abnormal returns
Evidence from the US
Author:
Mark Wijnholds
s1541897
University of Groningen
Faculty of Economics and Business
MSc Finance
Supervisor
2
Governance-based strategies and learning effects: the demise of
abnormal returns
Abstract
An investment strategy that bought well-governed companies and sold poorly-governed companies in the US equities market would have resulted in an abnormal return of 57 basis points during the 90s. However, it failed to do so in the 00s. It is hypothesized that the demise of abnormal returns is caused by market participants gradually learning the effects of differences between well-governed and poorly-governed companies on future profitability. Once a large number of market participants appreciate this difference trading based on differences in corporate governance does no longer yield abnormal returns. This is the first study that examines returns over two decades and that also incorporates the current financial crisis in its analysis. The learning factors that are hypothesized to explain the demise of abnormal returns consist of academic research, code and regulation issuance, media coverage and equity holders proposals submitted by institutional investors. While the majority is researched before, codes issuance is an important new variable subject to research. It is found that academic research and submitted proposals have a significant negative influence on abnormal returns. The effect of media coverage is puzzling and code issuance is insignificant.
JEL codes: G11, G30
3
Table of Contents
1. Introduction 4
2. Literature review 7
3. Hypotheses 10
3.1 Introduction to the Learning Hypothesis 10 3.2 Attention to Governance 11 3.2.1 Academic Research 11 3.2.2 Governance Codes and Guidelines 12 3.2.3 Media Coverage 14 3.2.4 Institutional Investors 16
4. Financial Crisis 18
5. Data 21
5.1 Sample Selection and Data Sources 21 5.2 Descriptive Statistics 23 5.3 Governance and Abnormal Returns 24
6. Empirical Results 28 6.1 Abnormal Returns 28 6.2 Academic Research 28 6.3 Code Issuance 28 6.3 Media Coverage 29 6.5 Institutional Investors 29 6.6 Financial Crisis 30 7. Conclusions 31
7.1 Summary of Key Findings 31 7.2 Suggestions for Future Research 32
8. References 33
4
1. Introduction
After the publication of a seminal article on corporate governance and company performance by Gompers, Ishii & Metrick (2003) (hereinafter GIM) a great deal of attention is given to their results and it spurred a growing body of academic research. Although this started already in the early 2000s governance and company performance still remain hotly debated in the aftermath of the US sub-prime mortgage rooted most recent financial crisis. The view adopted in this research is that of the individual and professional investor and his ability to be able to earn abnormal returns over
prolonged periods of time by utilizing a governance-based strategy. It can be argued that it is almost a natural instinct for these investors to hunt for opportunities to generate abnormal returns from their trading strategies. In academic literature this is translated into an ever expanding body of journal articles on asset pricing and the presence of abnormal returns. Besides the asset pricing component also behavioral finance plays a role because this paper makes an effort to explain the governance-based strategy returns with the learning hypothesis. Therefore, not only does this research contribute to the literature on governance-based strategies, but also it is related to the large body of asset pricing and behavioral finance literature on the presence and vanishing of abnormal returns associated with trading strategies based on public information.
Bebchuk, Cohen & Wang (2010) define the learning hypothesis as “a state in which initially market
prices do not accurately reflect the expected effects of differences between well-governed and poorly-governed companies on future profitability, but over time a sufficient number of market participants have learned to appreciate the significance of these differences, making trading on the basis of such differences no longer profitable”. The market participants reflect the individual and institutional
investors and future profitability is measured through equity returns. The time frame in which learning could occur runs from 1990 to 2011, which matches the time in which the strategy is employed.
5
Fourth, the hypotheses will be tested followed by a section devoted to the strategy’s performance during the most recent financial crisis. At last, conclusions will be given and suggestions for future research.
Apart from the contributions already mentioned this paper is the first to include governance codes and regulations as a learning factor. The learning hypothesis is first drafted by Bebchuk, Cohen & Wang (2010) and it takes place through increased attention by media, institutional investors and academic research. This leaves room for improvement and hence it is argued that it does not
represent a complete set of factors that cause gradual learning. Therefore, institutions are added and more precisely all institutions from governments to equity exchanges that issue governance codes and regulations. This at least theoretically can improve the explanatory power of the combined learning factors, because the issuance of codes has several beneficial features. Namely, they help explain corporate governance practices to investors and increase the attention paid to governance. Thus, the beneficial effects of code issuance are both in learning and attention.
There is only limited coverage of abnormal returns after 2000 even in the US, therefore the third contribution is performing a longitudinal study which includes the late 2000s. It is important to perform a longitudinal study, because the profitability of investment strategies based on historical equity returns is notoriously vulnerable to time periods selected by the researcher. Second, the demise of abnormal returns that have been reported so far is confined to small time periods in the early 2000s. Bebchuk, Cohen & Wang (2010) take a larger time period until 2008, but still little is known about the late 2000s and thus about periods of financial crisis.
Relating the governance-based strategy to the most recent crisis is the fourth contribution. The current financial crisis put the governance quality of companies on the map again. The OECD (2012) argues that the financial crisis revealed severe shortcomings in corporate governance. When most needed, existing standards failed to provide the checks and balances that companies need in order to cultivate sound business practices.
6
7
2. Literature Review
While there is no sole definition of corporate governance, the one from Shleifer & Vishny (1997) fits the most with governance-based strategies. It reads as follows: “corporate governance deals with the
ways in which suppliers of finance to companies assure themselves of getting a return on their investment”. An investor who invests using a governance-based strategy is the supplier of finance
who seeks return, and in this case abnormal returns from selecting equities that score relatively well and bad on corporate governance. Also, this definition implies that equities from companies with a governance structure in favor of equity holders are more likely to yield a return or, put in the words of Shleifer & Vishny (1997), investors are more assured of getting a return. Furthermore, a research with corporate governance as its subject is not complete without first turning to agency costs. In an article by Jensen & Meckling (1976) the principal-agent problem is discussed that gives rise to agency costs. These costs arise because the agent or manager will not always act in the best interests of the principal or investor. Agency costs are made up of monitoring and auditing costs at the expense of the investor. Good governance, thus low agency costs, may lessen investor’s monitoring and auditing costs, plus it has the ability to lower expropriation by managers thereby increasing return on
investment. On the other hand weak governance increases these agency costs and therefore lowers return on investment.
8
1990 to December 1999 an investment strategy that bought the Democracy Portfolio and sold the Dictatorship Portfolio would have yielded an abnormal return of 8.5 percent per year.
The article of GIM spurred greater attention to governance-based strategies in academic literature. Core, Guay & Rusticus (2006) question the results of these authors and investigate whether their results are time-period specific by employing the same long-short strategy over the period 2000 to 2003. They conclude that differences in equity holder rights cannot cause higher returns and that most likely time-period specific returns play a role. Also, Bebchuk, Cohen & Wang (2010) state that for a US sample between 2000 and 2008 no abnormal returns are yielded. It is striking that
governance-based strategies are unable to yield abnormal returns in the first decennium of the 21st century. There are however studies that do find abnormal returns when periods in the early 2000s are looked upon. Despite the fact that these studies include non-US samples still developed countries are included. An article by Drobetz, Schillhofer & Zimmermann (2004) covered the German market and abnormal returns were observed. Furthermore, Bauer, Guenster & Otten (2003) observed economically large excess return in the UK and somewhat smaller excess returns in EMU countries. However, the research from both Drobetz, Schillhofer & Zimmermann (2004) and Bauer, Guenster & Otten (2003) suffers from look-ahead bias initiated by the lack of multi-year data on company level governance quality. This lowers internal validity and therefore it is argued that abnormal returns in the 2000s are unlikely, given previously cited and look-ahead bias free work of Core, Guay & Rusticus (2006) and Bebchuk, Cohen & Wang (2010). Subsequently, it can be argued that it is fair to state that a long-short governance-based strategy using a US sample only provides abnormal returns for the 1990s while failing to do so for the 2000s.
This is puzzling and explanations are needed. In his work about efficient markets Fama (1970) states that current prices fully reflect all publicly available information. Therefore, a governance-based strategy that employs publicly available information cannot according to this theory produce
abnormal returns. Although the information needed for the long-short governance is not for free, the vendor the Investor Responsibility Research Centre (IRRC) from which the governance information is obtained started to provide information on corporate governance to institutional investors,
9
profits will disappear over time. Because technology and again information to pursue a profitable strategy becomes accessible to all investors after a while. Concerning the governance-based strategy this might have happened in the early 2000s.
However, the latter explanation focuses on technology and data availability and does not take into account an investor’s cognitive ability to correctly interpret data and to extract relevant
informational contents in data to come up with a governance-based strategy that can yield expected abnormal returns. In line with this cognitive ability notion a possible explanation could be that investors in the 1990s underestimated the effect that weak governance has on agency costs. Weak governance is associated with both higher capital expenditures and higher acquisition activity (GIM, 2003). And thus weak governance fuels agency costs. If corporate governance and its effect on return is not properly understood by investors it could lead to such an underestimation. Increased attention to governance in the 2000s can lead to investors correctly pricing the value effects of corporate governance and weed out the abnormal returns from governance-based strategies. Bebchuk, Cohen & Wang (2010) state that the lack of abnormal return in the 2000s were due to market participants’ gradually learning to appreciate the difference between companies scoring well and poorly on their governance rating. Included in this learning hypothesis is the condition that a sufficient number of investors must have learned the difference between well-governed and poorly-governed companies. And that they collectively perform trades based on this knowledge which in turn makes a
10
3. Hypotheses
3.1 Introduction to the Learning Hypothesis
The observation that the governance-strategy yields abnormal returns in the 1990s and fails to do so in the 2000s is presented as an anomaly. The efficient market hypothesis is based on the assumption that markets distill new information with speed and that prices accurately reflect this information. This is at odds with the observation that during the whole 90s abnormal returns existed. The role of learning is typically ignored, but it might be critical for understanding anomalies (Lewellen & Shanken, 2002). Therefore, the main concern of this section is that this anomaly is easier to understand if it becomes apparent that the long-term profitability or non-profitability of an investment strategy is subject to learning.
Before turning to the learning hypothesis it can be argued that legal developments shaped the significance of the governance provisions used by GIM in the 1990s. Therefore, one would observe abnormal returns in this time period. However, Bebchuk, Cohen & Ferrell (2010) argue that this cannot be the case, because the legal development that shaped the governance provision’s significance were largely in during the beginning of the 1990s. By which they mean that during the 1980s Delaware courts issued rulings expanding the power of boards to use governance provisions to “just say no” to acquisition offers they view as undesirable. This argument reinforces the need to put effort into researching the effects of learning.
11
mentioning that corporate governance has an impact on company profitability thereby attracting investors’ attention. When the media started to cover corporate governance investors have an opportunity to gradually learn the difference between well- and poorly-governed companies. Learning can make prices move to their “fundamental values” over time as investors update their beliefs (Lewellen & Shanken, 2002). In this particular situation “fundamental values” refers to equity prices in the 1990s moving to a state in the 2000s in which they fully reflect company level
differences in governance.
The learning hypothesis rests on a certain assumptions. First, an investor must exert sufficient effort to find news that can reveal relevant information. Second, a sufficient number of market participants must perform trades on company specific governance information so that prices are influenced. Third, sufficient capital must be available to benefit from the strategy. When these assumptions are met it is no longer possible to profit from a governance-based strategy in the long-term. To test whether these assumptions are met the measures of attention in the subsequent section will be directly related to the hedge portfolio’s abnormal return. It must be noted that this research deviates from earlier research on the learning hypothesis by Bebchuk, Cohen & Wang (2010) in that it does not put all learning factors in an attention index. This research deals with the factors individually to be able to make conclusions about which factors influence abnormal returns and which do not. This to contribute to future research on learning and investors trading.
3.2 Attention to Governance
In this section selected quantitative measures will be given that theoretically should influence gradual learning by investors. It is argued that learning can only happen if first these measures capture investor attention to corporate governance in order to create learning effects. These measures include: academic research, code issuing institutions, media coverage and submitted proposals by institutional investors.
3.2.1 Academic Research
12
Published academic journal articles that have as its subject corporate governance are looked upon. A search was performed in the well renowned databases Academic Source Premier, Business Source Premier and Econlit. Figure 1 presents the time series.
Figure 1
The time series shows a gradual increase in journal articles during the 1990s. Followed by a steep increase in the early 2000s and ultimately ending in a steady state in the late 2000s. Therefore, the hypothesis reads as follows:
H1: The number of journal articles published is negatively related to the hedge portfolio abnormal
returns.
Support for this hypothesis is given by Bebchuk, Cohen & Ferrell (2010). Academic research is one of the learning factor present in their attention index. They found that when the level of attention rises abnormal returns decrease. Therefore, a negative relation is hypothesized.
3.2.2 Governance Codes and Guidelines
13
Given that the focus lies on US governance ratings Haxhi & Van Ees (2009) make a relevant point in that individualist cultures have a strong tendency to develop codes, the US being one of such individualistic cultures. Also, governments and other institutions do not take static action when introducing codes, but rather engage in a dynamic process. Aguilera & Cuervo-Cazurra (2004) mention that institutions may introduce new corporate governance practices into the existing corporate governance system in order to respond to legitimation demands. They also state that Common-law legal systems are more prone to continue improving their systems and to develop codes. Furthermore, they conclude that during the 1990s US institutions have increased the number of codes issued over time. However, the increased issuance of codes over time alone cannot explain a possible increase of attention to corporate governance from the 1990s to the 2000s. And hence cannot legitimize the inclusion of institutional effects in the attention index. However, according to Gregory & Simmelkjaer (2002) codes help explain both governance related legal requirements and common corporate governance practices to investors and codes stimulate discussion of corporate governance issues among investors. The former relates directly to learning and the latter to
attention. Taking this into account a theoretical basis exists for the inclusion of code and guidelines. Table 1 in the Appendix givesan overview of issued codes and principles. The fast majority of US codes and regulations can be viewed as soft regulation, this is true for the 1990s as well as the 2000s. No US code demands disclosure on a ‘comply or explain’ basis. However, the New York Stock
Exchange is a notable exception in that it requires listed companies to adopt and publish governance guidelines. However, these are only related to the responsibilities of directors. Despite this exception the majority of issuing institutions are of the opinion that no code or regulation fits every company equally well and therefore adopt a soft approach. This view is reflected in the goals of the issuing institutions. These goals can be summarized as: to make recommendations, to promote a culture of professionalism, promotion of access to capital and to aid in the development of guidelines.
14
Figure 2
This figure shows quite an erratic pattern, however the highest peak is observed in the early 00s. Despite the erratic patterns the peak in the early 00s is seen as being a important contributor to the demise of the profitability of the hedge portfolio. Therefore, the hypothesis reads as follows: H2: The number of codes issued is negatively related to hedge portfolio abnormal returns. As it is the first time that code issuance is researched in a governance-based strategy setting the hypothesized relation relies solely on the work of Gregory & Simmelkjaer (2002). Who state that codes help explain corporate governance to investors and thus cause a learning effect. Along with Lewellen & Shanken (2002) who state that learning moves prices to their fundamentals over time a negative relation is hypothesized.
3.2.3 Media Coverage
The role of the media is to collect, select, certify, and repackage information (Dyck, Volchkova, & Zingales, 2003). Here the focus lies solely on newspaper articles. Reading these articles provides certain benefits to investors. It leads to greater attention to corporate governance issues, provided that they are extensively covered. It can also lead to greater interest in the topic trough information diffusion. This comes about in two ways, one the media may themselves reflect a greater interest and two, bring about greater interest in the investment community when corporate governance issues are cited as being a novel. Furthermore, subscription to newspapers is a low cost alternative to become informed, therefore making investors more susceptible to become acquainted with
15
A quantitative proxy for media coverage of corporate governance is obtained through the use of the Lexis-Nexis Academic database. In each calendar year between 1990 and 2011 the number of unique newspaper articles that cite the words “corporate governance” are recorded. Not all US newspapers are subject to this search. They have to comply with certain criteria. On the content side corporate governance articles must obviously be included in the papers. The newspapers must be widely followed because media impact is greater when there is a large audience. At last, the newspapers must be known as being a credible source, otherwise their stories will not be believed. The US newspapers that fit these criteria are: Financial Times, New York Times, USA Today and Washington
Post. Figure 3 reports the number of unique governance articles in each year. Figure 3
During the 1990s the number of articles citing corporate governance stays roughly between 500 and 1000 a year. In the beginning of the new millennium a rise followed by a fall in the late 00s is
observed. The hedge portfolio’s returns started to go down in the early 2000s,so they move in opposite direction. Therefore, the hypothesis reads as follows:
H3: Media coverage is negatively related to hedge portfolio abnormal returns.
Support for this hypothesis comes from Fang & Peress (2009) and Engelberg & Parsons (2011) who conclude in a US setting that media coverage affects equity pricing. However, this only explains that there is a relation between media and equities, but it does not explain the hypothesized negative relation between media and abnormal returns. The latter can be explained through the work of Barber & Odean (2008) who argue that individual investors are the net buyer and sellers of attention grabbing equities. These are equities that are extensively covered in the media. The rise in media
16
coverage of corporate governance issues in the 2000s may make investors aware of the value effect and abnormal returns vanish through their buying and selling behavior. This is confirmed by Bebchuk, Cohen & Ferrell (2010). Media coverage is one of the learning factor present in their attention index. They found that when the level of attention rises abnormal returns decrease. Therefore, a negative relation is hypothesized.
3.2.4 Institutional Investors
Corporate governance has become an important investment criterion for institutional investors for reasons of fiduciary responsibilities, monitoring costs, and liquidity (Chung & Zhang, 2011). The attention given by these professional investors is consistent with the learning hypothesis’ assumption that a sufficient number of market participants must perform trades on company specific governance information. Institutional investors influence financial markets worldwide through their
predominance as buyers, holders and sellers of corporate securities. Furthermore, the majority are willing to engage in equity holder activism (McChary, Sautner & Starks, 2010). Such activism is needed because it functions as proof that institutional investors act to address their governance issues to managers in the real world. These institutional investors are therefore “governance-sensitive”, they are institutions that explicitly consider company’s governance mechanisms in their investment decisions (Bushee, Carter & Gerakos, 2010). It are the corporate governance equity holder resolutions of these “sensitive” investors that function as a proxy for the attention paid to corporate governance by institutional investors. To obtain the resolutions the annual proxy season reviews of Georgeson Shareholder are used. These reviews list the types of investors that make governance-related proposals. Four different groups are distinguished, namely labor union pension funds, public pensions, other equity holder groups and religious organizations. All are labeled as being “sensitive” investors. The total number of proposals submitted by the “sensitive” set of investors in the early 1990s is dominated by labor unions pension funds which have a 39% share of proposals submitted followed by other equity holder groups with 37%. Which amongst others consists of the Investors Right Association of America. Around the turn of the century labor unions pension funds still dominate at 36% with other equity holder groups still in second at 30%. In 2011 labor union pension funds are still ahead with 63% while public pensions are now the second largest proposal sponsor with 18%.
Georgeson Shareholder is used as a source because it has a couple of virtues. First, it covers more or less the same provisions as in the G-Index. Therefore, a uniform set of corporate governance
17
attention by investors corresponds to key governance provisions used by GIM , namely staggered boards, poison pills and golden parachutes. Reinforcing the practical applicability of this index. The number of submitted proposals are shown in Figure 4.
Figure 4
Striking is the sharp increase in the number of submitted proposals during the early 2000s. This can be attributed to the Enron and WorldCom fraud scandals, which are labeled as corporate governance disasters. From that moment on the number declines to a level that is considerably higher than in the 90s. Due to the increase and sustained higher levels of submitted proposals the hypothesis reads as follows:
H4: The number of submitted proposals is negatively related to hedge portfolio abnormal returns. As is the case for academic research and media coverage support for this hypothesis is given by Bebchuk, Cohen & Ferrell (2010). Equity holder proposals are one of the learning factor present in their attention index. They found that when the level of attention rises abnormal returns decrease. Therefore, a negative relation is hypothesized.
18
4. Financial Crisis
The game changing nature of the financial crisis makes a valuation of the governance-based strategy during this time period a necessity. First, a discussion about the role of corporate governance in relation to company performance during financial crises must be provided to assess whether there are any distinctive features. Mitton (2002) and Beak, Kang & Park (2004) found that company-level differences in variables related to corporate governance had a strong impact on company
performance during the East-Asian financial crisis of 1997-1998. Companies that have higher
disclosure quality, serving as a proxy for good governance, suffer less during times of severe turmoil. Despite the fact this is true for the Asian crisis and not necessarily for the current financial crisis previous discussion about Asia is incorporated because the current crisis has not yet given rise to a large body of peer reviewed articles.
Despite this a recent article by Francis, Hasan & Wu (2012) about the current crisis comes to similar conclusions. These authors find that company performance during a crisis is a function of company-level differences in corporate boards. Furthermore, they argue that during crises, the quality of corporate governance is likely to attract more scrutiny. Thus, any preexisting weaknesses are more visible, thereby leading to a flight to quality by investors. The observed flight to quality might in this setting mean that investors sell their stakes in Dictatorship companies and buy Democracy
companies, leading to the hedge portfolio becoming profitable again after the absence of a abnormal returns in the beginning of the 2000s.
Corporate governance matters during crisis periods at least for non-financial companies which are subject of research in the work of previously mentioned authors. Erkens, Hung & Matos (2012) investigate financial companies and they conclude that financials with more independent boards experienced worse equity returns during the crisis period, due to more equity raisings that
transferred wealth from equity holders to debt holders. The independent board served as a pre-crisis channel for implementing the risk-seeking behavior of equity holders. This is contrary to the
19
However, the governance proxies cited in above mentioned articles may not necessarily coincide with the governance provisions used in the work of GIM. In the words of these authors almost every provision gives management a tool to resist different types of equity holder activism, such as calling special meetings, changing the company’s charter or bylaws, suing the directors, or just replacing them all at once. It is about a company’s internal power balance between equity holders and
manager. Rather than companies internal functioning of corporate governance mechanisms vis-à-vis other companies, such as disclosure quality and the independence and monitoring quality of
corporate boards. The peculiarities of this particular governance strategy do therefore not fit with the governance related conclusions of previously cited authors per se.
Despite this notion a lot of value is put on the argument of Francis, Hasan & Wu (2012) that there is a flight to quality during crises. Quality can be perceived as companies that have a balance of power tilted towards equity holders. Therefore, it is expected that the hedge portfolio will regain its abnormal returns after losing it in the early 2000s. The analysis will begin in 2007 when the first cracks in the US housing market began to show. March 2007 will function as a starting point when New Century Financial Corporation saw its equity value plummet due to heavy investments in the US sub-prime mortgage market and financial markets took a turn for the worst.
Figure 5 shows the hedge portfolio’s yearly average raw (unadjusted for risk) returns during the crisis as of the beginning of 2007. Figure 5 -1,5 -1 -0,5 0 0,5 1 1,5 2 2,5 2007 2008 2009 2010 2011
20
Holding the hedge portfolio from 2007 to 2011 would have resulted in a positive yearly average raw return in three of the four years. For that reason, and given the flight to quality argument the hypothesis reads as follows:
H5: Holding the hedge portfolio during the current financial crisis would have resulted in a positive
mean abnormal return.
21
5. Data
5.1 Sample Selection and Data Sources
The IRRC periodically releases Corporate Takeover Defenses in which US companies are tracked based on 24 governance provisions. The provisions in these publications are divided into five groups: tactics for delaying hostile bidders (Delay); voting rights (Voting); director/officer protection
(Protection); other takeover defenses (Other); and state laws (State). All these groups have their own provisions as summarized in Table 1.
Table 1
Governance Provisions
Delay Protection Voting Other State
Blank check Compensation plans Bylaws Antigreenmail Antigreenmail law Classified board Contracts Charter Directors' duties Business combination law Special meeting Golden parachutes Cumulative voting Fair price Cash-out law
Written consent Indemnification Secret ballot Pension parachutes Directors' duties law Liability Supermajority Poison pill Fair price law
Severance Unequal voting Silver parachutes Control share acquisition law Because the sample includes all years from 1990 to 2011, the following IRRC volumes are used for their governance provisions: September, 1990; July, 1993; July, 1995; February, 1998; November, 1999; February, 2002; January, 2004; and January, 2006. Although there is also a 2008 issue the data collection methodology differs from earlier volumes in terms of changes made in the set of
provisions included and definitions of provisions. Therefore, these material changes makes the 2008 volume unsuitable for inclusion.
22
Using data from the IRRC publications GIM provide a Goverance-Index (G-index) available on Andrew Metrick’s website (http://faculty.som.yale.edu/andrewmetrick/data.html). The G-Index ranks
companies based on the governance provisions. Annual time series of the G-indexare created using a forward-fill method which assumes that the governance ratings remain unchanged between IRRC publishing dates. Usually there is a time period of 2 to 3 years between subsequent volumes, except after the January 2006 volume. From 2006 the series is forward-filled for 5 years until December 2011 due to the exclusion of the IRRC 2008 issue. All the IRRC provisions on which the G-index is based are weighted equally, furthermore they are either “present” or “not present”. Some precision is lost in this process, this relates to the notion that there is no a priori reason to expect that all the provisions contribute to a correlation between the provisions and equity returns. Some provisions might have little relevance, and some provisions might be positively correlated with equity returns. Among those provisions that are negatively correlated with equity returns, some might be more so than others (Bebchuk, Cohen & Ferrell, 2008) (Jiraporn, 2005). The G-index favors simplicity in constructing the index over precision.
From the G-Index scores, the companies with a score of 5 or lower are put in the Democracy
portfolio and company that score 14 or higher are put in the Dictatorship portfolio. Therefore, a low score corresponds to good governance and a high score to poor governance. From these two extreme portfolios a hedge portfolio is created which means going long in the Democracy portfolio and short in the Dictatorship portfolios. Furthermore, the portfolios are market value-weighted, despite the fact that prior research also utilizes equal-weighted portfolios. The former provides the investor with the highest abnormal returns, observed in previous literature. Given the fact the general investor would favor a value maximizing investment strategy over a suboptimal one only market value-weighted portfolio are looked upon. The portfolios do not suffer from look-ahead bias. The potential problem of look-ahead bias is tackled by means of buying a portfolio’s equities the first day of the month following the month in which an IRRC issue is published. This gives a real-world investor enough time to calculate scores and construct portfolios.
23
Not all companies present in the G-Index are present in the governance portfolios. For reasons of not being able to obtain the return history after a merger or acquisition and in some cases DataStream could just not provide any data. The total percentage of missing companies is 7%, which can be considered as acceptable.
5.2 Descriptive Statistics
24 Table 2 The Governance-Index 1990 1993 1995 1998 2000 2002 2004 2006 Governance-Index Minimum 2 2 2 2 2 2 2 2 Mean 9 9,3 9,4 8,9 9,2 9,2 9,2 9 Median 9 9 9 9 9 9 9 9 Mode 10 9 9 10 9 10 8 9 Maximum 17 17 17 17 19 18 18 18 Standard Deviation 2,9 2,8 2,8 2,8 2,7 2,6 2,5 2,5 Number of companies G ≤ 5 158 139 120 215 145 119 124 129 G = 6 119 88 108 169 149 128 129 143 G = 7 158 140 127 186 187 185 182 196 G = 8 165 139 152 201 198 242 281 260 G = 9 160 183 183 197 239 236 255 283 G = 10 175 170 178 221 229 244 256 221 G = 11 149 168 166 194 191 204 212 204 G = 12 104 123 142 136 143 121 145 134 G = 13 84 100 110 106 102 95 91 77 G ≥ 14 85 93 87 83 87 94 85 66 Total 1357 1343 1373 1708 1670 1668 1760 1713 5.3 Governance and Abnormal Returns
The raw returns (unadjusted for risk) in US dollars for the market value-weighted hedge portfolio are summarized in the Table 3.
Table 3
Average Raw Returns
1990 1993 1995 1998 2000 2002 2004 2006
10/90-7/93 8/93-7/95 8/95-2/98 3/98-11/99 12/99-2/02 3/02-1/04 2/04-1/06 2/06-12/11 -0.27 0.34 0.22 3.07 -0.60 -0.57 -1.02 0.47 In the 1990s one can see that the market value-weighted hedge portfolio yields positive average raw returns. But it fails to do so in the 2000s until the year 2006 and onwards. The latter is striking considering that the financial crisis started in March 2007. At first glance the hedge portfolio seems to be crisis resistant.
25
the Democracy and Dictatorship portfolios. Several equity characteristics are known for their ability to significantly forecast future returns. These are the market factor, a company’s market
capitalization, book-to-market ratio and immediate pas returns. If the Dictatorship Portfolio differs significantly from the Democracy Portfolio in these characteristics, then style differences may explain at least part of the difference in returns. Although there is an ongoing debate about the forecasting ability of the style factors, no position is taken and this method of performance attribution is chosen because it is so widely used in financial literature.
The previously discussed performance attribution method will be put into a regression to be able to derive test results. It is the classical Fama & French three-factor model augmented by a the
momentum factor also defined by these authors. The following regression will be estimated:
(RDemocracy – RDictatorship) = α + β1(RMRF )t + β2SMBt + β3HMLt + β4MOMt + εt, (1)
where (RDemocracy – RDictatorship) is the return difference between the Democracy and Dictatorship
portfolios, the hedge portfolio. The four independent variables are the excess return on the market portfolio (RMRF)t, the difference between the returns to portfolios of small and big company equities (SMB)t, the difference in returns to portfolios of high and low book-to-market equities (HML)t, and
(MOM)t is the average return on two high prior return portfolios minus the average return on two
26 Table 4 ALPHA RMRF SMB HML MOM 1990-1999 Coefficient 0.57** -0.12 -0.08 -0.54*** 0.31*** Standard error 0.28 0.08 0.09 0.11 0.08 t-statistic 2.03 -1.51 -0,85 -4.79 3.76 N 111 R2 0.34 Adjusted R2 0.31 F-Statistic 1.36 Prob. (F-stat) 0.00 2000-2011 Coefficient -0.01 -0.02 0.02 -0.42*** -0.07 Standard error 0.27 0.06 0.08 0.08 0.05 t-statistic -0.04 -0.39 0.23 -5.38 -1.59 N 144 R2 0.20 Adjusted R2 0.18 F-statistic 8.67 Prob. (F-stat) 0.00 1990-2011 Coefficient 0.34* 0.01 -0.06 -0.48*** 0.01 Standard error 0.20 0.05 0.06 0.06 0.04 t-statistic 1.71 0.19 -1.07 -7.51 0.16 N 255 R2 0.20 Adjusted R2 0.19 F-statistic 1.60 Prob. (F-stat) 0.00
*Significant at 0.10 level; **significant at 0.05 level; ***significant at 0.01 level
27
Figure 6
As can be seen there is a steep and persistent increase in average abnormal returns during the 1990s. These returns peak in 2001, but afterwards decrease till 2005. From 2005 and on, and thus during financial turmoil, again an increase is observed.
-1 -0,5 0 0,5 1 1,5 199510 199604 199610 199704 199710 199804 199810 199904 199910 200004 200010 200104 200110 200204 200210 200304 200310 200404 200410 200504 200510 200604 200610 200704 200710 200804 200810 200904 200910 201004 201010 201104 201110
28
6. Empirical Results
6.1 Abnormal Returns
In this section the regression results that link the hedge portfolio rolling average abnormal returns to the learning factors will be discussed. Given the fact that the current financial crisis has a distinct impact on financial markets the learning factors will be regressed till the start of the crisis in the beginning of 2007. The result are provided in Table 5.
Table 5
INTERCEPT ARES CODES MEDIA PROP
1995-2007/2 Coefficient 0.86* -0.00* 0.01 0.00* -0.00* Standard error 0.08 0.00 0.01 0.00 0.00 t-statistic 1.09 -6.40 0.99 8.06 -3.90 N 137 R2 0.62 Adjusted R2 0.61 F-Statistic 5.31 Prob. (F-stat) 0.00 *significant at 0.01 level
Except for code issuance (CODES), the other learning factors are all significant. Submitted proposals by investors (PROP) and academic research (ARES) meets the hypothesized sign, however media coverage (MEDIA) is positively related . Overall model quality can be considered good, adjusted R2 is 0.61. In following sections each factor will be discussed individually.
6.2 Academic Research
H1 cannot be rejected, there is a significant relationship between peer reviewed journal articles and returns. Despite the criticism that there is a lag between observation of a phenomenon and the publishment of peer reviewed articles. This lag of a couple of years could potentially disentangle the relationship between articles and return. A possible reason that it does not do so could be that professional investors are the sole readers, because academic research is likely to be too costly for individual investors. Professional investors meet the learning hypothesis criteria of being able to influence prices because they are with many and have sufficient capital to move prices.
6.3 Code Issuance
29
fact that institutional investors themselves develop codes, like the California Public Employees’ Retirement System and the Council of Institutional Investors. This means that the only investors that benefit from discussion and learning are individual investors. Their total share in the US equity markets is relatively low and so is their influence on equity returns.
6.4 Media Coverage
Concerning media coverage, there is a significant relationship between corporate governance related articles in US newspapers and hedge portfolio returns. However, the expected negative sign in H3 is not met. The observed positive sign is puzzling. Focus should be placed on individual investors. Barber & Odean (2008) find that individual investors are net buyers of attention-grabbing equities, which are equities receiving media attention. Institutional investors do not display such behavior. They have the time and resources to follow large quantities of equities, therefore they do not resort to the subset of equities that receive media attention. Plus, they are likely to concentrate on
particular sectors or follow equities based on certain investment criteria. It could therefore be important to assess how individual investors conduct trades based on media information and how that impacts the equities prices of good and poorly governed companies. But conclusions are left for future research.
6.5 Institutional Investors
30
investors to engage in activism is largely restricted to exit (selling equities) and voice (e.g. submitting proposals). Helwege, Intintoli & Zhang (2012) argue that some institutions continue to sell their holdings of underperforming companies, while others, especially those that are known to be activists, pursue strategies designed to bring about change. Yet, these activists do not dominate the institutional investor landscape. The findings contradict previous research, this is likely to be the result of the strategy under consideration. In a setting where the abnormal returns are the result of a strategy that explicitly takes corporate governance into account equity holder proposals geared towards changing company governance do have a relationship with returns.
6.6 Financial Crisis
In Table 6 the regression results are presented. This displays whether the hedge portfolio is able to give the investor a positive alpha or in other words an average abnormal return over the period 2007-2011. Table 6 ALPHA RMRF SMB HML MOM 2007-2011 Coefficient 0.53 -0.21* -0.06 0.06 -0.07 Standard error 0.43 0.09 0.20 0.18 0.08 t-statistic 1.24 -2.31 -0,29 -0.35 -0.88 N 60 R2 0.11 Adjusted R2 0.05 F-Statistic 1.74 Prob. (F-stat) 0.15 *significant at 0.05 level
31
7. Conclusions
7.1 Summary of Key Findings
The seminal article of GIM spurred a great deal of attention towards governance-based strategies in the academic world. The particular governance strategy by GIM produces a hedge portfolio that goes long in equities from companies that have good governance and goes short in equities from
companies with poor governance. This strategy is designed to provide the investor with abnormal returns. GIM (2003) find abnormal returns in the 1990s, but subsequent research saw the abnormal returns fade in the beginning of the 2000s. Only recently have learning effect been studied to explain the existence and vanishing of abnormal returns in the first half of the 2000s. The learning hypothesis was only recently drafted by Bebchuk, Cohen & Wang (2010). These authors put their learning factors in a single index. In order to provide a better understanding of the effect of learning on abnormal returns this paper is first to draft and test hypotheses for each learning factor individually. Another novelty is the inclusion of the current financial crisis. This study takes a longitudinal view, it examines abnormal returns from 1990 to 2011.
32
From a regulators point of view it must disappointing that the issuance of codes and regulations is not related to equity return. If this was the case it would have been the ultimate demonstration that codes and regulations work even if companies do not adhere to them. Because then equities from good governance companies would have been bought and equities subject to poor governance would have been sold by the majority of market participants. Making them behave as regulators themselves. Furthermore, it is striking that the average governance score based on takeover defenses for US companies is the same in 2006 as it was in 1990. Despite the fact that codes and regulations are increasingly issued during this time. This is likely due to the soft regulation approach that US code issuing institutions adopt. Along with the fact that market participants themselves do not take part of the role as regulators, soft code issuance seems to have no real impact.
7.2 Suggestions for Future Research
33
8. References
Aguilera, R.V. & Cuervo-Cazurra, A. 2004, ‘Codes of good governance worldwide: What is the trigger?’ Organization Studies, vol. 25, pp. 415–443.
Baek, J.S., Kang, J.K. & Park, K.S. 2004, ‘Corporate governance and firm value: Evidence from the Korean financial crisis’ Journal of Financial Economics, vol. 71, pp. 265-313.
Barber, B.M. & Odean, T. 2008, ‘All that glitters: The effect of attention and news on the buying behavior of individual and institutional investors’ The Review of Financial Studies, vol. 21, pp. 785-818.
Bauer, R., Guenster, N. & Otten, R. 2004, ‘Empirical evidence on corporate governance in Europe: The effect on stock returns, firm value and performance’ Journal of Asset Management, vol. 5, pp. 91-104.
Bebchuk, L.A., Cohen, A. & Ferrell, A. 2008, ‘What matters in corporate governance?’ The Review of
Financial Studies, vol. 22, pp. 783-827.
Bebchuk, L.A., Cohen, A. & Wang, C.C.Y. 2010, ‘Learning and the disappearing association between governance and returns’ Working Paper, National Bureau of Economic Research.
Bodie, Z., Kane, A. & Marcus, A.J. 2005, ‘Investments’, 6th ed., Boston: McGraw-Hill.
Bushee, B. J., Carter, M.E. & Gerakos, J. 2010, ‘Institutional investor preferences for corporate governance mechanisms’ Working Paper, University of Pennsylvania.
Chung, K.H. & Zhang, H. 2011, ‘Corporate governance and institutional ownership’ Journal of
Financial and Quantitative Analysis, vol. 46, pp. 247-273.
Core, J.E., Guay, W.R. & Rusticus, T.O. 2006, ‘Does weak governance cause weak stock returns? An examination of firm operating performance and investors’ expectations’ Journal of Finance, vol. 61, pp. 655-687.
Cremers, K. J.M, Nair, V.B. & John, K. 2009, ‘Takeovers and the cross-section of returns’ Review of
Financial Studies, vol. 22, pp. 1409-1445.
34
Dyck, A., Volchkova, N. & Zingales, L. 2003, ‘The corporate governance role of the media: Evidence from Russia’ The Journal of Finance, vol. 3, pp. 1093-1135.
Engelberg, J.E. & Parsons, C.A. 2011, ‘The causal impact of media in financial markets’ Journal of
Finance, vol. 66, pp. 67-97.
Erkens, D.H., Hung, M. & Matos, P. 2012, ‘Corporate governance in the 2007-2008 financial crisis: Evidence from financial institution worldwide’ Journal of Corporate Finance, vol. 18, pp. 389-411. Fama, E.F., 1970, ‘Efficient capital markets: A review of theory and empirical work’ Journal of
Finance, vol. 25, pp. 383-417.
Fama, E.F. & French, K. 1993, ‘Common risk factors in the returns on stocks and bonds’ Journal of
Financial Economics, vol. 33, pp. 3-56.
Fang, L. & Peress, J. 2009, ‘Media coverage and the cross-section of stock returns’ Journal of Finance, vol. 64, pp. 2023-2052.
Francis, B.B., Hasan, I. & Wu, Q. 2012, ‘Do corporate boards matter during the current financial crisis?’ Review of Financial Economics, vol. 21, pp. 39-52.
Gillan, S.L. & Starks, L.T. 2000, ‘Corporate governance proposals and shareholder activism: The role of institutional investors’ Journal of Financial Economics, vol. 57, pp. 275-305.
Gillan, S.L. & Starks, L.T. 2007, ‘The evolution of shareholder activism in the United States’ Journal of
Applied Corporate Finance, vol. 19, pp. 55-73.
Gompers, P., Ishii, J. & Metrick, A. 2003, ‘Corporate governance and equity prices’ Quarterly Journal
of Economics, vol. 118, pp. 107-155.
Gregory, H.J., 2009, ‘Comparison of Corporate Governance Guidelines and Codes of Best Practice’
Weil, Gotshal & Manges LLP.
Gregory, H.J. & Grapsas, R.C., 2012, ‘Comparison of Corporate Governance Principles & Guidelines: United States’ Weil, Gotshal & Manges LLP.
Gregory, H.J. & Simmelkjaer R.T., 2002, ‘Comparative study of corporate governance codes relevant to the European Union and its member states’ Weil, Gotshal & Manges LLP.
35
Helwege, J., Intintoli, V.J. & Zhang, A. 2012, ‘Voting with their feet or activism? Institutional investors’ impact on CEO turnover’ Journal of Corporate Finance, vol. 18, pp. 22-37.
Jensen, M.C. & Meckling, W.H. 1976, ‘Theory of the firm: Managerial behavior, agency costs and ownership structure’ Journal of Financial Economics, vol. 3, pp. 305-360.
Jiraporn, P. 2005, ‘An empirical analysis of corporate takeover defenses and earnings management: evidence from the US’ Applied Financial Economics, vol. 15, pp. 293-303.
Lewellen, J. & Shanken, J. 2002, ‘Learning, asset-pricing tests and market efficiency’ Joumal of
Finance, vol. 57, pp. 1113-1145.
McCahery, J. A., Sautner, Z. & Starks, L.T. 2010, ‘Behind the scenes: The corporate governance preferences of Institutional Investors’ Working Paper, University of Amsterdam.
Mitton, T. 2002, ‘A cross-firm analysis of the impact of corporate governance on the East Asian financial crisis’ Journal of Financial Economics, vol. 64, pp. 215-241.
OECD, viewed 3 August 2012, from
http://www.oecd.org/daf/corporateaffairs/corporategovernanceprinciples/corporategovernanceand thefinancialcrisis.htm.
Parrino, R., Sias, R.W. & Starks, L.T. 2003, ‘Voting with their feet: Institutional ownership changes around forced CEO turnover’ Journal of Financial Economics, vol. 68, pp. 3-46.
Peng, L. & Xiong, W. 2006, ‘Investor attention, overconfidence and category learning’ Journal of
Financial Economics, vol. 80, pp. 563-602.
Shleifer, A., & Vishny, R.W. 1997, ‘A survey of corporate governance’ Journal of Finance, vol. 52, pp. 737–783.
Thomas, R. & Cotter, J. 2006, ‘Shareholder proposals post-Enron: What’s changed, what’s the same?’
36
9. Appendix Table 1
List of Codes and Principles 1990
Business Round Table ("BRT"), Statement on Corporate Governance and American Competitiveness
1994
The American Law Institute (“ALI”), Principles of Corporate Governance: Analysis and Recommendations, Vol. 1 (revised, 2002)
General Motors Board of Directors, Corporate Governance Guidelines (revised, 2007 and 2008) National Association of Corporate Directors (“NACD”), Report of the NACD Blue Ribbon Commission on Performance Evaluation of Chief Executive Officers, Board and Directors
1996
National Association of Corporate Directors (“NACD”), Report of the NACD Blue Ribbon Commission on Director Professionalism (reissued 2001, 2005 and 2011)
1997
American Federation of Labor and Congress of Industrial Organizations (“AFL-CIO”), Exercising Authority, Restoring Accountability – AFL-CIO Proxy Voting Guidelines (revised, 2003)
American Society of Corporate Secretaries, Suggested Guidelines for Public Disclosure and Dealing with the Investment Community
Business Round Table ("BRT"), Statement on Corporate Governance
Teachers Insurance and Annuity Association–College Retirement Equities Fund (“TIAA-CREF”), TIAA-CREF Policy Statement on Corporate Governance (revised, 2007 and 2011)
1998
Business Sector Advisory Group on Corporate Governance, Corporate Governance: Improving Competitiveness and Access to Capital in Global Markets
California Public Employees’ Retirement System (“CalPERS”), Corporate Governance Principles and Guidelines (revised, 2007 and 2011)
Council of Institutional Investors (“CII”), Corporate Governance Policies (revised, 2007, 2008 and 2011)
1999
Blue Ribbon Commission on Improving the Effectiveness of Corporate Audit Committees, Report and Recommendations (revised, 2004)
California Public Employees’ Retirement System (“CalPERS”), Global Corporate Governance Principles Organisation for Economic Co-operation and Development (“OECD”), Principles of Corporate Governance (revised, 2004)
2002
Business Roundtable, Principles of Corporate Governance (revised, 2005 and 2010)
Council of Institutional Investors (“CII”), Core Policies, General Principles, Positions and Explanatory Notes Corporate Governance Center, Kennesaw State University, 21st Century Governance and Financial Reporting Principles
The Conference Board Commission on Public Trust and Private Enterprise, Findings and Recommendations (Part 2 & Part 3, 2003)
2003
Institutional Shareholder Services, ISS Corporate Governance: Best Practices User Guide & Glossary (revised, 2007)
National Association of Securities Dealers, Inc. (“NASD”), NASDAQ Marketplace Rules (revised, 2004) New York Stock Exchange (“NYSE”), Corporate Governance Rules (amended, 2004)
The United District Court for the Southern Distric of New York, Restoring Trust - The Breeden Report on Corporate Governance for the Future of MCI, inc.
2004
37
Table 1 continued 2005
The Conference Board, Corporate Governance Handbook 2005: Developments in Best Practices, Compliance, and Legal Standards
2008
National Association of Corporate Directors (“NACD”), Key Agreed Principiles to Strenghten Corporate Governance for US Publicly Traded Companies
2009
The Conference Board, Corporate Governance Handbook: Legal Standards and Board Practices (2009)
2010
New York Stock Exchange (“NYSE”), Report of the NYSE Commission on Corporate Governance
2011
ISS, U.S. Proxy Voting Guidelines Summary; ISS, Governance Risk Indicators 2.0 Technical Document
Adopted from: Gregory (2009) and Gregory & Grapsas (2012)
Table 2 shows the monthly returns from the Democracy portfolio (DEM), Dictatorship portfolio (DIC) and the hedge portfolio (DEM-DIC). Rolling average (RAVE), plus the Fama & French factors.
Table 2
38
Table 2 continued
39
Table 2 continued
40
Table 2 continued
41
Table 2 continued
42
Table 2 continued
43
Table 2 continued
