Amsterdam school of economics
Bachelor thesis Economics
Track Finance & Organization
The market reaction of delisting a cross-listed position from the
United States
Author: Nina Maier
Student number: 10881042
Supervisor: Jan Lemmen
Date: January 17
th2017
ii Abstract
A sample of 111 cross-listed foreign firms that voluntary delisted from the US organized exchanges between 2005-2015 is examined. The delisting announcement had an overall significant negative effect on market value, but there is a wide variety in abnormal returns. The variety is found to be partially explained by the governance level of the firm’s home country, leverage level of the firm, P/E
ratio and the way the firm was listed in the United States.
JEL Keywords: F21 International investments, G34 Corporate governance, G12 Asset pricing
Statement of Originality
This document is written by Student Nina Maier who declares to take full responsibility for the contents of this document.
I declare that the text and the work presented in this document is original and that no sources other than those mentioned in the text and its
references have been used in creating it.
The Faculty of Economics and Business is responsible solely for the supervision of completion of the work, not for the contents.
iii LIST OF FIGURES
Figure 1 total sponsored and unsponsored ADRs in period 2010-2014 5 Figure 2 number of delistings in dataset from period 2005 – 2015 12 Figure 3 Average abnormal return for each individual day in the event window 18 Figure 4 Delisting by continent 26 LIST OF TABLES
Table 1 explanation of variables used in Regression 14 Table 2 Cumulative average abnormal return different event windows 17 Table 3 Significant individual abnormal returns 19 Table 4 stock price reaction in 51 and 25 day event window around announcement 20 Table 5 stock price reaction in 11 and 3 day event window around announcement 22 Table 6 delistings by country 25
iv
TABLE OF CONTENTS
Abstract………..ii
List of Tables & Figures ………...ii
Table of contents ………...iii
1. Introduction………..1 2. Literature review 2.1 Cross-listing………...2 2.2 Cross-delisting………...8 3. Methodology 3.1 Dataset………..11
3.2 Model specification & hypothesis………12
4. Results 4.1 Market value effect………...17
4.2 Abnormal return explaining factors………..20
5. Conclusion and discussion………..24
Bibliography………..26
1
1.Introduction
In the current wave of globalization, companies are not only expanding business operation internationally but also seeking new ways to diversify their funding channels. A way firms use to attract new investors is by cross-listing their shares on a major foreign stock exchange in addition to their home listing. The motivation of firms to choose a secondary listing (cross-listing) on a foreign exchange has been studied extensively in the financial literature and it is generally viewed that cross-listing has a positive impact on firm value. The US capital markets, due to its high level of governance control and visibility, has traditionally been a very popular place to cross-list stocks. By cross-listing, a firm subjects itself to the United States regulation laws, which increases the level of the governance control (often cited as one of the main purposes of cross-listing, Coffee, 2002). Cross-listings also reduce trade barriers, increase firm visibility, and help with price discovery.
Considerably less research exists on the motivation and effect of voluntarily ending a cross-listed position. More recent research into the effect of cross-listings also mentions the disadvantages and questions some of the high positive effects found by research in the 1990s. The discovered disadvantages in addition to a higher compliance burden due to the Sarbanes-Oxley Act of 2002 and low trading volume of cross listed stock lead to a substantial increase of delistings in the last two decades (Karolyi, 2006). The increased amount of delistings raises the importance of clearly understanding the rationale behind delisting from the US exchanges and the related impact on a firm’s valuation.
This paper studies the impact of voluntary delistings from the United States stock exchanges on companies’ market value in their home country. The studied sample contains 111 companies from a diverse set of countries and industries that delisted from the US exchanges in the period 2005-2015 but kept their home country listings. All of these delistings were voluntarily, i.e. the firms chose to leave the exchanges and did not leave because of takeovers, acquisitions or financial problems. The market model is used to find abnormal returns in 4 different short and medium term event windows around the
announcement of the delisting ranging from the day before and after delisting to 25 days before and after.
Results show an overall significant negative effect on market value in the medium term and an insignificant return in the direct days around the announcement. Vast differences between companies are found, where various companies show a significant positive abnormal return.
2 OLS regression is used to further explain the abnormal returns of the individual firms. The key results are the following: (i) the governance level of the home country is found to have a significant positive effect on the market value of a firm in the medium term; (ii) leverage and the P/E ratio also have a significant effect on the abnormal returns in the 25 days before and after delisting announcement; (iii) firms that delist a level 2 American depository receipt have on average a significant lower abnormal return in the 25 day period after the delisting announcement; (iv) in the two short term event windows no factors are found to significantly explain the abnormal returns.
There have been some papers that discussed the impact of a cross-listing delisting on market value, which however has seldom been the main topic. This paper contributes by specifically looking at voluntarily delistings and not delistings in general, by observing the returns around the announcement date of the delisting and not the time frame around the actual delisting implementation, by only considering delistings from organized exchanges (many papers include OTC delistings), and by taking a closer look at factors that might impact the abnormal returns observed. It also includes a newer time frame and a different methodology to find the abnormal returns.
Managers of firms that are currently cross-listed in the United States could use these results to better estimate the effect a potential voluntarily delisting could have on their firms’ market value.
In part 2 cross-listing is introduced followed by a short overview of previous research about the benefits and costs of cross-listing and the effects cross-listing and delistings have on market value. Part 3 states the hypothesis, data set and research methodology. Part 4 describes the empirical results and part 5 concludes and gives suggestions for further research.
3
2. Literature review
This part explains the concept of cross-listing, the potential benefits of cross-listing described by the academic literature and the effect cross-listing has on a firm’s market value. It is followed by a discussion of delisting, firm’s motivation for delisting cross-listed positions and previous found effects of a delisting on a firm’s market value.
2.1 Cross-listing
A cross-listed company is a single legal entity that lists its shares in multiple markets. This is different to a dual listed company which consists of two different legal entities who agree to share profits but retain independent listings in their own markets. A company can cross-list on many platforms worldwide but especially the exchanges in the United Kingdom and the United states are popular places to cross-list due to the high level of governance, esteem and visibility. Companies can cross-list in the United States in two different ways; They can have an ordinary listing or list by using an American Depository receipt (ADR). Ordinary listing is in general only possible for Canadian firms.
In the case of a ADR, a depository receipt certificate is issued by a depository bank that holds the foreign shares denominated in the home country currency and then issues U.S. shares denominated in dollars. The ADR ratio represents the relation between the number of ordinary shares and ADRs. ADRs trade like any regular US security, pay-out dividends in dollars and must register with the US Securities & Exchange commission (SEC). The holder of a ADR can redeem the receipt for the underlying share and has the same voting rights as shareholders owning the actual shares. Transaction costs are paid to the depository bank for the creation and cancellation of ADRs. There are only 4 depositary banks; BNY Mellon, J.P. Morgan, Citibank and Deutsche Bank, of these BNY Mellon has with a market size of 58% sponsored programs by far the biggest market capitalization (BNY Mellon, 2015).
There are 4 different forms of American depository receipts depending on the extend a firm wants to be listed in the United States.
American depository receipts under Rule 144A: US private placements, which are exempt
from American disclosure regulation or GAAP accounting. Not listed on an organized exchange and therefore excluded in this study. This form allows for issuing new capital but can only be bought by Qualified institutional investors. Qualified institutional investors manage at least $100 million in securities or are broker-dealers investing at least $10 million
4 in outside securities. Due to the limited investor pool 144A receipts are relatively illiquid compared to the other forms of ADRs, but they are still used to raise international capital without meeting the SEC disclosure or organized exchange regulations. Excluded in this research.
Level 1 American depository receipts only trade over the counter (OTC). As they are not
listed on an organized exchange they are not required to adhere to American disclosure regulation or accounting standards and do not have to register with SEC. As Level 1 ADRs are not listed on an organized exchange they are excluded in this research. Figure 1 indicates that this form of cross-listing is the most common. By listing OTC, firms can still attract investors who prefer to have shares denominated in dollars while avoiding the high costs of listing on organized exchanges.
Level 2 American depository receipts: Listing on a major American organized exchange such
as the NYSE, AMEX or NASDAQ. Firms must follow American disclosure regulations and use GAAP accounting.
Level 3 American depository receipts: Listing on organized exchanges with full disclosure
requirements. In contrarily to a level 2 ADR, this form allows issuing of new capital in the US through a public offering of ADRs. It is used frequently as a tool for companies to raise international capital.
In the begin days of depository receipts ADRs were often non sponsored, meaning they were initiated by an individual investor or investor group without prior consent of the company. In 1983 the SEC forbid this and started to only allow sponsored depository receipts, a signed agreement between company and depository bank to be the sole agent for its receipts. Since 2008 the SEC allows unsponsored level 1 ADRs once again. Figure 1 displays a substantial and increasing amount of unsponsored ADRs in the American market.
To list on an organized exchange such as the NYSE (whether in depository or ordinary form) the firm needs to meet the strict requirements set by the specific exchange which among others includes minimum number of shareholders, market makers and minimum market capitalization. They also need to pay the (ongoing) listing fees.
5 Figure 1: Total sponsored and unsponsored DRs in the period 2010-2014 (BNY Mellon, 2015)
Figure 1 shows that the amount of worldwide DRs is substantial but the majority is
unsponsored or unlisted. Many of the sponsored DRs only trade OTC. All DRs together are valued at $3.1 trillion US dollars (BNY Mellon, 2016) indicating the substantial size and importance of this mechanism.
Potential Benefits of cross-listing
Potential benefits of cross-listing are discussed frequently in the academic literature; this part gives a quick overview of the most commonly mentioned. Possible disadvantages are
described in part 2.2.
Cross-listing can remove market segmentation which can arise from direct (ownership reduction, taxes, high transaction costs) or indirect barriers (different rules, information availability) which lowers the cost of capital. It allows US investors to diversify their stock portfolios while still buying US denominated shares and receiving dividend in dollars and bypasses the trouble to directly invest in companies located in countries with restricted access to the equity market (Foerster & Karolyi, 1999). The market segmentation hypothesis has been studied in detail but has also been frequently criticized in more recent years. Stulz (1999) was the first to openly discuss the potential problems of the market segmentation theory. He mentions that if the main benefit of cross-listing would be to reduce market segmentation, all foreign firms that will gain a sufficiently decrease in cost of capital to justify the costs of listing abroad would do so. In reality many firms around the world do not cross-list. Another criticism involves the many event studies on abnormal return that still find a large positive effect for countries such as Canada and Western Europe that are well
6 integrated in the world markets, which is not in line with the segmentation theory. Bekaert and Harvey (1995) also show that certain segmented markets become less segmented because of increased globalization and development of financial markets.
Research is moving away from the segmentation hypothesis and instead focuses more on the potential benefit cross-listing in the United States has on corporate governance due to the higher level of disclosure and governance control. This is studied in the bonding theory that states that because firms have to follow U.S. laws, foreign firms credibly bond
themselves to avoid actions that might decrease the benefits to minority shareholders and ensure a more independent board of directors. Coffee (2002) discusses how the SEC enforces this bonding with its high disclosure regulations and mentions the many reputable
intermediaries in the United States including rating agencies, auditors and the organized exchanges themselves.
Lel & Miller (2008) find that cross-listed firms from countries with less regulations are more likely to fire poorly performing CEOs than non-cross-listed firms. Cross-listings on exchanges that do not require high level of disclosure (OTC, London listings) do not have this positive increase of corporate governance.
Reese & Weisbach (2002) explain that by cross listing in the United States the firm loses out on potential corruption gains, but benefit from increased credit ratings which allow them to finance investing opportunities at better terms. They find that firms with higher growth opportunities cross-list on average more than firms with lower growth numbers. The effect a US cross-listing has on the firms’ credit score depends on the regulation in the firms’ home country, where firms from well regulated countries see a smaller increase in credit ratings.
The increased governance also reduced firms cost of capital by 0.7% to 1.2% (Hail & Leuz, 2008). They find that the reduction in cost of capital accounts for more than half of the increase in value around cross-listings. The cost of capital reduction is smaller (0.3-0.6%) for firms that list OTC or for firms from countries with stronger governance regulation.
The higher level of disclosure in a format US investors understand reduces the information asymmetry between investor and firm. Firms can now more creditable signal their quality to investors. Listing on a big organized exchange such as the NYSE also
increases the visibility of the firm to investors by increased media and analyst coverage. Lang et al. (2003) find that firms cross-listed in the US have on average 2.64 more analysts
7 for the cross-listed firms which is related to the higher analyst following. Baker et al. (2002) also find a significant increase in analysts following the firm after cross-listing.
Merton (1987) develops the investor recognition hypothesis and discusses that investors have incomplete information and only invest in companies that they are aware off. If more investors become aware of the firm the “shadow costs” of not being aware lowers and the investors expected return reduces, which increases the market value of the firm.
Ahearne et al. (2004) look at information asymmetry in relation with the home bias, the tendency of investors to invest in companies listed in their home country. They show that countries with a greater share of firms with US cross-listings are less underweighted in US equity portfolios and thus show a decrease of US investors’ bias against its stock.
Cross-listing also increases price discovery, finding the right equilibrium price of a stock based on demand and supply. Eun and Sabherwal (2003) consider Canadian stocks that are ordinarily cross-listed in the US. The market keeps the prices of both exchanges equal by regularly adjusting the stock price on one exchange to the other. They find that while US prices in general adjust more to the Canadian prices, the Canadian prices sometimes also change based on the US prices which shows that the US cross-listing contributes to the price discovery of the Canadian stocks. They also discover that when the share trading volume in the US is higher, total adjustment in prices of US on Canadian prices increases (and thus an increased price discovery effect by the US cross-listing).
Firm specific factors also impact the total share of US trading volume. Pulatkonak, and Sofianos (1999) research non-US firms cross-listed on the NYSE in 1996 and find that the amount of trading on the NYSE as percentage of total trading volume greatly differs from anywhere between 1% and 95%. They find that companies from countries that trade around the same time-zone as the US are likely to have a greater trading volume on the NYSE. This would indicate that countries in Latin America and Canada would have a stronger price discovery effect and a greater effect on overall stock return.
There is also some evidence that cross-listing reduces the cost to acquire companies in the cross-listed market. Burns (2004) find that cross-listed foreign firms acquiring US
companies by using equity, pay on average 10% less than non-cross-listed firms paying cash. Companies from less regulated countries have to pay significantly more than firms from more developed economies.
The cross-listing also effects the home country market. Levine and Schmukler (2006) look at liquidity of emerging markets and find that trading of cross-listed firms that move to the international market reduce their domestic trading. Cross-listed firms facilitate the
8 migration of trading out of the domestic market, which hurts the liquidity of domestic firms. Karolyi (2004) also reports a negative effect on liquidity in domestic markets which hurts overall market capitalization and number of listed firms.
Effect of cross-listing on market value
Several papers found a link between the decision to cross-list and a significantly increase in the firms’ market value.
Miller (1999) looks at the cross-listing announcement time instead of the listing time and is therefore aligned with the goal of this study. He studied short and long term stock price effects after announcement. He valuates the effects of the different ADR levels and finds a significant positive stock price effect and a significant difference between countries with higher governance standards and those with lower standards. There is a significant higher announcement-day share price reaction for exchange listings then for SEC Rule 144a and OTC listings. This is in line with the bonding theory that states the exchange functions as an extra bonding mechanism.
Foerster and Karolyi (2000) observe the long run equity performance of 333 non-US stocks that raise capital in the United States (so firms under rule 144a or ADR level 3). They consider both initial and secondary public offerings. They find that while abnormal returns are on average 20% in the year before the capital raise, there is no significant abnormal return in the 3 years following the raise. They observe that private placements and ADR capital raisings from developed countries significant outperform those from developing nations and that a relatively higher volume of US trade positively influences the stock performance. Level 3 ADRs’ raisings perform better then private placements which is in line with the bonding theory.
You et al. (2012) also found a positive effect of cross-listing on market value. Hail & Leuz (2008), Foerster & Karolyi (1999) and Lang et al. (2003) mentioned earlier all found a positive effect of cross-listing on market value.
2.2 Cross-delisting
Delisting is the process of ending a listed position on a stock exchange. Delisting of cross-listed stocks can happen for a multitude of reasons. They can stop existing as a result of a company takeover or merger or they can be removed by the exchange as the firm does not comply to exchange rules or requirements, this often happens when a firm is in serious financial difficulty or is bankrupt. Voluntarily delistings occur when a firm decides
9 themselves to switch to another organized exchange or end their listing completely. Only in the case of voluntary delistings it makes sense to look at the potential arguments of leaving. Involuntarily delisting’s contain many factors which might impact the stock price aside from the actual delisting, such as a financially bad state of the firm. In this paper the focus is on delistings of voluntarily cross-listed positions, to avoid this other factors that might influence the studied abnormal stock return and to take a close look at the reasoning of the firms to leave the United States.
Possible Reasons to delist
There are high costs associated with a US listing. Adhering to the SEC reporting and
compliance requirements, requires firms to create new reports following US GAAP. The US organized exchanges and the depository bank also charge fees. The Sarbanes-Oxley Act (SOX) adopted in 2002 expanded requirements for all U.S. publicly listed company boards and increased the cost to comply with US regulations. It sets out the responsibilities of the board of directors, extends accounting rules and adds criminal penalties for misconduct. The higher requirements and costs of SOX are one of the possible reasons to delist. SOX might also lead firms to never list in the US and instead cross-list on a different developed exchange with much of the same benefits as listing in the US (reduce market segmentation, increased governance etc.). Piotroski & Srinivasan (2008) look at the cross-listing choice between the UK and US stock exchanges before and after introduction of the Sarbanes-Oxley Act. They found that while listing preferences of large foreign firms did not change after the adaption of SOX, smaller firms preferred UK exchanges afterwards. The cost of SOX compliance is relatively higher for these firms.
Rule 12h-6 established in March 2007 made it a lot easier for firms to delist from US exchanges. Firms can now easily qualify for deregistration if less than 5% of worldwide average daily trading volume take place on U.S. markets. Given that they have listed in the US for a minimum of 1 year, not issued US capital in the last year and kept an exchange listing in their home market. Many firms that already wanted to leave finally left and more foreign firms delisted in 2007 and 2008 than between the period from 2002 to the adoption of rule 12h-6 (Doidge et al. 2010). Fernandes et al. (2010) connect the level of home country governance and the stock price reaction after announcement of rule 12h-6. They find a negative stock price reaction for firms in low governance countries and an insignificant change for other firms. They also find a significant link between lower governance and higher rates of delisting right after the rule got in place.
10 The bonding theory states that the firm benefits from better growth financing terms. This suggests that firms should delist if they do not see good growth opportunities or have sufficient internal funds to finance growth.
Many of the delisted companies have low trading volume in the US market and therefore benefit little from a potential decrease of market segmentation or increase in
governance. The increasing globalization of financial markets and decrease of overall market segmentation also make cross-listing less important nowadays (Bekaert and Harvey, 1995).
Prior research cross-delisting has on market value
Doidge et al. (2010) look at voluntarily delisting of cross-listed firms in the US in the aftermath of the Sarbanes–Oxley Act. They found a negative stock price return after
announcement of the delisting and found that foreign firms with more agency problems have worse stock-price reactions which is in line with the bonding theory. Most of their data is from pre-rule 12h-6, the limited data they gathered from after 2007 shows a lower stock price decrease then before.
Bessler et al. (2012) find no significant effect of cross-listing in the US on the market value of German firms, they do find a positive market value effect after announcement of rule 12h-6 making it easier to leave. They mention a higher integration of financial markets and improvement of multilateral trading facilities as possible reasons for 13 of the 18 cross-listed German firms to leave the US exchanges after 12h-6. You et al. (2012) find a negative effect of the actual delisting on the firm market value. Fernandes et al. (2010) mentioned earlier also found a negative stock price effect around rule 12h-6 for firms with low governance standards.
11
3. Methodology
This methodology chapter will first discuss the dataset used in the paper followed by an explanation of the research method and the formulation of the research hypothesis.
3.1 Dataset
Foreign firms that delisted from the US organized exchanges in the period 2005-2015 were found in the Center for Research in Security Prices (CRSP) database. For this paper it is important that these delistings are on a voluntary basis and not because of for example financial distress, mergers or takeovers. First CRSP was used to find companies that delisted during the studied period. Then the foreign companies were identified by using share codes 12 and 30-32. 12 stands for foreign ordinary listed shares while 30-32 are ADRs. To see whether the companies left for a voluntary reason delisting codes 520 and 570 were used (codes 500, 514 and 519 were also considered but did not result in any foreign delistings during the studied period). Companies were no share code or delisting code was found were removed from the collected data. All companies were cross-checked on EDGAR for SEC file listings 15F or 25, to discover whether the company really voluntarily delisted and to obtain information about the announcement date, level of ADR, reason for firm to voluntarily delist and country of origin. Certain companies had a very long time span between the
announcement date and delisting date (especially in the period 2005-2007), companies with a time span of more than 3 months were removed.
185 companies were found initially, 74 Companies were removed as they did not delist in reality, were in serious financial problems at time of the delisting announcement, were acquired or merged within the next half year or were data could not be found. Leaving a total sample of 111 companies from 28 countries, with the majority from Europe and other western countries. See table 5 and figure 3 at the end of this paper for more information about the country and continent composition. See appendix 1 for a list of all delisted companies with announcement date, delisting date and country of origin. All the Canadian firms are ordinary listings, all other companies were cross-listed through an American depository receipt. The delisted companies are active in various industries including mining, consumer products, telecommunication, pharmaceutics and electronics. A few financial and utility companies are also included in the sample but taken out for part of the regressions.
The vast majorityof the companies that left the organized exchanges kept their depository receipts and instead sponsored over-the-counter trading (79 out of 99 depository listings).
12 The majority of the Company press announcements explaining the delisting decision mentioned costs and complexity associated with American reporting requirements and the substantial increase of costs after the Sarbanes-Oxley Act coupled with low trading volume as reasons for delisting. Various firms mention the good corporate governance principles and accessible market in their home country and tried to convince their shareholders that the American governance control therefore is not necessary. Costs associated with OTC are lower and as firms do not want to lose their American investors many firms keep their ADRs trading in OTC form.
49 out of the 111 companies found were delisted in 2007 (all after March), indicating the importance of rule 12h-6. The other years are similar to each other in amounts of delistings.
Figure 2: Number of delistings in dataset from period 2005 – 2015
Many of the companies in the sample were not only cross-listed on the US exchanges but also at some point in time on other foreign exchanges.
3.2 Model specification.
As this paper is studying share price differences before and after an event (a delisting), event study methodology is used, as is common in virtually all the research on cross-listing and cross-delisting (e.g. Miller 1999; Doidge 2010; Bessler et al. 2012). Abnormal stock return is used as event proxy. Abnormal return is the difference between the actual return of a security and the expected return. Warner & Brown (1985) mention that daily event data usually violates the assumptions of the central limit theorem and are non-normal. They find that in events with a big sample size (n>50) the mean excess return approximates the normal distribution. In this research the sample size is therefore sufficient big. They also show that the market model, which is used in this research outperforms simpler models such as the mean adjusted model. Campbell et al. (2009) provide extra support for the market model in
0 10 20 30 40 50 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015
13 multi country event studies. They find that the model in combination with local market
indexes are sufficient to produce well-specified and powerful tests.
In cross-listing and delisting papers it is unclear whether you should study the announcement day of the (de)listing or the (de)listing date itself, both options are used frequently to describe abnormal behaviour of company fundamentals. In this research the focus is on the announcement day of the delisting. As Kadlec & McConnell (1994) mention. in an efficient market the stock price reflects full information available and thus the valuation effects should be included in the stock price starting at the initial announcement date.
To estimate the abnormal returns the Market model is used, the market model estimates a time window before the event as normal period and uses the found regression intercept and beta correlation with a stock market index to calculate expected return during the event period. The used time window and event methodology is similar to Miller (1999), who looked at the effects of cross-listing on market value. The estimation period taken is 125 days till 25 days before the event {-125,-25}. 4 different event windows are considered to find out whether there is a short and or medium term effect around the delisting
announcement on stock returns. The different event windows are 25 days before and after the announcement of the delisting {-25,25}, the 25 days after announcement {0,25} , the 5 days before and after announcement {-5,5} and 1 day before and after {-1,1}.
Of the firms in the final dataset daily stock prices in their home country are collected on Datastream and Compustat. The market index based on the respective major stock
exchange in the different countries is received from the websites of the different indices and Compustat. All firm and index data has been double checked on Bloomberg. Returns are calculated on closing prices in local currency adjusted for dividends and stock splits. Then the returns of stock and market of firm i on day t of the estimation period are used in the market model:
!",$ = '"+ )"!*,$ + +",$
to estimate firm beta ()) and intercept (a) for the whole estimation window of firm i. The standard error of the regression (-) and the R-squared (R2) are also calculated. The predicted
beta and intercept from the estimation window are used to calculate the daily expected returns .(!I,t) during the event window; .(!I,t) = 'I + )I Rm,t Then the daily abnormal return ARi,t is calculated for firm i on day t:
14 0!",$ = !",$− .(!",$)
Where Ri,t is the actual return of the stock during the individual event days.
The standard t-test: 234
5 is used to calculate if the individual abnormal returns are significantly
different to zero. Then the abnormal returns for a stock are added together to get the cumulative abnormal return (CAR) of the specific company during the event window:
CAR = 6 0!
$78
The t-test is calculated again 923
5/ ; for the stock as a whole, were n stands for the amount of
days (individual abnormal returns) in the event window. The t-test for the individual AR and the CAR will tell whether the different stocks have a significant difference in stock price before and after the delisting announcement which could be explained by the delisting. Finally all CARs of the different stocks are averaged together to get the cumulative average abnormal return (CAAR) of all firms during the event window and the t-test is calculated one more time to find the overall abnormal stock price behaviour around announcement.
Explaining the differences in abnormal returns of the companies in the dataset
The cumulative abnormal returns of the different time periods will be used to perform multiple OLS regressions to see what exactly impacts the abnormal return. It tries to explain the returns based on the sort of stock that is delisted (ordinary, level 2 or level 3), whether it is a value or growth firm, on whether the company still holds a sponsored Level 1 ADR and whether the company delisted before or after March 2007 and could use rule 12h-6. The regressions control for various firm characteristics that might affect the abnormal returns; size, governance level of the home country, debt levels and return on assets.
The regression will have the following form:
<= = ' + !>?@)8+ A.)B+ !C0 )D+ AE)F+ GCH)I+ A2)K+ C!L)M+ CNO)P+ Q.)R+ +"
where <I stands for the CAR of the company. A description of the regression variables
15 Table 1 Explanation of variables used in Regression
Company accounting & stock data including assets, liabilities, net income, dividents on preferred stock and average outstanding shares are retrieved from Compustat Global & Compustat North America for the fiscal year in which the company delisted, missing data was obtained from the company’s annual reports. Information about announcement date and way of cross-listing was gathered through the US Securities & Exchange commission EDGAR database. Governance
indicators were obtained from the Worldbank website.
Dummy Level 3 ADR is omitted to avoid perfect multicollinearity. There are no significant correlations found between other variables that could lead to mulicollinearity.
Rule LN LE GOV L2 ORD OTC ROA
Dummy: 1 for a delisting after March 21st 2007 (passage of rule 12h-6), 0 if before.
The natural logarithm of total assets (in US$ millions) at the end of the fiscal year in which the cross-listing announcement occurs as a proxy for company size.
Leverage: ratio of total debt to total assets: STUV$ $WV* XWY$Z[U;\ $WV* XWY$$U$][ ]SSW$S The cost of US compliance is an extra burden for an already indebt company. By leaving the US this cost is reduced and a positive reaction of investors on the delisting announcement is expected.
weighted average of two of the Worldwide Governance Indicators, created by the World Bank, as a proxy for the level of governance of the home country. The governance indicators considered are Regulatory quality and control of corruption. The first is regarded as more important for governance level of a company and gets a weight of 2, the latter a weight of 1. The original individual indicator scores are on a -2.5 to +2.5 scale, but for this variable put on a 0-5 scale. The optimal governance level attainable for a country is 15, a maximum score of 10 for disclosure level and 5 for corruption control. The United states had a score of 11.64 in 2007, 11.48 in 2010 and 11.84 in 2015. Governance levels are
calculated for the year of delisting.
Dummy: 1 for a Level 2 American depository receipt, 0 if an ordinary listing or level 3 American depository receipt.
Dummy: 1 for companies that have an ordinary share listing, 0 if ADR. In the studied dataset only Canadian firms have ordinary listings so this dummy is equivalent to 1 Canada 0 other countries. This dummy might therefore also say something about the costs of GAAP.
1 if company left US organized exchanges but still trades OTC, 0 if company left US completely. As US investors, can still easily trade stock if it stays OTC a more positive effect on abnormal return is expected then for firms that leave the US completely.
Return on Assets: 6U$][ 2SSW$S^W$ ";_U*W Investors might have more confidence and higher valuation expectation in companies having higher returns on assets and an announcement of delisting from the US exchanges is expected to be more bearable for investors, thus a positive relationship between ROA and abnormal returns is expected.
16 PE Price-earnings ratio (P/E ratio) which measures the share price of a
company relative to its earnings per share (EPS) at the end of the fiscal year of the delisting announcement. A firm with a high P/E ratio is often a growth firm. If company induces net loss then P/E ratio is assumed 0. Reese & Weisbach (2002) mentioned in part 2 described that companies with higher growth rates have a potential higher benefit from the
increased credit ratings generated by cross listing as they can borrow more easily and at a cheap rate. If a firm that announces a delisting is consider a growth firm by investors they probably react more negatively to the announcement.
Most of the regression variables are based on potential significant factors found in earlier research articles on cross-listing and delistings (e.g. Doidge, 2010). For 2 firms; Espirito Santo Financial Group and Tenon LTD, full variable data could not be found and this two companies are left out of part 4.2.
Hypothesis:
Hypothesis 1: The announcement of a voluntary delisting lead to a decrease in the market value of the firm in the home country. Papers such as Miller (1999), You et al. (2012) and Hail & Leuz (2008) have found a positive effect between cross-listing and stock value, so a delisting announcement should logically lead to the opposite.
17
4. Results
4.1 Stock price effect
In this section the abnormal returns and significance of the individual days and companies are described as well as the average and cumulative abnormal returns for the 4 different event windows. The part has a top down approach, considering the total cumulative effects first. It starts with an analysis of the overall observed effects in table 2 by explaining the cumulative average abnormal return in different event windows for all firms together. Figure 3 shows the average abnormal return for each individual day, while table 3 considers the significance of individual stocks by event day.
Table 2: Cumulative average abnormal return different event windows This table displays the cumulative average abnormal returns (CAAR) for 5 different event windows. CAAR is calculated by averaging the abnormal returns for the 111 different stocks at each day in the event window and adding them together to obtain the cumulative average abnormal returns. ** indicates significance from the t-test at the 5% level.
event window CAAR
{-25,+25} -0.020865**
{-5,+5} -0.015866
{-1,+1} 0.002893
{+1,+25} -0.029460**
{+2, +5} -0.007331
As shown in Table 2, in the periods {-25,25} and {+1,+25} all individual abnormal returns taken together are significantly different to 0 and negative. Which means that in the medium term around the delisting announcement abnormal returns significantly deviate from the expected returns and individual stock prices decrease by on average -0.020865 in period {-25,+25} and -0.029460 in period {+1,+25}. The significant negative returns in this two periods are in line with hypothesis 1 that expected negative abnormal returns around the delisting announcement. The abnormal returns in the three shorter periods {-5,+5}, {-1,+1} and {+2, +5} are insignificant from 0, so on average stock prices do not deviate from what was expected under the market model.
That only the medium term event windows are significant could indicate that the stock market needs time to react to the delisting. Press statements announcing the delisting and reasons to leave the US exchanges need to be translated in the local language and
18 investors need time to get to know about the American delisting of their investments before they can react. The potential time lag indicates an inefficiently in the market and deviates from the efficient market hypothesis.
In summary there is a cumulative significant abnormal return in the medium term periods, and insignificant abnormal returns in the short term.
Table 2 however is too generic and in reality the overall stock behaviour is less obvious, In the 51 day period around the delisting announcement {-25,25} only 54 of the 111 companies have on average a negative abnormal return, while 56 of the companies have an on average positive abnormal return. The negative abnormal returns are greater than the positive abnormal returns thus creating the significant negative CAAR in the medium term periods. In the shorter term the amount of negative abnormal returns increases (61 in 11 day period) but become smaller.
The significant differences between individual companies and days ask for a more detailed description which follows in figure 3 & table 3.
Figure 3: Average abnormal return for 111 stocks for each individual day in the event window
Figure 3 presents average abnormal returns (AAR) for all companies combined for each individual day in the 51 day {-25,25} event window. There is great fluctuation between different days in the event window and no real clear direction of overall average abnormal return before or after the event window. This can be partly explained by the different ways firms react on the announcement. Some show a highly positive effect while others have a
-0.02 -0.015 -0.01 -0.005 0 0.005 0.01 0.015 0.02 0.025 0.03 -25 -23 -21 -19 -17 -15 -13 -11 -9 -7 -5 -3 -1 1 3 5 7 9 11 13 15 17 19 21 23 25 av er ag e ab nor m al re tu rn
AAR over the {-25,25} days event window
19 negative share price effect, for certain firm returns are only affected a few days while for other companies the effect takes place over a longer period. Other events effecting individual companies/industries that happened during the event window also have an effect on abnormal return and questions the overall relationship between the delisting announcement event and abnormal returns displayed in figure 3 & table 2. A possible time lag between the
announcement and investors reaction mentioned earlier could be present, day 0 and day 1 have a small average abnormal return followed by days with a more significant abnormal return.
As there are considerable differences between firms it is important to study significant abnormal returns of individual companies.
Table 3: Significant individual abnormal returns
Firms from the dataset of 111 companies with significant positive or negative abnormal returns of at least the 10% level (t-test) on the particular event day. Significant abnormal return companies are calculated by dividing the abnormal return for that day for company i with the standard deviation calculated during the estimation period for company i. The first number in the significance column indicates the amount of total (negative or positive) significant abnormal returns out of the 111 companies on that particular day in the event window. The numbers in parentheses indicate the number of companies with significant negative abnormal returns while the first numbers indicate total amount of companies with significant abnormal returns (negative and positive).
DAY SIGNIFICANCE DAY SIGNIFICANCE DAY SIGNIFICANCE DAY SIGNIFICANCE
-25 13 (5) -12 18 (11) 1 10 (5) 14 12 (5) -24 14 (4) -11 14 (7) 2 16 (8) 15 14 (5) -23 13 (6) -10 7 (3) 3 17 (9) 16 12 (6) -22 7 (1) -9 17 (7) 4 18 (7) 17 12 (5) -21 9 (5) -8 16 (8) 5 23 (14) 18 19 (6) -20 12 (7) -7 10 (2) 6 19 (8) 19 16 (5) -19 13 (5) -6 15 (5) 7 16 (9) 20 13 (5) -18 13 (4) -5 11 (5) 8 17 (10) 21 7 (4) -17 11 (5) -4 11 (4) 9 17 (11) 22 9 (4) -16 5 (1) -3 13 (5) 10 11 (8) 23 9 (4) -15 14 (5) -2 14 (9) 11 11 (3) 24 8 (3) -14 10 (4) -1 21 (8) 12 17 (6) 25 8 (5) -13 9 (6) 0 14 (9) 13 17 (6)
Table 3 tells more about the distribution of the significant abnormal returns and the differences between companies.
The amount of significant abnormal returns increases closer to the delisting announcement and more of them are negative. The differences between days is wide, were at day 1 only 10 out of 111 companies have a significant abnormal return, on day 5 this are 23.
From Table 3 you can also see that on any particular day only up to a 1/5 of the companies have a significant abnormal return. Certain companies do not have any significant abnormal
20 returns during the {-25,25} event window and an effect of the delisting announcement on stock prices can therefore not be proven for these firms.
After the delisting announcement (day 0) more firms have a negative abnormal return than before but there are still various companies that show positive returns, again great differences between companies.
The many significant abnormal returns before the announcement raises serious questions about the causality of the announcement and the abnormal returns. It could be that the effect is limited and many of the abnormal returns are caused by other firm or industry independent factors, It could also be that for certain firms the information of the prospective delisting may leak to certain informed investors before the actual announcement by the firm. If this is the case investors will react before the actual delisting announcement affecting abnormal returns before day 0 in the event window.
In summary Section 4.1 reveals that while the cumulative average abnormal returns in the medium event windows are significantly negative which support the hypothesis of this thesis, the abnormal returns and the potential delisting announcement effect greatly differs between companies. The wide variety of abnormal returns between different companies ask for a further study of firm specific measures, part 4.2 will discusses this further.
4.2 Abnormal return explaining factors
This section will discuss various firm and delisting factors that might have had an impact on the size of the individual abnormal returns. Regression outcomes for the medium term event windows {-25,25} and {0,25} around the delisting announcement are discussed first followed by an analysis of the short term event windows {-5,+5} and {-1,+1}.
Table 4: Stock price reaction {-25,+25} & {0,+25} event window around announcement Regressions estimation of the medium term periods around the announcement of the delisting. Where regressions 1 – 4 observe the 25 day period before and after announcement, regressions 5-8 consider only the period afterwards. 1 and 5 consider all variables. For 2 companies full data could not be found and they are taken out of the regression, total number of observations N = 109. Regressions 3,4 and 7,8 exclude financial and utility companies and companies with big outliers (in both directions) in either size, ROA or leverage, leaving a sample of 93 companies. Regressions 2, 4, 6 and 8 look at the more significant variables. All regressions test significant for heteroscedasticity (Breusch-Pagan test) so robust standards errors are used. Regressions are Ordinary least squares. *, ** and *** indicate t-test significance at the 10%, 5% and 1% level respectively. Standard errors are in parentheses.
21 Period {-25,+25} Period {+0,+25} (1) (2) (3) (4) (5) (6) (7) (8) constant -0.563 (0.387) -0.716** (0.34) -0.210 (0.120) -0.193 (0.106) -0.348 (0.271) -0.334 (0.233) -0.130 (0.116) -0.056 (0.042) LN 0.003 (0.02) 0.007 (0.009) -0.007 (0.015) 0.018* (0.009) 0.015* (0.008) LE 0.226 (0.177) 0.187 (0.15) 0.234*** (0.086) 0.254*** (0.084) 0.100 (0.124) 0.066 (0.108) -0.061 (0.041) ROA 0.165 (0.33) 0.280 (0.177) 0.334* (0.195) -0.002 (0.230) -0.0002 (0.126) Rule 0.038 (0.143) -0.028 (0.048) 0.078 (0.099) 0.048 (0.047) GOV 0.048* (0.028) 0.58** (0.026) 0.011 (0.010) 0.010 (0.009) 0.034* (0.020) 0.031* (0.018) -0.001 (0.009) OTC -0.089 (0.123) -0.077)* (0.042) -0.064 (0.039) -0.056 (0.086) -0.047 (0.041) -0.056 (0.042) L2 -0.137 (0.122) -0.121 (0.096) 0.009 (0.045) -0.159* (0.085) -0.146** (0.067) -0.061 (0.041) -0.078** (0.036) ORD -0.051 (0.192) -0.025 (0.098) -0.060 (0.134) 0.094 (0.100) PE -0.003 (0.003) -0.003** (0.001) -0.003* (0.001) 0.001 (0.002) -0.001 (0.0008) Adjusted R2 F-test model 0.0019 F(9,99) 1.02 0.042 F(3,105) 2.58 0.1159 F(9,83) 1.69 0.1465 F(5,87) 2.80 0.035 F(9,99) 1.04 0.050 F(3,105) 2.91 0.031 F(9,83) 0.91 0.060 F(3,89) 2.05 Prob>F 0.4271 0.0577 0.104 0.0214 0.4123 0.0381 0.52 0.11
Table 4 presents 8 different regressions in two medium term event windows around the announcement of the delisting. The governance is significant in the period {-25,25} and less but still significant in the after-announcement period. A significant governance effect
corresponds with the bonding theory, but the direction is surprising. Expected is a decrease in abnormal return when the company is from a country with higher corporate governance standards. This is however explained by the fact that although the overall Cumulative average abnormal return in the period is negative many of the individual stocks have a positive return. Individual stock data also shows a relation between higher governance level and positive abnormal return. Investors are aware that the US listing cost the firm a lot of money which they could spent on other things. They are also not afraid of a governance problem as the home country is well regulated, and know they get little in return of extra investors as US trading volume is almost always very low (<5%) on time of delisting thus investors react positively on the announcement.
Regressions 1,2, 5 and 7 however have insignificant F-statistics and low adjusted R2 doubting the quality of these regressions output and indicating that potential explanatory
22 variables (if any) are missing. It cannot be proven that the output is not just chance and we should be extremely careful with drawing valid conclusions out of these regressions. The non-significance of governance in the other regressions put doubt on the certainty of the bonding effect.
The regressions with outliers and financial & utility companies removed seem to have a better explanatory value. They show a highly positive significant effect of Leverage on abnormal return in the {-25,25} event window. Companies listed in the United States induce high listing and compliance costs which are a great burden for companies who already suffer from high debt. An announcement of delisting reduces the fixed burden of the company and could result in a positive reaction by shareholders. It is however surprising that there is no significance effect for leverage in the event window after the announcement {0,25}. This proves that the leverage effect is only strong in the period before the announcement and if you assume that the delisting is not known to any investor before the official announcement, the causality between high leverage & the delisting announcement seems to be missing. The causality of the significant PE variable in the period {-25,25} should similarly be questioned. The direction is as expected negative, if a firm has a higher P/E ratio, investors anticipating higher growth rates in the future which could benefit from the increased credit ratings gained through cross-listing.
In the 25-day period after announcement there is a significant negative effect on the abnormal return if the firm is delisting a level 2 American depository receipt. There is a negative correlation of -0.2 between governance levels and being a level 2 depository receipt, As Governace levels are lower it could be that the investors of many of the firms who have a level 2 receipt benefit more from the increased compliance requirements of the US cross-listing and reacting negative when the decross-listing is announced.
OTC and size are also found to have a small significant effect on abnormal returns.
Table 5: Stock price reaction in {-5,5} & {-1,1} event window around announcement Regressions estimations of the short term periods around the announcement of the delisting. Where regressions 1 – 4 observe the 5 day period before and after announcement, regressions 5-8 only consider a 3 day period. Regressions 1 and 5 consider all variables. For 2 companies full data could not be found and they are taken out of the regression, total number of observations N = 109. Regressions 3,4 and 7,8 exclude financial and utility companies and companies with outliers (in both directions) in either size, ROA or leverage, leaving a sample of 93 companies. Regressions 2,4,6 and 8 look at the more significant variables. All regressions test significant for heteroscedasticity (Breusch-Pagan test) so robust standard errors are used. Regressions are Ordinary least squares. *, ** and *** indicate t-test significance at the 10%, 5% and 1% level respectively. Standard errors are in parentheses.
23 Period {-5,5} Period {-1,1} (1) (2) (3) (4) (5) (6) (7) (8) Constant -0.277 (0.225) -0.302 (0.218) -0.099 (0.122) -0.122 (0.084) -0.148 (0.096) -0.131 (0.098) -0.089 (0.054) -0.062* (0.036) LN -0.004 (0.006) -0.004 (0.006) 0.004 (0.004) 0.007 (0.006) 0.007 (0.005) LE 0.069 (0.089) 0.051 (0.071) 0.007 (0.092) 0.030 (0.035) 0.045 (0.028) -0.008 (0.030) ROA 0.057 (0.077) 0.017 (0.110) -0.052 (0.049) -0.106 (0.086) -0.093 (0.089) Rule 0.028 (0.046) 0.002 (0.040) 0.011 (0.021) -0.004 (0.020) GOV 0.023 (0.019) 0.025 (0.016) 0.007 (0.008) 0.009 (0.007) 0.009 (0.008) 0.009 (0.007) 0.002 (0.003) OTC -0.002 (0.041) 0.025 (0.026) 0.002 (0.038) 0.013 0.018 0.022 (0.014) 0.022* (0.013) L2 -0.051 (0.045) -0.051 (0.041) 0.007 (0.032) -0.021 (0.022) 0.013 (0.014) ORD 0.024 (0.117) 0.057 (0.111) -0.020 (0.041) 0.008 (0.036) PE -0.001 (0.006) -0.0001 (0.001) -0.0001 (0.003) -0.0001 (0.0003) Adjusted R2 0.022 0.031 -0.01 0 -0.001 0.023 0.032) 0.089 F-test model F(9,99) 0.90 F(3,105) 0.96 F(9,83) 0.44 F(2,90) 0.84 F(9,99) 1.15 F(2,106) 1.37 F(9,83) 1.2 F(3,89) 3.06 Prob>F 0.53 0.41 0.91 0.43 0.3362 0.25 0.31 0.032
In the shorter periods around the announcement of the delisting there is very little
significance. This supports the data found in part 4.1 and could be because the market needs time to accurately respond to the announcement. Adjusted R2 and F-statistics are low in all but regression 8 so there is very little explanatory value.
In short the explanatory value of the regression variables on abnormal return seems to be weak overall but stronger in the two medium term event windows around the delisting announcement.
24
5. Conclusion
This paper considered a sample of 111 companies that announced a voluntarily delisting from the US organized exchanges in the period 2005-2015. The results indicate an overall
significant negative abnormal return around the time of announcement in the medium term, but no significant abnormal return in the short term. The conclusion on the medium term is in line with the hypothesis of this paper and the result of Doidge (2010) and You et al. (2012). The result indicates that on average a delisting of a cross-listed position has the opposite effect of a cross-listing. Governance level of the home country is found to have a significant positive effect in the medium term on the absolute return. The importance of governance on cross-listing value is also described in various other research papers such as Doidge (2010), Foerster and Karolyi (2000) and Miller (1999). Leverage, P/E ratio and the kind of American depository receipt that gets delisted also show significance.
Managers of cross-listed firms could look at research such as this to estimate the effects a potential (voluntarily) delisting has on the firm and to evaluate whether the various benefits of cross-listing outweigh the high costs of being listed in the United States.
We should however be careful with the results of this study, as averaged abnormal returns fluctuate greatly between individual days in the event window and the abnormal returns of the individual companies are diverse and found to be insignificant for the majority of the companies. The individual significant abnormal returns might also (partly) be caused by other factors influencing the individual companies which are not related to the delisting. The many significant abnormal returns displayed in table 3 for the days before the
announcement of the delisting are also questionable. Most of the performed regressions have a low explanatory power with insignificant f-statistics and we should therefore be critical of the found results. Various factors mentioned earlier that are affected by cross-listing such as the visibility level and home bias, might also impact the abnormal return after announcement of a delisting but these variables are very hard to measure in the internet age.
Further research could try to find factors that better explain the abnormal return differences between firms. It is also worthwhile to research the market returns around the actual delisting implementation and observe the effects over a longer event window.
25 Table 6 – Dataset: delistings by country in period 2005-2015
Country data is retrieved from the US Securities & Exchange commission EDGAR database.
COUNTRY AMOUNT USED MARKET INDEX
AUSTRALIA 9 ASX 200
AUSTRIA 1 Austrian Traded Index (ATX)
BELGIUM 1 Bel 20
CANADA 12 S&P/TSX Composite index
CHILE 1 IPSA SANTIAGO DE CHILE
FINLAND 2 OMX Helsinki 25
FRANCE 13 Cac 40
GERMANY 12 DAX
GREECE 1 Athens Stock Exchange General Index HONG KONG 3 Hong Kong Hang Seng Index
INDIA 2 BSE SENSEX Index
ISRAEL 4 TA-100 index
ITALY 3 S&P/MIB
JAPAN 6 Nikkei 225
SOUTH KOREA 1 Kospi index
LUXEMBOURG 1 Luxx
MEXICO 4 indice de Precios y Cotizaciones
NETHERLANDS 6 AEX index
NEW ZEALAND 1 NZX 50
NORWAY 2 OBX index
PORTUGAL 1 Used Spain: IBEX 35
RUSSIA 1 RTSI
SOUTH AFRICA 3 FTSE/JSE all-share
SPAIN 2 IBEX 35
SWEDEN 3 OMX STOCKHOLM 30 INDEX
SWITZERLAND 5 Swiss performance index (SPI)
TAIWAN 1 TSEC weighted index
26 Figure 4: Dataset: delisting by continent in period 2005-2015
Continent composition of the 111 companies in the dataset that announced a delisting from the US organized exchanges in the period 2005-2015.
Bibliography
Ahearne, A., Griever, W., & Warnock, F. (2004). Information costs and home bias: An analysis of US holdings of foreign equities. Journal of International Economics, 62(2).
Baker, H., Nofsinger, J., & Weaver, D. (2002). International Cross-Listing and Visibility. The
Journal of Financial and Quantitative Analysis 37(3).
Bekeart, G., & Harvey, C. R. (1995). Time-Varying World Market Integration. The Journal of
Finance, 50(2).
Bessler, W., Kaen, F. R., Kurmann, P., & Zimmermann, J. (2012). The listing and delisting of German firms on NYSE and NASDAQ: Were there any benefits?
Journal of International Financial Markets, Institutions and Money, 22(4).
BNY Mellon. (2015). 'The depositary receipt market review', Market revieuw. BNY Mellon. (2016). 'The depositary receipt market review ', Market review.
Brown, S. J., Warner, J. B. (1985). Using daily stock returns: the case of event studies. Journal
Of Financial Economics, 14(1).
Burns, N. (2004). The role of cross-listed stocks as an acquisition currency: Evidence from takeovers of U.S. firms, working paper, University of Georgia, Athens, GA.
Campbell, C. J., Cowana, A. R., & Salotti, V. (2010). Multi-country event-study methods, Journal of
Banking & Finance, December, 34(12).
Coffee, J. (2002). Racing towards the top? The impact of cross-listings and stock market competition on international corporate governance, Columbia Law Review, 102(7).
Delistings by continent
27 Doidge, C., Karolyi, G. A., & Stulz, R. M. (2010). Why do foreign firms leave US equity markets?,
The Journal of Finance, 65(4).
Eun, C., & Sabherwal, S. (2003). Cross-border listings and price discovery: Evidence from US-listed Canadian stocks, Journal of Finance, 58(2).
Fernandes, N., Lel, U., & Miller, D.P. (2010). Escape from New York: the market impact of loosening disclosure requirements, Journal of Financial Economics 95(2).
Foerster, S. R., & Karolyi G. A., (1999). The effects of market segmentation and investor recognition on asset prices: Evidence from foreign stocks listing in the United States, The Journal of
Finance, 54(3).
Foerster, S. R., & Karolyi, G. A. (2000). The long-run performance of global equity offerings,
Journal of Financial and Quantitative Analysis, 35(4).
Hail, L., & Leuz, C. (2008). Cost of Capital Effects and Changes in Growth Expectations Around U.S. Cross-Listings, Journal of Financial Economics, 93(3).
Kadlec, G., & McConnell, J. (1994). The Effect of Market Segmentation and Illiquidity on Asset Prices: Evidence from Exchange Listings, The Journal of Finance, 49(2).
Karolyi, G. A. (2004). The Role of ADRs in the Development of Emerging Equity Markets, AFA
2004 San Diego Meetings.
Karolyi, G.A. (2006). The World of Cross-Listings and Cross-Listings of the World: Challenging Conventional Wisdom, Review of Finance, 10(1).
Lang, M., Lins, K., & Miller, D. P. (2003). ADRs, analysts and accuracy: Does cross-listing in the U.S. improve a firms information environment and increase market value? Journal of
Accounting Research, 41(2).
Lel, U., & Miller, D. P. (2008). International Cross-Listing, Firm Performance, and Top
Management Turnover: A Test of the Bonding Hypothesis, The Journal of Finance, 63(4). Levine, R., & Schmukler, S. (2006). Internationalization and stock market liquidity, Review of
Finance, 10(1).
Merton, R. C. (1987). A Simple Model of Capital Market Equilibrium with Incomplete Information, The Journal of Finance, 42(3).
Miller D. P. (1999). The market reaction to international cross-listings: evidence from depository receipts, Journal of Financial Economics, 51(1).
Piotroski, J. D., & Srinivasan, S. (2008). Regulation and Bonding: The Sarbanes-Oxley Act and the Flow of International Listings. Rock Center for Corporate Governance at Stanford
University, Working Paper 11.
Pulatkonak, M., & Sofianos, G. (1999). The distribution of global trading in NYSE-listed non-U.S. stocks, New York stock exchange working paper 99(3).
Reese, W., & Weisbach, M. (2002). Protection of minority shareholder interests, cross-listings in the United States, and subsequent equity offerings, Journal of Financial Economics, 66(1).
28 Stulz, R. M. (1999). Globalization, corporate finance, and the cost of capital, Journal of Applied
Corporate Finance 12(3).
Witmer, J. L. (2005). Why do Firms Cross-(de)list? An Examination of the Determinants and Effects of Cross-delisting, Bank of Canada.
You, L., Parhizgari, A., & Srivastava, S. C. (2012). Cross Listing and Subsequent Delisting in Foreign Markets, Journal of Empirical Finance 19(2).
Appendix 1: Full list of all delisting companies
Delisting time and delisted companies are retrieved from Wharton research data services (WRDS). Announcement date and home country is gathered through the US Securities & Exchange commission EDGAR database.
delisting time company name home country
announcement date
29/06/2007 Amcor Limited Australia 02/05/2007
21/06/2007
NATIONAL AUSTRALIA BANK
Limited Australia 10/05/2007
31/07/2007 AUSTRALIA & NEW ZEALAND bank Australia 20/06/2007
31/07/2007 Atlas Pearls and Perfumes Limited Australia 11/07/2007
08/07/2009
CHEMGENEX PHARMACEUTICALS
LTD Australia 29/06/2009
30/07/2009 Santos Limited Australia 30/06/2009
22/07/2009 PHARMAXIS LTD Australia 13/07/2009
31/10/2013 SIMS METAL MANAGEMENT LTD Australia 30/09/2013
28/02/2014 ALUMINA LTD Australia 18/02/2014
04/06/2007 TELEKOM AUSTRIA AG Austria 24/04/2007
06/11/2006 Icos Vision Systems Corp Belgium 26/10/2006
14/06/2006 FNX Mining Company Inc Canada 23/05/2006
26/09/2006 Alliance Atlantis Communications Canada 26/09/2006
12/03/2007 PoLYAIR INTER PACK INC Canada 20/02/2007
28/02/2007 Dectron internationale Canada 28/02/2007
21/10/2007 WESTAIM CORP Canada 20/10/2008
01/04/2009 LUNDIN MINING CORP Canada 13/03/2009
02/12/2009 INTERTAPE POLYMER GROUP Canada 12/11/2009
04/08/2010 CORUS ENTERTAINMENT INC Canada 15/07/2010
31/07/2012 HELIX BIOPHARMA CORP Canada 19/07/2012
19/02/2015 QUEST RARE MINERALS LTD Canada 09/02/2015
31/07/2015 ATLATSA RESOURCES CORP Canada 20/07/2015
27/10/2015 ALDERON IRON ORE CORP Canada 02/10/2015
15/07/2005 Cristalerias de Chile Chile 19/04/2005
17/09/2007 METSO CORP Finland 26/07/2007
31/12/2007 STORA ENSO CORP Finland 05/12/2007
23/09/2005 Transgene SA France 21/09/2005
04/06/2007 SCOR SE France 03/04/2007
29
31/05/2007 GENESYS S A France 10/05/2007
16/07/2007 SODEXHO ALLIANCE S A France 30/05/2007
06/08/2007 TECHNIP France 25/07/2007
28/09/2007 RHODIA France 31/07/2007
07/09/2007 PUBLICIS SA NEW France 07/09/2007
29/02/2008 AIR FRANCE KLM France 22/11/2007
31/10/2008 DASSAULT SYSTEMES INC France 31/07/2008
25/03/2010 AXA France 25/01/2010
31/03/2011 TECHNICOLOR France 24/03/2011
31/12/2014 VEOLIA ENVIRONNEMENT France 23/12/2014
26/06/2007 SGL CARBON AKTIENGESELLSCHAFT Germany 26/03/2007 31/05/2007 ALTANA AG Germany 26/04/2007 28/09/2007 BASF AG Germany 30/07/2007 28/09/2007 EON AG Germany 21/08/2007
04/10/2007 PFEIFFER VACUUM TECH AG Germany 30/08/2007
28/09/2007 BAYER AG Germany 05/09/2007
30/11/2007 EPCOS AG Germany 08/11/2007
23/04/2009 INFINEON TECHNOLOGIES A G Germany 03/04/2009
23/10/2009 ALLIANZ SE Germany 22/09/2009
27/11/2009 EVOTEC AG Germany 10/11/2009
18/06/2010 DEUTSCHE TELEKOM AG Germany 21/04/2010
04/06/2010 DAIMLER AG Germany 14/05/2010
17/09/2010 HELLENIC TELECOMMUNICATIONS Greece 12/05/2010
17/05/2007 PCCW LTD Hong Kong 27/04/2007
25/01/2008
ASIA SATELLITE TELECOM HLDGS
LTD Hong Kong 08/01/2008
06/08/2008 APT SATELLITE HOLDINGS LTD Hong Kong 17/07/2008
31/12/2012 MAHANAGAR TELE NIGAM LTD India 31/12/2012
28/06/2014 TATA COMMUNICATIONS LTD India 26/06/2014
07/12/2005 Robogroup Israel 08/09/2005
29/06/2007 KOOR INDUSTRIES LTD Israel 14/05/2007
26/03/2008 DELTA GALIL INDUSTRIES INC Israel 04/03/2008
06/01/2010
ELRON ELECTRONIC INDUSTRIES
LTD Israel 12/11/2009
29/06/2007 DUCATI MOTOR HOLDING SPA Italy 14/05/2007
03/11/2007 FIAT SPA Italy 03/08/2007
31/10/2007 BENETTON GROUP SPA Italy 12/09/2007
18/12/2006 Pioneer Corp Japan 18/12/2006
08/08/2008 NIS GROUP CO LTD (Nissin) Japan 14/07/2008
08/01/2009 Daiei INC Japan 24/12/2008
30/04/2013 MAKITA CORP Japan 22/04/2013
30/04/2013 WACOAL HOLDINGS CORP Japan 25/04/2013
24/04/2015 KONAMI CORP Japan 01/04/2015
