed that nowadays firms tend to bypass less developed home country stock exchanges and pursue IPOs in more developed foreign capital markets (Bell, Moore, et al., 2012; Ding, Nowak, & Zhang, 2010; Moore et al., 2012).

It is an important strategic choice that cannot be made without proper decision-making processes conducted by the firm. It seems logical that sometimes firms choose foreign stock exchanges over the ones in their home countries, as usually the capital markets of more developed countries provide more opportunities. Therefore, listing on a foreign stock exchange of a more developed economy brings new opportunities that result in positive shifts in growth levels. Not only does it immediately move a company to a new level, but it also provides different long-term benefits (Ding, Nowak & Zhang 2010).

Long-term benefits are one of the main reasons for foreign IPOs, as companies tend to be more underpriced on host stock exchanges than on home exchanges, which diminishes short-term benefits (Bell et al., 2008).

Underpricing

Underpricing is a well-known attribute of equity markets (Eckbo, 2008). As Ritter and Welch (2002) point out, Stoll and Kerley (1970), Logue (1973), Reilly (1973) and Ibbotson (1975) were among the first scholars to notice the tendency of stocks to rise in price by the end of the first day of trading. Eckbo (2008) defines underpricing as the difference between the initial price of the stock and the price of the stock at the first closure of the stock exchange.

(11)

Underpricing is so complex and dependent onso many factors, that it is not a surprise that it has been a center of scholars’ attention for quite a while, providing ample time for the development of many different theoretical explanations of the phenomenon.

Rock’s (1986) “winner’s curse” hypothesis was one of the first explanations of the reasons behind underpricing. In Rock’s model, there are two kinds of investors: perfectly informed about the future price of the stock they are buying, and not informed at all about the price. Whereas informed investors only buy shares of companies they know are underpriced, uninformed investors invest only part of their money in underpriced securities and the other part in overpriced shares, as clueless as to which shares are underpriced and which are not. That is why uninformed investors face the winner’s curse: they can get all the shares they want if and only if informed investors are not interested in those shares. Because of this selection problem, the uninformed investor only buys shares if there is a sufficient average rate of underpricing that can compensate for his investment in bad stocks (e.g. overpriced).

Benveniste and Spindt (1989) suggested another reason for the existence of underpricing with their “costly information acquisition” hypothesis. They argue that investment bankers (underwriters) intentionally underprice IPOs in order to stimulate regular investors to share all the information they have about the true value of the company.

In 1992 Welch presented another theory. His “Cascades” (1992) hypothesis claims that when investors are considering buying stocks, not only do they use the information they have about the company, but they also check to see whether other investors are buying the stocks of the company they are considering investing in. If no one is buying, it is probable that they will not buy the shares even if they have favorable information. It is for this reason that issuers tend to underprice their offerings in order to stimulate investors to buy shares.

Baron and Holstrom (1980) proposed the “Investment Banker’s Monopsony Power” hypothesis, which claims that, when investment bankers underwrite the IPO, they use their

(12)

superior equity market conditions knowledge to underprice offerings, thus pandering to the interests of buying investors.

Other scholars (e.g. Allen & Faulhaber, 1989; Grinblatt & Hwang, 1989; Welch, 1989) proposed that companies use underpricing as a way to signal their high value. Welch (1989) noticed about 30% of the companies follow their IPOs with seasoned offerings, which nullifies the issue of underpricing because they sell securities with their seasoned offerings for a market price. However, underpricing signals the high value of the company to investors, which raises the price and improves the company’s reputation.

Brennan and Franks (1997) argue that issuers may intentionally underprice their offerings so they stimulate high demand, which in turn results in better liquidity for their shares and disperses ownership. With disperse ownership it is harder for shareholders to control management and make coordinated decisions.

The theories discussed above are not the only explanations of underpricing in the field. Structurally speaking, it is useful to point out, that scholars tend to divide all of these theories into four broad groups: “asymmetric information”, “institutional reasons”, “control considerations”, and “behavioral approaches” (Eckbo, 2008).

The flotation process is a very complicated phenomenon, with many different actors involved. Thus, in order to properly understand motives, actions and consequences an appreciation of underlying theories is necessary. Many scholars (Allen & Faulhaber, 1989; Bell et al., 2013; Bell, Moore, et al., 2012; Bruton, Filatotchev, Chahine, & Wright, 2010; Certo et al., 2001; Moore et al., 2010) underline the importance of three theories: institutional theory (DiMaggio & Powell, 1983), signalling theory (Certo et al., 2001; Certo, 2003), and agency theory (Jensen & Meckling, 1976). Thus, the sections below briefly review each one.

(13)

Institutional Theory

North (1991, p. 97) defines institutions as “humanly devised constraints that structure political, economic and social interaction”. Institutions can be divided in two groups: informal institutions (e.g., customs, traditions etc.) and formal institutions, such as laws (North, 1991). One of the main contributions of the institutional theory is the notion that organizations are considered legitimate when they follow the rules (both formal and informal) of a given environment (Dowling & Pfeffer, 1975).

Suchman (1995, p. 574) defines legitimacy as “a generalized perception or assumption that the actions of an entity are desirable, proper, or appropriate within some socially constructed system of norms, values, beliefs, and definitions”. Legitimacy is critical in attracting resources for the firm (Deeds et al., 2004). The Capital Market is not an exception to the rule. In fact, Capital Market legitimacy is one of the most important factors of IPO Success (Bell et al., 2008; Bell, Moore, et al., 2012; Moore et al., 2010).

As Moore et al. (2010) state, it is harder for companies to be perceived legitimate if they are undergoing a foreign IPO, as they have to simultaneously comply with the different and sometimes conflicting norms of two different environments. In order to achieve acceptable levels of legitimacy, companies use signalling to show their value and prove their reliability.

Signalling Theory

As a well-known phenomenon of IPO, asymmetric information between insiders and outsiders, plays an huge role in the process of Securities Offering, so signalling theory continues to be an important part of IPO research (Brau & Fawcett, 2006). Signalling theory is mainly concerned with diminishing information asymmetry between actors (Spence, 2002).

(14)

Spence (1973) used the job market to construct his signalling hypothesis. He claimed that, as it is not that easy for an employer to formulate the right opinion about future employee’s skills, job seekers use education as a way to signal their value. The same rationale applies to capital markets: it is hard for outside investors to obtain all the information about the company. Thus, issuers employ different strategies to signal certain things about the company (e.g., good corporate governance) by hiring independent directors (Chahine & Filatotchev, 2008).

Agency Theory

When a company undergoes an IPO, the structure of ownership changes and corporate governance issues arise. Agency theory is the most popular theory used to explain the principal-agent (shareholders-managers) relationship. Jensen and Meckling (1976, p. 308) define an agency-principal relationship as “a contract under which one or more persons (the principal(s)) engage another person (the agent) to perform some service on their behalf which involves delegating some decision making authority to the agent”.

The problem that arises when the ownership and control are divided is managers’ opportunistic behavior. This is because sometimes what is good for managers is bad for the owners and vice versa (Tirole, 2010).

It is harder to manage agency costs if a company is listed abroad. That is why companies employ transparent corporate governance practices that help them signal the firm value to foreign investors. There is a significant amount of research analyzing the effects of corporate governance practices on IPO success (Bell et al., 2013; Bell, Moore, et al., 2012; Certo, 2003; Dahya, Dimitrov, & McConnell, 2009; Filatotchev & Bishop, 2002).

(15)

Country of Origin Effects

The IPO process is often associated with uncertainty, for both investors and issuing firms. Investors face uncertainty because of information asymmetry, which makes the proper evaluation of a company’s value a very complex task (Higgins & Gulati, 2006). It is even more complicated, when the company undertaking an IPO is from another country because, in addition to firm-level factors, investors have to evaluate the environment the company operates in. This phenomenon is often called a Country of Origin Effects (COO) (Lampert & Jaffe, 1996).

COO effects are known to have a significant influence on shaping firms’ international strategies (Lampert & Jaffe, 1996; M. S. Roth & Romeo, 1992). Despite the fact that this field of research has great potential, scholars continue getting mixed results, which is a serious limitation of the COO effects research (Obermiller & Spangenberg, 1989).

Even though the limited explanation power of COO research is well-known, COO effects rationale found a wide acceptance in IPO research (Bell et al., 2008; Bell, Moore, et al., 2012; Moore et al., 2010) because scholars showed the importance of home-country effects on foreign IPOs. Bell et al. (2008) show that COO effects in terms of home-country economic freedom levels are important predictors of foreign IPO success.

Economic Freedom

Economic freedom is an important proxy of evaluating legitimacy of a foreign company because it helps measure the level of protection of property rights, freedom of competition and transparency vitally important aspects for any investor (J. Gwartney, Lawson, & Norton, 2008). Thus, other factors taken equal, the higher the level of economic freedom, the more legitimate a company is considered by actors of developed capital markets (Bell et al., 2008).

(16)

Miller et al. define (2014) economic freedom as “the fundamental right of every human to control his or her own labor and property”. They claim, “In an economically free society, individuals are free to work, produce, consume, and invest in any way they please. In economically free societies, governments allow labor, capital and goods to move freely, and refrain from coercion or constraint of liberty beyond the extent necessary to protect and maintain liberty itself” (Miller et al., 2014).

The Heritage Foundation Economic Freedom index is measured by combining four groups of factors, such as “rule of law” (property rights, freedom from corruption), “limited government” (Fiscal Freedom, Government Spending), “Regulatory Efficiency” (Business Freedom, Labor Freedom, Monetary Freedom), and “Open Markets” (Trade Freedom, Investment Freedom, Financial Freedom) (Miller et al., 2014). Such a thorough analysis aims to capture all aspects of economic states of the countries.

Economic Freedom’s effects on IPOs have has been a recent topic of investigation. Most recently, Bell et al. (2008) proved a negative relationship between economic freedom and foreign IPO underpricing. Home-country Economic Freedom levels affect foreign IPO underpricing such that companies from emerging markets are more underpriced than companies from developed markets (Bell et al., 2008).

However, Economic Freedom is a very broad concept that mostly reflects developments on a country-level. For example, there is a lot of evidence that economic freedom stimulates economic growth (de Haan & Sturm, 2000; J. D. Gwartney et al., 1999). When an investor is considering buying shares of a newly listed company operating in a foreign market, he might not be interested in economic freedom as a whole, finding some particular aspects of the concept more important than others. Investment Institutions’ level of development surely is one of them.

(17)

Investment Freedom

Investment Freedom is a term used by Miller et al. (2014) to measure levels of development of capital markets and institutions that are supposed to support equity markets. Scholars believe than in a free economy investments would not be limited by any constraints and would flow in and out of the country without any government intervention.

Even though there is no such concept as an investment freedom in the literature, this measure can be used as a good proxy of COO effects connected with investments. This is because the measure includes many important factors that influence IPO success. Some of these factors are: “foreign investment code,” “levels of transparency and burdensomeness of bureaucracy,” “policy implementation efficiency,” “transparency of investment laws and practices”.

Many scholars have demonstrated the importance of these factors. Bell et al. (2012) argue that home-country investment protection is positively related to IPO Success. In line with previous research, Moore et al. (2010) find that investment protection is negatively correlated with IPO underpricing. These relationships can also be moderated by firm-level factors, such geographic scope, insider ownership, board independence, underwriters’ reputation etc. (Bell, Filatotchev, et al., 2012; Bell et al., 2008; Bell, Moore, et al., 2012; Moore et al., 2010).

Board independence

Strong corporate governance is almost as important as financial results because it significantly decreases the level of investors’ uncertainty in different ways (Gillan & Starks, 2003). Bell et al. (2012) contend that employing good corporate governance in a company is a good way of signalling foreign firms’ value. It is also known to reduce Liability of Foreignness in Foreign Capital Markets.

(18)

Scholars treat board independence as a very strong signal because firms with independent boards tend to be better governed (Aggarwal, Erel, Stulz, & Williamson, 2007; Bell, Moore, et al., 2012). Independent directors are known to effectively monitor managers’ actions and prevent boards from making wrong and harmful decisions, as well as forcing boards to take unpleasant actions, such as firing non-performing CEOs (Weisbach, 1988).

Board independence also means that a company is transparent and is ready to protect interests of shareholders. In a recent study, Dahya et al. (2009) found a strong positive relation between board independence and corporate value, especially in countries with weak investor protection.

Geographic Scope

Scholars mostly agree that the more countries a company operates in, the stronger the positive signals it sends to market participants (Goerzen & Beamish, 2003). However, there are inconsistent results as to whether geographic scope has a positive or negative influence on companies’ performance (Denis, Denis, & Yost, 2002; K. Roth, 1992).

Geographic Scope is believed to bring a number of benefits to a company, such as increasing proprietary assets’ value appropriation levels (Goerzen & Beamish, 2003), improving organizational learning (Delios & Henisz, 2003; Kogut & Zander, 1993), and supporting competition on an international scale (Karnani & Wernerfelt, 1985).

2.2 Linking Investment Freedom, Economic Freedom, Board Independence, Geographic Scope, and IPO underpricing

Institutional theory argues that organizations are seen as legitimate when their inner environment follows the rules of the environment in which they are functioning (Deephouse, 1996; DiMaggio & Powell, 1983; Dowling & Pfeffer, 1975).

(19)

Some scholars also underline the importance of external characteristics of an organization as affecting how legitimate a company is perceived to be (Hybels, 1995; Kostova & Zaheer, 1999). Legitimacy plays an important role in Capital markets because, when investors see a company as legitimate, they are more likely to invest their money in the firm (Deeds et al., 2004). In theory, investors have access to all of the information and act rationally at all times. However, in reality capital markets are associated with imperfect information, which results in the fact that more often than not they base their decisions on their experience and on whatever amount of limited information they can get (Westphal & Zajac, 1998).

Information asymmetry is an inherent part of the IPO process, because insiders have more insights into the future of the company than outsiders (Certo et al., 2001). It is even more relevant for firms listed on foreign exchanges because they bring Liability of Foreignness into the mix (Bell, Filatotchev, et al., 2012).

In the flotation process, a company actually earns money only from the first transaction, so even if the price of stocks goes up later, this created value is captured not by the company, but by other market actors (Bell et al., 2008). In order to reduce this underpricing effect, firms should consider factors that influence the level of legitimacy they will receive when choosing foreign IPO capital market because it is known, that legitimacy reduces underpricing effects (Bell, Moore, et al., 2012).

One of the measures of investment institutions, investment freedom (Miller et al., 2014), combines a number of institutional factors (e.g., “foreign investment code,” “levels of transparency and burdensomeness of bureaucracy,” “policy implementation efficiency,” “transparency of investment laws and practices”). As many of these factors were studied independently (Bell et al., 2013; Bruton et al., 2010; Moore et al., 2010) and were found to

(20)

influence IPO Underpricing, it is assumed that investment freedom is a good predictor of Foreign IPO Underpricing.

Keeping in mind the enormous uncertainty investors face when considering investing in a foreign company, and following institutional logic (DiMaggio & Powell, 1983; North, 1991), it is assumed that investors in the American market, which is known to have high levels of Investment Freedom, see companies that originate from countries with high levels of Investment Freedom as more legitimate. Because they have to comply with higher institutional standards, they are considered more reliable, which in turn reduces underpricing levels.

Existing literature on IPO performance employs a number of measures of institutional development, such as the strength of minority shareholders protection (Bell et al., 2013; Moore et al., 2010), legal institutions (Bruton et al., 2010) and economic freedom (Bell et al., 2008) and suggests that these measures have an effect on IPO underpricing. Therefore, we propose:

Hypothesis 1: There is a negative relationship between home-country Investment Freedom

and foreign IPO Underpricing such that higher home-country Investment Freedom causes a lower rate of foreign IPO Underpricing.

As previously mentioned, legitimacy is an important concept for capital markets, where asymmetric information prevails and investors’ decision making processes are often biased (Deeds et al., 2004). However, when a company issues shares abroad it automatically functions in several environments simultaneously. In order to function properly and enjoy the benefits of trust and reputation among market actors, it has to be seen as legitimate in every environment in which it functions.

(21)

Moreover, the environments can be seen differently, depending on the angle from which they are examined. For instance, Economic Freedom combines a broad variety of factors, which allows it to operate as a proxy of the state of the economy on a country level. Economic Freedom can be used as a very good proxy of institutional development of a certain country. Thus, investors can use it as a way to reduce uncertainty when they are considering investing in a foreign company. This notion finds support in existing literature, as Bell et al. (2008) show a negative relationship between Economic Freedom and IPO Underpricing. Their research used a sample of 105 foreign companies that underwent an IPO on American Capital Market.

Investment Freedom, however, is a somewhat narrower proxy. But, despite being so specialized, it reflects many vitally important factors that are aimed to reduce the uncertainty of investors. The capital market in The US is known for its short-term orientation (Laverty, 1996; Marginson & McAulay, 2008). This is not easily applied to the concept of Economic Freedom because economic freedom has to do more with the stimulation of a country’s growth long-term (J. D. Gwartney et al., 1999). Investment Freedom, at the same time, provides measures of institutional factors that are vitally important for investors both in the short-term in the long-term. This, in theory, makes it more influential for Foreign IPO Underpricing. Thus, the second hypothesis is formulated as follows:

Hypothesis 2: Investment Freedom is a better predictor of foreign IPO underpricing

than Economic Freedom.

One of the most upsetting aspects of COO is that companies often have very little control over which countries they come from and operate in. However, a company can compensate for environmental imperfections by signaling that the company is implementing

(22)

measures to overcome this imperfection to other actors (i.e., investors). There is a significant amount of research that used signalling theory to find different ways in which managers can reduce the information asymmetry (Bell, Filatotchev, et al., 2012; Bell et al., 2008; Filatotchev & Bishop, 2002; Francis et al., 2010; Moore et al., 2012).

In the case of Investment Freedom, Corporate Governance Factors can be treated as these signals. As it was mentioned above, scholars treat board independence as a very strong positive signal because firms with independent boards tend to be better governed (Aggarwal et al., 2007; Bell, Moore, et al., 2012). Independent directors are known to effectively monitor managers’ actions and prevent boards from making wrong and harmful decisions, as well as forcing boards take unpleasant actions, such as firing non-performing CEOs (Weisbach, 1988).

Board independence also means that a company is transparent and is ready to protect the interests of shareholders, which is especially important when a government is not protecting them (i.e. in countries with low Investment Freedom). In a recent study, Dahya et al. (2009) found a strong positive relationship between board independence and corporate value, especially in countries with weak investor protection. Therefore, it is proposed that companies originating from countries with low levels of Investment Freedom can partially compensate for LoF by increasing the number of independent directors in BD. Thus, combining signaling theory (Connelly, Certo, Ireland, & Reutzel, 2011; Spence, 1973) and agency theory (Eisenhardt, 1989; Jensen & Meckling, 1976) the following hypothesis is proposed:

Hypothesis 3: Board independence positively moderates the relationship between Investment

(23)

Bell et al. (2008) argue that increased geographic scope and operations in different countries signals that a company is well established in a global market. Furthermore, it also can be a proxy of a company’s experience (Delios & Henisz, 2003). In addition, as it was mentioned earlier, Geographic Scope is believed to bring a number of benefits to companies, such as increasing proprietary assets’ value appropriation levels (Goerzen & Beamish, 2003), improving organizational learning (Delios & Henisz, 2003; Kogut & Zander, 1993) , as well as supporting competition on an international level (Karnani & Wernerfelt, 1985).

Thus, high geographic scope should signal to investors that a company is well established and can compete and function on an international level. Moreover, large geographic scope can also help to deal with home-country market’s volatility and imperfections.

Positive firm-level signals (e.g., international experience) are very important for a successful foreign IPO because they compensate for the low-level development of home-country institutions by demonstrating that the company can operate successfully in more institutionally developed foreign countries (Bell et al., 2008). It is therefore hypothesized that Geographic Scope positively moderates the relationship between Investment Freedom and foreign IPO Underpricing:

Hypothesis 4: Geographic Scope positively moderates the relationship between Investment

Freedom and IPO Underpricing.

Figure 2.1 shows a framework that was developed in order to better illustrate the relationships stated earlier.

(24)

Figure 2.1 Conceptual Framework

H1: There is a negative relationship between home-country Investment Freedom and foreign

IPO Underpricing such that higher home-country Investment Freedom causes a lower rate of foreign IPO Underpricing.

H2: Investment Freedom is a better predictor of foreign IPO underpricing than Economic

Freedom.

H3: Board independence positively moderates the relationship between Investment Freedom

and IPO Underpricing.

H4: Geographic Scope positively moderates the relationship between Investment Freedom

(25)

3. Methodology

3.1 Data Sources

The dataset consists of foreign firms (non-American) that did an IPO listing on American stock exchanges (either NYSE or NASDAQ) between 1999 and 2012. In order to collect data about these companies, the Thomson One database was used. Data needed for computing independent variables (i.e., Total Score of Economic Freedom and Investment Freedom Score) was gathered through Heritage Index of Economic Freedom website (Miller et al., 2014).

Additional information, necessary for computing moderating variables (i.e., board independence and geographic scope) such as information about board composition and amount of assets abroad (placed not in a firm’s home-country) at the time of IPO, was collected through the new issue prospectus of each firm from the sample as well as from Thomson Reuters DataStream database. The submission of a new issues prospectus is mandatory for each company that is undergoing a procedure of Initial Public Offering on American stock exchanges. A prospectus of each firm was examined in order to check if the company was raising capital in The United States for the first time.

Following previous research, stock listings that were done because of mergers and acquisitions, spin-offs, and privatizations were excluded from the sample, as well as were units, warrants, and rights (Bell et al., 2008; Bell, Moore, et al., 2012; Kadiyala & Subrahmanyam, 2002). Countries were sorted based on the level of country market development in general, using the study written by Hoskisson et al. (2000), as a guide to understanding whether companies were from emerging markets or not.

(26)

3.2 Sample

The sample consists of 198 foreign (non-American) firms that underwent an IPO in The United States (either on the NYSE or the NASDAQ) from 1999 to 2012. Such a long time frame was chosen due to the fact that not so many foreign companies choose the American Capital Market for their Initial Public Offering. It can be partially explained by the implementation of Sarbanes-Oxley Act (SOX), as from 2006 foreign companies with market capitalization exceeding $75 million have to comply with it. Another reason is the financial crisis of 2007-2008, which sufficiently disrupted many companies’ plans of going public.

The initial sample consisted of 372 companies. However, after eliminating all IPOs that occurred as a result of mergers, acquisitions, or privatizations, as well as units, warrants, and rights, the sample was narrowed down to 198 companies. According to Hoskisson’s (2000) classification, 76% of the companies in the sample were from emerging markets. The distribution of the firms by their country of origin can be found in Appendix 1.

American capital market turned out to be a very attractive destination for Chinese companies, as 96 (48%) companies in the sample were from China. In order to ensure the robustness of the analysis, a second sample was constructed excluding Chinese companies, and the analysis was run again. Thus, the second (non-Chinese) sample consisted of 102 companies.

3.3 Dependent variable

Underpricing

This study has only one dependent variable: underpricing. Underpricing was chosen as a dependent variable due to its high explanatory power of IPO success in general, as indicated by research conducted by scholars in the financial field (Beatty & Ritter, 1986; Daily, Trevis Certo, Dalton, & Roengpitya, 2003; Filatotchev & Bishop, 2002). In line with previous

(27)

research (Bell et al., 2008; Pollock & Rindova, 2003), underpricing was calculated by the following formula:

Underpricing= (price first day – price initial)/ price initial

As can be seen from the formula, in order to calculate the level of Underpricing only two pieces of data are required: initial stock price and the price of a stock at the end of the first day of trading. When an outcome is greater than zero, a price of a stock is greater than the prospectus price, which means that following amount of money (in percent of an initial stock price) was lost by the company and gained by other actors.

The data needed to compute the level of underpricing, was gathered through multiple sources: Thomson One Database, Thomson Reuters DataStream, firms’ prospectuses as well as NASDAQ’s and NYSE’s websites.

3.4 Independent Variables

This thesis has two independent variables: Total Score of Economic Freedom and the Score of Investment freedom.

Economic freedom

The data concerning economic freedom was obtained through Heritage Foundation Index of Economic Freedom (Miller et al., 2014). This index is calculated by using ten different dimensions of freedom: Property Rights, Freedom from Corruption, Fiscal Freedom,

Government Spending, Business Freedom, Labor Freedom, Monetary Freedom, Trade Freedom, Investment Freedom and Financial Freedom.

Each dimension is measured on a scale from 0 to 100, where 0 stands for no freedom at all, and 100 stands for absolute freedom. After all data is obtained, these ten economic freedoms are averaged with every dimension having equal weight (Miller et al., 2014).

In order to simplify the research, Economic Freedom scores were transferred from the scales from 0 to 100 to the scales from 0 to 5, where 0 stood for no freedom at all, and 5

(28)

stood for complete freedom. Moreover, following previous research (Bell et al., 2008), the scores of Economic Freedom for the sample period were collected and the arithmetic means were calculated for each country in the sample.

Investment Freedom

Another independent variable used in this study is Investment Freedom. The data for the variable was collected through Heritage Index of Economic Freedom website (Miller et al., 2014).

The score of Investment Freedom is calculated by combining several factors (e.g. national treatment of foreign investment, transparency and burdensome of bureaucracy, transparency of investment laws, etc.).

As was said earlier, Investment Freedom is measured on a scale from 0 to 100, where 0 stands for no freedom at all, and 100 stands for complete freedom.

In order to ensure the consistency of research, Investment Freedom scores were transferred from a scale from 0 to 100 to a scale from 0 to 5, where 0 stood for no freedom at all and 5 stood for complete freedom. In addition, the scores for the period of the sample were collected and the arithmetic average was calculated for each country.

3.5 Moderating Variables

This study has two moderating variables- Geographic Scope and Board independence.

Geographic Scope

The data needed to calculate this variable was collected through Thomson Reuters DataStream and from offering prospectuses of each firm. In line with previous research, Geographic Scope was measured as a percentage of firm assets abroad at the time of IPO (Bell et al., 2008) by the following formula:

(29)

Board independence

Due to the fact that the majority of databases (e.g. WRDS, Thomson Reuters DataStream, Thomson One) do not have information about corporate governance practices of foreign companies, data for computing Board Independence variable was collected manually from the Management section of offering prospectuses of the firms from the sample. Board independence was measured as a percentage of independent (external) directors in a board at the time of IPO (Chahine & Filatotchev, 2008; Daily et al., 2003; Sanders & Boivie, 2004) by the following formula:

Board Independence= (number of independent directors/size of a board of directors)*100

3.6 Control Variables

Following the research in the field, the analysis was controlled for firm age, firm size and high-tech industry of a firm (Bell et al., 2008; Moore et al., 2012).

Firm Age

The data needed to measure the Firm Age variable was collected using Thomson Reuters DataStream database in combination with offering prospectuses of the firms from the sample. Firm age was measured as the difference between the date of IPO and the date of foundation of a company plus one year. Additionally, following previous research (Bell et al., 2008; Daily et al., 2003), in order to ensure normal distribution of the variable, a natural logarithm was taken, thus, the firm age variable is calculated by the following formula:

Firm Age= Ln(year of IPO-year of foundation of the company+1)

Firm Size

Following previous research, Firm Size was measured as the natural logarithm of revenues at the time of IPO (Bell et al., 2008).

(30)

The information about revenues was mainly collected in Thomson One database. Missing data was taken from the offering prospectuses of these firms.

High Tech Industry

As it is known, firms in High-Tech industries struggle more than other firms from high level of underpricing (Deeds et al., 2004) and due to the fact that scholars in the field find it important to control their research for industry effects (Bell et al., 2008; Daily, Certo, & Dalton, 2005) the analysis is controlled for industry effects.

High-Tech industry is coded as a dummy variable where 1 stands for companies that operate in high-tech industries, and 0 stands for companies that operate in low-technology industries. The data about industries was obtained through the Thomson One database.

3.7 Data Analysis

In order to test the proposed hypotheses, regression analysis was used. Regression analysis is a common technique used to predict an outcome variable from one predictor variable (independent variable) in case of simple regression and from several (independent variables) in case of multiple regression (Field, 2013). A multiple regression can be explained by the following general equation:

Where is an outcome variable, are the coefficients of the first independent variable , second independent variable , and n-th independent variable respectively. The symbol stands for the difference between the predicted and observed values of the dependent variable (Field, 2013). The model is designed to determine which linear combination of predictors (independent variables) can best predict the outcome (dependent variable).

(31)

Following previous research (Bell et al., 2008), hierarchical regressions were conducted. As Julie (2010) argues, hierarchical regression is used when a scholar is using previous research to justify the order in which predictors are entered into the model.

To test the first hypothesis, a hierarchical regression was run in order to test the effect of Investment Freedom on foreign IPO underpricing. In order to do that, control variables (Firm Age, Firm size, High-tech Industry) were entered into the model first, and then the Investment Freedom variable was added second (see Table 3.2). Hence, the regression equation was formulated as follows:

Table 3.1 Hierarchical regression on Investment Freedom

Model

To test the second hypothesis a hierarchical regression on economic freedom was conducted. In the first model (see table 3.1) the effect of Economic Freedom on Underpricing was tested. In doing so, control variables (Firm Age, Firm size, High-tech Industry) were entered into the model first and then the Economic Freedom variable was added second. Hence, the regression equation was formulated as follows:

(32)

After the regression was run, the R2 _{of regressions on Economic Freedom and Investment}

Freedom were compared in order to determine which of the variables had a bigger explanatory power and to determine whether Hypothesis 2 was supported or not.

Table3.2 Hierarchical regression on Economic Freedom

Model

In order to test remaining hypotheses, two additional hierarchical regressions were conducted. The moderating effect of the degree of internationalization was measured by running a hierarchical regression with four models. In the first model, control variables were added (Firm Age, Firm size, High-tech Industry); in the second model, the independent variable of Investment Freedom was entered into the model; in the third model, the moderating variable Geographic scope was added. Following previous research (Bell et al., 2008) a computed variable equal to Geographic scope score multiplied by Investment Freedom (IF x Geographic Scope) was entered in the last model (see Table 3.3). Hence, the regression equation for this model was formulated as follows:

(33)

Table 3.3 Hierarchical Regression on Geographic Scope

Model

Similar to the way the moderating effect of Geographic Scope on the relationship between Investment Freedom and IPO underpricing was measured, the moderating effect of Board Independence was measured by a hierarchical regression consisting of four models. In model 1, control variables (Firm Age, Firm size, High-tech Industry) were entered; in model 2 the independent variable Investment freedom was added; in model 3 Board Independence was entered. Finally, in line with research conducted by Bell (2008), a computed variable ( Investment Freedom X Board Independence) was entered in model 4 (see table 3.4).

(34)

Table 3.4 Hierarchical Regression on Board Independence

(35)

4. Results and analysis

This chapter covers information about statistical analysis and results. First, descriptive analysis are executed, including information about main variables descriptive statistics, such as the mean, standard deviation, minimum and maximum values. Second, Variation Inflating Factors (VIFs) are exhibited. Third, a correlation matrix is presented along with a discussion of main correlation coefficients. Finally, the hierarchical regression analyses are presented.

4.1 Descriptive Analysis

The level of economic freedom varied from 2.6 (companies from the Russian Federation and companies from India) to 4.5 (companies from Hong Kong). Investment Freedom ranged between 1.6 (firms from China) and 4.5 (firms from Hong Kong) (see Table 4.1). Therefore, it can be concluded that Hong Kong is truly one of the most economically free countries in the world.

A total of 118 companies from the sample (61%) operate in low-technology industries and 80companies (49%) operate in high-technology industries.

Levels of board independence also varied significantly, with some companies having no independent (external) directors on their board of directors when their offering prospectuses were written, and with other companies having board independence levels of 78%.

Geographic scope, which is measured as the percentage of a company’s assets abroad, ranged from 0% to 83%. Such a high degree of economic scope can be explained by the fact that some companies were focused on foreign markets from the very beginning.

(36)

Table 4.1 Descriptive Statistics (Main Sample)

Variable Observations Mean SD Min Max

HighTech 198 0.40 0.49 0 1 Ln(Revenues) 198 11.66 1.62 7.90 15.11 Ln(Age) 198 2.02 0.60 0.69 3.37 Economic Freedom 198 3.11 0.52 2.6 4.5 Inv. Freedom 198 2.59 1.13 1.6 4.5 Geographic Scope 198 0.20 0.30 0.00 .83 Board Ind. 198 0.38 0.18 0.00 .78 IFxGeo.Scope 198 0.67 1.02 0.00 3.74 IF.xBoard.Ind. 198 1.18 0.60 0.00 3.43 Underpricing 198 0.23 0.21 -0.19 0.98

(37)

Table 4.2 Descriptive Statistics (non-Chinese sample)

Variable Observations Mean SD Min Max

HighTech 102 0.37 0.49 0 1 Ln(Revenues) 102 11.82 1.91 7.90 15.11 Ln(Age) 102 2.08 0.68 0.69 3.37 Economic Freedom 102 3.48 0.48 2.6 4.5 Inv. Freedom 102 3.47 0.87 1.6 4.5 Geographic Scope 102 0.31 0.32 0.00 0.83 Board Ind. 102 0.38 0.19 0.00 0.78 IFxGeo.Scope 102 1.11 1.16 0.00 3.74 IF.xBoard.Ind. 102 1.30 0.71 0.00 3.43 Underpricing 102 0.16 0.19 -0.19 0.98

4.2 Variation Inflating Factors test

In order to check the sample for reliability, multicollinearity tests were conducted for each regression individually. As a part of this process, Variation Inflating Factors (VIFs) were calculated because VIF scores show whether the correlation between variables harms the statistical results of multiple regression (Field, 2013). Another way to check for multicollinearity is to calculate tolerance: if the value is close to 0, it means that there is no problem with multicollinearity; if it is close to 1, it means that there is multicollinearity between some variables ’ . The basic rule of VIF scores is that, if the VIF value is greater than 10, there are some reasons for concern (Myers, 1990).

(38)

The VIF test showed no multicollinearity for first two regressions (see Table 4.3, Table 4.4). However, the regression on Geographic Scope VIF test (see Table 4.5) showed high VIF and low tolerance values for Geographic scope (VIF- 45.936, Tolerance- 0.022) and IFxGeographic Scope (VIF- 50.258, Tolerance- 0.020). The high multicollinearity of these variables can be explained by the nature of IFxGeographic Scope, as it is a product of Investment Freedom and Geographic scope. Thus, high multicollinearity is predictable and does not compromise the integrity of the analysis. A similar pattern can be seen for the regression on Board Independence (see table 4.6). The VIF test revealed low tolerance and high VIF scores for Board Independence (VIF- 36.809, Tolerance- 0.027) and IFxBoard Independence (VIF- 42.229, Tolerance- 0.024). The high multicollinearity of these values can also be explained by the nature of IFxBoard Independence variable, as it is computed by multiplying Investment Freedom and Board Independence. Therefore, the high multicollinearity of these variables is logical and is not a threat to the reliability of the analysis.

The VIF test for Non-Chinese sample showed the same results. No multicollinearity was found for the first two regressions (See Table 4.7, Table 4.8). At the same time, as it was found in the main sample, the VIF test found high multicollinearity and low tolerance for regressions on both Geographic Scope (See table 4.9) and Board independence (See Table 4.10): Geographic Scope (VIF- 62.088, Tolerance- 0.016), IFxBoard Independence (VIF- 64.117, Tolerance- 0.016), Board Independence (VIF- 53.566, Tolerance- 0.019), IFxBoard Independence (VIF- 61.153, Tolerance- 0.016). As variables from the non-Chinese sample were calculated in a manner following the main sample, similar rationale can be employed in order to explain such high VIF scores.

(39)

Table 4.3 Collinearity Statistics on Economic Freedom (Main Sample)

Variable

Standardized

Beta t Tolerance VIF

High Tech Industry 0.254 3.930*** 0.978 1.023 Ln(Revenues) -0.014 -0.217 0.926 1.080 Ln(Age) -0.022 -0.326 0.937 1.067 Economic Freedom -0.376 -5.868*** 0.994 1.006

Dependent Variable: Underpricing * Significant at the 0.1 level ** Significant at the 0.05 level *** Significant at the 0.01 level

Table 4.4 Collinearity Statistics on Investment Freedom (Main Sample)

Variable

Standardized

High Tech Industry 0.225 3.511*** 0.972 1.029

Ln(Revenues) -0.024 -0.361 0.927 1.079

Ln(Age) -0.064 -0.978** 0.939 1.065

Investment

Freedom -0.399 -6.286*** 0.992 1.008

Dependent Variable: Underpricing * significant at the 0.1 level ** significant at the 0.05 level *** significant at the 0.01 level

Table 4.5 Collinearity Statistics on Geographic Scope (Main Sample)

Variable Standardized Beta t Tolerance VIF

Ln(Revenues) -0.019 -0.294 0.928 1.078 Ln(Age) -0.040 -0.616 0.932 1.073 Investment Freedom -0.511 -6.150*** 0.559 1.788 Geographic Scope -0.592 -1.404* 0.022 45.936 IF x Geographic Scope 0.630 1.430* 0.020 50.258

(40)

Table 4.6 Collinearity Statistics on Board Independence (Main Sample)

Variable

Standardized

Ln(Revenues) -0.030 -0.452 0.901 1.110 Ln(Age) -0.034 -0.514 0.890 1.124 Investment Freedom -0.276 -1.813* 0.172 5.830 Board Independence 0.450 1.176 0.027 36.809 IF x Board Independence -0.483 -1.179 0.024 42.229

Table 4.7 Collinearity Statistics on Economic Freedom (Non-Chinese Sample)

Variable

Standardized

Ln(Revenues) -0.051 -0.520 0.887 1.128

Ln(Age) -0.044 -0.460 0.903 1.107

Economic Freedom -0.254 -2.738*** 0.984 1.017

Table 4.8 Collinearity Statistics on Investment Freedom (Non-Chinese sample)

Variable Standardized _Beta t Tolerance _VIF

Ln(Revenues) -0.058 -0.612 0.886 1.129

Ln(Age) -0.105 -1.099 0.877 1.140

Investment Freedom -0.333 -3.623*** 0.949 1.054 Dependent Variable: Underpricing

* significant at the 0.1 level ** significant at the 0.05 level *** significant at the 0.01 level

(41)

Table 4.9 Collinearity Statistics on Geographic Scope (Non-Chinese Sample)

Variable

Standardized

Ln(Revenues) -0.068 -0.707 0.862 1.161

Ln(Age) -0.068 -0.716 0.868 1.152

Investment Freedom -0.474 -3.912*** 0.538 1.860

Geographic Scope -0.518 -0.740 0.016 62.088

IF x Geographic Scope 0.608 0.854 0.016 64.117

Table 4.10 Collinearity Statistics on Board Independence (non-Chinese sample)

Variable

Standardized

High Tech Industry 0.196 2.132** 0.924 1.082

Ln(Revenues) -0.083 -0.870 0.858 1.166 Ln(Age) -0.071 -0.717 0.811 1.233 Investment Freedom -0.057 -0.284 0.197 5.073 Board Independence 1.297 2.002** 0.019 53.566 IF x Board Independence -1.385 -2.001** 0.016 61.153

4.2 Correlations

Following the guidelines of statistical research, the levels of correlation between variables were calculated. Table 4.11 summarizes data concerning the correlation of variables for the main sample; a correlation matrix for the non-Chinese sample is summarized in table 4.12. The results of the correlation analysis did not find any threats to the robustness of the research.

(42)

Economic Freedom was found to be negatively correlated with the dependent variable, Underpricing (-0.383, p < 0.01). The correlation analysis of the non-Chinese sample shows similar results, with a significant negative correlation between Economic Freedom and Underpricing (-0.217, p < 0.05). Another independent variable, Investment Freedom, was found to be negatively correlated with Underpricing, (-0.414, p < 0.01). Similar to the main sample, there was a significant negative correlation for the non-Chinese sample (-0.319, p < 0.01). This indicates that, with increasing levels of Investment Freedom, Underpricing diminishes, which goes in line with hypothesis 1.

Economic Freedom and Investment freedom are also highly correlated between each other, r=0.891 (p < 0.01) for the main sample, and r=0.731 (p < 0.01) for the non-Chinese sample. This is explained by the fact that Economic Freedom is calculated by taking an average of 10 different dimensions of freedom, and Investment Freedom is one of them.

Significant levels of IFxGeo Scope’s correlation with Investment Freedom, r=0.515 (p < 0.01) for the main sample and r=0.305 (p < 0.01) for the non-Chinese sample, and with Geographic Scope, r=0.985 (p< 0.01) for the main sample and r=0.986 (p< 0.01) for the non-Chinese sample, also have a logical explanation. This is because IFxGeo Scope is computed by multiplying Investment Freedom by Geographic scope.

High levels of IFxBoard Ind’s correlation with Investment Freedom, r=0.345 (p < 0.01) for the main sample and r=0.339 (p < 0.01) for the non-Chinese sample, and with Board Independence, r=0.927 (p < 0.01) for the main sample and r=0.956 (p < 0.01) for the non-Chinese sample, have same nature, as this variable was also calculated by multiplying two variables- Investment Freedom and Board Independence.

Significant level of correlation between Underpricing and High-Tech dummy variable, r=0.265 (p < 0.01) for the main sample and r=0.298 (p < 0.01) for the non-Chinese

(43)

sample, goes in line with previous research (Bell et al., 2008; Daily et al., 2005), as high-tech firms are considered to be more underpriced.

High levels of negative correlation between underpricing and IFxGeo Scope (-0.198, p < 0.01), and between underpricing and IFxBoard Ind. (-0.314 p < 0.01) mean that both Geographic Scope and Board Independence Level can reduce underpricing, which is in line with hypotheses 3 and 4.

(44)

Table 4.11 Descriptive Statistics and Correlation Matrix (Main Sample)

Variable Mean SD 1 2 3 4 5 6 7 8 9

1. High Tech industry 0.40 0.49

2. Ln(Revenues) 11.66 1.62 -0.143* 3. Ln(Age) 2.02 0.60 -0.074 0.239** 4. Investment Freedom 2.59 1.13 -0.079 0.004 -0.038 5. Economic Freedom 3.11 0.52 -0.019 0.045 0.072 0.891** 6. Geographic scope 0.20 0.30 0.045 0.023 0.081 0.380** 0.453** 7. Board Independence 0.38 0.18 -0.054 0.136 -0.124 0.005 -0.008 -0.019 8. IFxGeo Scope 0.66 1.06 0.024 0.020 0.077 0.515** 0.468** 0.985** -0.013 9. IFxBoard Ind. 0.98 0.67 -0.099 0.135 -0.146* 0.345** 0.296** 0.114 0.927** 0.152* 10. Underpricing 0.23 0.21 0.265** -0.073 -0.071 -0.414** -0.383** -0.159* -0.010 -0.198** -0.314** * Correlation is significant at the 0.05 level (2-tailed).

(45)

Table 4.12 Descriptive Statistics and Correlation matrix (Non-Chinese Sample)

Variable Mean SD 1 2 3 4 5 6 7 8 9

1. High Tech industry 0.37 0.49

2. Ln(Revenues) 11.83 1.90 -0.114 3. Ln(Age) 2.08 0.68 -0.027 0.311** 4. Investment Freedom 3.48 0.48 0.098 -0.092 -0.203* 5. Economic Freedom 3.47 0.87 -0.056 -0.134 0.069 0.731** 6. Geographic scope 0.31 0.32 0.121 -0.105 -0.023 0.132 0.246* 7. Board Independence 0.38 0.19 -0.097 0.123 -0.171 0.082 0.078 -0.038 8. IFxGeo Scope 1.11 1.16 0.088 -0.097 0.051 0.305** 0.239* 0.986** -0.016 9. IFxBoard Ind. 1.31 0.71 -0.135 0.084 -0.232* 0.339** 0.265** 0.009 0.956** 0.059 10. Underpricing 0.16 0.19 0.298** -0.077 -0.063 -0.319** -0.217* 0.050 -0.049 0.008 -0.182 * Correlation is significant at the 0.05 level (2-tailed).

Country of origin effects and foreign IPO success : investigating the relationship between investment freedom and foreign IPO underpricing