DEGREE PROJECT
REAL ESTATE AND COSTRUCTION MANAGEMENT REAL ESTATE ECONOMICS AND FINANCE
MASTER OF SCIENCE, 30 CREDITS, SECOND LEVEL STOCKHOLM, SWEDEN 2020
TECHNOLOGY
DEPARTMENT OF REAL ESTATE AND CONSTRACTIO ROYAL INSTITUTE OF TECHNOLOGY
DEPARTMENT OF REAL ESTATE AND CONSTRUCTION MANAGEMENT
Diversification Attributes of Dutch REITs
During Recessions:
Return, Standard Deviation and Liquidity Characteristics.
Tom Bergstrรถm
Patrik Carlsson
Master of Science Thesis
Title Diversification Attributes of Dutch REITs During
Recessions: Returns, Standard Deviation, and Liquidity Characteristics.
Authors Tom Bergstrรถm & Patrik Carlsson
Department Real Estate and Construction Management
TRITA-ABE-MBT-20590 TRITA number
Supervisor Bertram Steininger
Keywords REITs, Netherlands, Real Estate, Mixed-Asset
Portfolio, Liquidity, Performance, Diversification
Abstract
The objective of this thesis is to determine the performance of Dutch REITs and liquidity aspects during recessions and economic upswings as well as correlation with other asset classes to gain further knowledge in the field of real estate investment and asset performance during certain time periods. This is achieved through a quantitative analysis of historical daily returns, standard deviation and transaction volume of shares regarding REITs and other assets that usually pertain to an investorโs portfolio. The analysis covers the time-period just prior to the global financial crisis up until the beginning of the financial crisis caused by Covid-19. Analyzed data display some correlation between REITs and other assets. However, the data still implies some diversification benefits of incorporating REITs in a portfolio during all economic states through the time periods with both the objective to minimize risk i.e. standard deviation, and to maximize return. The liquidity results on offset-time efficiency is comparable with other assets stock, which suggests that REITs are as liquid as other stock and thus is more liquid than direct real estate investments. In conclusion the data does support some diversification benefits of including Dutch REITs in a Netherlands-based investment portfolio. However, to what extent cannot be determined, in part because of individual investors preferences, beliefs, and behavior, but also because of additional factors, such as dividends, that affect the value of REITs to an actual investor.
Acknowledgement
We would like to thank our thesis advisor Assoc. Prof. Bertram Steininger at the Department of Real Estate and Construction Management of the Division of Real Estate Economics and Finance at KTH, The Royal Institute of Technology. He guided us through the research process and aided us whenever we ran into difficulties while still allowing this thesis to be our own work. Thank you.
KTH, Royal Institute of Technology, Stockholm, May 2020.
Tom Bergstrรถm & Patrik Carlsson
Examensarbete
Titel Diversifieringsattribut hos nederlรคndska REITs under
lรฅgkonjunkturer: Avkastning-, standardavvikelse-, samt likviditetsegenskaper
Fรถrfattare Tom Bergstrรถm & Patrik Carlsson
Institution Fastigheter och Byggande
TRITA-ABE- MBT-20590 TRITA nummer
Handledare Bertram Steininger
Nyckelord REITs, Nederlรคnderna, Fastigheter,
Investeringsportfรถlj, Likviditet, Diversifikation
Sammanfattning
Syftet med arbetet รคr att faststรคlla Nederlรคndska REITs prestanda och likviditet under lรฅgkonjunkturer och ekonomiska uppgรฅngar samt korrelationen med andra tillgรฅngsklasser fรถr att fรฅ ytterligare kunskap inom omrรฅdet fastighetsinvesteringar och investeringstillgรฅngars prestanda under vissa tidsperioder. Detta uppnรฅs genom en kvantitativ analys av historisk marknadsutveckling, standardavvikelse samt transaktionsvolym av antalet aktier.
Det รคr REITs tillsammans med tillgรฅngar som vanligtvis hรคnfรถr sig till en investerares portfรถlj som har undersรถkts.
Analysen behandlar ett tidsspann frรฅn perioden strax innan finanskrisen 2008 fram till bรถrjan av den finansiella krisen orsakad av Covid-19. Analyserad data visar viss korrelation mellan REITs och de andra tillgรฅngarna, men innebรคr fortfarande vissa fรถrdelar med att inkludera REITs i en portfรถlj under samtliga ekonomiska tillstรฅnd under samtliga tidsperioderna. Bรฅde avseende syftet att minimera risken, som i detta fall utgรถrs av standardavvikelse, samt fรถr att maximera avkastningen. Likviditetsstudien visar att avyttringshastigheten รคr jรคmfรถrbar med andra aktietillgรฅngar, vilket antyder att REIT รคr lika likvida som andra aktier och dรคrmed รคr mer likvida รคn direkta fastighetsinvesteringar. Sammanfattningsvis stรถdjer resultaten vissa diversifikationsfรถrdelar med att inkludera Nederlรคndska REITs i en Nederlรคndskt baserad investeringsportfรถlj. Men i vilken utstrรคckning kan inte faststรคllas, delvis pรฅ grund av enskilda investerares preferenser, รถvertygelser och beteende, men ocksรฅ pรฅ grund av ytterligare faktorer, sรฅsom utdelning, som pรฅverkar vรคrdet av REIT till en faktisk investerare.
Fรถrord
Vi vill tacka vรฅr handledare Lektor Bertram Steininger vid institutionen fรถr Fastigheter och Byggande, avdelningen fรถr fastighetsekonomi och finans, vid KTH, Kungliga Tekniska Hรถgskolan. Han vรคgledde oss genom forskningsprocessen och hjรคlpte oss nรคr vi stรถtte pรฅ svรฅrigheter medan han fortfarande tillรคt denna avhandling att vara vรฅrt eget arbete. Tack.
KTH, Kungliga Tekniska Hรถgskolan, Stockholm, Maj 2020.
Tom Bergstrรถm & Patrik Carlsson
Table of Content
Introduction 1
Delimitations 1
Research Questions 1
Disposition 2
Background 3
History of the Dutch Financial Landscape 3
Pre Covid-19 Real Estate Market Outlook 4
The Global Economy 2006-2020 4
Characteristics of REITs 7
Qualifications and Regulations for REIT status in the Netherlands 7
Literature Review 10
Emergence of REITs in Europe 10
Real Estate as an Investment 11
REITs in Relation to the Real Estate Market 11
Behavioral Differences Between REITs and Stocks 11
Portfolio Diversification 12
Financial Crises 13
Theoretical Framework 14
Portfolio Theory - Mean-Variance Optimization 14
Behavioral Finance 14
Liquidity 15
Methodology 16
Quantitative Method 16
Data Sources 16
Mathematical Analysis 17
Results 21
Research Question 1 21
Research Question 2 24
Research Question 3 29
Analysis and Discussion 30
Conclusion 32
References
1
Introduction
The real estate market sector accounted globally for approximately US$ 1,8 trillion in 2018 (โGlobal real estate investment H1 2019,โ 2019). Making it undeniably one of the biggest investment sectors in the world. To further expand current research, this study examines risk aspects during recession periods using measurement derived from standard deviation, historical daily returns and daily traded volume of shares. More specific, the study intends to investigate the diversification benefits and liquidity attributes of Dutch REITs during the financial crisis of 2008 and the global downturn of the economy in 2020 with regards to the pandemic outbreak of Covid-19. Apart from the 2020 financial crisis, the last few yearsโ low interest- and inflation rates real estate makes for an attractive investment, commercial real estate in particular. Commercial real estate generally generates a quite constant cash flow as well as a historically high potential for capital gain on divestments (Lee, 2005; Newell and Marzuki, 2018). What sets traditional real estate investing to a disadvantage to private as well as institutional investors is the risk associated with low transparency (Hordijk and Teuben, 2008; Newell, 2016), illiquidity and the subsequent deficiency in valuations (Clayton et al., 2009; Clayton and MacKinnon, 2002; Fisher et al., 2003; Lin and Fuerst, 2014; Scofield, 2013). Obviously, these mentioned inefficiencies pertain in particular to direct investments of real estate. In comparison to for example stocks, direct real estate investing see less frequent transactions (Lin and Fuerst, 2014). Direct real estate investments are generally also quite large, which is of no issue to institutional investors, but could for the individual investor be a hurdle to enter the market. Furthermore, the high transaction costs generally linked with direct real estate investing could be seen as discouraging. In contrast, the stock market does not suffer from high transaction costs and is highly liquid, which are both desirable traits for investors (Fisher et al., 2003; Lin and Fuerst, 2014). REITs regimes allows for merging of features between real estate investing and the stock market. REITs are seen as highly liquid and are exposed to only negligible transaction costs. Subsequently the demand for exchange traded equity REITs is high and makes for a sensible component of an investors asset portfolio (Buttimer et al., 2012, 2005; Ciochetti et al., 2002). In times of financial crises, as the one faced caused by the 2020 novel Coronavirus, Covid-19, a diversified portfolio with stable assets is without doubt of importance for many investors. The results of this study provide further insight on the efficiency of Dutch REITs as an investment vehicle and its benefits to investors, as creating a greater understanding of REITs general risk-adjusted performance and diversification benefits in a mixed-assets portfolio.
Furthermore, the study indicates how the liquidity attributes are affected in recessions for Dutch REITs and in relation to other major listed assets in the Netherlands.
Delimitations
To uphold a sufficient level of reliability and preciseness in the study, limitations are defined to frame the research questions. The quantitative portion of this research is focused on the Dutch market. However, no analysis is conducted on comparable foreign markets, because of regulation dissimilarities between countries. The research data subject to analysis will be contained to publicly accessed financial data from the Dutch market, i.e. listed REITs, stocks, and bonds. Furthermore, direct property investments will not be included in the analysis because of limitations in the coherence of data over time and in transparency of financial processes. All five existing REITs in the Netherlands are equity REITs which is why this study exclude mortgage REITs and hybrid REITs from the research. Thus, the mortgage and debt markets are excluded. Dividends from the REITs will not be included in the quantitative analysis.
Research Questions
RQ1: How do the Dutch REITs returns, with regards to their standard deviation, perform in comparison to other major asset classes in the Netherlands during recessions?
RQ2: Do Dutch REITs provide diversification benefits in a mixed-asset portfolio during recessions?
RQ3: How do Dutch REITs compare to other asset classes in terms of liquidity before, during, and after a period of recession?
2
Disposition
Chapter 1 โ Introduction
Chapter 2 - Background
In this chapter, some foundations of Dutch finances are explained. From the development of financials establishments and the significance of various institutions along the way until today's financial landscape in the Netherlands. The current market situation is described as well as the characteristics of REITs, their structure, regulations, and differences with U.S. REITs. Lastly, the global economy through specific time periods with recessions and upswings up until the recession due to Covid-19.
Chapter 3 - Literature Review
Previous research relevant to the research questions is reviewed. The development of REITs in Europe as well as REITs in relation to the rest of the real estate market is explained. Furthermore, the work upon which the theories utilized in this thesis paper is presented.
Chapter 4 - Theoretical Framework
The research is based on the theoretical framework of Modern Portfolio theory, behavioral finance and liquidity. These are commonly recurring concepts within the investment industry and are very useful in understanding investment strategy combined with investors behavior.
Chapter 5 - Methodology
This chapter present methods used in order to answer the research questions. The focus on historical return and standard deviation of the study requires an objective approach when analyzed, hence the structure of a quantitative research. Central contribution is derived from Modern Portfolio Theory but also individual performance and liquidity attributes. Each component and procedure are further described.
Chapter 6 - Results
The processed data is displayed and presented. Initial analysis of the data and stating of the research question answers follows.
Chapter 7 โ Analysis and Discussion
A presentation of the analyzed results as well as the actual analysis and discussion, linking back to background and theory.
Chapter 8 - Conclusion
A presentation of conclusions and answers for the research questions together with suggestions for further research.
3
Background
Chapter 2 - Background. In this chapter, some foundations of Dutch finances are explained. From the development of financials establishments and the significance of various institutions along the way until today's financial landscape in the Netherlands. The current market situation is described as well as the characteristics of REITs, their structure, regulations, and differences with U.S. REITs. Lastly, the global economy through specific time periods with recessions and upswings up until the recession due to Covid-19.
History of the Dutch Financial Landscape
The Netherlands had a lagging industrial development yet being inventive in producing new financing arrangements. When the time came to develop the industrial economy in the end of the nineteenth century, the by then advanced and active market did not initially have any significant impact. Even though the market was overcrowded in addition to being upheld to the high standard of unrestricted public access by the authorities. What partially drove the development was colonial securities, international industrial stocks and shares (American in particular) as well as foreign state loans. The financing from external sources made it possible for providing credit in anticipation for the issuing of public industrial shares, and thus the subsequent repayment (de Jong and Roell, 2005; Zanden, 1998, 1987). The larger part of the financing of the initial industry development is credited, as in many other countries, to the local cities earlier up risen trading elite. Cross business entrepreneurs in different dynasty groups stimulated trade within the dynasty. The government and the current monarchical governor, King William I, gave ease with credit lines. Infusions from private equity and acquaintances retained earnings laid ground for the larger part of the source of risk capital for industry in the Netherlands, and Amsterdam, during the latter part of the nineteenth century and provided the foundation for the markets that grew during the twentieth century. The only exception is some capital-intensive project, such as rail networks, that relied on government support to an extent and general issued public shares to be put into motion. (de Jong and Roell, 2005)
When it comes to the role of banks and banking on the development of Dutch finance, and particularly the development of the earlier mentioned industrial ventures, one could argue that the impact was limited. Not only during the earlier days of the Dutch financial landscape, but well into the twentieth century. There were initiatives for banking during the nineteenth century, but few were fruitful. One of the most attributing factors was the
โProlongatieโ- financing system. A financing method comprised of callable short-term margin loans from wealthy individuals through local agents, typically a mix of lawyers, brokers, and notaries. The supply of wealthy individuals and financial agents were plentiful due to the strong stock market trading culture developed during the prior centuries of trade. Commonly, and considering the name, the short-term loans that often was directly made available to the borrower and thus not through securities, prolonged. Still, they were callable on a short term as to the prior arrangement. For collateral, the most common approach was to back with exchange traded assets such as commodities and securities, (de Jong and Roell, 2005).
Apart from banks and private investors there are a number of institutions that has contributed to the development of industrial finance in the Netherlands, especially during the twentieth century. Pension funds and different sorts of insurance companies played a significant role in the absorption of issued bond as well as for taking equity stakes. They also offered long-term loans. During the last century restrictions of how insurers are allowed to invest has loosened, moving investments from primarily bonds issued by the Dutch government but also from other considerably safe and liquid assets. The trend of investments in non-government bonds and equity didn't really pick up speed until the second half of the twentieth century when more investors moved their interest towards corporate equity (de Jong and Roell, 2005; Jonker, 1995).
4
Pre Covid-19 Real Estate Market Outlook
The Netherlands has over the last years seen declining vacancy rates and yields in most market segments. The changes in consumer behavior towards online shopping has brought with it record high investment volumes in the logistics and warehouse market with especially high demand for high-grade large-scale facilities. Also, last-mile distribution warehouses have seen increased demand. Take-up volume in logistics saw a peak in 2018 with a slight decrease in 2019 because of the limited supply. However, there is approximately 2 million square meters of logistics facilities under construction with more than half being pre-let. With further globalization and structural consumer behavior changes some predict a high demand for logistics facilities going forward, with some trade flow disruptions as a result of the Coronavirus (Bertens, 2019a). On the office market side, the Netherlands has seen a very high increase in demand and subsequently rents over the last years. Record low prime yields and vacancy rates in most submarkets has also been recorded. Some slowdown in take-up volume has also been seen in the office market, but as with the logistics market a result of lack of supply. The office market is expected to grow slower than the previous years, not only because of the Coronavirus, but because of little addition of new supply in both prime and subprime markets (Bertens, 2019b). The residential market currently faces a demand that outnumbers the supply much because of low numbers of new building permits, but also because of new strict regulations of pollution and land contamination. The constricted market for housing makes for increased prices, especially in the larger cities. (Bertens, 2019c). The one market segment that doesn't follow the trend as well as the others is retail. Even though investor focus is on a continuous high level for prime high streets and neighborhood shopping centers along with high prime retail rents in the top 15 cities, properties outside these criteria is on a substantial decline. Retail expansion is modest to none. With the changing consumer behavior the market development is not a surprise and retail actors are subsequently required to adapt to new trends (Bertens, 2019d).
The Global Economy 2006-2020
A holistic approach in portraying the global economy is summarized in Figure 1 to 5 by using the indices of various portfolios around the world during specific time periods of one year. Chosen benchmarks are โS&P 500โ,
โS&P Europeโ, โS&P Topix 150 yenโ, โS&P Asia 50โ, โS&P Latin America 40โ and โFTSE 100โ (see Table 1).
Their daily return changes are shown in relation to the starting point of the time period in each figure. Assisting in greater understanding of the economic consequences during a recession. Where the return for example in both GFC and Covid drops as high as 50-60 percent in relation to the beginning of the time period.
Table 1. Global Indices
Name Included assets
S&P 500 500 large listed companies in the U. S.
S&P Europe 350 leading โblue-chipโ companies from 16 developed European markets S&P Topix 150 yen 150 large โblue-chipโ corporations from the Japanese market
S&P Asia 50 50 leading โblue-chipโ companies that are listed in Asian markets (Hong Kong, Korea, Singapore and Taiwan) S&P Latin America 40 40 leading โblue-chipโ companies that capture almost 70 percent of the regions total market capitalization (Brazil, Chile,
Colombia, Mexico and Peru)
FTSE 100 100 listed companies with the highest market capitalization of the London stock exchange
5 The market in Latin America is particularly showing a positive trend prior to the global finance crisis as seen in Figure 1. Other markets also display somewhat positive trends with minor fluctuations i.e. volatility, although not as distinct as in Latin America. Noticeable drop on all markets at the end of August 2007.
Pre GFCDaily Return Indices
Figure 1. 2006-09-26 to 2007-09-26 daily return data from Thomson Reuters Datastream visualized as changes with regards to the chosen base point at 0 percent.
The global effects on all markets is evident as they all illustrate drops between 30-60 percent during the last quarter of 2008. Latin America performance exhibiting the most volatile behavior (Figure 2).
GFC Daily Return Indices
Figure 2. 2008-12-30 to 2008-12-31 daily return data from Thomson Reuters Datastream visualized as changes with regards to the chosen base point at 0 percent.
The referred time period โUpswingโ in Figure 3 displays a relatively more stable market in comparison to other time periods as the values only fluctuate between negative 20 percent to positive 30 percent. The majority also have a positive trend.
6
2015-2016 โUpswingโ Daily Return Indices
Figure 3. 2015-09-29 to 2016-09-30 daily return data from Thomson Reuters Datastream visualized as changes with regards to the chosen base point at 0 percent.
Various behavior with no clear general trend even if the โwesternโ markets suggest a more positive outcome.
Compared to the outlook before the global financial crisis there do not seem to be any recurring patterns connected to a downfall. Asian market negative pattern could be affiliated with ongoing demonstrations in Hong Kong and as the outbreak continent epicenter of Covid-19 (Figure 4).
Pre Covid-19 Daily Return Indices
Figure 4. 2018-03-27 to 2019-03-27 daily return data from Thomson Reuters Datastream visualized as changes with regards to the chosen base point at 0 percent
7 The spread of Covid-19 is expanding exponentially over the world resulting in a pandemic disease as seen in Figure 5. The repercussions of this are evident as all markets display a significant drop in a short period of time.
Covid-19 Daily Return Indices
Figure 5. 2019-03-27 to 2020-03-27 daily return data from Thomson Reuters Datastream visualized as changes with regards to the chosen base point at 0 percent.
Characteristics of REITs
Real estate investment trust is an investment concept constructed during the 1960s in the United States, with the objective to facilitate investment in real estate and profit mainly from their yearly revenue income (Grybauskas and Pilinkienฤ, 2019). These are called equity REITs and another common form is mortgage REITs whose aim is instead to focus on mortgage obligations. It also exist a combination of the two, which is known as a hybrid REIT (Han and Liang, 1995). Public REITs are listed on a national stock exchange as a company. The Equity REIT own real estate that provides yearly cash flows through the rents of their tenants (EPRA, 2019)
Private real estate is often associated with negative characteristics for an investor e.g. liquidity and information asymmetry. By instead investing in listed/public real estate reduces the chance of such factors and thereby also reducing risk. From an international point of view, REITs structures address these problems together with tax efficiency for investors, creating unique investment opportunities in the listed real estate sector. However, over a one year holding period, public real estate tends to be more correlated with the stock market. Also, since listed securities are traded on a daily basis they display a more volatile behavior (Hoesli and Lekander, 2008).
Qualifications and Regulations for REIT status in the Netherlands
Certain criterion must be met to qualify as a REIT in accordance to both law and local regulations. The REIT must comprise of a corporation or an association. It could also be a business trust. Furthermore, said corporation, association or trust must be managed by a board. The board ought to be compiled of directors and trustees. A REITs source of income has to be derived from dividends, rents, interest and other gains related to real estate and real estate sales. On the investment side the requirements for โREIT statusโ and subsequently the assets in a REITs portfolio, real estate assets, government bonds or shares in other REITs (maximum of 10 percent, including other cash items) have to be invested into a degree of 75 percent of the portfolio. Ownership and shares also comes with restrictions if REIT status are to be reached. To qualify as a REIT, there has to be a minimum of one hundred unitholders. Also, five or fewer of the unitholders are not allowed to own above fifty percent of the total outstanding shares. There is one exception, which is that as of December 31, 1993, pension funds in the Netherlands are permitted to acquire an unlimited percentage of the shares without disqualifying the tax status of the REIT (Schuurman and de Lange, 2019).
8 There are differences on who is eligible for REIT status in the Netherlands compared to who is eligible in the U.S.
The distinction is important to make since the majority of REIT literature focuses of the regulations of U.S REITs, (Grybauskas and Pilinkienฤ, 2019). In the U.S any corporation with a predominantly tangible assets in real estate can apply for REIT status without being met with an initial taxation, although already earned profits must be distributed as dividends. Assets with incorporated gain will be taxed at entity level if recognized within five years.
Because of the relatively easy conditions and quite acceptable costs, many RETIs stems from converted corporations (Taylor and Vermulen, 2013). Regarding the Netherlands, the requirements are stricter. Any accumulated profits must, before the application of REIT status, be taxed under the normal Dutch corporate income tax, which is a 25 percent. Thus, there is no significant tax incentive for conversion to REIT for regular corporations. In the United Kingdom, Italy, and Belgium, the rules are the same as for the US, in converting to REIT (Taylor and Vermulen, 2013).
Depending on where the REIT is registered and operative, there are local regulations and requirements that takes effect. In the Netherlands, the requirements for a fiscal investment institution (FBIs) must be met to reap benefit of the zero percent corporate tax which characterized a REIT along with the above-mentioned qualifications. To attain FBI-status, and thus be regulated, the shares or units are to be admitted to and in accordance with the Dutch Financial Markets Supervision Act (FMSA). Another way to be considered regulated under FMSA, the FBI, or if applicable its manager, is required to hold a license as referenced under said act. The third way is to be exempted completely from the license requirements set up by the FMSA. Listing on a stock exchange is not a requirement for FBI-status (Havenga and Lekkerkerker, 2019).
Entrance to the FBI regime is only admitted for Dutch public companies, mutual funds and limited liability corporations. Non-Dutch entities may also enroll under the condition that the entity is established in an EU member state and under that stateโs laws. Additionally, the former โDutch Antillesโ-islands or a nations with tax treaties, may also be able to qualify for FBI-status (Schuurman and de Lange, 2019).
Investor Requirements
Investor quantity requirements differ between regulated public FBIs and Private FBIs. Private FBIs are required to have no less than 75 percent of its shares held by entities that does not stand subject to taxation of its profits or are exempted from tax, or other regulated FBIs. Furthermore, no single individual is allowed to hold direct or indirect interest of more than 5 percent. Residents of the Netherlands are not allowed, in a FBI, to hold more than 25 percent interest though non-resident entities of mutual funds with capital completely or partly divided into shares or units (Schuurman and de Lange, 2019).
Assets and Income Activities
Regarding assets and income activities there are certain requirements as well. The nature of an FBI, and thus Dutch REITs, must be of passive investment strategies. This entails restrictions that disallows active investments as well as trades in real estate development. However, the FBI is allowed to take part, i.e. mange and hold shares, of a company dealing in real estate development activities. The entity or subsidiary will in that case be subjected to standard corporate income tax up to a degree of 25 percent (Schuurman and de Lange, 2019).
Dividends
All taxable income, not including capital gains for future investments, must be paid out to the unit holders. FBIs is by rule required to pay out the entirety of its taxable profit within the span of eight months of the financial end of the year. This does not apply to capital gains that are assigned for a reinvestment reserve (Schuurman and de Lange, 2019). In addition, the capital gains generated by the FBI have the possibility to be distributed tax-free to non-domestic investors (Van Oostrom et al., 2015). Dividends paid out from the FBI are regulated with a 15 percent withholding tax, with reservation for reductions under applicable tax treaties (Schuurman and de Lange, 2019). Shareholders pay income tax on dividends received from the FBI. Depending on local regulations the amount vary regarding both dividends and tax (EPRA, 2019).
9 In Contrast to U.S REITs
Although REITs can be found in both the United States and the Netherlands, there are regulations that sets them apart. The legal requirements for REIT status in each country differs in whether foreign entities are allowed to qualify as a REIT in the respective country. The approach in the United States is that foreign entities are not allowed to become a REIT, even if all other requirements are met. The reason for this seems to be that there are tax rules that differs for U.S and non-U.S corporations in such a way that the exclusion of foreign entities does not fall under the discrimination act (Lang et al., 2010). In comparison, the Dutch REIT regime does allow for foreign entities to qualify for REIT status on the ground of the Dutch discrimination act. Under the Netherlands- United States Tax Treaty, U.S based, i.e. under U.S law, companies or residents can qualify for Dutch REIT status (Neth.-U.S., 1992).
Regarding regulation for allowed activities the Netherlands apply considerably harder restrictions even though both REIT regimes were designed for the same purpose i.e. for collective passive investments in the principle of tax neutrality (Taylor and Vermulen, 2013). The Dutch apply activity requirements while in the U.S income and asset tests are conducted instead. Because of interpretations, Taylor and Vermulen (2013) argues that, in the U.S, the rules have evolved and now enables activities that surpass the scope of passive investment activities, especially when incorporating taxable subsidiaries. The subsidiaries can perform any activity. In the Netherlands, however, whilst there is a form for taxable subsidiaries for the REITs, they are required to keep within activities in the scope of project development performed by REIT unitholders or another related party. The regime allows for taxable subsidiaries engaged in customary services to the REIT, which allows for activities within the business sphere of the REIT. Although, the subsidiary is not exempt from corporate tax like the parent REIT (Taylor and Vermulen, 2013).
In the Netherlands there are a set of rules that acts as a safeguard for the taxable basis of REITs and restricts debt financing. The reason is for attracting dividend withholding tax connected to distribution. Furthermore, the Dutch government may experience decreased income from dividend withholding tax from higher leverage. The first rule restricts deductions of applied interest. The second rule limits activities to the scope of investment activities. No restrictions exist in the case of the U.S, except for REIT subsidiaries and debt financing. In the aspect of profit distributions, the rules are less strict in the U.S compared to the Netherlands. If Dutch REITs do not distribute all profit (excluding capital gains) their REIT status is lost (Taylor and Vermulen, 2013). If a REIT in the U.S would fail to meet a requirement, for example generating โbad incomeโ, the consequences would be further taxation.
According to Taylor and Vermulen (2013), the constant threat of disqualifying for REIT status because of minor error, may be a disproportionate punishment for the magnitude of the fault.
10
Literature Review
Chapter 3 - Literature Review. Previous research relevant to the research questions is reviewed. The development of REITs in Europe as well as REITs in relation to the rest of the real estate market is explained.
Furthermore, the work upon which the theories utilized in this thesis paper is presented.
Emergence of REITs in Europe
In 2017 there was more than 477 REITs all over the world, although only 13 out of 28 European countries had established legal framework and adopted REITs as an alternative investment strategy (Grybauskas and Pilinkienฤ, 2019). Studies on REITs and their performance have been conducted since 1990 where they investigated trades in the 70โs and if stocks are segmented from the real estate market. Result showed that there is an integration between REITs and the stock market. Since then there has been a lot of research concerning the U.S REITs. Mostly multi-factoring models are used in order to find factors that explain the various behavior and outcome of the REIT performance over time (Coletta and Busato, 2019).
REIT is a relatively new investment vehicle in Europe, yet an important aspect for growth within commercial real estate. Creating a market beyond the domestic nation and stimulating international investments (Marzuki and Newell, 2019). France and Germany are great examples of countries that have benefited from this (Wijburg, 2019;
Wijburg and Aalbers, 2017). There were a number of European countries that introduces REITs around the global financial crisis, 2007 - UK, Germany and Italy and 2009 - Finland and Spain, that constrained continued growth.
Although due to undervalued assets during the crisis, asset portfolio value in e.g. Spain greatly expanded post- crisis (Marzuki and Newell, 2019). Europe's small percentage of REITs also reflects previous studies that analyze companies located in Europe (Grybauskas and Pilinkienฤ, 2019). Research in both France and Spain states that the REITs contribute with good risk-adjusted returns in comparison to bonds, which is very helpful tool in constructing a diversified portfolio (Marzuki and Newell, 2018; Newell et al., 2013). Real estate is an asset allocation that is often used to diversify risk and hedge for inflation (Lekander, 2017). As a result of the cost- competitive and liquidity advantages of REITs, large institutional investors, as for example pension funds, are likely to be interested (Andonov et al., 2012).
Belgium is one of the European countries that embraced the REIT structure and is in fact the second oldest market for REITs in Europe, it has been around since 1995. The REIT investment vehicle has since then overgone multiple structural transformations in order to achieve todays stature of a sophisticated investment opportunity. With characteristics of transparency, professionalism and fiscal efficiency. The performance analysis of REITs in Belgium present superior risk-adjusted returns and robust diversification possibilities in a long-term investment perspective (Marzuki and Newell, 2019).
Germany acquired law structure to form REITs in 2007 and as of 2017 there is only four REITs on the market.
One explanation to the slow expansion of Germany REITs could be that there is a lack of incentives for already established listed property companies to convert to a REIT. Another factor is the regulations that prohibits REITs to compose of residential real estate. Attracting global real estate investors to Germany derives from the concept of liquid assets, that easily can be realized using REITs. To fully develop and improve current REIT regulations would create effective investment opportunities both local and global, resulting in increased exposure and interest for the German real estate market (Marzuki and Newell, 2018).
In the future for international equity REITs there is five separate major directions that is identified. Global regulations, โflagship propertyโ, focus on people not property as the REIT portfolios expand and increased focus on the duration and quality of cash flows rather than specific asset classes. However, it is unlikely that they evolve independently whereas interaction effects are possible if not unavoidable (Parker, 2019). The structure and regulations regarding REITs and their uniqueness creates motivations for expanding research in the field (Dogan et al., 2019).
11
Real Estate as an Investment
Incorporating real estate into a mixed-asset portfolio is well known to be beneficial (MacGregor and Hoesli, 2014).
Direct holdings are a frequently used method of diversifying risk within real estate investments. However, it is also often affiliated with lack of liquidity and information asymmetries. The market was until recently also quite limited to non-listed real estate funds or listed real estate companies. The objective of transparency and professionalism for many investors is made possible by institutional and legal changes, creating investment opportunities in former limited markets and countries (Hoesli and Lekander, 2008). Demand also derives from benefits received from the long duration within property assets together with contractual rents that include inflation indices (Chun et al., 2000; Craft, 2001). Growth in commercial real estate market is one of many factors that is of importance when investing globally. Factors of local demand, liquidity, legislations, governance and political stability also affects the decision (Fiorilla et al., 2012). Most frequently used method to achieve real estate market exposure have historically been through direct investments. However, this trend is shifting towards third party allocations where indirect investments are used to achieve a well-balanced real estate portfolio. This include REIT structures and the derivative market (Hoesli and Lekander, 2008).
Previous research on weather REITs is exposed to stock market risk and risk associated with changes in interest rates show mixed results. One aspect concerning financing real estate which often derives from heavily borrowed funds, is that that the value can be influenced by cost associated with different type of financing. An upswing in interest rate could thereby affect the affordability and demand for real estate in a negative way. The various and individual characteristics of REITs are fundamental factors that affect the empirical results of systematic risk and interest-rate risk exposure. Although, the general conclusion is that REITs are more sensitive to stock-market conditions than interest-rate changes (Allen et al., 2000).
REITs in Relation to the Real Estate Market
To compare REITs to the general real estate market, one must decide on certain factors to analyze. A common comparative method is to look at risk, return, and the structure of distribution. With a starting point in the S&P/Case-Shiller (SCS) index and related data, Cotter and Roll (2015) observed that REITs over the last thirty years in the United States have had a slightly lower degree of market risk than equity but a significantly higher volatility. Furthermore, REITs seemed to have had a threefold higher return than of the SCS returns. Some risk in REITs may be accounted for by REITs tendencies to require liquidity, generally provided by banks, which comes with sensitivity to interest rate risk from the loans (Nguyen and Steininger, 2019). Although there are benefits to the REIT structure, there are still opposing views that is concerned about competition incentives being altered with increasing number of REIT funds. Combined with high yields, derived from increased risk, show potential for extreme volatility (Geltner and Miller, 2007).
REITs are capital intensive endeavors and thus need a frequent access to capital markets to sustain the high dividend ratios. Together with high expectations on REITs to continuously pay out high levels of dividends and that a significant number of REITs have a capital structure incorporating quite substation amount of debt, a liquidity crisis with limited availability to capital markets could theoretically have significant consequences which could impact dividend policies. Even though many REITs apply high leverage to their financing strategy, REITs are usually considered a low-risk investment, like comparable real estate markets (Case et al., 2012).
Behavioral Differences Between REITs and Stocks
REITs have been examined in detail in a number of ways, from financial performance to organizational structure (Wang et al., 1995). The relationship between the general stock market and the REIT stock market has been researched. According to Wang et al. (1995) the previous observations in research suggests that there is difference in behavior between the REIT market and the general stock market. However, a question asked in earlier research is whether the markets effects each other and how. Howe and Shilling (1988) presents evidence that debt offering announcements from REITs have a positive response on the general stock market, whilst debt offerings from industrial firm necessarily doesn't. To understand how the overall interdependence is configured between the
12 REIT stocks and the general stock market, studies has been conducted to see which assets REITs performance correlate with the most. Studies suggests, according to Vakalopoulos (1993), that the performance of REITs align to a higher degree with capital market vehicles such as bonds and stock, rather than with real estate assets.
Furthermore, Vakalopoulos (1993) goes on saying that REITs strongest correlation is with small stock indices. It should be noted that REITs went under a structural change during the 1990s and that there are differences in conclusion about how long-run pricings likeness to other assets varies (Clayton and MacKinnon, 2002). On one hand, and in line with the common industry angle, some say that REITs converge towards private real estate rather than with stock (Lee and Chiang, 2010; Murphy et al., 2003). On the other hand, there are those who have found contradicting evidence with different analysis methods (Glascock et al., 2000). The most common conclusion seems to be that REIT performance follow private real estate and small stock indices in the same industry (Lee and Chiang, 2010; Vakalopoulos, 1993). Studies made by (Clayton and MacKinnon, 2002; Ghosh et al., 1996;
Khoo et al., 1993) reported that REIT-market derived from the same economic factors as large capitalization stocks. This linkage declined during the 1990 showing both economic factors of small capitalization stocks and real estate factors. This was supported by results from an international perspective (Hoesli and Serrano, 2006).
Portfolio Diversification
The start of what is considered modern portfolio theory, MPT, is of course the work made by Markowitz in 1952 (Mangram, 2013; Statman, 2019). His theorems explained the maximizing the expected return whilst holding the portfolio variance constant, and also holding the expected return constant while minimizing the portfolio variance.
Subsequently these two theorems together form the basis for the efficient frontier method (Elton and Gruber, 1997). Furthermore, it is established that together with Miller and Modiglianiยดs arbitrage principles, CAPM theory made by Sharpe, Black, and Lintner, and the option-pricing theory formulated by Merton, Black, and Scholes, Markowitz constitutes the foundation in what is known as standard finance(Statman, 2019). It is the body of knowledge and springboard of which many modern economists have developed their methods and theorems on.
Their theories, according to Statman, are compelling because of their ability to, with simple tools and methods, comprise unified theories on most questions formulated in finance. But as with any theory, all empirical data and anomalies does not conform these theories (Statman, 2019).
Tobin (1958) was early with responding to Markowitz theories with conditions necessary to prove them more accurate. More precisely the conditions were regarding the utility function (Bernoulli, 1954) for investors and the asset return distribution. Other portfolio theories were also proposed during the following years, trying to account for other variables and moments. Prominent theories incorporated skewness or in other ways building upon Markowitz theories to more accurately portray reality (Elton and Gruber, 1997; Fama, 1965; Kraus and Litzenberger, 1976; Lee, 1977). When considering the Markowitz framework, 1960s contemporary literature e.g.
Lintner (1965a, 1965b), Sharpe (1964), and Tobin (1958) deals with the affiliated โSeparation Theoremโ
proposed by Tobin (1958), and it seems that some describes the discussions of quadratic preference function in portfolio optimization at a given return as under the โTobin-Markowitz frameworkโ (Breen, 1968). The Separation theoremsays that the investorโs preferences in return and variance is independent from an optimal portfolio of risky assets, if the investor has riskless assets available (Tobin, 1958; Elton and Gruber, 1997), Furthermore, the theory implicates that when the ratio between the standard deviation of the asset and the expected return subtracted by return of the riskless asset is maximized, the portfolio is a so called tangency portfolio. A tangency portfolio subsequently entails the portfolio problem calculations can be conducted through the ray within the standard deviation area in tangent with the non-risk assets expected return. The separation theorem also paves way for the mutual fundsโ theorem, of which investors and financial institutions at times use as guidance on what portfolio structures is favorable or ought to be attractive to others. The mutual funds theorem combines funds, one riskless and one comprised of a tangency portfolio, to generate the investors desired portfolio characteristics. Assumption such as the one stating constant relaxed riskless borrowing and lending rate make way for more and different types of funds to include in the portfolio make up decisions (Elton and Gruber, 1997).What should be payed attention to is that the different variants of the mean variance portfolio theory are that it is designed to calculate a maximizing return portfolio, or minimized variance portfolio, or many other derivatives of MPT, for a single time period.
13 When looking at multiple period portfolio analyses several researchers has proposed different modification to single-period methods to better suit investment strategies over multiple periods on different assumptions (Fama, 1970; Gressis et al., 1976; Hakansson, 1974, 1970; Merton, 1990; Mossin, 1968). Furthermore,when dealing with multi-period problems one must assess whether the portfolio returns over periods are independent from each other or not. In many studies the independence between periods is assumed (Elton and Gruber, 1997), even though research suggests that metrics such as variance and mean returns can quite easily be proven to be related over time (Campbell and Shiller, 1988; Fama and French, 1989).
There are also a body of literature not looking at a discrete-time perspective but on continuous-time problem formulation where portfolio problems are solved simultaneously with consumption investment problems (Merton, 1990). It seems like the main purpose of exploring a continuous-time perspective of portfolio management is to verify or confirm discrete-time results. Although this area of research has saw some significant attentions following the 1990s, practical implementation and impact has been scarce. Another areawhich have seen little practical applications is how and if transactional and holding costs play a part or should play a part in portfolio rebalancing according to MPT. Transactional costs in the sense of either a proportion to a change in a portfolioโs risky assets or as a fixed fraction of the total portfolio value, seems to affect portfolio choice according to some, but not to others (Constantinides, 1986; Davis & Norman, 1990; Morton & Pliska, 1993; Schroder, 1995, Balduzzi
& Lynch, 1999).
Financial Crises
Financial crises are without a doubt a phenomenon that has affected most of the world's economies and sectors.
Subsequently a plethora of academic research has been conducted to assess the reasons behind the crises and their respective aftermaths, especially when looking at the period from the beginning of the twentieth century until now. There seems to be two opposing side in early views of the causes of financial crisis, as outlined by Mishkin (1992). The first view being from a monetarist perspective based on the work of Friedman and Schwartz (1963).
In summary a financial crisis has a strong linkage with banking panics. More precisely, banking panics pave way for contractions in money supply and thus subsequent contractions in aggregate economic activity, at least in the United States. Furthermore, the monetarists argue that increase in business failure and bankruptcy applications together with sharp declines in assets prices does not constitute since there is not necessarily a risk for banking panic and therefore not necessarily a risk for a contracted money supply. These non-banking panic events merely qualifies as โpseudo financial crisesโ by this view and does not require interventions from governmental agencies.
An intervention could even be harmful since it could hinder economic efficiency since firms that were going out of business would get bailed out. The opposing view, as presented and based on Kindleberger, CP (1978) and Minsky (1972), defines financial crises in broader terms, in comparison to Monetarists. What constitutes a financial crisis in their view could be some or a combination of the following factors; A sharp decline in asset prices. disruption in foreign exchange markets, or business failures either institutional or non-institutional companies. In contrast to the Monetarists, a belief that governmental interventions are key to mitigate negative consequences for the aggregate economy following earlier stated factors. Furthermore, there is a wide array of literature describing individual financial crises and the respective effects on businesses in different sectors. It is well known and widely established within the science of investment and portfolio management that risk can to an extent be mitigated by diversification. But in the light of a financial crisis the same assets in various countries, or in some cases even different assets, has a tendency to decrease in value in tandem which can lead to a narrower window of opportunity for diversification. This is what is called contagion and it can be defined by the as increased cross-country correlations during โcrisis timesโ relative to โtranquil timesโ (Dornbusch et al., 2000).
14
Theoretical Framework
Chapter 4 - Theoretical Framework. The research is based on the theoretical framework of Modern Portfolio theory, behavioral finance and liquidity. These are commonly recurring concepts within the investment industry and are very useful in understanding investment strategy combined with investors behavior.
Portfolio Theory - Mean-Variance Optimization
Portfolio theory is often referred to as one of the cornerstones of financial economics and corporate finance. Harry M. Markowitzโs that is thought of as the founder of the theory got the Nobel Prize in economics 1990 for his findings 1952 (Mangram, 2013). The model primarily impact was its statements about portfolio diversification regarding number of included securities within a portfolio and the relationship between their covariance (Megginson, 1996). This model is the foundation to Modern Portfolio Theory (MPT) that has evolved with contribution from CAPM-model that was developed by William Sharpe, John Lintner and Jan Mossion (Mangram, 2013). The MPT framework is used in order to minimize investment risk and simultaneous maximize expected returns. This concept of diversification aims towards allocations of asset in specifics weights in order to exhibit lower risk than investing in the assets separately, more commonly known as not to put all of your eggs in one basket (Fabozzi et al., 2002).
REITs is compounded of various real estate segments such as residential, commercial, offices, to mention a few, but also diversified according to geographical aspects. The construction of REITs is therefore arguably designed according to MPT, which is why there is a central component of this report to test their diversification benefits during extreme conditions on the financial market and comparing them to the bond- and equity market during the same time period. By combining all of the assets in a portfolio generate conclusions about their diversification benefits and weights to achieve the efficient frontier. The efficient frontier is described in the book Corporate Finance (Berk and Demarzo, 2017) as the specific weighted portfolio that offers the highest possible expected return at a given level of risk.
Behavioral Finance
Modern portfolio theory by Markowitz (1952) within the financial field is used in various extent to achieve optimal asset and portfolio allocation. Although, often used with unrealistic behavioral conditions/assumptions.
Perception, belief and experience are included in investment decision making, but excluded from the model, which generates an unfair impression of reality (Antonides and Van Der Sar, 1990). Overconfident and experienced investors do not always hold a diversified portfolio due to the reliance in their own abilities, even if it is a fundamental approach to diversify in portfolio theory (Bondt and Thaler, 1985). A better understanding of behavioral factors that influence decision making, both emotional and psychological, is critical in finding the optimal asset allocation (Chira et al., 2008; Riaz et al., 2012).
The behavior of investors is often considered rational, derived from historical data and calculated assumptions about the future. Although, the impact of psychological factors can affect the behavior and be observed as irrational. Aspects of interpreting data and acting on specific information are individual characteristics that modifies investors behavior which may directly affect the market. Identifying these patterns is the key elements used in the behavioral portfolio theory (BPT) in order to reduce the occurrence of stock market anomalies (Antony, 2019). BPT is a complement to standard finance theories, however, it shows that investors have various combined objectives, for example, retirement savings, emergency savings and future requirement of family. Highlighting the presence of crowd psychology and behavioral decisions in asset allocation (Shefrin and Statman, 2000).
15 Irrational behavior has been categorized into two subgroups, theory of cognitive bias and prospect theory (Antony, 2019). Festinger (1957) studied cognitive bias, a theory of how behavior and emotions is derived from an individual's own mind i.e. reflection and attitude towards information. The other subgroup, prospect theory, is stated by Kahneman & Tversky (1979), which illustrate behavior modifications of individuals based on how they perceive profit and loss. The theory does not assume that investors are risk averse, which is the case in for example the utility theory, whereas additional wealth is valued less than previous wealth units added. According to Tversky and Kahneman (1973), this is not always true. Their theory incorporates human behavior and show results that contradict such actions. Concluding that individuals tend to gamble for a higher loss than obtaining a certain smaller one. The integration between theoretical finance and meaningful psychological elements was supported by Thaler (1999) as an exemplary behavioral economic theory. Additional studies on different type of investors show scientific evidence that confirm the linkages between behavioral characteristics and investment decisions (Pasewark and Riley, 2010; Ricciardi and Simon, 2001). Shanmugham & Ramya (2012) even found evidence showing that economic factors are inferior to sociological and psychological factors regarding investment decisions. There is also a distinction between speculating or โhobbyโ investing and building a financial buffer. As the first category is more connected to positive emotions and thereby more frequently traded (Barberis and Thaler, 2003).
Liquidity
Real estate markets if often seen to be fairly illiquid markets with high transaction costs (Krainer and LeRoy, 2002). One crucial variable within financial markets is liquidity because of the basic idea of easily being able to trade assets. Although, not easily measured since there are numerous ways to test for liquidity and no definite approach (Ramos and Righi, 2020). The idea of liquidity was already actualized in 1936 by the British economist John Maynard Keynes. Where cash were reviewed as the most liquid asset and measured in how fast other assets can be transferred into cash. Proposing that low liquidity is linked with risk, which means that investors would require higher premiums for illiquid assets (Keynes, 2018). The multidimensional phenomenon of liquidity is widely researched from different aspects, such as asymmetric information that occur in transactions (Brogaard et al., 2014), liquidity premiums (Amihud, 2002) and volatility (Chordia et al., 2005) etc. Although they are mainly defined in three areas; transaction cost, depth and resiliency (Foucault et al., 2013; Hasbrouck, 2007). Challenges have emerged among both researchers and the industry to determine liquidity for each of the dimensions (Ramos and Righi, 2020).
Sarr and Lybek (2002) breaks down the dimension in to characteristics that often to some extent overlap.
Complicating measurements since available data of qualitative factors does not always correspond to all dimensions. Their stated characteristics are tightness, immediacy, depth, breadth and resiliency, where no method unambiguous can combine all five measures.
The real estate market is slow in regard to transaction occurrence and due diligence, making it a quite illiquid market (Lin and Fuerst, 2014). Some of the reasons for the relatively low liquidity is the characteristics of assets transaction and holding periods. It is not uncommon that real estate investors tend purchase and then manage a property over a long-term basis. Transactions for high value assets brings with high transaction costs. From due- diligence reports and valuations to technical surveys, combined, it all takes time and money. The high threshold resulted from the previously mentioned reasons, is making it more difficult for smaller investors to take part which even further reduced liquidity and subsequently raises the risk premium (Vinell, 1996).
Ekbom and Sandberg (2005) argues that there are two areas that should be taken into consideration in the discussion on REIT liquidity. The first one being the possible effect REIT investment capital flows have on the property market. The direct property market in particular. The second area is whether the REITs in themselves are a liquid form of investment asset.
16
Methodology
Chapter 5 - Methodology. This chapter present methods used in order to answer the research questions. The focus on historical return and standard deviation of the study requires an objective approach when analyzed, hence the structure of a quantitative research. Central contribution is derived from Modern Portfolio Theory but also individual performance and liquidity attributes.
Each component and procedure are further described.
Quantitative Method
Quantitative research aims to use or generate numerical data is often referred to as techniques of data collection or data analysis procedure. It is a standard measure to ensure validity of collected data (Saunders et al., 2016). The objective of the analysis is focused on conclusion about the population rather than explaining specific outcomes (Goertz Gary and Mahoney James, 2012). Research that require more of an interpretive philosophy incorporate a qualitative method (Denzin and Lincoln, 2011). Studying the historical return and their standard deviation, diversification and liquidity attributes in REITs in comparison with other major asset classes, such as equities, bonds, commodities and property companies during the same time-period involves numerical data sets. Deriving conclusion about the data set does not concern a subjective approach, hence the importance of a quantitative method approach in this study.
Data Sources
The primary aim is to examine previously stated research questions during extreme financial conditions. More specific, during the global financial crisis in 2008 and the present volatility experienced on financial markets all over the world as a result of the pandemic virus Covid-19. The pandemic and its repercussions are currently developing, which means that collected data is only covering a limited time period before its final fallout. Hence, the full extent of the crisis is not within the scope of this study. Time periods adjacent to the recessions are also included as well as a period that is the opposite of a recession i.e. โUpswingโ for comparison. Historical data was collected from Thomson Reuters Datastream database containing daily prices from the time periods 2007-2008 (GFC), 2015-2016 (โUpswingโ) and 2019 until end of March 2020 (novel Coronavirus). The various assets that have been incorporated into the study is stated in Table 2. All assets except the REITs are indices, which affect the results since they to some extent already have lowered volatility and thereby less risk according to the used framework. However, the commodities and government bonds are each combined with the same product which lowers the impact of using their indices. More significant deviation is plausible for the AEX index that represent equities. The AEX index include various sectors of the market and could be interpreted as its own portfolio with lowered standard deviation in comparison to the individual assets within the index would have been used. Meaning that the results of diversification benefits within a portfolio are to be interpreted as the minimum benefits possible, since using each individual asset within the indices would in comparison improve the REITs performance.
17
Table 2. Selected assets.
Class Name/asset ID Actuality Main Assets Types
REIT Eurocommercial Properties NV D:N4S(RI) From 2007 Retail, Office, Warehouse
REIT NSI NV H:NSI(RI) Whole period Office
REIT Wereldhave NV H:WH(RI) Whole period Office, Retail, Services
REIT WFD Unibail-Rodamco H:UBL(RI) Whole period Retail
REIT Vastned Retail H:VAST(RI) Whole period Retail
Bonds Daily 5-year government bonds BMNL05Y(RI) Whole period Bonds
Equities Amsterdam exchange index AMSTEOE(RI) Whole period Mixed
Commodities Crude oil GSCLSPT Whole period Oil
Commodities Gold GSGCSPT Whole period Gold
Commodities Base metals WCFIBMI Whole period Base Metals
Commodities Natural gas DJUBNGT Whole period Natural Gas
Mathematical Analysis
The research contribution is derived from historical data that have been analyzed with a variety of mathematical theories and formulas. In order to analyze the performance and attributes of the research questions, the following key measurements have been used.
Historical return
As a first measurement, this study examines all the assets individual performance using historical price data to calculate returns and standard deviation. Additional correlation analysis is used in order to provide information about the return movement pattern in relation to each other.
Return
๐ != #""!"#
! $ โ 1 (1)
๐ != ๐ท๐๐๐๐ฆ ๐๐๐ก๐ข๐๐ ๐#$%= ๐๐๐๐๐ ๐๐ก ๐๐๐ฆ ๐ก + 1 ๐#= ๐๐๐๐๐ ๐๐ก ๐๐๐ฆ ๐ก
Variance
๐&= โ,-((,)%$)(ฬ )& (2)
๐&= ๐๐๐๐๐๐๐๐ ๐ = ๐๐๐๐๐๐ ๐ ๐๐ง๐ ๐ฅ-= ๐ ๐๐ก๐ข๐๐ ๐๐ ๐๐ ๐ ๐๐ก ๐
๐ฅฬ = ๐ด๐๐๐กโ๐๐๐ก๐๐ ๐๐๐๐ ๐๐ ๐๐๐ก๐ข๐๐๐
18
Standard Deviation
ฯ = โ๐& (3)
ฯ = ๐๐ก๐๐๐๐๐๐ ๐๐๐ฃ๐๐๐ก๐๐๐
Covariance
๐ถ๐๐ฃ(๐ -, ๐ .) = ๐-.=/)%% โ (๐ ,# -,#โ ๐ L-)(๐ .,#โ ๐ L.) (4)
๐-.= ๐ถ๐๐ฃ๐๐๐๐๐๐๐ ๐๐๐ก๐ค๐๐๐ ๐๐๐ก๐ข๐๐๐ ๐๐ ๐ ๐๐๐ ๐ ๐ = ๐๐ข๐๐๐๐ ๐๐ ๐๐๐๐๐๐๐
๐ -= ๐ ๐๐ก๐ข๐๐ ๐๐ ๐ ๐ก = ๐๐๐๐๐๐ ๐ก
๐ L-= ๐๐๐๐ ๐๐ ๐๐๐ก๐ข๐๐ ๐
Correlation
๐ถ๐๐๐(๐ -, ๐ .) = ๐ผ =21$,&
$2& (5)
๐ผ = ๐ถ๐๐๐๐๐๐๐ก๐๐๐ ๐๐๐ก๐ค๐๐๐ ๐ ๐๐๐ ๐ ๐-= ๐๐ก๐๐๐๐๐๐ ๐๐๐ฃ๐๐๐ก๐๐๐ ๐
Mean-Variance Optimization
According to modern portfolio theory there are five measurements used to indicate an optimal portfolio; Expected return, Variance, Standard deviation, Covariance and Correlation (Markowitz, 1952). Typically required method in MPT is forecasting future values of return and volatility (Byrne and Lee, 1995). However, this study uses historical data, making it an ex-post analysis. According to Black & Litterman (1992) this form of study have a probability of creating so called โcorner solutionsโ in multi-asset portfolios, which means that extreme results could arise. Resulting in weights of zero and one being applied, which, in theory would generate the most optimal portfolio. Although, highly unlikely to be acceptable by portfolio managers (Jorion, 1985). Another implication with the method is misinterpretation of risk and return. Whereas risk is often underestimated and returns often overestimated. By inserting lower and upper boundaries for each asset class in the portfolio simulates closer to
โreal worldโ application (Frost and Savarino, 1988). In order to avoid the extreme results, the following limitations have been set; the weights for each asset are assigned minimum and maximum values (see Table 4), the sum of all weights are restricted to one, and no negative weights are allowed.
Expected Return of Portfolio
๐ธ[๐ "] = ๐ธ[โ ๐ค,- -๐ -] = โ ๐ธ[๐ค,- -๐ -]= โ ๐ค,- -๐ธ[๐ -] (6) ๐ธ[๐ "] = Expected return of portfolio
๐ธ[๐ -] = ๐ธ๐ฅ๐๐๐๐ก๐๐ ๐๐๐ก๐ข๐๐ ๐๐ ๐๐ ๐ ๐๐ก ๐ ๐ค-=Weight of asset i
๐ = ๐๐ข๐๐๐๐ ๐๐ ๐๐ ๐ ๐๐ก๐
