The Effects of Foreign Cross-listing on Firm
Performance – Evidence from Latin America
By
Leendert Schipper
University of Groningen
Faculty of Economics and Business
2 ABSTRACT
Due to financial globalization firms are able to enter foreign markets easier; therefore it is also easier for firms to list their shares on a foreign stock exchange. In this thesis the relationship between foreign cross-listing and firm performance is investigated. A comparison has been made between cross-listed firms from Latin America which are cross-listed in the U.S. and firms from Latin America which are not cross-listed. The countries used in this thesis are Brazil, Argentina and Mexico.
The literature suggests and the expectations were that there is a positive effect between foreign cross-listing and firm performance due to an increase of the corporate governance. In this thesis 1 hypothesis was supported, 2 hypotheses were not supported, and the results from the ad hoc analyses did not show a significant difference. Therefore it can be concluded that there was not enough evidence to support the theoretical assumption that foreign cross-listing has a positive effect on firm performance measured in the ROE, ROA and Tobin’s q. for all three countries; Brazil, Argentina and Mexico.
An explanation might be the time-frame used in thesis; the process to improve a corporate governance system might take longer than 1 year. Or the improvements with regard to the corporate governance system are made before cross-listing; consequently the influence is less than expected after cross-listing.
By looking at the ad hoc analyses the results do show differences between the three countries. The cross-listed firms from Mexico do have a higher mean increase for all the three
performance indicators, this might be due to differences between the corporate governance of each country, but still no significant results were found. By looking at the cross-listed firms from Brazil a decrease of ROE was found an explanation for this might be because of the sample size used in thesis might have been too small.
There is no supporting evidence for the theoretical assumption that foreign cross-listing has a positive effect on firm performance. It might be that the influence of better corporate
3
Overall it can be concluded that there is not enough evidence to support the theoretical assumption that cross-listing in the U.S. has a positive effect on firm performance for firms from Brazil, Argentina and Mexico.
Key words: Corporate governance, foreign cross-listing, firm performance
Research theme: The effects of foreign cross-listing on firm performance
4 Table of content:
1. Introduction & problem statement: 6
1.2 Research questions 7
2. Literature review
Corporate governance:
2.1 Definition 9
2.1.1 Aspects/components of corporate governance system 10 2.1.2 Benefits of good corporate governance 10
2.1.2 Ownership concentration 11
2.2 Differences between the Anglo-Saxon corporate governance system and the Latin American corporate
governance system 12
Foreign cross-listing:
2.3 What is foreign cross-listing? 16
2.3.1 Benefits of cross-.listing in U.S. 16
2.3.2 The bonding hypothesis 17
Firm performance:
2.4 Effects of foreign cross-listing on firm performance: 18
2.5 Conclusion literature review 20
3. Hypotheses 21
4. Research design
4.1 Research method 22
4.1.1 Measuring firm performance 22
4.1.2 Measuring corporate governance 23
4.2 Data collection 24
4.2.1 Sampling strategy 25
5 5. Results and discussion
5.1 Descriptive results 29
5.1.1 Cross-listed firms (CL-group) 29 5.1.2 Not cross-listed firms (Non CL-group) 30 5.1.3 Comparison between firms from Brazil
(CL-group – Non CL-group) 31
5.1.4 Comparison between firms from Argentina
(CL-group – Non CL-group) 32
5.1.5 Comparison between firms from Mexico
(CL-group – Non CL-group) 33
5.1.6 Overview of mean changes in performance indicators
(CL-group – Non CL-group) 34
5.1.7 Overview of mean changes in performance indicators
(Countries specific) 35
5.2 Hypotheses testing 36
5.2.1 Comparing return on equity (ROE) 37 5.2.2 Comparing return on assets (ROA) 38
5.2.3 Comparing Tobin’s q 38
5.3 Ad Hoc Analyses 39
5.3.1 Comparing CL-group and Non CL group of Brazil 39 5.3.2 Comparing CL-group and Non CL- group of Argentina 40 5.3.3 Comparing CL-group and Non CL-group of Mexico 41 5.3.4 Summary of hypothesis tests 41
5.4 Summary of the results 42
6. Conclusions & discussion
6
1. Introduction & problem statement
Globalization is a process which involves crossing or transcending national borders and boundaries, and is a process which has been investigated by researchers from different fields (Staples, 2008). Central players in the process of globalization are multinational corporations (MNC’s). The process of globalization made it easier for MNCs to operate abroad and more opportunities came forward, in addition national markets are getting more integrated in the world economy. Due to the process of globalization MNCs are operating in more countries and in different cultures and markets. Consequently, MNCs have to deal with different markets and cultures and MNCs becoming more complex.
Because of the complexity of MNCs, researchers have mentioned the importance of a good corporate governance system. Due to the complexity, the theory of corporate governance has expanded and become increasingly more complex. Corporate governance can take a wide variety of aspects, and can be defined in different ways (Erkens, Hung & Matos, 2009).
Foreign cross-listing is listing the shares of a firm on a foreign stock exchange. The world economy is getting more integrated and therefore is it easier for MNCs to cross-list in foreign countries. Foreign cross-listing is used to break away from the home country corporate governance system and to improve the corporate governance system
(Oxelheim & Randøy, 2001). Research has been done on the relationship between corporate governance and firms’ market value or performance. There is much more limited work that assesses the relationship between foreign cross-listing, and firms’ market value or
performance. Research that has been done on this topic focussed mainly on developed countries and not on developing countries (Erkens et al, 2009).
The listing of corporate stock on foreign stock exchanges have been one of the most fascinating aspects of financial globalization in the last two decades (Wójcik, Clark, and Bauer, 2004; Pagano, Roell and Zechner, 2002). Cross-listing for firms with growth
7
operating in countries with poor investor rights are more driven to cross-list in the U.S. These positive aspects might influence the performance of a firm in a positive way (Pagano et al, 2002; Doidge, Karolyi, and Stulz, 2004).
This thesis will be focusing on the relationship between foreign cross-listing and firm’s performance by looking at developing countries, mainly Latin America and investigate whether foreign cross-listing does have a positive effect on the performance of the MNCs. Therefore the following central research question has been formulated:
Does cross-listing on a foreign stock exchange have a positive effect on firm performance for firms from Latin America?
1.2 Research questions
In order to answer the central research question and to structure the thesis several research questions are formulated:
Corporate governance:
1. What is a corporate governance system? – Definition
2. What are the aspects or components of a corporate governance system?
3. What are the differences between the Anglo-Saxon corporate governance system and the Latin American corporate governance system?
Listing on a foreign stock exchange: 4. What is (foreign) cross-listing?
5. What are the benefits of listing on a stock exchange in the U.S.? – NYSE or NASDAQ
Firm performance:
6. What is the relationship between foreign cross-listing and firm performance?
The relationship between foreign cross-listing and firm performance has been an interesting topic in the academic literature, but still the existing literature is mainly focussing on
8
insight will be given on what the relationship is between foreign cross-listing and firm performance, in Latin America.
The thesis is structured as followed; the relevant literature is discussed in chapter 2, the literature review will first discus the different corporate governance systems and analyses the differences and similarities between them. The second part of the literature review discusses foreign cross-listing and analyses the effects of foreign cross-listing on firm performance. The hypotheses are deduced from the literature review and are discussed in chapter 3. The
9
2. Literature Review
The literature review will start to answer the research questions regarding corporate
governance and will follow with the research question regarding foreign cross-listing and firm performance, in the end the relationship between foreign cross-listing and firm performance is discussed.
2.1 Definition of corporate governance
The world economy is becoming more integrated. Corporate governance has become an interesting topic in an increasingly global economy. According to Carpenter & Sanders (2007) corporate governance generally addresses the distribution of rights and responsibilities among different participants in the organization, such as the board, managers, shareholder, and other stakeholders, and spell out the rules and procedures for making decisions on corporate affairs. More specifically they define corporate governance as the system by which owners of firm direct and control the affairs of the firm (Carpenter & Sanders, 2007). According to the Center of International Private Enterprises (CIPE) it is important to put in place good corporate governance, this is particular important for developing countries to remain competitive, to attract capital, ensure sustainability and to combat corruption (www.cipe.org).
2.1.1 Aspects/components of corporate governance system
10 Internal aspects;
The internal aspects refer to the internal control by defining the relationship between
managers, shareholders, board of directors and other stakeholders. Moreover it also entails the promotion of transparency within the company.
External aspects;
The external aspects refer to governmental laws and provisions; disclosure requirements, accounting standards and investor’s protection laws. Moreover other external aspects can be seen in regulations imposed by other institutions, such as listing requirements of stock exchanges.
2.1.2 Benefits of good corporate governance
It is important to have a good/ developed corporate governance system especially with regard to the external aspects of corporate governance. Oxelheim & Randøy (2001) for instance argue that differences between countries with regard to governmental laws and enforcement have resulted in a substantial variation in performance and financial development. This refers to the external aspects of corporate governance. According to CIPE good corporate
governance entails (www.cipe.org):
Shareholder select directors who represent them;
Directors vote on key matters and adopt the majority decision;
Decisions are made in a transparent manner so that shareholders and others can hold directors accountable;
The company adopts accounting standards to generate the information necessary for directors, investors and other stakeholders to make decisions; and
The company’s policies and practices should be applicable to local laws. Benefits of good corporate governance (Doidge et al,. 2007; www.cipe.org):
Improves access to capital and financial markets and attracts greater investment; Creates a system of internal controls that stimulates greater accountability and
transparency in dealing with investors and creditors;
11
Protects minority shareholders and their investments; Minimizes the rate of corruption;
Makes sure a smooth inter-generational transfer of wealth and divestment of family assets, as well as reducing conflicts of interest.
It is important and necessary to have an effective external system of institutions and markets. Otherwise the government and institutions, such as stock exchanges cannot set standards and promote efficiency if there is not an effective external system. Moreover the government and stock exchanges do encourage the flow of information and punish violations which can only be done if there is a good and effective external system. This is in line with Doidge et al (2007) they argue that better corporate governance gives firms the opportunity to access capital on better terms.
Transparency of a MNCs, influence directly to the stock market’s ability to observe a firms performance. When MNCs is not transparent, and has no effective disclosure of financial performance, investors are not able to evaluate the MNCs past performance, and consequently investors cannot forecast the corporation’s future cash flow and this (could) result in a lower share price (Gilson, 2000). When that happens, investors are less willing to invest in that company and the company will encounter problems when attracting capital.
In the absence of effective financial disclosure, a country’s capacity to support stock markets is compromised. It can be said that good corporate governance does require the capacity to make credible disclosure of financial results and that stock investments also requires good corporate governance. Cross-listing in the U.S. can be used to improve the corporate
governance system and break away from the home corporate governance system (Oxelheim & Randøy, 2001).
2.1.3 Ownership concentration
Effective corporate governance also requires a second form of transparency, ownership transparency (Gilson, 2000). It is important that firms disclose the identity of shareholders who own a significant amount of shares. Shareholders can suffer from poor corporate
12
more, have consequently much more voting power, compared to shareholders who have fewer shares. Therefore it should be publicly disclosed as it is imposed by the Securities Exchange Act (SEC) when having more than 5% of the shares. This is to minimize the change that a controlling shareholder takes opportunities to itself and takes private benefits from the
company at the expenses of the other less powerful shareholders (Carpenter & Sanders, 2007).
The concept of corporate governance is widely used, although different models of corporate governance can be distinguished. Corporate governance differs between countries and is organized in different ways (Ooghe & Vuyst, 2001). In this thesis two main corporate governance system are distinguished: The Latin American corporate governance system and the Anglo-Saxon corporate governance system. The Latin American corporate governance system is similar the stakeholders system. The Anglo-Saxon corporate governance system is known as the shareholders model (Ooghe & Vuyst, 2001).
2.2 Difference and similarities between the Anglo-Saxon corporate governance system and the Latin American corporate governance system
This section will discuss the Anglo-Saxon corporate governance system and the Latin American corporate governance system and will outline the differences and similarities between them.
13
In the paper of Cernat (2004) both corporate governance systems are described and two categories are used to describe the complexity of corporate governance. The two categories used are: capital-related aspects and labour-related aspects of corporate governance.
The capital-related aspects mainly refer to: ownership structure, corporate voting, and the role of institutional owners and the identity of owners and accounting standards and disclosure requirements. The labour-related aspects contain the stakeholder’s position of labour in corporate governance.
Labour-related aspects
In Anglo-Saxon countries, such as the U.S. trade unions are active, but their influence is considered to be low. By looking at how social partners co-operate with each other a distinction can be made. The co-operation between social partners in the U.S is minimal, consequently the influence of trade unions in the U.S. is low, compared to trade unions in Latin America were the influence is high, even though trade unions and labour unions are emerging in Latin America, which will give them more power and influence (Cernat, 2004). The unions, in Latin America, are strong and are getting more centralized; moreover the unionization in the U.S. is low and even declining whether it is increasing in Latin America. The biggest firms in the U.S. have delegated more decisions to the workforce, by using team decision making and employee involvement programs. This participation is limited to operation management and has no equivalent at strategic management level (Cernat, 2004; Cook, 2002).
14 Capital related aspects
Regarding the capital-related aspects of corporate governance, Cernat (2004) mentions that Anglo-Saxon countries are known to offer good developed mechanisms, such as ownership structure, corporate voting and/ or accounting standards. Ownership concentration in Anglo-Saxon countries is weak, which means that there are no major shareholders who have significant amount of shares. The few large shareholders are not significant bigger than the smaller shareholders. Because of that effective control of the management can only be
accomplished through coalitions and not by individual shareholders. Ownership concentration in the U.S. is low and individual shareholders influence on management is therefore weak. When a country has a high level of dispersed ownership it is important that minority
shareholders are well protected, which is the case in Anglo-Saxon countries such as the U.S. The influence of a coalition of shareholders is high on management and therefore the focus is mainly short-term. This is different in Latin American countries where ownership
concentration is high and therefore is the focus not short-term but long-term. When a
company has a major shareholder the interest of that particular shareholder is focussed mainly on long-term instead of a short-term.
According to Coffee (1999) and Doidge et al (2007) developing countries have a less developed corporate governance system compared to developed countries, especially on the part of the external aspects of corporate governance, such as listing requirements and
accounting standards. To stimulate the process of improving the corporate governance system in Latin America, the Latin America company circle was established. The members of the company circle are business leaders from Latin America and share experiences to improve the Latin American corporate governance system (www.oecd.org).
Latin American Companies circle
The companies circle in Latin America was introduced in 2005. The members of the
companies circle are firms whose shares are listed on local and international stock exchanges and who have a significant present in their capital markets. The Companies Circle was
15
progress, at both company and policy level. The progress made, entails regulatory and legal reforms, and the development in regard to country codes of best practice as well as heightened attention to corporate governance by many market participants. The Companies Circle
members have published a guide of good corporate governance (www.oecd.org).
The circle members believe strongly in good corporate governance and according to them it contributes substantially to the success of their firms and that good corporate governance is essential for good business prospects in the long term. Moreover, the Companies Circle members who have invested more in corporate governance policies and practises do have better operational and markets results compared to firms in Latin American who did not. It can be seen in higher levels of profitability, relative share price and liquidity and a reduced cost of capital. By looking at the financial crisis now, the Company Circle members mention that good corporate governance also helps firms to weather the severe consequences of an economic downturn with more equilibrium, as the results of the current global financial crisis show (www.oecd.org).
This is in line with Murphy & Topyan (2005) they found a positive relationship between good corporate governance and firm performance. Politicians and the business leaders, in Latin America, have acknowledged that changes need to be made in order to improve the corporate governance system (Aquilare, 2008). According to Aquilare (2008) business leaders and politicians mentioned 5 important reasons why to improve the corporate governance system:
Increase of corporate transparency requirements Strengthen the legal system
Prevent shareholder expropriation by developing effective minority shareholder protection
Provide adequate access to investment resources for small and medium-sized firms Allow more activism from institutional investor who are likely to demand higher
corporate governance standards
16
MNCs are therefore able to improve their corporate governance, and increase firm performance. According to Stulz (2009) the reward for switching to the Anglo-Saxon
corporate governance system from a less demanding system, in this case the Latin American system, appears in the shape of a lower cost of capital and a higher firm value.
Table 2.1
Overview of the differences and similarities between the Anglo-Saxon corporate governance system and the Latin American corporate governance system
Anglo-Saxon corporate governance system
Latin American corporate governance system Labour related aspects
Low Medium (increasing) Low Medium (increasing) Low Medium (increasing)
Low High High Low
High (short-term) Low (long-term) High Low
High Low High Low Trade unions (activity)
Labour organizations Co-determination
Capital related aspects Ownership concentration Influence of shareholders Financial orientation Financial disclosure Accounting standards Listing requirements 2.3 Foreign Cross-listing
This section will discuss and explain foreign cross-listing and gives an overview of the benefits of cross-listing in the U.S.
Firms that want to cross-list usually choose a country that is close to the home country and has similarities and historical ties. For example many of the European firms on U.S. exchanges are British and most cross-listed firms in Vienna are German. Abdallah & Ioannidis (2010) argue that factors such as common language, colonial ties, trade
17 2.3.1 Benefits of cross-listing in the U.S.
Risk premium reduction
When firms do make use of cross-listing in the U.S. the risk premium will reduce. Due to cross-listing there will be a greater risk-sharing which will reduce the risk premium investors require when they hold the shares of a firm (Doidge et al., 2004).
Access to more developed capital markets
Due to reduce of the risk premium, firms will be able to access capital easier and on better terms.
Better information disclosure/ better accounting standards
Firms that do cross-list in the U.S. have to do that according the U.S. law, and have to improve their standards, mainly standards regarding financial disclosure and accounting.
Bonding and monitoring
Firms that cross-list in the U.S have bond themselves to the U.S. law and stock exchanges. Consequently shareholders are able to monitor the company better. Bonding to the U.S. law and stock exchanges is a signal and a willingness to improve the corporate governance and therefore shareholders a more interested to invest in that company (Doidge et al., 2004; Leuz, 2003; Pagano et al., 2002).
18 2.3.2 The bonding hypothesis
The bonding hypothesis implies two main aspects. The bonding hypothesis implies that firms that are cross-listed on a U.S. stock exchange do have better corporate governance than firms that are not cross-listed on a U.S. stock exchange. Secondly the bonding hypothesis implies that the difference in corporate governance between cross-listed firms will be the greatest in countries with weak protection for investors and have a less developed financial system (Charitou et al., 2007; Coffee, 1999).
The literature suggests that firms from developing countries benefit more from cross-listing in the US than firms from developed countries. The reason for this mainly lies in the bonding hypothesis; firms from developing countries benefit more from the tighter control and regulations in the U.S. market (Coffee, 1999 & 2002; Stulz 1999). This is in line with Charitou et al. (2007); and Lel & Miller (2008) who have found evidence to support this theory.
2.4 Effects of foreign cross-listing on firm performance
The literature suggests that firms that choose to list their stocks in foreign markets realize substantive valuation gains. Such gains are generally attributed to a reduction of the risk premium associated with the disadvantaged features of the particular home stock market (Abdallah & Ioannidis, 2010). Abdallah & Ioannidis (2010) mention that early research has shown that capital market segmentation is a strong motive for cross-listing, segmented markets do have a high cost of capital. Therefore it could be beneficial for developing countries to cross-list in the U.S, which will reduce the cost of capital (Coffee, 1999; Pagano et al., 2002).
19
and disclosure requirements applied to all U.S. firms, and are subject to supervision and enforcement by the Securities and Exchange Commission (SEC). Earlier research mentions that the level of investor protection in the U.S. is higher than in developing countries either because of stricter supervision by the SEC or greater scrutiny by reputational financial analysts (Coffee, 2002).
When a MNCs cross-list in the U.S. and is originally from a country with weaker shareholder protection, it has to apply the new regulations and listing requirements and the bonding hypothesis is applicable. Moreover cross-listing in the U.S. can also be used to signal the shareholders and investors willingness to change (Cernat, 2004; Ooghe & Vuyst, 2001; Oxelheim & Randøy, 2001).
Doidge et al (2004) and Pagano et al (2002) have found evidence that cross-listing in the U.S. will result in better financial disclosure and better accounting standards. Consequently the reputation of the firm increases and it is easier to monitor a company’s management which will result in a decrease of agency costs. This is in line with Leuz’s (2009) study, which provides evidence that firms which are cross-listed in the United States have better information intermediation by financial analysts, and therefore they have a better market analyst and following forecast accuracy. This all together will result in better firm
20 2.5 Conclusions literature review
In this section the conclusions from the literature review are formulated. From the conclusion the hypotheses are derived. It can be concluded that the Anglo-Saxon corporate governance system is different in several aspects from the Latin American corporate governance system. Moreover is can be concluded that the Latin American corporate governance system is improving but still changes need to be made.
The most important difference between both the corporate governance systems lies in the external aspects of corporate governance. The difference is that the Latin American corporate governance system needs to change mainly in regard to the external aspects. The Anglo-Saxon corporate governance is developed and offers better mechanisms compared to the Latin American corporate governance.
Some of these external aspects refer to governmental laws and provisions; disclosure
requirements, accounting standards and investor’s protection laws and regulations imposed by other institutions, such as listing requirements of stock exchanges. Firms in Latin America can use these external aspects in the U.S. to improve the corporate governance system, by
conducting in a cross-listing in the U.S.
21 3. Hypotheses
In this chapter hypotheses will be formulated which are deduced from the literature review. Firms from developing countries are expected to benefit from cross-listing in the U.S. due to better shareholder protection, tighter control by the U.S. government and the SEC and more regulations imposed by stock exchanges (NYSE/ NASDAQ). I expect that firms from Latin America which are cross-listed in the U.S. have a higher performance, measured in terms of return on assets (ROA) and return on equity (ROE) than firms which are not cross-listed in the U.S. Hence the following hypothesis is formulated:
Hypothesis 1. Latin American MNCs, which are cross-listed in the U.S., have a higher increase of ROA than Latin American MNCs, which are not cross-listed.
Hypothesis 2. Latin American MNCs, which are cross-listed in the U.S., have a higher increase of ROE than Latin American MNCs, which are not cross-listed.
With regard to the external expects of corporate governance, improvements can be made for the Latin American corporate governance system. Due to foreign cross-listing firms will improve their corporate governance and will increase the transparency of the firm. And this signal, cross-listing in the U.S., can be seen as the first step towards investors to show their willingness to improve the corporate governance system and follow stricter accounting standards and regulations and this will result in a lower cost of capital and a higher share price. This will result in a higher market value/firm value. In the following hypothesis I expect that due to foreign cross-listing the market value of a firm will increase, measured in terms of Tobin’s q. Hence, the following hypothesis is formulated:
22
4. Research design
In this chapter the research design is discussed. First the research method is stated, following by how the data are collected and the sample used. In the end the analyses to test the
hypotheses are discussed.
4.1 Research method
The relationship between foreign cross-listing and firm performance is investigated in this thesis. Several different factors do influence firm performance, in this thesis the relationship between foreign cross-listing and firm performance is investigated. Therefore the first part of the research method starts with explaining how firm performance is measured, following how corporate governance is measured and how the data are collected.
4.1.1 Measuring firm performance
In this thesis firm performance is measured in three different ways. To be able to test the first H1 firm performance is measured in terms of return on assets (ROA). ROA is an indicator of how profitable a company is relative to its total assets. In the second hypothesis firm
performance is measured in terms of return on equity (ROE). ROE measures a firm’s
profitability by revealing how much profit a company generates with the money shareholders have invested. In the third hypothesis firm performance is measured in terms of firm value, by using the Tobin’s q ratio. The Tobin’s q ratio is calculated as the market value of a firm divided by the replacement value of the firm's assets, thus it implies that the market value of a firm should be about equal or more than the replacement value of firm assets.
The change in ROA, ROE and Tobin’s q is measured in terms of the percentage change which is calculated by the difference before listing and after listing. 1 year before cross-listing (t=1) and 2 year later (t=2) thus one year after cross-cross-listing. By collecting data before cross-listing and after cross-listing a change in the performance indicator can be measured, an increase/decrease or no change at all can be measured.
23
Groningen). DataStream is a database which is used to access firm financial records. The University of Groningen has a licence which gives students the opportunity to access financial information, such as firm performance indicators and other financial data.
Return on assets (ROA)
When ROA is not available, the following way will be used to calculate the ROA; total profit, divided by total assets and multiplied by 100.
ROA = Total profit
Total assets * 100
Return on equity (ROE)
When it is not available, the following way will be used to calculate the ROE; total profit, divided by total equity and multiplied by 100.
ROE = Total profit
Total Equity * 100
Tobin’s q
For the third hypothesis firm performance will be measured in terms of firm value, market value of a firm. In order to calculate the firm value the Tobin’s q ratio is used. When the Tobin’s q is not given and accessible in DataStream, a formula is used to calculate the Tobin’s q. In this thesis the Tobin’s q will be calculated as followed (Doidge et al., 2004).
Book value of total assets – book value of equity + market value of equity
Book value of total assets * 100
4.1.2 Measuring corporate governance
24
Brazil, Argentina and Mexico are considered to have a Latin American corporate governance model. In this case the geographical area is used a proxy-variable to determine the Latin corporate governance system. I assume that all three do have a Latin American corporate governance system as is specified at both the stock exchanges in the U.S. (NYSE/NASDAQ). All three countries are considered to be from the same geographical area, Latin America. (www.nyse.com)
For the U.S. the proxy-variable used to determine what of corporate governance system is used is the language which is spoken. In some cases the language spoken in a country can be used to determine what kind of corporate governance system is used. In the literate the U.S is referred to have an Anglo-Saxon corporate governance system. In this thesis the United States of America is assumed and implied to have an Anglo-Saxon corporate governance system, which corresponds with the literature and the assumption that English spoken countries have an Anglo-Saxon background.
4.2 Data collection
The data collection consists of several phases. To know the total population it is necessary to investigate how many firms from Brazil, Argentina and Mexico are cross-listed in the U.S. In the following table an overview is given on how many firms from these countries are cross-listed in the U.S., on the NYSE and or the NASDAQ.
TABLE 1 Total population
Brazil Argentina Mexico Total
Population (bruto) 54 23 18 95
Number of cross-listed firms on both the stock
exchanges (NYSE/NASDAQ)
13 5 0 18
25 Population (bruto)
From Brazil 54 firms are cross-listed in the U.S. on the NYSE and/ or the NASDAQ. From Argentina 23 firms are cross-listed on the NYSE and/ or the NASDAQ. From Mexico 18 firms are cross-listed on the NYSE. In total 95 firms are listed on stock exchanges in the U.S.
Number of cross-listed firms on both the stock exchanges (NYSE/NASDAQ)
From the 95 firms 18 firms are cross-listed on both the stock exchanges. Therefore from the 95 firms, 18 firms are subtracted otherwise 18 firms would have been counted twice, which gives a total number of 77 firms which are cross-listed in the U.S.
Population (netto)
The total population consists of 77 firms from Brazil, Argentina and Mexico which are cross-listed on one of the stock exchange in the U.S. the NYSE or the NASDAQ.
The data can be used from all the 18 firms, which are cross-listed on both the exchanges. The following reasons are given why the data can be used.
5 firms have been listed in the same year on both of the stock exchanges.
From 6 firms it is certain that they have a different listing year on both of the stock exchanges. The difference between both of the listing is more than 10 years therefore the data of the first listing can be used.
By 7 firms it is unknown whether or not they are listed on the stock exchanges in the same year, I expect that the given listing year is the first listing year or the same listing year on the other stock exchanges, therefore the data of these firms can be used as well.
4.2.1 Sampling strategy
The sampling strategy used in this study is probability sampling and specifically simple random sample strategy. A condition to use this strategy is that the total population should be known. The total population, the sampling frame, in this study is 77 firms from Brazil, Argentina and Mexico which are cross-listed in U.S. The stock exchanges (NYSE &
26
non-zero change of selection. This strategy of sampling is considered as a fair way of selecting a sample from a given population. By using this strategy the influences of the differences in industry and company size are minimized, because every company has the same change of being selected (Cooper & Schindler, 2008).
Sample cross-listed firms (CL-group)
By selecting the sample in this study, 14 firms from Brazil, 8 firms from Argentina and 8 firms from Mexico were selected. In an alphabetical way the firms were selected starting on the top of the list with A. When data was not available the next and following company is selected, in an alphabetical way starting with the following company. In total 30 firms were selected.
Sample non cross-listed firms (Non CL-group)
To investigate whether foreign cross-listing does influence firm performance non cross-listed firms were selected. In this thesis 30 firms were selected from the stock exchanges in Brazil, Argentina and Mexico. The firms selected were not cross-listed in the U.S. only listed on the stock exchange in their home country.
The same sampling strategy is used to collect the firms of the Non CL-group. The stock exchanges in Brazil, Argentina and Mexico give an alphabetically overview of all the listed firms (Cooper & Schindler, 2008). The indices used are:
The Bovespa index in Brazil represent 65 most traded stocks in Brazil, The Merval index in Argentina represent 20 most traded stocks in Argentina The IPC index in Mexico represent 37 most traded stocks in Mexico
14 firms from Brazil, 8 firms from Argentina and 8 firms from Mexico are selected.
Outliers
27
The kurtosis indicates the ‘flatness’ or the ‘peakedness’ of a distribution. A value near zero is considered to be very good (+/-1) and does indicate that the sample has a normal distribution. A kurtosis value between -2 and +2 is usually acceptable.
The skewness indicates the extent to which a distribution of values deviates from symmetry around the mean. A value near zero indicates that the distribution is symmetric. The same values for the skewness are acceptable as they are for the kurtosis.
The table below is generated to give an overview on how the total sample is generated.
Table 4.2 Total sample Cross-listed firms
from Brazil, Argentina and Mexico
Not cross-listed firms from Brazil, Argentina and Mexico
Total of both groups
Total firms 30 30 60
Outliers 2 4 6*
Total sample (n=) 28 26 54
* For the specific outliers see appendix A and B
Collecting data of the firm performance indicators; ROA, ROE and Tobin’s q
When was known which firms were selected, the data regarding firm performance were collected. Collecting the data of ROA, ROE and Tobin’s q ratio DataStream was used. The data were collected one year before the cross-listing (t=1) and one year after the cross-listing (t=2). The following information was necessary to have, to obtain the performance indicators and to attract it from DataStream:
Company name or code which is provided at the stock exchange it is listed on. The year the company was cross-listed
28
The company (name or code) was not known by DataStream;
The performance indicator (ROA, ROE or Tobin’s q) was not available or was not possible to calculate;
In some cases one of the indicators was missing and/ or was not possible to calculate
4.3 Data analysis
The data used in this thesis were analysed by using the statistical program SPSS. SPSS can be used to test the hypotheses; first the descriptive analyses are explained followed by the
independent sample t-tests.
Descriptive analyses
In order to get a good overview of the change in the performance indicators descriptive analyses were conducted. The mean increases/decreases were collected of the ROA, the ROE, the Tobin’s q, and the standard deviation. Moreover the minimum/ maximum were
determined of the performance indicators, for the CL-group, the Non CL-group and for the countries specific. Next to that for the CL-group, the Non CL-group and for every country specific the normality indicators were determined, which gives an indication whether there is a normal distribution or not.
Independent sample t-tests
This test was used in order to analyse whether there was an increase or decrease in the firm performance between the CL-group and the Non CL-group. Next to that the independent sample t-tests were also conducted to test whether there was an increase or decrease in firm performance indicators between the specific countries (Brazil- Brazil; Argentina- Argentina; and Mexico- Mexico).
29
5. Results and Discussion
In this chapter the results of the analyses are presented. First, results of the descriptive
analyses are presented and discussed. Next, results from the tests are reported and discussed.
5.1 Descriptive results
In this paragraph the descriptive results of the total sample as well as the results of the three different countries are presented. The number of observations, the mean change in ROE, ROA, Tobin’s q, the minimum and maximum, the standard deviation and the normality figures are given, the outliers are deleted.
5.1.1 Cross-listed firms (CL-group)
TABLE 5.1
Descriptive statistics of firms which are cross-listed in the U.S
ROE ROA Tobin’s q
Number of observations 29 30 29 12,53 32,14 9,15 -157,19 -148,87 -9.11 301,14 331,84 49,55 96,46 116,82 15,93 1,25 1,21 1,45 2,59 ,90 1.14 Mean increase in % Minimum change in % Maximum change in % Standard Deviation Skewness Kurtosis
From all the observations from the Latin American firms which are cross-listed in the U.S. table 5.1 shows that the percentage mean increases of all three firm performance indicators are positive. The mean increase of ROE is 12,53 percent, the mean increase of ROA is 32,14 percent and the mean increase Tobin’s q is 9,15 percent.
30
is 331,84 percent which gives a range of 480,71 percent. The minimum decrease in Tobin’s q is 9,11 percent and the maximum increase is 49,55 percent, which gives a range of 50,46 percent.
The standard deviation for the three performance indicators (ROE, ROA and Tobin’s q) are respectively 96,46 and 116,82 and 15,93 percent. The mean plus the standard deviation and the mean minus the standard deviation represents 68 percent of all observations. The mean minus twice the standard deviation and plus twice the standard deviation represents 95 percent of all the observations.
5.1.2 Not cross-listed firms (Non CL-group)
TABLE 5.2
Descriptive statistics of firms which are not cross-listed in the U.S
ROE ROA Tobin’s q
Number of observations 30 29 29 4,60 -12,54 1,36 -106,04 -105,33 -24,18 149,38 69,87 25,61 61,24 45,74 8,66 ,56 -,08 ,21 ,43 -,43 3,90 Mean increase in % Minimum change in % Maximum change in % Standard Deviation Skewness Kurtosis
According to table 5.2 the ROA shows a negative mean, respectively 12,54 percent. The ROE and the Tobin’s q show a positive mean, respectively 4,6 and 1,36 percent. The minimum decrease in ROE is 106,04 percent and the maximum increase in ROE is 149,38 percent. This means that the range between the minimum and the maximum is 255,42 percent. The
31
5.1.3 Comparison between firms from Brazil (CL-group – Non CL-group)
TABLE 5.3
Descriptive statistics of firms from Brazil which are cross-listed in the U.S. and firms which are not cross-listed
CL- group Non CL-group
ROE ROA Tobin’s q ROE ROA Tobin’s q Number of observations 13 14 13 14 14 12 -13,41 25,38 11,13 12,21 7,03 1,80 -112,98 -113,95 -5,35 -97,62 -96,81 -5,99 114,04 271,45 49,55 149,38 48,24 10,97 65,09 115,89 16,77 63,86 60,67 4,81 ,34 1,28 1,27 ,42 ,60 ,42 -,31 ,50 ,88 ,62 1,33 -,12 Mean increase in % Minimum change in % Maximum change in % Standard Deviation Skewness Kurtosis
For the Brazilian firms which are cross-listed both the ROA and Tobin’s q do have a positive mean increase, respectively 25,38 and 11,14 percent. The ROE does have a mean decrease, which is -13,41 percent. Furthermore table 5.3 shows as well three positive means (ROE, ROA and Tobin’s q) regarding the firms from Brazil which are not cross-listed, respectively 12,21, 7,03 and 1,8 percent.
For the firms from Brazil which are cross-listed the minimum decrease of ROE is -112,98 and the maximum increase of ROE is 114,04 percent. The minimum decrease of ROA and
Tobin’s q is -113,95 and -5,35 percent. For firms which are not cross-listed table 5.3 shows that the minimum and the maximum of ROE are -97,62 and 149,38 percent. The minimum and maximum changes in percentage of ROA are -96,81 and 148,24. The minimum and maximum of the Tobin’s q are respectively -5,99 and 10,97 percent.
32
5.1.4 Comparison between firms from Argentina (CL-group – Non CL-group)
TABLE 5.4
Descriptive statistics of firms from Argentina which are cross-listed in the U.S. and firms which are not cross-listed
CL- group Non CL-group
ROE ROA Tobin’s q ROE ROA Tobin’s q Number of observations 8 8 8 8 8 7 12,53 11,92 1,66 3,76 -27,18 4,63 -62,53 -62,14 -5,91 -106,04 -105,33 -8,82 87,39 99,60 7,56 142,45 15,24 25,61 53,98 53,95 4,52 72,60 39,30 12,03 -,30 ,13 -,38 ,66 -1,07 1,02 -,92 -,42 ,24 1,59 1,39 -,10 Mean increase in % Minimum change in % Maximum change in % Standard Deviation Skewness Kurtosis
According to table 5.4 firms from Argentina which are cross-listed do have a positive mean increase for all the three performance indicators (ROE, ROA and Tobin’s q), respectively 12,53, 11,92 and 1,66 percent. The Non CL-group on the other hand, 2 performance
indicators are positive, a ROE of 3,76 percent and a Tobin’s q of 4,63 percent. The ROA has a negative mean decrease of 27,18 percent.
For the firms form Argentina which are cross-listed the minimum decrease of ROE is -62,53 and the maximum increase of ROE is 87,39. The minimum decrease of ROA and Tobin’s q is -62,14 and 62,14 percent and the maximum increase of ROA and Tobin’s q ratio are 99,6 and 7,56 percent. By looking at the firms which are not cross-listed table 4 shows that the
33
5.1.5 Comparison between firms from Mexico (CL-group – Non CL-group)
TABLE 5.5
Descriptive statistics of firms from Mexico which are cross-listed in the U.S. and firms which are not cross-listed
CL- group Non CL-group
ROE ROA Tobin’s q ROE ROA Tobin’s q Number of observations 8 8 8 8 8 8 84,31 64,20 4,01 -7,86 -12,04 ,64 -157,19 -148,87 -55,53 -60,95 -67,99 -3,25 485,23 331,84 46,90 72,89 62,48 5,98 209,49 64,19 31,41 48,94 56,67 3,56 1,16 ,71 -,54 ,79 ,52 ,76 ,79 -,50 1,34 -,54 -1,17 -,80 Mean increase in % Minimum change in % Maximum change in % Standard Deviation Skewness Kurtosis
For the firms from Mexico which are cross-listed all three performance indicators have a positive mean increase, respectively 84,31, 64,20 and 4,01 percent. By looking at firms from Mexico which are not cross-listed, the mean increase regarding ROE and ROA are negative, a ROE of -7,86 percent and a ROA of -12,04 percent. The Tobin’s q does have a positive mean increase of 0,64 percent.
For the firms from Mexico which are cross-listed the minimum decrease of ROE is -157,19 and the maximum increase of ROE are 485,23 percent. The minimum decrease of ROA and Tobin’s q are -148,87 and -55,53 the maximum increase of ROA and Tobin’s q are 331,84 and 45,90 percent. By looking at the firms which are not cross-listed table 4 shows that the minimum and the maximum of ROE are -60,95 and 72,89 percent. The minimum and
34
5.1.6 Comparison of mean changes in performance indicators (CL-group – Non CL-group)
FIGURE 5.1
Comparison of performance indicators between CL-group and the Non CL-group
Figure 5.1 shows the mean change in performance indicators (ROE, ROA and Tobin’s q) for the CL-group and the Non-CL group. Figure 5.1 shows that the firms from Latin America which are cross-listed in the U.S. do have higher firm performance; all three indicators do show a higher increase. By looking at the Non CL-group, two indicators (ROE and Tobin’s q) are positive as well, but the percentage increase is lower compared to the CL-group. Next to that the Non CL-group does have a percentage decrease in ROA of more than 10 percent.
-15 -10 -5 0 5 10 15 20 25 30 35
ROE ROA Tobin's q
35
5.1.7 Comparison of mean changes in performance indicators
FIGURE 5.2
Comparison of performance indicators between the specific countries
Figure 5.2 shows the mean changes in performance indicators (ROE, ROA and Tobin’s q), for the CL-group compared with the Non CL-group and gives specific descriptive figures
regarding the specific countries.
The performance indicators of firms which are cross-listed do have a higher mean increase in percentage, with the exception of Brazilian firms which are cross-listed which do have a negative mean change, a decrease in ROE of almost 13 percent. Moreover cross-listed firms from Mexico do have a higher increase of ROE and ROA compared to firms from Argentina and Brazil which also cross-listed in the U.S.
36 5.2 Hypotheses testing
In this part of the chapter the hypotheses are tested and the results are presented. Independent sample t-tests are conducted in order to test whether the mean increase of the CL-group (M1) is significant different from the mean increase of the Non CL-group (M2). A confidence level of 95% is used which indicates an alpha of 0.05.
5.2.1 Comparing return on equity (ROE)
In the table below the results are presented. An independent sample t-test is conducted in order to test the first hypothesis, which is:
Hypothesis 1: Latin American MNCs, which are cross-listed in the U.S., have a higher increase of ROE than Latin American MNCs, which are not cross-listed.
TABLE 5.6
Independent sample t-test ROE Number of observations M1 increase in % SD M2 increase in % SD t-value df Significance (2 tailed) ROE 59 12.53 96.46 4,6 61,24 .376 47.16 .707
According to table 5.6 M1 increase is 12,53 percent and M2 increase is 4,6 percent. There was no significant difference in the scores for the CL-group (M1=12.53, SD=96.46) and the Non CL-group (M2=4.6, SD=61.24) conditions; t(47.16)=376, p=.707
37 5.2.2 Comparing return on assets (ROA)
In the table below the results are presented. An independent sample t-test is conducted in order to test the second hypothesis, which is:
Hypothesis 2. Latin American MNCs, which are cross-listed in the U.S., have a higher increase of ROA than Latin American MNCs, which are not cross-listed.
TABLE 5.7
Independent sample t-test ROA Number of observations M1 increase in % SD M2 increase in % SD t-value df Significance (2 tailed) ROA 59 32.14 116.83 -12,54 45,74 1.95 37.94 .06
According to table 5.7 M1 increase is 32,14 percent and M2 decrease is 12,54 percent. There was no significant difference in the scores for the CL-group (M1=32.14, SD=116.83) and the Non CL-group (M2=-12.54, SD=45.74) conditions; t(37.94)=1.95, p=.0.06.
38 5.2.3 Comparing Tobin’s q ratio’s
In the table below the results are presented. An independent sample t-test is conducted in order to test the third hypothesis, which is:
Hypothesis 3. Latin American MNCs, which are cross-listed in the U.S., have a higher increase of Tobin’s q than Latin American MNCs, which are not cross-listed.
TABLE 5.8
Independent sample t-test Tobin’s q Number of observations M1 increase in % SD M2 increase in % SD t-value df Significance (2 tailed) Tobin’s q 58 9,15 15,93 1,36 8,66 2.311 43.22 .026 Table 5.8 shows us that M1 increase is 9,15 percent and that M2 increase is 1,36. There is a significant difference in the scores for the group (M1=9.15, SD=15.93) and the Non CL-group (M2=1.36, SD=8.66) conditions; t(43.22)=2.311, p=.0.026.
39 5.3 Ad Hoc Analysis
In this part of the chapter independent sample t-tests are conducted in order to test the differences in means between specific countries. Two of the three hypotheses were not significant at an alpha of 0.05, independent sample t-tests are conducted whether to find if there is a significant difference between the CL-group and the Non CL-group of a specific country (Brazil, Argentina and Mexico). I might be interesting to find whether a specific country is significant, by doing so the differences between the countries will be gone. A confidence interval of 95% is used, which indicates an alpha of 0.05.
5.3.1 Comparison between the CL-group and the Non CL- group of firms from Brazil
TABLE 5.9
Independent sample t-test Number of observations M1 increase in % SD M2 increase in % SD t-value df Significan ce (2 tailed) ROE 27 -13,41 65,09 12,21 63,86 -1,031 27,77 ,312 28 25,38 115,89 7,03 60,67 ,525 19,63 ,606 25 7,32 12,37 1,8 4,81 1,45 23 ,162 ROA Tobin’s q
The results from table 5.9 show that for all three indicators (ROE, ROA and Tobin’s q) there is no significant difference between the means. ROE; (M1=-13.41, SD=65.09, M2=12.21, SD=63.86); t (df=27.77) =- 1.031, p= .312. ROA; (M1=25.38, SD=115.89, M2=7.03,
SD=60.67); t (df=19.63) = .525, p= .606. Tobin’s q; (M1=7.32, SD=12.37, M2=1.8, SD=4.81); t (df=23) = 1.45, p=.162.
Table 5.9 shows that firms from Brazil which cross-listed in the U.S. do have a higher mean increase in ROA and Tobin’s q compared to firms which are not cross-listed. Even though the mean increase is higher, the results in table 5.9 indicate that the differences in mean increase are not significant.
40
5.3.2 Comparison between the CL-group and the Non CL- group of firms from Argentina
TABLE 5.10
Independent sample t-test Number of observatio ns M1 increase in % SD M2 increase in % SD t-value df Significanc e (2 tailed) ROE 16 12,53 53,98 3,76 72,60 .274 12,93 ,788 16 11,92 53,95 -27,18 39,30 1.657 12,80 ,122 15 4,11 8,09 1,63 9.22 .550 12,10 ,592 ROA Tobin’s q
The results from table 5.10 show that for all three indicators (ROE, ROA and Tobin’s q) there is no significant difference between the means. ROE; (M1=12.53, SD=53.98, M2=3.76, SD=72.60); t (df=12.93) =.274, p= .788. ROA; (M1=11.92, SD=53.95, M2=-27.18, SD=39.30); t (df=12.80) = 1.657, p= .122. Tobin’s q; (M1=4.11, SD=8.09, M2=1.63, SD=9.22); t (df=12.10) = .550, p=.592
41
5.3.3 Comparison between the CL-group and the Non CL- group of firms from Mexico.
TABLE 5.11
Independent sample t-test Number of observation s M1 increase in % SD M2 increase in % SD t-value df Significance (2 tailed) ROE 16 84,31 209,49 -7,86 48,94 1,21 14 .246 16 64,19 164,19 -12,04 51,67 1,25 14 .231 16 4,01 31,41 ,64 3,56 ,30 14 .767 ROA Tobin’s q
Table 5.11 shows that for all three performance indicators (ROE, ROA and Tobin’s q) there is no significant difference between both the groups. ROE; (M1=84.31, SD=209.49, M2=-7.86, SD=48,94); t (df=14) =1.21, p= .246. ROA; (M1=64.19, SD=164.19, M2=-12.04, SD=51.67); t (df=14) = 1.25, p= .231. Tobin’s q; (M1=4.01, SD=31.41, M2=0.64, SD=3.56); t (df=14) = .30, p=.767
According to table 5.11 the results do indicate that the mean increase in ROE, ROA and Tobin’s q of the CL-group is higher compared to the Non CL-group. However the differences are not significant.
5.3.4 Summary of hypothesis tests
The following table below gives an overview on the hypotheses that were tested. 1 hypothesis was supported, at an alpha of 0.05 and 2 hypotheses were not supported. In all three
hypotheses a positive change in performance indicator were observed.
Table 5.12
Overview of the results of the hypotheses
Hypothesis Expected change in Observed change in Significant performance indicator performance indicator
Hypothesis 1 Positive change Positive No (ROE)
Hypothesis 2 Positive change Positive No (at alpha of .05)
(ROA) Yes (at alpha of 0.1)
42 5.4 Summary of the results
In this paragraph a summary of the results of the analyses are presented. First the results of the hypotheses are discussed followed by a discussion of the ad hoc analyses.
Hypothesis 1: Latin American MNCs, which are cross-listed in the U.S., have a higher increase of ROE than Latin American MNCs, which are not cross-listed.
The first hypothesis has an expectation that firms from Latin America benefit from a cross-listing in the U.S. which will result in a relative higher increase of return on equity (ROE) compared to Latin American firms which are not cross-listed in the U.S. The results do show a higher mean increase in percentage for firms which are cross-listed in the U.S., but the differences between both groups is not significant. Therefore there is not enough evidence to support hypothesis 1.
Hypothesis 2. Latin American MNCs, which are cross-listed in the U.S., have a higher increase of ROA than Latin American MNCs, which are not cross-listed.
The second hypothesis has an expectation that firms from Latin America benefit from a cross-listing in the U.S. which will result in a relative higher increase of return on assets (ROA) compared with firms from Latin America which are not cross-listed in the U.S. It can be concluded that firms from Latin America which are cross-listed in the U.S. do have indeed a higher increase in ROA compared to firms which are not cross-listed in the U.S.
43
Hypothesis 3. Latin American MNCs, which are cross-listed in the U.S., have a higher increase of Tobin’s q than Latin American MNCs, which are not cross-listed.
The third hypothesis expects an increase in firm value measured in Tobin’s q. This hypothesis has an expectation that firms from Latin America which are cross-listed in the U.S. have a higher mean increase in Tobin’s q compared with firms which are not cross-listed in the U.S. The results indicate that firms from Latin America which are cross-listed in the U.S. do have a higher mean increase in Tobin’s q, compared to firms which are not cross-listed in the U.S.
This difference is significant at an alpha of .05 therefore there is enough evidence to support hypothesis 3.
Ad hoc analyses
44
6. Discussion & Conclusion
In this chapter the results are discussed and conclusions are given, moreover further research and limitations are discussed.
6.1 Discussion
Overall the results show that there is not enough support for the theory that foreign cross-listing has a positive effect on firm performance. By looking at the total sample the results do show a higher mean increase in performance indicators, but only 1 hypothesis was supported. By looking at the ad hoc analyses the results show no significant difference between the means of each country for all three performance indicators. The theory suggests that foreign cross-listing has a positive effect on firm performance, due to an improvement of corporate governance; this theory is not supported in this study. There is not enough evidence to support the theoretical assumption that foreign cross-listing has a positive effect on firm performance.
The third hypothesis is supported at an alpha of 0.05, the results show a significant difference between the means of the cross-listed firms from Latin America and firms from Latin
America which are not cross-listed, measured in Tobin’s q. This is not in line with the results from the ad hoc analyses; the results from the ad hoc analyses show no significant outcomes for all three countries. The theory suggests that firms which operate in a developing country with a less developed corporate governance system can benefit from a cross-listing in the U.S. The bonding hypothesis implies that the effect on firm performance will be the greatest when the difference between both countries is big, in regard to their corporate governance system. A reason why there are no significant outcomes might be that the corporate governance in Latin America is more developed than expected.
45
used in this study might have been too small. After all the bigger the sample the possibility of significant results are higher.
Another explanation which can explain the results is that the data which were collected were from one year before cross-listing and 2 year later, thus 1 year after cross-listing. By
conducting in a foreign cross-listing the expectation is that cross-listed firms benefit from it by improving their corporate governance, because of better financial disclosure and
accounting standards, which result in higher and better firm performance. Improving a corporate governance system might take more time than the expected 1 year. It is a process which might take longer than 1 year. Or it could be that the improvements of the corporate governance system are made before cross-listing consequently the level of influence might be less than expected, after cross-listing.
By looking at the ad hoc analyses cross-listed firms from Mexico have the highest mean increase for all three performance indicators compared with cross-listed firms from Brazil and Argentina. The reason for this might be that the corporate governance system in Mexico compared with the two other countries is less developed and therefore firms in Mexico can benefit more from a foreign cross-listing than firms from Brazil and Argentina. But still the differences between both the means of the cross-listed firms and the not cross-listed firms from Mexico are not significant.
The theoretical assumption that foreign cross-listing has a positive effect on firm performance due to an improvement of the corporate governance system is not supported. The assumption is that due to foreign cross-listing the corporate governance will improve which will result in better firm performance. It might be that the influence of better corporate governance on firm performance has not such an influence as expected.
46 6.2 Conclusion
It can be concluded that there is not enough evidence to support the theory that foreign cross-listing has a positive effect on firm performance. 1 hypothesis was supported, 2 hypotheses were not supported. The results from the ad hoc analyses do indicate that there is not enough evidence to support the theory, none of the ad hoc analyses showed a significant difference.
Therefore it can be concluded that there is not enough evidence to support the theoretical assumption that foreign cross-listing has a positive effect on firm performance for firms from Latin America. The results do show a higher mean increase for firms which are cross-listed in the U.S. but only 1 hypothesis was significant.
Several reasons can be given why there were no significant results. The theory suggests that foreign cross-listing in the U.S. has a positive effect on firm performance especially for firms who operate in a less developed corporate governance system. A reason might be that the corporate governance system for all three countries is more developed than expected, even though the ad hoc analyses do show different results between the three countries. Firms from Mexico which are cross-listed do show a higher mean increase compared to the two other countries, an explanation for this might be that between the three countries there are still some differences in corporate governance and firms from Mexico are able to benefit more from a cross-listing in the U.S than firms from Brazil and Argentina. It might be that the corporate governance system in Mexico is less developed in comparison to the two other countries. But still not significant outcomes were found in Mexico as well.
Another explanation might lie in the time-frame used in this thesis which might have been too short. The data were collected 1 year before cross-listing and 1 year after cross-listing. It might take longer to improve the corporate governance system after cross-listing within 1 year; and it might be that firm performance increases after a couple years instead of 1 year after cross-listing. Or it might be that the improvements are made before cross-listing and therefore the influence after cross-listing is less than expected.
47
governance on firm performance has not such an influence as expected. There is no
supporting evidence found that this theory is right, it might be that this theory is not right after all, even though there is no evidence to support the either.
6.3 Future Research
This thesis did not find enough evidence to support the theoretical assumption that foreign cross-listing has a positive effect on firm performance for firms from Latin America. The effect of foreign cross-listing is measured and is mainly focussed on financial indicators, non-financial indicators are not in the picture, and a reason for this might be that in-depth research is necessary to find a relationship between foreign cross-listing and non-financial indicators. An interesting topic for future research might be to investigate what the relationship is between foreign cross-listing and non-financial indicators such as product quality and employee satisfaction.
The theory suggests that firms which operate in high tech sector are more driven to cross-list in the U.S. due to the expertise and knowledge in the market. This could influence the quality of the products in a positive way. Employee satisfaction on the other hand might be
influenced in a negative way due to a foreign cross-listing because the financial focus of a firm could change towards a more short term focus, the decisions could be made in the interest of the shareholders and not in the firm’s interest and that could influence employee satisfaction in negative way.
In order to find out if foreign cross-listing has a positive effect on firm non-financial indicators a case study can be conducted of four firms from Latin America and investigate what the effect is between foreign cross-listing and firm non-financial indicators. By doing so it is necessary to use in-depth interviews with managers of all four firms to find whether there is a relationship between foreign cross-listing and firm non-financial indicators. It might also be useful to use questionnaires in order to find out whether employee satisfaction has been changed or not.
Moreover this thesis investigated the short-term relationship between foreign cross-listing and firm performance; a topic for future research might be to do the same research and to
48
long-term, 5 or 10 years after cross-listing instead of 1 year after cross-listing. Moreover it might be interesting to investigate whether foreign cross-listing effects also the internal expects of corporate governance. By doing so it might be necessary to do also a case study and use in-depth interviews to find a relationship.
6.3 Limitations
A limitation could be on how the corporate governance systems are measured. I have used a proxy-variable to measure the corporate governance systems. It might be that the corporate governance systems between the countries might differ between each other, which could bias the results. I have chosen three countries which are most similar and are considered to be from Latin America. The stock exchanges (NYSE/NASDAQ) do support the fact that Brazil, Argentina and Mexico are from the same geographical area. According to the literature a geographical area can be used to determine what kind of corporate governance is used, therefore I have used a proxy variable which replaces the variable which is not measurable. By doing so and with regard to a short time-frame I have minimized these effects.
