Legislative easing for Canadian REITs and its impact on shareholders wealth

Master Thesis

University of Amsterdam

Legislative easing for Canadian REITs and its impact on shareholders wealth

Daan Visser 10074449 July 7, 2015 University of Amsterdam

Master’s in Business Economics, Real Estate Finance

Supervisor Dr. Martijn Dröes

Second reader Dr. Marcel Theebe

2 Statement of Originality

This document is written by Student Daan Visser, 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.

3

Abstract

According to the literature, different legislation for Real Estate Investment Trust, like mandatory dividend payout, leverage ratios and trading constrains, should have an impact on the performance of these companies. This paper analyzes the impact of a change in legislation in Canada, by calculating the abnormal returns, and the change in systematic risk. The Technical Tax Amendments Act in 2012-2013 (C-48 bill), included a few easing legislative changes for REITs in Canada. The calculation is done by analyzing ten event-dates, with the use of a regression-based event methodology. A portfolio is constructed of 30 REITs in Canada, using the inverse of the variance of the returns as weights. The S&P/TSX SmallCap Index is used as benchmark. The actual signing of the so called C-48 bill, gives a positive significant effect on shareholders wealth. The change in legislation did not have a significant effect on the systematic risk of the REIT market.

4

Table of Content

2.3 General REIT legislation ... 10

2.4 Policy ... 11

2.5 Constrains and performance ... 12

3 Change in REIT legislation Canada ... 12

3.1 Non Qualifying REIT assets ... 13

3.2 Eligible resale property... 14

3.3 Passive revenue test ... 14

3.4 Foreign property ... 15 4 Data ... 15 4.1 Return data ... 15 4.2 Event-dates ... 16 4.3 Descriptive statistics ... 18 5 Methodology ... 20 5.1 Event study ... 20 5.2 Risk ... 21 6 Results ... 22 6.1 Empirical results ... 22 6.1.1 Shareholders wealth... 22 6.1.2 Systematic Risk ... 25 7 Conclusion ... 27 8 References ... 29

5

1 Introduction

The popularity of Real Estate Investment Trust has resulted in an increase in number of REITs, and the number of countries that established REITs. For the United States alone, the REIT market now has a value of over $2 trillion in real estate assets (REIT, 2015). REITs are trusts that experience a special tax advantage on corporate level. Because the trusts distribute large parts of their income or profits (this differs among countries), which is in some cases mandatory, these dividend distributions are not taxed. According the European Public Real Estate Association (2014), there are 37 countries worldwide that have REIT or REIT-like legislation. This legislation differs among countries, and it changes over the years. In some countries major changes in legislation have occurred. In the United States for instance, a major amendment happened in 1999. This legislative change can be seen as relaxation, in the way that the mandatory dividend payout was lowered, and this had an positive impact on the market (Howe & Jain, 2004).

In November 2012, the Technical Tax Amendments Act, called the C-48 bill, which included changes in legislation for Canadian REITs, was first proposed to the House of Commons in Canada. The changes include less restrictions on property and revenue gains for existing REITs. It also made it easier to qualify for a REIT as a currently non-REIT company. The eased legislation on property held, gave the opportunity to REITs to restructure, without disqualifying as a REIT. The proposals were desired by the market, to keep up with the United States’ and Australian REIT market. These governments have made easing amendments to the REIT market, because the market had grown substantially and had become more integrated between countries. In June 2013, the C-48 bill in Canada, was passed.

This paper will investigate the impact of this legislative changes in Canada on REIT returns. The reaction of the market will show if such a change in regulation is seen positive or negative, and for this paper, the aim is to show if relaxation programs of legislation for REITs will give a positive market reaction, by analyzing the abnormal returns around the event-dates that are associated with this period of legislative change. Besides that, the impact of these relaxations on the systematic risk for a countries’ REIT market will be analyzed.

6 Since the first implementation of REIT legislation in the United States in 1960, several papers have been written about the effect of legislation on the performance of the REIT market. Different choices on the structure of legislation for REITs have different impact (Sirmans and Campbell, 2002) on the REIT market. The legal system (Edelstein et al., 2011), as well as trading constrains (Mühlhofer, 2013) have a significant effect on the performance of REITs.

The study of Howe & Jain (2004) is an important and relevant research for this thesis. Howe & Jain (2004) investigate the impact of legislative change in the United States on shareholders wealth. The legislative change in the United States in 1999, consisted of two major amendments. The first one was the possibility for REITs to set up a Taxable REIT Subsidiary, which is used to provide specific services to tenants. The second major change is the decrease in mandatory dividend payout from 95 percent to 90 percent. This implies that REITs had more freedom in choosing its favorable dividend payout policy, and therefore had more influence on their cash balance. Howe & Jain (2004) find a positive significant reaction on the market, during the event-window around the day that the amendments were announced. When analyzing the effect of the legislative change on the systematic risk, they find a significant decrease in risk in the period after the change.

This thesis contributes to the existing literature of legislation for REITs, by focusing on a different country, in a different time period and on different changes in legislation. The last two differences between this thesis and the study of Howe and Jain (2004), are of particular importance and will therefore be discussed in more detail. The market for REITs faces periods of growth and of decline, it gets more globalized and integrated between countries. The fact that the REIT Modernization Act (RMA) in the United States happened in 1999, and the legislative change in Canada occurred in 2012, should cause a different perception of the changes by investors, because they base their perceptions on, for instance, the state of the real estate market. The amendments in Canada occurred during the economic crisis. Hence, more efficiency and less restricting rules for REITs, could free the market and will most likely have a positive impact on REIT performance. The other reason why it is interesting to study the effect of the changes in legislation in Canada in particular, is more obvious. The major amendments in Canada are different from the ones in the United States. As mentioned above, the RMA in the United States had two major amendments, the first was the possibility to provide special

7 services to tenants, the second was the decrease of the mandatory dividend payout percentage. In general, the first amendment had impact on the income of REITs, and the second on the cash balance of REITs. In Canada, the possibility to hold Non Qualifying REIT Assets (NQRA), can not be compared with the amendments in the United States. This Canadian amendment focusses on holding assets in your portfolio, that were not allowed before, and has nothing to do with either dividend payout or services to tenants. Though, it can be seen as an easing change in legislation. The inclusion of eligible resale and foreign property to the list of permitted activities in Canada, can be compared with the inclusion of the Taxable REIT Subsidiary in the United States. These aspects all have impact on the income of REITs. But because the easing amendments differ, it is interesting to examine whether these different types of legislative easing still had a positive effect on REIT returns. It’s expected that easing legislation for REITs should improve the performance of the trusts. For example, forced divestments of assets, which are needed when a REIT may not hold NQRA, will cause that the price of these assets is not the true price, because the timing of the sale is not optimal (Mühlhofer, 2013). This thesis will provide a clear explanation of how the markets values the legislative changes and what impact it had on the systematic risk for REITs in Canada.

To analyze this reaction of the market on the legislative change in Canada, an event methodology is used, which is also used by Howe & Jain (2004). The returns of 36 REITs are obtained from Datastream, with a time window from March 2011 until December 2013. In combination with the return of a benchmark, the S&P/TSX SmallCap Index, an event-study is constructed to see if the amendments had a significant impact on the Canadian REIT market. This study includes 10 different event-dates, that are important for the period of legislative change. Examples of events are the introduction of the proposal to the public, the introduction to the House of Commons, and the actual signing of the C-48 bill.

In the situation of legislative change, the event-dates are the same for all entities in the dataset and this results in correlated abnormal returns, which will cause biased standard errors. The extensive literature about the different event-methodologies, gives a very clear solution for this problem (Thompson (1985) and Brown & Warner (1980 and 1985)). A regression-based technique is used to

8 calculate the abnormal returns for the different event-dates. The change in systematic risk is calculated by separating the market beta before the amendments, and the change afterward, using a regression.

The results of the regression on the returns of the REIT portfolio, the returns of the market index and the event-dates, show that Event 10, the event of the Royal Assent, which is the actual signing of the bill, shows a positive significant abnormal return of 2 percent. When only Event 10 is included in the regression, the abnormal return is 1.3 percent. In comparison to the paper of Howe & Jain (2004), where they find an cumulative abnormal return of 1.17 percent, the findings of this thesis are of the same magnitude. The difference is that Howe & Jain (2004) find the abnormal returns during the introduction of the proposals.

In this paper, the first event, the introduction of the proposals, has not shown a significant effect. Efficient market theory suggest that new information is incorporated in the returns of stocks rapidly. This event-study shows that this is the case for the REIT market in Canada, but not in the period of announcing, but in the period of the actual signing of the bill. This can be possibly explained by the absence of the proposals of 2010 in this thesis. These proposals were not converted into laws, but were changed and completed with new rules. Because of this new form of proposals in 2012, the proposals in 2010 are not taken into account, but it can be that the expectation of investors of a change in legislation, resulted in an insignificant effect on the day of the announcement. Another possibility is that the market participants did not react on the introduction of the proposals, so they did not speculate on a change in legislation. The different findings of this paper compared to the paper of Howe & Jain (2004), can possibly be explained by the economic situation of the two legislative amendments. The period between these two legislative changes, is known for its economic crisis. This crisis could have had impact on the aversion of investors. Guiso et al. (2014) find that investors are more risk-averse after a period of crisis, even if a particular investor did not suffer from losses.

The second aspect of shareholders wealth that is analyzed, is the impact of the legislative change on the systematic risk of the sector. The coefficient for the change in risk has the expected negative sign, but it is not statistically significant. A possible explanation could be that the different amendments had different effects on the systematic risk, and cancelled each other out.

9 The remainder of this thesis is structured as follows, in Section 2, the existing literature about REITs is dicussed, and in Section 3, the legislative change in Canada is explained. After that, the data that is used for this master thesis is described in Section 4. Section 5 consist of the methodology, and Section 6 discusses the results. This paper ends with a conclusion in Section 7.

2 Previous literature

2.1 Efficient Markets

An event-study captures the reaction of the market after new information is known to market participants. This study will do so in the case of legislative change in Canada. To construct a solid conclusion, it is important to understand what efficiency is and it which ways it can be explained. The efficient market hypothesis is based on three forms of efficiency (Fama, 1970). The first one is the weak form, where efficiency is based on past returns and prices. The second one is the semi-strong form, where the focus is on the anticipation of future events and the information that is available. The strong form suggest that it is impossible to beat the market consistently.

In real estate, a discussion exist about the efficiency of the market. Property asset market seem to be a priori inefficient, because the trading of property is not frequent, the information cost are high and the market is dispersed (Keogh & D’Arcy, 1999). However, the market for REITs has different characteristics. The behavior of REITs can be compared to the behavior of partly stocks and partly bonds (Benjamin, Sirmans, & Zietz, 2001). This suggests that returns for REITs incorporate new information rapidly, as it is in the “efficient” stock market (Fama et al., 1969).

2.2 Event studies

To point out the effects of relaxation of legislation for REITs, this study will use an event-methodology. The existing literature for event studies is extensive. One of the first papers for this sort of methodology, that is still used at the moment, is introduced by Dolley and Clay (1933). Other pioneers of this method are Myers & Bakay (1948), Fama, Fisher, Jensen and Roll (1969), and Brown

10 & Warner (1980). The event-study methodology aims to investigate the effect of events, which can be economical or firm-specific, on firm-value. These events could be earnings announcements, mergers or legislative changes. The first two are firm-specific, and the event-dates are different for each entity. To calculate the abnormal returns for these events, an event-window is chosen around the event, where the expected returns are calculated, with the use of an index. The difference between the expected returns and the actual returns, is the abnormal return. In the case of regulatory changes, the event date is the same for all entities. This results in correlated abnormal returns, and therefore a different technique needs to be used in comparison to the firm-specific events. The use of a regression, where the returns of portfolio of the firms in the dataset are regressed on the returns of a chosen index and event-dummies, controls for this correlation between abnormal returns. The coefficient that is calculated for the dummies, will capture the abnormal returns. Important studies about event-studies with a firm-specific character are written by Schwert (1981), Schipper and Thompson (1983) and Brocket, Chen and Graven (1999).

2.3 General REIT legislation

Since the first legal structure for REITs in the United States in 1960, and the implementation of a REIT structure in The Netherlands in 1969, a lot has been written about the performance, structure, IPOs and risks of REITs. REITs nowadays are quite different than in the early days. Chui et al. (2003) describe the difference in structure of REITs before and after the year 1990. The main difference is the self-managed structure. Another aspect in which the REIT market is different, is the legislation for REITs. Each country that has adopt legislation for REITs, has its own view on which environment is the best, and therefor differences in legislation exists between different countries.

In contrast to the early adoption of REIT regimes in the United States and The Netherlands, the implementation of the regime for Canada took quite a long time. In 1995, the Income Tax Act made it possible for trusts to qualify as a REIT. The growth of number of REITs in the first years was not astonishing, where there were three REITs in 1994 and five in 1997 (EPRA, 2014). But from that

11 point in time, the market for REITs grew significantly. In 1997, seven new REITs were listed on the Toronto Stock Exchange and the total assets grew from $ 80 million in 1993 to $ 4 billion in 1997.

2.4 Policy

In the paper of Sirmans and Campbell (2002), it’s discussed how European countries can learn from the legislation in the United States, and what they should implement in their own REIT regime. They discuss the possible advantages and disadvantages of implementing REIT structures in general. Some of the advantages are that such a REIT structure helps the real estate market by creating more liquidity, which will cause more efficiency for capital movements, and more price stability. On the other side, a country will receive less taxes, and it can cause unfair competition between real estate companies. Furthermore, the authors try to give answer to the questions that governments face, when implementing a REIT structure. The restrictions on the number of owners of a REIT, prevents them from avoiding taxes. Easing changes in the Unites States regime about the size of institutional investors in REITs, have caused growth in size of REITs and the liquidity of the market. Another important aspect for legislation is the mandatory payout requirements. REITs will have the tax advantages on corporate level if they payout their profit to their shareholders, but a discussion can be held about the mandatory percentage of this payout. A high payout requirement can cause inflexibility in a downturn market and therefor the volatility will be higher. High payout requirements are on the other hand in favor of the shareholders. Lowering the requirements will not immediately cause lower payout, because this strategy is tax inefficient.

The legislation in the United States, concerning the management of properties, is changed extensively since the start of the regime. Starting with the required outsourcing of the management of a REITs own property, to the current status, where REITs can manage properties of others. Easing legislation on this topic again will increase flexibility, but it can be seen as unfair competition in respect to taxed operating companies.

The issues that are described above are not only present for a new REIT regime, but for an existing as well. Besides the different aspects described above, the REIT regime has many more rules to fit the market. Because the market changes, the legislation for REITs need to change as well.

12

2.5 Constrains and performance

The discussed paper of Sirmans and Campbell (2002) gives insight in the different possibilities of policy for governments concerning REITs. The study is a policy paper and does not focus on data and in-depth research. This part of this paper will discuss some studies done on the effect of different constrains, institutional factors and structures on performance.

In a paper of Edelstein et al. (2011) the effect of institutional factors on real estate is studied by using multivariate regression models. They argue that a countries legal system has a significant impact on the risk premium of real estate. The effect is negative, so if the system is better, the risk premium, which is the difference between the return on real estate for a specific firm and the risk-free rate of the country where the firm is located, will decrease. The factor legal system in this paper, is originally build by La Porte (1998) and contains factors such as mandatory dividend payout, restrictions on reorganization, and the efficiency of judicial system. This means that changes in legal system can have impact on the performance of REITs, and therefore should be seen on market level.

Mühlhofer (2013) finds in his research that trading constrains for REITs have effect on the timing of REITs in the market. Therefore, REIT returns only reflect income returns and it will miss the effect of appreciation of property. The author suggest that REITs are, due to trading constrains, property income vehicles, instead of being property investment vehicles. The difference is that the property held by REITs are only generating income for the return of REITs, but not the combination of income and appreciation of the property.

3 Change in REIT legislation Canada

In Canada, major players in the REIT market and the investors desired changes in REIT legislation after the economic crisis. A first proposal was made in 2010, but it was not implemented in the law directly. These proposals in 2010 are not taken into account, for the reason that it was amended and completed with different new rules. In 2012, the new proposals were presented. These proposals are implemented in the C-48 bill by the Canadian government in 2013 and went effective as of 2011, so trust could make use of these new laws retroactively. This means that when trusts had no REIT-title

13 under the old rules, they could make use of the special tax advantages when they could comply with the new ones. The major changes of the C-48 bill are discussed by the Real Property Association of Canada, KPMG, Goodmans and others. An overview of the most important changes is listed below. With every amendment, the expected impact is discussed, with the help of papers from Mülhofer (2013), Sirmans and Campbell (2002), and Edelstein (2011).

3.1 Non Qualifying REIT assets

The old rules for REITs in Canada imply that the trust may not hold any Non-Qualifying REIT assets in their portfolio. The new rules give the REITs the opportunity to hold 10 percent of Non-Qualifying REIT assets in their portfolio (KPMG, 2013). That the threshold is increased, gives an indication of lower requirements for the trusts and can be seen as an easing amendment. Non-portfolio assets can now be hold, in comparison with the old rules, and this will increase the flexibility of the portfolio construction of a REIT.

In the previous situation, REITs were not able to time their sales, because NQRA had to be sold immediately, to keep the REIT status, and this can be seen as a trading constrain. The effects of trading constrains on the returns of REITs, studied by Mülhofer (2013), is discussed above, and indicates that trading constrains have a negative impact on the performance of REITs, because the REIT will miss the revenue of the appreciation of the property. Under the new rules, REITs are less constrained, and therefore, the hypothesis is that this amendment will be seen as positive by the market.

The companies have more freedom in the selection of assets, and this will reduce the risk in two ways. First, the risk of losing the REIT-title will decrease, because the rules can be followed more easily. Second, the possibility to reallocate the assets of the company with less restricted boundaries, and therefore better efficiency, can result in a decrease of the systematic risk. But on the other hand, the assets that are Non-Qualifying REIT Assets, are possibly high risk assets. An increase of these assets can cause a serious upward shift in systematic risk. Therefore, the effect on risk of this legislative amendment is unknown.

14

3.2 Eligible resale property

Before the amendments, REITs were not able to develop property for resale, but according to the new rules for REITs in Canada, eligible resale property is seen as qualified property (in some circumstances). Eligible resale property is defined as “eligible resale property held by the entity or of another entity affiliated with the entity, and is ancillary to the holding of that particular property, or property that is not capital property, held by an entity in which the REIT holds a security, contiguous to a particular real or immovable property that is capital property,” (KPMG, 2013).

The inclusion and description of this sort of activity will increase the distinctness for the market participants, and it creates new opportunities for REITs to increase revenue Monroe (2009). These opportunities arises when REITs are not only focusing on existing property, and exploit these, but also consider the possibilities to develop properties on their own and sell them afterwards. The expectation is therefore, that it will have a positive impact on the market.

3.3 Passive revenue test

Before the change in legislation, 95 percent of the gross revenues had to come from the following activities;

 rent from real or immovable properties  interest

 capital gains from dispositions of real or immovable properties  dividends

 royalties

The amendments decrease the percentage of the “revenue test’’ to 90 percent. Another aspect which is included in the list of possible revenue sources, is the gain from dispositions of eligible resale properties, as described above (Goodmans, 2010). The decrease of this revenue threshold, can be seen as a great relaxation of legislation. With this amendment, REITs were able to increase their revenue, coming from other than the five activities that are listed above, without losing their title as a REIT. Another positive aspect of this amendment is that a REIT now has the ability to manage their

revenue-15 producing assets better. The companies have more freedom in the selection of assets, and this will reduce the risk in two ways. First, the risk of losing the REIT-title will decrease, because the rules can be followed more easily. Second, the possibility to reallocate the assets of the company with less restricted boundaries, and therefore better efficiency, will decrease the systematic risk.

3.4 Foreign property

The rules before the C-48 bill was passed, didn’t allow revenue from foreign currency exchange, as qualified revenue. If a property was sold in exchange for a foreign currency, this would than needed to be exchanged for the domestic currency. Gains from this exchange was not seen as qualified revenue. The amendments now include foreign currency fluctuations, as qualifying revenue from: “currency hedging contracts in respect of foreign property that can reasonably be considered to have been entered into to reduce the trust's risk to fluctuations in the foreign currency; foreign currency debt incurred to acquire foreign real property for the purposes of earning rental revenue, and rent from foreign real property” (Goodmans, 2010, and KPMG, 2013). This inclusion is again an opportunity for REITs to increase their revenue. In that sense, it’s an easing amendment as well.

According to Monroe (2009), this change in legislation gives the opportunity for REITs to benefit from currency exchange income, because this revenue isn’t taxable and can therefor increase the performance of the REIT. On the other hand, exchange rates are very volatile, and therefore an expectation of the reaction of the market is hard to define and will mainly be an empirical question.

4 Data

4.1 Return data

The daily total adjusted percentage returns of 36 REITs in Canada are gathered from Datastream. On the moment of the start of the period of change, this was the number of REITs active in Canada. These returns are adjusted for capital movement such as dividends, so the total return includes dividend payments. In the market for REITs, these dividends are of great extent, so it is of high

16 importance that they are taken into account. The sample ranges from March 1st 2011 until December 31st 2013. This sample period should be extended enough to measure the effect of the change in legislation on REIT performance and is also used in a previous study of legislative change in the United States (Howe & Jain, 2004).

From the 36 REITs that are initially in the dataset, six of them are dropped due to incorrect or missing data. This leaves 30 REITs with daily returns, which generates more than 20,000 observations.

In order to calculate the abnormal returns around the event dates, an index will be used as benchmark for the market. For Canada this will be the S&P/TSX SmallCap Index. According to Howe and Jain (2004) and Clayton and MacKinnon (2001), a small cap index is most suitable for benchmarking REIT returns, because the correlation between the two is higher than for indices with big stocks.

4.2 Event-dates

This part of the paper will clarify the different event dates, that are related to the C-48 bill in Canada. A brief explanation of what occurred on the listed dates will be given. Table 1 shows the different dates and these dates are numbered. The numbers will refer to a specific event, which are obtained from the website of the Parliament of Canada. The period of legislative change can be divided into three parts, where it’s expected that they cause different market movements. The first part is the announcement of the proposals, the second part are the different stages of the C-48 bill passing the numerous checks and readings, and the third part is the actual signing of the bill. As mentioned before, the C-48 bill is based on proposals that were introduced to the public in 2010. These proposals weren’t converted into new laws. This event, the proposals of 2010, is not included in the list of events. This is because the proposals of 2011 included more and other amendments, so this can be seen as a new period of legislative change. The possibility of underestimation by investors on the day of the announcement in 2011 is present, because investors will expect a period of legislative change in the future.

17 On the 24th of October, 2012, the proposals were introduced to the public by the Department of Finance (Event 1). This introduction is new information for the market, so according to the literature, this could have an impact on the returns around this date. Because these proposals include the proposals of 2010 (these were not implemented as new legislation), plus additional amendments, the possibility of underestimation is present, because market participants already expect a period of legislative change. This can already be factored into the price of REITs.

On the 21st of November, 2012, the amendments of the rules concerning REITs, were presented to the House of Commons in Canada (Event 2). The fact that the announcement to the public is less formal than the introduction to the House of Commons, this presentation of the amendments could have an impact on the market for REITs. This expectation comes from the fact that the introduction to House of Representatives of the RMA in 1999, which is studied by Howe and Jain (2004) had a significant impact on the market in the United States.

The second part of the legislative change (Event 3-9), are the dates on which the C-48 bill passes the different stages of the governments legal departments. In general, these dates should not be of high interest if one is looking for positive abnormal returns. If the amendments face a hold-up in the between the stages, this off course can have a negative impact on the expectations of the market participants.

The last step in this period of change, is the actual signing of the bill. According to the paper of Guiso et al. (2014), market participants are more risk-averse during or after an economic crisis. This means that the actual signing of the bill, which takes the risk of a no-signing away, could have an impact on the returns on that day.

18 Table 1: Event dates

Event Dates

1 10/24/2012 First introduction of Proposals by Department of Finance 2 11/21/2012 Presentation of amendments to house of Commons

3 3/8/2013 Second Reading by House of Commons and Referral to Committee 4 3/27/2013 Committee Reporting the Bill without Amendment

5 5/21/2013 Concurrence at Report Stage

6 5/29/2013 Third Reading by Committee and First Reading by Senate 7 6/6/2013 Second Reading by Senate and referral to Committee 8 6/13/2013 Committee Report Presented

9 6/25/2013 Fourth Reading 10 6/26/2013 Royal Assent

4.3 Descriptive statistics

In Table 2a, the average daily returns and the accompanying standard deviations are listed for the returns of the REITs. The returns for the years 2011 and 2012 are positive, with decreasing standard deviation. The average daily return in 2013 was negative, but with a lower volatility than the previous years. In contrast to the returns of the 30 REITs in the dataset, the return of the S&P/TSX SmallCap Index have different signs. The average daily returns for 2011 and 2012 are negative, where it’s positive in 2013. The standard deviations of the market index are lower than that of the REITs in Canada, and is decreasing over the years.

Table 2a: Descriptive Statistics; Average returns

Year REITs Market

Average

daily return Std. Dev.

Average

daily return Std. Dev.

2011 0.035 3.203 -0.116 1.504

2012 0.076 2.476 -0.016 0.935

2013 -0.010 1.481 0.020 0.809

Observations 20610

The correlation between the returns of the REITs and the market index in the period of interest is shown in table 2b. The correlation is the highest in the first year of the period 2011-2013, and lowest in 2012. According to Clayton and MacKinnon (2001), a small-cap index needs to be used when analyzing the REIT market. The S&P/TSX SmallCap Index is used as benchmark, and the correlation-table gives prove that there actually is a correlation between the REIT-market and the index.

19 Table 2b: Correlation; REIT/Market returns REITs/Market 2011 0.2124 2012 0.1002 2013 0.1830 Overall 0.1690

In Figures 1 and 2 the returns of the portfolio and the S&P/TSX SmallCap Index around the event dates of Event 1 and Event 10 are shown in a graph. The expectation is that these two events have the most impact on the returns of the REITs in Canada. The REIT portfolio is constructed as follows. The variance of the returns for every REIT is inversed, and normalized to a percentage of the total. This percentage is the weight for every REIT in the data set (Thompson (1985) and Howe & Jain (2004)). Event 1 shows an increase in return on the moment of the event. The index moves in the same direction, but seems to be lagged. The returns around Event 10, where the C-48 bill was signed, show a positive return the day before, a negative return on the actual event-day and a positive return on the day after the signing. The index returns have the same sign, but have a more extreme value. This seems the case for the whole period.

Figure 1 Introduction Figure 2 Actual signing

-1.5 -1 -0.5 0 0.5 1 -5 -4 -3 -2 -1 0 1 2 3 4 5

Event 1

Portfolio Return Index Return

-4 -3 -2 -1 0 1 2 3 -5 -4 -3 -2 -1 0 1 2 3 4 5

Event 10

20

5 Methodology

5.1 Event study

The events that have occurred around the legislative change in Canada are discussed above. These include the introduction of the Proposals, the different reviews of the Senate and other layers of the government, and eventually the date of the signing of the bill. These different dates will be the event-dates, and the aim is to search for the effect of these different stages of the inclusion of the C-48 bill.

The event-study methodology has several different techniques, dependent on the moment of the event-dates. If the event dates are firm specific, which means that they are not on the same date for each firm in the dataset, a different technique is used, than when the event-dates are on the same date. The latter is the case for this study, where the legislative change occurs at the same time for all the REITs in Canada. When this is the case, the abnormal returns are correlated, and therefore the standard errors will be biased (Jaffe (1974), and Brown & Warner (1980, 1985). Also, no control group is available, because the legislative change has effected all of the REIRs simultaneously in Canada. To overcome the problem of correlating abnormal returns, the event-study will be done with the use of a particular adjusted regression. This regression is not perfect, where there are no control variables included. Following the examples of Jaffe (1974), Brown & Warner (1980, 1985) and Howe & Jain (2004), this thesis will use the same regression to calculate the cumulative abnormal returns. In this regression, the use of the index will partly account for these control variables. The following regression will be estimated:

𝑅_𝑝𝑡= 𝛼_𝑝+ 𝛽_𝑝𝑅_𝑚𝑡+ ∑ 𝛾_𝑝𝑘𝐷_𝑘𝑡+ 𝜀_𝑝𝑡 (1)

Here, the daily returns of the countries REIT portfolio, 𝑅_𝑝𝑡, are regressed on the daily returns of the used S&P/TSX SmallCap index, 𝑅_𝑚𝑡, and the ten dummies for the event days, 𝐷_𝑘𝑡. For each event date, 𝐷_𝑘𝑡 will be equal to 1/3, and zero otherwise. This ratio is chosen because of the event-window,

21 which is three days. The coefficient 𝛾_𝑝𝑘 will capture the immediate impact, which is valued as the cumulative abnormal returns, for the specific event and its event-window.

Because the event dates are on the same moment for each firm, the regression converted to a portfolio, as in (1), which is weighted. This weighting can be done in various ways, and the technique that is chosen for this thesis is justified by papers of Thompson (1985), Howe and Jain (2004), and Lee (2011). The weighting is done by calculating the variance of the returns of each REIT. The inverse of these variances of the REITs are normalized to one, and this weight is used to create the portfolio. The inverse is taken, in order to give the lowest weight to the firms with the highest variance in the dataset. This method, because of the weighting and regression technique, accounts for heteroskedasticity (Thompson, 1985), which is mainly a concern to calculate correct standard errors.

5.2 Risk

Another interesting aspect that will be investigated, is whether there is a decrease or increase of systematic risk, after the amendments. If one mentions the impact on shareholders wealth, not only the returns should be analyzed. For investors, the change is systematic risk is of great importance as well. Returns can be higher, but then the question is if the risk is higher too. This paper is doing so, by analyzing the market beta, which is the volatility in comparison to the market, before and after the legislative change. Howe & Jain (2004) have set up a regression model, which separates the market beta before the amendments, and the change afterwards. This thesis will use a similar approach. The following regression is estimated:

𝑅𝑡𝑖𝑡 = 𝛼 + 𝛽0 𝐷𝑢𝑚𝑀𝑜𝑛𝑡ℎ𝑡+ 𝛽1 𝑀𝑟𝑘𝑡𝑡+ 𝛽2 𝑅𝑖𝑠𝑘𝑡+ 𝜀𝑡 (2)

The daily returns of the REITs in a country, Rt_it, are regressed on a month dummy, which is one in the months after the legislation is changed, the market return, Mrktt, and the product of the month dummy, Risk_t. The change in risk for REITs in the specific country is displayed by β₂.

22

6 Results

6.1 Empirical results

6.1.1

The purpose of this study is, to investigate the effect of legislative change for the REIT market in Canada, on shareholders wealth. This effect can be judged by the effect of the market on ten different important event dates, controlling for the general market movement. The event-study will show the abnormal returns for these specific days, if there are any.

The changes in Canada contained several relaxing amendment, such as the increase of the possibility to hold Non Qualifying REIT Assets, the decrease of the percentage revenue that has to come from a specified list of operations and the inclusion of foreign operations. Linking these amendments to the existing literature, the expectation is that the market will view these changes as positive.

Table 3 shows four different regressions of the portfolio return on the index returns and the different event dates. Regression (1) shows the results of the event of the announcement of the proposals. Regression (2) shows the effects of Event 10, the actual signing, because this event was expected to have a significant effect. The third regression shows the results of Event 6. This is done because this event has an unexpected negative significant effect, and it’s not clear why this is. The last regression, regression (4), will includes all the variables.

Where it was expected that the announcement of the proposal to the public was an important event, the results in regression (1) suggest otherwise. The coefficient of Event 1, which is the cumulative abnormal return, shows the expected positive sign, but is insignificant. This is not in line with the previous findings of Howe & Jain (2004), who find significant effects during the introduction to the House of Representatives. The absence of significant effects on the first event-dates, can probably be caused by underestimation of the market participants. This can be a result from the previous proposals in 2010, which were introduced, but not implemented. This event is not included in this thesis, because it is a different period of legislative change. Because investors expect a proposal in

23 the future, the actual introduction doesn’t show a significant reaction in returns, possibly because the expectation of a legislative change is already factored into the prices. Another possibility is the change in risk-aversion of investors during the period of legislative change in the United States in 1999 and Canada in 2011. Probably the participant are not speculating on the possibility of a change, but are waiting till the actual inclusion, which can be seen as risk-averse behavior. This behavior can also be fed by the period of economic crisis (Guiso et al., 2014).

Regression (2) shows the cumulative abnormal returns of the event-window around Event 10. As discussed above, the actual signing of the C-48 bill could have a positive impact on the returns of REITs in Canada. The result of including only Event 10 in the regression (besides the index return and a constant), is a significant cumulative abnormal return of 1.3 percent. This means that Event 10, in total, has a positive effect on shareholders wealth. As discussed above, it is not a total surprise that this event shows a positive significant effect. The actual signing takes the risk of a no-inclusion away, which is positive for the market participants.

Figure 3: Anomaly at event 6

The third regression is included because of the surprising results. Including only Event 6 in the regression, the cumulative abnormal returns yield a negative 3 percent. This can’t be explained by the literature in this paper. The reason for this is unclear, and only some hypothetical explanations can be given. Because the portfolio is weighted by taking the inverse of the variance of the returns of REITs, one REIT with a high variance, and a possible huge negative return on the three days around the event-date, should not have much impact on the portfolio. This means that for all the REITs in the dataset, the return was low or negative on this date. One possibility is that an unidentified event happened on

-2 -1 0 1 -5 -4 -3 -2 -1 0 1 2 3 4 5

Event 6

24 the days of the event-window, which only influenced the REIT market, and not the whole market. This seems to be the case, because Figure 3, shows that the returns of the index are positive around Event 6. The fourth and last regression includes all the events. The easing legislation, where thresholds are lowered, and the number of sources for assets and revenue are increased, have caused a positive reaction of the market. This is shown in a positive cumulative abnormal return of 2 percent over the three days during Event 10, which is significant at the 1 percent level. This means that changing the legislation for REITs is necessary in the eyes of market participant, and it can be discussed that easing legislation has a positive impact on the returns of REITs. The difference in findings of Howe and Jain (2004) and this paper, suggest that the period of legislative change can have impact on the behavior of market participants. In 1999, a period of economic growth, the market reacted immediately, and in 2013, a period of economic recovery after a recession, the reaction of the market was likely more risk-averse.

The F-statistic of regression (4) is 11.63, which is significant at the 1 percent level. This implies that the coefficients of the cumulative abnormal returns are together different from zero. This statistic increases, when the regression contains less coefficients in regressions (1-3). The R2 is 0.32 and this is a satisfying result with only the return of the market index and the event-dummies as explanatory variables. This statistic is approximately of the same value as the statistic found in the study of Howe & Jain (2004). In the regressions (1-3), the R2 decreased by a small number, so it can be concluded that the index return has the most explanatory power.

25 Table 3: Impact on legislative change on REIT returns

(1) (2) (3) (4)

Return Return Return Return

Announcement Signing Event 6 All Index return 0.349*** 0.348*** 0.350*** 0.349*** (0.036) (0.036) (0.036) (0.037) Event 1 0.580 0.565 (0.572) (0.575) Event 2 -0.334 (0.395) Event 3 -0.635 (0.520) Event 4 0.151 (0.257) Event 5 -0.717 (0.556) Event 6 -3.109*** -3.106*** (1.071) (1.077) Event 7 -0.184 (1.020) Event 8 0.697 (2.698) Event 9 -1.178 (0.738) Event 10 1.253*** 2.018*** (0.468) (0.736) Constant 0.044** 0.043* 0.049** 0.048** (0.022) (0.051) (0.022) (0.028) N 687 687 687 687 R2 0.306 0.308 0.316 0.320 F 47.00 54.86 50.68 11.63

Standard error in parentheses *p<0.1,

**p<0.05,

***p<0.01

Notes: Based on data from 2010-2013, portfolio of 30 REITs in Canada

6.1.2

In addition to the search for abnormal returns on the Canadian REIT market, an interesting aspect of legislative change is, if it has an impact on the risk of the REIT market. One of the amendment was the change in the possibility to hold 10 percent of Non-Qualifying REIT Assets, and this can change

26 the risk for REITs in different ways. With this amendment, REITs are able to better manage their portfolio, repel properties/assets, or attract properties/assets without losing the status of a REIT, but on the other hand, these assets can be of high risk. Table 4 gives an overview of the regression outcomes.

Table 4: Change in risk due to legislative change in Canada Return Market Return 0.376*** (0.017) Dummy -0.0838*** (0.047) Risk -0.0173 (0.064) Constant 0.0607*** (0.020) N 20610 R2 0.0283

Standard error in parentheses * p<0.05

** p<0.01 *** p<0.001

Notes: Based on returns of 30 REITs in Canada, 2010-2013

As can be seen from the table, the market beta before the legislative changes was 0.376. The Risk variable indicates the change in market risk after the changes. This change has the expected negative sign, but statistically insignificant different from zero. This implies that the legislative changes made by the government had no impact on the risk of the Canadian REIT market. Explanations for this can be that the different amendments cancel each other out. If one amendment causes a decrease in risk, and another amendment increases the risk, like the implementation of eligible resale property, the total effect is zero.

27

7 Conclusion

The purpose of this thesis has been to analyze the effect on shareholders wealth of a legislative change in Canada. In 2013, the C-48 bill was signed by the government. This bill contained a few easing amendments for the REIT market in Canada, such as the possibility of holding Non Qualifying REIT Assets to 10 percent of the REITs assets, the lowering of the required percentage of revenue which has to come from an selected list of activities, the addition of a new activity to this list, and the possibility to trade in a foreign currency.

Using a regression based event methodology, the cumulative abnormal returns for ten different event dates concerning the period of change were analyzed. A regression based event-study was chosen, because the events were the same for the REITs in the data set and this can cause biased standard errors. The methodology chosen solves this problem and controls for heteroskedasticity. The regression resulted in a positive significant effect of 2 percent on the day that the bill was signed, Event 10. This implies that the market participants were not speculating on the possibility of the inclusion, but actually waited until the legislative change was certain. This is different than the previous findings of Howe & Jain (2004), who find a significant result on the day of the announcement. This difference can possibly be explained by the expectation of a legislative change, or by the increased risk-aversion, due to the economic crisis.

The second aspect concerning shareholders wealth, is the impact on the systematic risk after the change. With the use of a regression, the market beta before the change is calculated, where the change in risk is analyzed using the interaction of the market return and a before-after dummy variable. The effect on systematic change is negative, but insignificant. This may be caused by the different amendments of the legislative change, which can be offsetting.

The two findings show that easing legislative change for REIT markets is seen as positive by market participants. The REIT markets have grown substantially and evolved over time. It is more globalized than before, and its complexity has increased. Easing legislation will give more freedom in decision making, and this increases REIT returns. The economic situation in a period of legislative change can have impact on the way the market reacts, but it can be concluded that because the REIT

28 markets are changing in growth and complexity, it needs to be accompanied with changing legislation, in the eyes of investors.

The limitations of this paper lie within the short time-window. This paper aims to find the abnormal returns around different important dates of the legislative change. This gives a snapshot of how the market judges the amendments. For a more detailed analyses, an long-run event-study would be interesting. These type of studies aim to find the long run abnormal returns after specific events. Important literature about this topic is written by Fama, Fisher, Jensen, and Roll (1969), Jaffe (1974) and Mandelker (1974).

29

8 References

Benjamin, J., Sirmans, S. & Zietz, E. (2001) Returns and Risk on Real Estate and Other Investments: More Evidence. Journal of Real Estate Portfolio Management: 2001, 7(3), 183-214

Brockett, Patrick L., Hwei-Mei Chen & James R. Graven (1999). A new stochastically flexible event methodology with application to Proposition 103. Insurance: Mathematics and economics 25, 197—217.

Brown, Stephen J. & Jerold B. Warner (1980). Measuring security price performance. Journal of Financial Economics 8(3), 205-258.

Brown, Stephen J. & Jerold B. Warner (1985). Using daily stock returns: The case of event studies. Journal of Financial Economics 14(1), 3-31.

Chui, A.C.W., Titman, S. & Wei, K.C.J. (2003) The cross section of expected REIT returns. Real Estate Economics, 31, (3), 451-479

Clayton, J. & Mackinnon, G. (2001). The time-varying nature of the link between REIT, Real Estate and financial asset returns. Journal of Real Estate Portfolio Management, 43-54.

Dolley, James Clay. (1933). Characteristics and procedure of common stock split-ups. Harvard Business Review, 316—326.

Edelstein, R., W. Qian, and D. Tsang (2011). How Do Institutional Factors affect International Real Estate Returns? Journal of Real Estate Finance and Economics, 43, 130-151.

Fama, E. F., Fisher, L., Jensen, L.C. & Roll, R. (1969) The adjustment of stock prices to new information. International Economic Review, 10(1), 1-21.

Fama, E. F. (1970) Efficient capital markets: a review of theory and empirical work. Journal of Finance, 25, 383-417.

Feng, Z., Ghosh, C. and Sirmans, C.F. (2007). On the capital Structure of Real Estate Investment Trusts (REITs). Journal of Real Estate Finance and Economics, 34(1), 81-105.

Goodmans LLP. (2010). Finance Releases Proposed Amendments to the Qualifying REIT Rules. Guiso, L, Sapienza, P & Zingales, L. (2014). Time Varying Risk Aversion. Working paper, NBER. Howe, J.S., Jain, R. (2004). The REIT Modernization Act of 1999. The Journal of Real Estate

Finance and Economics, 28, 369-388.

Jaffe, Jeffrey F. (1974). The effect of regulatory changes on insider trading. Bell Journal of Economics and Management Science 5:1, 93—121.

Keogh, G. and D’Arcy, E. (1999) Property market analysis: an institutional approach, Aberdeen Papers in Land Economy (forthcoming).

Kim, J., Jang, S. (2012). Do hotel REIT companies face investment constraints? A comparison with C-corporation hotel companies. International Journal of Hospitality Management, 31, 573-578. KPMG LLP. (2013). Comparing REITs - US vs. Canada.

30 La Porta, R., Lopez-de-Silanes, F., Shleifer, A., & Vishny, R. W. (1998). Law and finance. Journal of

Political Economy, 106(6), 1113–1155.

Lee, W. “Risk-Based Asset Allocation: A New Answer to an Old Question?” The Journal of Portfolio Management, Vol. 37(4), 11-28.

Mühlhofer, T. (2013). Why Do REIT Returns Poorly Reflect Property Returns? Unrealizable Appreciation Gains due to Trading Constraints as the Solution to the Short-Term Disparity. Real Estate Economics, 41(4), 814–857.

Myers, John H. & Archie J. Bakay (1948). Influence of stock split-ups on market price. Harvard Business Review 26(2), 251—265.

REIT.com, History of REITs. Retrieved June 22, 2015, from https://www.reit.com/investing/reit-basics/history-reits.

Robert D. Campbell and C.F. Sirmans, (2002). Policy implications of structural options in the development of real estate investment trusts in Europe. Journal of Property Investment & Finance, 20(4), 388 - 405

Schipper, Katherine & Rex Thompson (1983). The impact of merger-related regulations on the shareholders of the acquiring firms. Journal of Accounting Research 21(1), 184-121.

Schwert, G. William (1981). Using financial data to measure the effect of regulation. Journal of Law and Economics 24, 121—157.

Stevenson, S. (2006). International Real Estate Diversification: Empirical Tests using Hedged Indices. Journal of Real Estate Research 19(2), 106- 131. `

Thompson (1985), Conditioning the return-generating process of firm-specific events: a discussion of event studies methods. Journal of Financial and Quantative Analysis, 20, 151-168.

Worzala, E., and C.F. Sirmans (2003). Investing in International Real Estate Stocks: A Review of the Literature. Urban Studies 40(5), 1115- 1149.