Dissertation template bcu_format_belinda -sample
Download as DOCX, PDF1 like288 views
This chapter presents a literature review focusing on financial theories related to risk and return, particularly in real estate management and S-REITs. It discusses the Capital Asset Pricing Model (CAPM) and the Arbitrage Pricing Theory (APT), exploring how these models aid investors in evaluating risk-return trade-offs and performance. Various studies and analyses contribute to the understanding of real estate returns, linking macroeconomic factors and investor behavior.
![4
i. The risk-free rate in the market rf
ii. The stock’s market risk is measured by beta (β),
iii. The expectedreturnof the market rM
Formula 1: CAPMModel
Where,
E(rit) = Expected return of security i at time t
rft = Risk-free rate of return at time t
βit = Beta of security at time t
rmt = Return of the market at time t
[rmt - rft] = Market risk premium
The Formula 1 above is the CAPM model.
Formula 2: Single-index model
Where,
Rit = Expected return of security i at time t
α = Risk-free rate of return at time t
βimt = Market Beta of security i at time t
Rmt = Return of the market at time t
eit = Non-systematic risk or idiosyncratic error term of security i at time t
Bordie et al. (2008) explain that the CAPM is a model about expected returns, “whereas I
practice all anyone can observe directly are ex post or realised return” (Bodie et al., 2008,
p.308). The author use the CAPM model from a single factor model point of view
because the purpose of this research is not to examine the expected returns but the
influences of factors on Singapore Real Estate Investment Fund Trust returns. The
Formula 2 above presents the index model which can be interpreted as a regression
equation through which estimates of the alpha and beta can be obtained by Ordinary
Least Squares (OLS). OLS is a statistical technique which attempts to find the function](https://image.slidesharecdn.com/dissertationtemplatebcuformatbelinda-sample-150905185841-lva1-app6891/85/Dissertation-template-bcu_format_belinda-sample-4-320.jpg)
Dissertation template bcu_format_belinda -sample
- 1. Chapter Three: Literature Review In today’seconomicenvironment,organizationsnormallyneedtooptimize all the resourcesand assetsat theirdisposal,tomanage themefficientlyinordertoperformbetter.(Varcoe,2001, p.117). Thisstatementexactlyinterpretthe real estate managementandespeciallyforthe S- REITs. Inthis dissertation,the author’sresearchcompriseof the propertyof return relatedto investorbehaviour,differentsectors’activityandmajoreconomicfactors.Iwill presentmy literature reviewinthischapterin termsof the academicresearch of my field. 3.1 Financial Theories In well-known finance literature, academics combine return and risk as two main interests. “There are only two theories that provide a rigorous foundation for computing the trade-off between risk and return: the Capital Asset Pricing Model (CAPM) and the Asset Pricing Theory (APT).”according to Burmeister, Roll and Ross (2003, p.2), Both the single factor model (CAPM) and multifactor model (APT) assist to investors to make financial decisions when investors evaluate the risk-return performance in terms of the systematic and unsystematic risks. A number of journals and articlesscrutinised the pricing of real estates in the risk- return performance and the macro-economic context relating to the literature on real estate return. Lizieri etal.(2003) examine the REITs’underlyingreturn-generatingfactorsapplyingaprinciple componentsanalysisapproach. Lingand Naranjo(1997) focustheiranalysisonthe economic riskfactors and commercial real estate returnsutilizingamultifactorassetpricingmodel accordingto the CAPMmodel andthe APT model.Additionally,Chaudhryetal.(2004) started withthe CAPMtheoryand decomposingthe CAPMtheoryintosystematicandunsystematicrisk to investigatethe idiosyncraticriskof REITs. Liowet al.(2006) analyse the influence andrelationsbetweensome majormacroeconomic factors andthe expectedriskpremiaonpropertystocksbythree stepmethods:principal componentanalysis,GARCHandGMM. Kimetal. (2007) studythe REITs’ dynamicsbetween microeconomicsvariable andfinancialmarkets. However,onthe contrary,inorderto determine the explanatorypoweronmovementsin real estate return,theyprefertouse the vectorauto regression(VAR)model.Inthisdissertation,the authorpresentsanoverview of the REITs’ literature indicatingthatthe disputingaboutrisk-returnassessmentmethodare normallymixed and isnot close as the results.
- 2. 2 3.1.1 Modern Portfolio Theory and Single Factor Model The basic Markowitz portfolio theory derives the expected return rate of return for a portfolio of assets and measure of expected risk, which is the standard deviation of the expected rate of return. Markowitz showed that the expected rate of return of a portfolio is the weighted average of the expected return for the individual return investment in the portfolio. The standard deviation of portfolio is a function not only of the standard deviations for the individual investment but also of the covariance between the rates of return for all the pair of assets in the portfolio. The Modern Portfolio Theory (MPT) was developed by Harry Markowitz. He assumed that most investors want to be cautious when investing and that they want to take the smallest possible risk in order to obtain the highest possible return, optimizing return to the risk ratio. MPT states that it is not enough just to look at the expected risk and return of one particular stock. By investing in more than one stock, an investor can obtain the benefits of diversification, a reduction in the volatility of the whole portfolio (Markowitz,1959). The CAPMis builtona setof assumptions: Individual investors o Investors evaluateportfoliosbythe meanandvariance of returnsovera one periodhorizon o Preferencessatisfynon-satiation o Investorsare riskaverse Tradingconditions o Assetsare infinitelydivisible o Borrowingandlendingcanbe undertakenatthe risk-free rate of return o There are notaxesor transactioncosts o The risk – free rate isthe same for all o Informationflowsperfectly The set of investors o All investorshave the same horizon o Investorshave identical expectations The CAPMmodel assumesthatriskisa functionof onlyone factor,whichisthe relationship betweenasecurity’sreturnandthe marketreturn.Thisrelationshipisdefinedbythe securities beta.It alsoassumesinvestorsfullydiversifiedtherefore onlysystematicneedsconsidering. The CAPMmodel alsorefertothe efficientmarkethypothesisthatassumesthe investorsare
- 3. 3 rationallyandact ina predictable way.The CAPMarguesthatthese assumptionsimplythatthe tangencyportfoliowill be avalue-weightedmix of all the assetsinthe world. The proof is actuallyan equilibriumargument.Itbeginswiththe assertionthatall riskyassetsinthe world may be regardedas “slices”of a global wealthportfolio. The major factor that allowedportfoliotheorytodevelopintocapital markettheoryis the conceptof a risk-free asset.Followingthe developmentof the Markowitzportfoliomode. Several authors considered the implications of assuming the existence of a risk-free asset, that is, an assetwithzerovariance.Aswe will show,suchan assetwouldhave zerocorrelationwithall otherriskyassetsandwouldprovide the risk-free rate of return(RFR).Itwouldlie onthe vertical axisof a portfoliograph. (TEXTBOOK,P232). The directimplicationsare: i. All investorsface the same efficientset of portfolios ii. All investorschoose alocationonthe efficientfrontier iii. The locationdependsonthe degree of riskaversion iv. The chosenportfoliomixesthe risk-free assetsandportfolioMof riskyassets Thisassumptionof a risk-free assetallowsustoderive ageneralizedtheoryof capital asset pricingunderconditionsof uncertaintyfromthe Markowitzportfoliotheory.Thisachievement isgenerallyattributedbyWilliamSharpe (1964),butLinter(1965) and Mossin(1966) derived similartheoriesindependently.Consequently,we seereferencestothe Sharpe-Lintner-Mossin (SLM) capital assetpricingmodel. The CAPMmodel alsointroducedtwofundamental conceptsthatare the Capital MarketLine (CML) and SecurityMarketLine (SML).The Capital Market Line indicatesthatall optimal investmentportfoliosshouldbe splitbetweenapercentage investmentinthe risk-freeasset and percentage investmentinmarketportfolioM, thisline definedbyeverycombinationof the risk-free assetandthe marketportfolio,presentingthe superiorreturnyouearnfortakingeach extrarisk.(http://www.nasdaq.com/investing/glossary/c/capital-market-line). An investor is only willing to accept higher risk if the return rises proportionally. The optimal portfolio for an investor is the point where the new CML in tangent to the old efficient frontier when only risky securities were graphed. (http://www.researchgate.net/publication/264547651_Capital_market_line_based_on_efficient_fr ontier_of_portfolio_with_borrowing_and_lending_rate). SML is a linear(straight) line showing the relationshipbetweensystematicriskandexpectedratesof returnforindividual assets (securities).Accordingtothe capital assetpricingmodel the returnabove the risk-free rate of returnor a riskyassetisequal tothe riskpremiumforthe marketportfoliomultipliedbythe betacoefficient. (http://www.lse.co.uk/financeglossary.asp?searchTerm=&iArticleID=969&definition=security_m arket_line).The CML only deals with composition of optimal investment portfolios. But Security Market Line (SML) says that the expected return of any stock or portfolio is related to three factor.
- 4. 4 i. The risk-free rate in the market rf ii. The stock’s market risk is measured by beta (β), iii. The expectedreturnof the market rM Formula 1: CAPMModel Where, E(rit) = Expected return of security i at time t rft = Risk-free rate of return at time t βit = Beta of security at time t rmt = Return of the market at time t [rmt - rft] = Market risk premium The Formula 1 above is the CAPM model. Formula 2: Single-index model Where, Rit = Expected return of security i at time t α = Risk-free rate of return at time t βimt = Market Beta of security i at time t Rmt = Return of the market at time t eit = Non-systematic risk or idiosyncratic error term of security i at time t Bordie et al. (2008) explain that the CAPM is a model about expected returns, “whereas I practice all anyone can observe directly are ex post or realised return” (Bodie et al., 2008, p.308). The author use the CAPM model from a single factor model point of view because the purpose of this research is not to examine the expected returns but the influences of factors on Singapore Real Estate Investment Fund Trust returns. The Formula 2 above presents the index model which can be interpreted as a regression equation through which estimates of the alpha and beta can be obtained by Ordinary Least Squares (OLS). OLS is a statistical technique which attempts to find the function
- 5. 5 which most closely approximates the data (a “best fit”). In general terms, it is a method to fitting a model to the observed beta. This model is specified by an equation with “free” parameters. In technical terms, the Least Squares method is used to fit a straight line through a set of data0points, so that the sum of the squared vertical distances (called residuals) from the actual data-points is minimised. (http://www.strath.ac.uk/aer/materials/4dataanalysisineducationalresearch/unit4/ordinaryl eastsquaresmethod/) Formula 3: Calculation of the Beta (β) with CAPM model Where, βimt = Market Beta of security i Ri = Expected return of security i Rm = Return of the market Beta can be viewed as a standardized measure of systematic risk because it relates this covariance to the variance of the market portfolio. (text book p240). Beta measuresthe sensitivityof the stock’sreturntothe market’sreturn.If a stock has a highbeta,thenwhenthe marketgoesup,the stockgoesup evenmore (andvice versa).The price movementsof alow betastock are lesssensitivetovariationsinthe market.Asconvention,betaonthe marketis one and stocksare thoughtof as beingmore or lessriskythanthe market,accordingto whether theirbetaislarger or smallerthanone (Eltonetal.,2007, p137). Therefore,the betafluctuates negativelyorpositively,abetacoefficientof 1 presentsthatthe stockhas the same risk as the overall market,andwill notearnmore extrareturnthan market.A coefficientbelow1suggests the riskand returnof the stock will be lessthanthe average (where the average meansthe overall market). Onthe otherhand,the coefficienthigherthan1 suggests the riskof the stock will be more riskyandprofitable thanthe overall marketrisksandreturn. (http://accountingexplained.com/misc/corporate-finance/beta-coefficient) Throughoutour presentationof the CAPM,we notedthatthe marketportfolioincludedall the riskyassetsinthe economy.Further,inequilibrium, the variousassetswouldbe includedinthe portfolioinproportiontotheirmarketvalue.Therefore,thismarketportfolioshouldcontainnot onlyU.S. stocksand bondsbutalso real estate,options,art,stamps,coins,foreignstocksandso on,withweightsequal totheirrelativemarketvalue.(textbook,p257) Althoughthisconceptof a marketportfolioof all riskyassetisreasonable intheory,it’sdifficult to implementwhentestingorusingCAPM.Most studieshave beenlimitedtousingastock or bondseriesalone due toitis difficulttoderive the monthlyfinancial dataina timelyfashionfor
- 6. 6 numerousotherassets. Mostacademiciansrecognizethispotential problembutassume that the deficiencyisnotserious.SeveralarticlesbyRoll (1977a,1978, 1980, 1981), however, concludedthat,onthe contrary,the use of these indexesasaproxyfor the marketportfoliohad veryseriousimplicationsfortestsof the modelsandespeciallyforusingthe model when evaluatingportfolioperformance.Roll referredtothisproblemasabenchmarkerrorbecause the practice is to compare the performance of a portfoliomanagertothe returnof an unmanagedportfolioof equal risk –thatis,the marketportfolioadjustedforriskwouldbe the benchmark.Roll’spointitthat,if the benchmarkismistakenlyspecified,youcannotmeasure the performance of a portfoliomanagerproperly.(testbook,p257) The CAPMhas beenone of the most useful andmostfrequentlyusedfinancialeconomic theorieseverdeveloped.However,manyempirical studiescitedalsopointoutsome of the deficienciesinthe model asanexplanationof the linkbetweenriskandreturn.Forexample, assumingthe sample periodsare longenoughandthe tradingvolume isadequate,testsof the CAPM presented thatthe betacoefficientsforportfoliogenerallywere stablewhile the beta coefficientforindividualsecuritieswere notstable.Anotherchallengetothe CAPMwas that itis possible touse knowledge of certainfirmorsecuritycharacteristicstodevelopprofitable tradingstrategies,evenafteradjustingforinvestmentriskasmeasuredbybeta. Banz(1981) showedthatportfolioof stockswithlow marketcapitalizations(i.e.,“small”stocks) outperformed“large”stockportfoliosonarisk-adjustedbasis,andBasu(1977), who documentedthatstockswithlow price-earnings(P-E) ratiossimilarlyoutperformedhighP-E stocks. Fama and French(1992) demonstratesthat“value”(i.e.,those withhighbookvalue-to- marketprice ratios) tendto produce largerrisk-adjustedreturnsthan“growth“stocks(i.e., those withlowbook-to-marketratios).(textbookp270) 3.1.2 Modern Portfolio Theory and Multifactor Model In the early1970s, the academiccommunitysearchedforan alternative assetpricingtheoryto the CAPMthat was reasonablyintuitive,requiredonlylimitedassumptions,andallowedfor multiple dimensionsof investmentrisk.Theresultwasthe arbitrage pricingtheory(APT),which was developedbyRoss(1976, 1977) inthe mid -1970s. Unlike the CAPM,it doesnot depend criticallyonthe notionof an underlyingmarket,whichRoll (1977) critique of the CAPM. In Formula2 single-factormodelintroducedamannerof breakingupthe marketor systematicrisk due to macroeconomicsfactors,againstthe firm-specificrisk of idiosyncraticeffects(Chaudhry et al.,2004). The single-factorgeneratesthe multifactorbythe integrationof severalsourcesof systematicrisk.Thismodel dividesthe risksintosystematicandunsystematicriskwhere systematicriskisnon-diversifiable andunsystematicriskisdiversifiable,inthe same wayasthe single model does(Burmeisteretal.,2003, p.2).GroeneworldandFraser(1997) empirically examinedthe CAPMandAPTmodelsonthe Australianmarketandas resultindicatedAPT model outperformsCAPMaswritteninBurmeisteretal.(2003, p. 16) the multifactormodel “has far greaterexplanatorypowerthanthe CAPM”.
- 7. 7 Chen, Roll and Ross (1986) was the first study to select macroeconomic variables to estimate U.S. stock returnsandapplythe APT models.Theyemployedsevenmacroeconomicvariables, namely:termstructure,industrialproduction,riskpremium, inflation,marketreturn, consumptionandoil pricesinthe periodof Jan1953-Nov 1984. Duringthe testedperiodintheir research,theyfoundapositive relationshipbetweenthe macroeconomicvariablesandthe expectedstockreturns.Theynote thatindustrial production,changesinriskpremium, twistsin the yieldcurve,measure of unanticipatedinflationof changesinexpectedinflationduring periodswhenthese variablesare highlyvolatile,are significantexplainingexpectedreturns. Consumption,oil pricesandmarketindex are notpricedbythe financial markethasbeen discovered.Theyconcludeassetpricesreactsensitivelytoeconomicnews,especiallyto unanticipatednews. Readmore:http://www.ukessays.com/dissertation/literature-review/literature-review-of- arbitrage-pricing-theory.php#ixzz3iu2skI91 The relationships between the Singapore stock index and chosen macroeconomic variables over a seven-year period from 1988 to 1995 were experimented by Maysami and Koh (2000). It resulted in existence of a positive relationship between stock returns and changes in money supply but negative relationships between stock returns with changes in price levels, short- and long-term interest rates and exchange rates. Read more: http://www.ukessays.com/dissertation/literature-review/literature-review-of- arbitrage-pricing-theory.php#ixzz3itsOz81A To examine the interdependence betweenstockmarketsandfundamentalmacroeconomic factors inthe five SouthEast Asiancountries(Indonesia,Malaysia,Philippines,Singapore,and Thailand) wasthe mainpurpose of WongbangpoandSharma (2002). Monthlydata from 1985 to 1996 isusedin thisstudyto representGNP,the consumerprice index,the moneysupply,the interestrate,andthe exchange rate forthe five countries.Theirresultsshowedthathigh inflationinIndonesiaandPhilippinesinfluencesthe long-runnegative relationbetweenstock pricesand the moneysupply,asthe moneygrowthinMalaysia,Singapore,andThailandinduces the positive effectfortheirstockmarkets.The exchange rate variable ispositivelyrelatedto stock pricesinIndonesia,Malaysia,andPhilippines,yetnegativelyrelatedinSingapore and Thailand. Readmore:http://www.ukessays.com/dissertation/literature-review/literature-review-of- arbitrage-pricing-theory.php#ixzz3iu2l8HSi
- 8. 8 In contrastto the CAPMmodel,the APTadvocatesthatthe risksnotonlyfromthe suggested market-systematicriskof the CAPMbut frommanyothersystematicrisks.APTassertsthatan asset’sexpectedreturndependsonalinearcombinationof asetof factorswhose identifymust be determinedempirically.Examplesof suchfactorsmightinclude suchmacro-economic variables as real economic growth,exchange rate, inflation, interest rates, employment level etc, or such financial variableasdividendyield,capital structure etc. AccordingtoRoll andRoss (1980) the fewconditionsinthe use of APTsuch as randomassetreturnfollowsamultivariate normal distributionandinvestors behave rationallyinthe market(Roll andRoss,1980, p. 1074- 1075). As writtenbyBodie: “Theprice of this generality is thatAPT doesnotguaranteethis relationship forall securities at all times” (Bordie etal.,2008,p.350). Arbitrage PricingTheory has three majorassumption: Capital marketare perfectlycompetitive Investorsalwaysprefermore wealthtolesswealthwithcertainty The stochasticprocessgeneratingassetreturnscanbe expressedasalinearfunctionof a set of K riskfactors (orindexes). The operational formof the APT can be expressedinFormula4 as follows; Formula 4: Multifactor APT model Ri = Return on asset i, for i = 1 … n βij =Sensitivity parameters of asset i to risk factor j, for i = 1 … k and for j = 0 … n Fj = j-th risk factor for j = 1 … n ei = Non-systematic risk or idiosyncratic error term of asset i The advantage of this approach is that the investor knows precisely how many and what things need to be estimated to fit the regression equation. However, the major disadvantage of a multifactor model is that it is developed with little theoretical guidance as to the true nature of the risk-return relationship. (test book p280) Although the APT is considered newer than the CAPM, it has undergone numerous empirical studies. Roll and Ross produced one of the first large-scale empirical tests of the APT. Cho, Elton and Gruber (1984) tested the APT by examining the number of factors in the return-generating process that were priced. Because APT model contends that more factors affect stock returns than are implied by the CAPM, they examined different sets of data. Dhrymes, Friend, and Gultekin (1984) re-examined the methodology used in prior studies and contended that these techniques have several
- 9. 9 major limitations. Roll and Ross (1984) acknowledged that the number of risk factors differ with 30 stocks versus 240 but contended that the important consideration is whether the resulting estimates are consistent because it is not feasible to consider all of the stocks together. Dhrymes, Friend and Guitekin (1985) repeated the prior tests for larger groups of securites. They found that the unique or total standard deviation for a period was as good at predicting subsequent returns as the factor loadings. These results are not favourable to the empirical relevance of APT because the model depends on group size and the number of observations. Finally Cornnor and Korajczyk (1993) argued that most tests for the number of priced risk factors are valid only for strict factor models in which diversifiable returns are uncorrelated across the set of stocks in the sample. Reinganum (1981) addressed the APT’s ability to account for the differences in average returns between small firms and large firms. The small-firm portfolio experienced a positive and statistically significant average excess return, whereas the large-firm portfolio had a statistically significant negative average excess return. The mean difference in excess returns between the small and large firms was about 25 percent a year. Also, the mean excess returns of smallest through largest portfolios were perfectly inversely ordered with firm size. Given the so-called January effect, where returns in January are significantly larger than in any other month. Gultekin and Gultekin (1987) tested the ability of the APT model to adjust for this anomaly. The APT model was estimated separately for each month, and risk premia were always significant in January but rarely priced in other months. It was concluded that the APT model can explain the risk-return relation only in January. Burmeister and McElroy (1988) estimated a linear factor model (LFM), the APT, and a CAPM. They found a significant January effect that was not captured by any of the models. They rejected the CAPM in favour of the APT. Kramer (1994) shows that an empirical form of the APT accounts for the January seasonal effect in average stock returns while the CAPM cannot. (Test book P276-278) 3.2 Economics Theories In respect to volatility, both capital market and real estate market approach are seems to be in the contradiction while both of them lead to a perfect equilibrium (Brown and Matysiak, 2000). According to Marty (2008), the long-term strategy are more secured investment which can avoided, however, the short-term strategy is dependent from the daily variation of market and cannot be avoided. Singapore REITs, underlying the properties on the stock-exchange market, is supposed as “long-term” oriented strategy of the investors. Investors have to take the economic conjunction along the period into consideration when they concern about the performance of these investments. The author believe that non- anticipated event and fundamental economic aggregates affect asset prices at different levels from financial theories perspective. (Roll and Ross, 1980; Chen et al., 1986; Burmeister et al., 2003; Lizieri et al. 2007). In this section, I will review
- 10. 10 the specific risks’ factors of holding a real estate assets’ portfolio and the main factors that influence the performance of Singapore real estate investment trust. 3.2.1 Differentiation between Macro and Microeconomics A wide variety of empirical factor specifications have been employed in practice. A hallmark of each alternative model that has been developed is that it attempts to identify a set of economic influences that is simultaneously broad enough to capture the major nuances of investment risk but small enough to provide a workable solution to the analyst or investor. (test book p280) Two general approaches have been employed in this factor identification process. First, risk factors can be macroeconomic in nature; that is, they can attempt to capture variations in the underlying reasons an asset’s cash flows and investment returns might change over time (e.g., changes in inflation or real GDP growth). Macroeconomics measures the natures and actions of the economy as a whole by businesses or government usually. On the other hand, risk factors can also be identified at a microeconomic level by focusing on relevant characteristics of the securities themselves, such as the size of the firm in question or some of its financial ratios. Microeconomics measure how individuals or institutions make their financial decisions. (test book p280). In order to assist investors, firms, financial institutions to make financial decisions, both macroeconomics and microeconomics aim to forecast aggregation. This table describes the difference between the two economic fields: real estate market and the sources of risks.The specific risk of the overall market is macro variable, while all risks specific to assets are micro variable. Heidra and Van Der Ploeg (2002) entails many concepts relating to the demand for money and
- 11. 11 aggregate labour market and on the opposite behaviour of individual. However,the differentiationof isnotsoeasyin termsof the case of Singapore REITs.As BrueggmanFisher(2008, p. 625) state,mostof the equityREITsspecialise bygeographic location,propertytype,andsometime bybothof them.Assuch,the risk of geographyand sectoris notclear and can be consideredasbothmicro – and macroeconomicfactors. Initially,economistsdividedthe economicenvironmentintotwodistinctacademicsfields: macro – and microeconomics.Butnowadaysthese twofieldsinfluence eachotherandthe distinctionbetweenmacroandmicrovariable can’tbe establishedisolated becausethey influencedeachotherasthe table 2 above.The assetrisksandthe overall marketriskscan interpretthe returnsof Singapore REITsinrespecttothe geographicrisksandthe sector. Singapore REITs’returnsare expectedtochange trendsas the changingof the economicand businessconditions.(Ducoulombier,2007) The purpose of reviewmacro – andmicroeconomicfactorsseparatelyistoanswerthe second hypothesis whetherthe performance of Singapore REITsare influencedbythe economicfactors. The author will use the mostcommoneconomicfactorsrelatedtoreal estate marketusedin the academicresearch. 3.2.2 Macroeconomic Factors and Real Estate Market Ball,Lizieri andMacGregor (2001), among others,foundthateconomicactivityisamajor driver of demandforreal estate.Inmystudy,controlsfor macroeconomicconditionsinclude GDP growth,inflationandinterestrate.Generally,whenthe Singaporedoingwell,GDPwill grow, and investorswillhave more confidencetoinvestinSingapore.ManyinvestorsinAsiaare keen to investinreal estate,andwouldlike tobuypropertiesinacountry where the economyis growing,asthiswouldensure ahealthystreamof demandforthe real estate inthe countryto boostthe investmentyield. Astheypurchase more real estate assets,thiswouldpushup Source:Ducoulombier(2007,p.41)
- 12. 12 propertypricesandconsequentlyresultinrisingrentals,akeyingredientof netproperty income.Furthermore,asthe economyflourish,more companieswouldinvestinthe country, pushingupdemandforcommercial real estate space,leadingtohigherrentalsaswell.The reasonfor choosingunemploymentiscloselyrelated,ashigherunemploymentreflectsabad state of economy,whichwouldlike meanlowerbusinessconfidence andcorrespondinglylower demandforcommercial real estate. Chen,Roll andRoss (1986) testthe influenceof aset of economic “state variables”on the US stock marketreturnsbyappyingthe Fama-MacBethtechnique,assumingthatpricesof assets respondsensitivelywhenthe economicnewsare released. Theyuse manyfactorstoperform theiranalysis,forexample,short-termtreasury-bill,long-termgovernmentbonds,inflation, value weightedequities,equallyweightedequities,industrial production,low-gradebonds, consumptionandoil prices,andsoon. Theyfoundthat the expectedstockreturncanbe explainedbymostof these variables. Basedonthe researchresultfromDeutsche Bank Research (2008, p. 23), Gross DomesticProduct(GDP) growthtrendisthe major indicatorsfor the real estate marketas well asGPD per capita,population,medianage,populationgrowth, legal system,financial marketdevelopmentandaverage inflation.Furthermore,Ducoulombier (2007) discoveredEmployment,unexpectedinflationandinterestratesare the othersourcesof systematicrisk.Generallyspeaking, almostall of themagree onthe use or ona variantof GDP, interestrate,real wage,rate of employmentand tax rateswhenmacroeconomiststrytostudy whatvariablesinfluence the macroeconomics.Inaddition,Liow etal.(2006, p. 301) gave a seriesof analysisonthe influence of macroeconomics,he foundthatthe mostrelevant indicatorsare:inflation,GDPandinterestrate.Inthisdissertation,the authorwill chooseGDP, CPIinflaction,shorttermtrade bill interestrate andlongtermgovernmentbondinterestrate as macro-factorsto illustrate. GDP is the mostimportantmacroeconomicsmeasure whichmeasuresthe total value of economicactivitywithinanation.Tobe more specific,GDPisthe sum of the marketvaluesor prices,of all finical goodsandservice producedinaneconomyduringaperiodof time. (http://www.sparknotes.com/economics/macro/measuring1/section1.html).GDPisthe reflectionof the growthof the economy,ahighGDP indicatesthatthe economicconditionis healthycause todrive the Singapore REITspositively.Governmentspendmoneyin infrastructure,Investorsandinstitutionsinnew constructionwhile individualsinowningand rentinghouses. The difference betweenReal GDPandnominal GDPis intermsof the inflation whetherbeenfactorsin.Generally,Real GDPmeasure the value of the goodsandservices producedexpressedinthe pricesof some base yearwhichnominal GDPmeasuresthe value of the goodsand servicesproducedexpressedinthe currentprices. Forexample,Real GDPtake five yearstimeline intoconsideration. (http://economics.about.com/cs/macrohelp/a/nominal_vs_real.htm) As one of the main macroeconomic factor, inflation is commonly accepted by academics. Inflation is defined as the rate at which price
- 13. 13 rise for goods and services. However, when economic calculate the risk relating to the inflation, they preferred to use Consumer Price Index (CPI) as a proxy (Chen et al., 1986; Ling and Naranjo, 1997). The CPI allows appreciating the movements in prices of products on a constant basis as the official instrument for measuring the inflation. Brueggma and Fisher (2008) compared the CPI and the performance of real estate and found that real estate exceed the growth rate of inflation from each category. In addition, from their research, they found that the values of inflation and real estate return are the opposite resulting in the irrelevant correlation. Nevertheless, the highlight “that a positive correlation with inflation is desirable because it indicates that the asset is an inflation hedge” (2008, p.666). Interest rate is another important factor of macroeconomics accepting by academics. Investors and financial institutions notably through the interest rate and use the relatively long period to finance to the cost of purchase the real estate in order to make the financial investment decision. The interest rate is the relevant factor in the real estate market. Like the real GDP, the real GDP has a better prospective of the real cost of fund for the borrower because it removes the effects of inflation which is preferred by the investors and financial institutions. The real interest rate when a borrower pays a lender, the percentage which increasing in purchasing power. In reality, Researchers and Academics usually prefer to divide the interest rate factor into short- term and long-term rate when they study the influence of multiple factors on stocks. Three-months treasury bills and ten-years government bonds are commonly use in this purpose (Bodie et al., 2008; Chen et al., 1986). Chapter Four: Methodology My literature review and research methodologyare from secondary sources, the text books, study notes, statistical databases and scientific journal s are the main study source. The methodology chapter outlines the business research strategy, data selection, specifications on the regression model and lastly the dataset. 4.1 Scientific Point of Departure Many alternativesandorientationscanbe selectedwhenchoose aspecificbusiness strategyfor dissertation.The readerare informedthese assumptionsandviewpointsbythese specifications that the authorshave taken.
- 14. 14 4.1.1 Business ResearchStrategy The business research strategy entails all the methodological choices done by the author. In this section, the blueprint aims at presenting the three different steps followed in order to fulfil our business research problem: the major influential factors on S-REITs performance. In reference to traditional business research methods, two general methods of reasoning exist and are known as inductive and deductive approaches. The first one starts from specific observations to broader generalizations and theories, whereas the second one starts from hypotheses and theories to achieve the research purpose (Bell and Bryman, 2003, p. 9). Hence, I utilize a deductive approach which requires, as premise, to state the hypotheses related to our research problem. As a reminder from part one (Section 1.2.3.), the two hypotheses that need to be scrutinized are respectively linked to financial, economic and theories and they are synopsized below: � Hypothesis 1: Some categories of S-REITs generate superior performance than others. � Hypothesis 2: Some economic factors affect the S-REITs performance. 4.1.2 ResearchDesign The research design specifies the process that will be followed in the data collection. The case study approach and comparative design were the two most appropriate for our paper as they fulfil our objectives. As indicated by its name, the case study is an intensive analysis of a single variable whereas the comparative one comprehends at least two different cases with distinctive sets of observations and are compared. In reference to earlier sections I intend to conduct an intensive case-study analysis by examining one specific country, Singapore, in the real estate market and for one specific class of assets within seven years. I narrowed down my field of research questions to closely determine the circumstances in which our two hypotheses will and will not be validated (Bell and Bryman, 2003, p. 55). This choice of focus is firstly on the major influences that I defined as a classification of real estate, Singapore listed companies, then on economic factors and financial behaviour. 4.1.3 Choices of the Sources Our main sources of information are based on secondary sources. After a comparison on main specialized websites in finance such as Bloomberg, Morningstar, Reuters, Yahoo Finance, Straits Time, Singapore News Paper, SGX.com, DBS Vickers.
- 15. 15 4.1.4 Data To foster the quality of our research, the literature review and theoretical framework have been updated in a continuous flow depending on the empirical data and literature that I get. Furthermore, as the structure of S-REITs is new in Singapore (seven years), the research available on the property stock market and S-REITs may not be peer-reviewed or relevant enough. Thus, I prefer reliable sources and decided to mostly use the scientific articles for our literature review. This practical consideration has been initiated as I am fully aware of limitations due to the amount of data available, and also by using a case analysis method combined with a sampling process which delimited our field of research even more. The choice of the model and the explanatory factors may be seen as restricted. The author is aware of biases that may occur when it comes to interpreting the collected data, notably the performance, due to the short time-period and small sample of 14 S-REITs chosen due to the limited available data. I believe that additional factors should be examined on a wider scale. Furthermore the choice of the multifactor model can be criticized as it is used predominately in academic research rather than in practice. While it is acknowledged that other alternatives exist I feel more familiar and confident with the multifactor model derived from the APT and I employed it by preference. The sources of information were difficult to obtain and required costs that I could not afford. In order to overcome this problem, I used, Datastream database and national statistics as quasi-unique resources in the extractions of the stock prices, classification, index and market trend indicators. The overall credibility of the paper dealing with the reliability, replication and validity (Bell and Bryman 2003, p. 33) will be developed in our last part. 4.2 Data Selection Process As a result of the literature research, the data has been carefully selected. In this section, I explain all the considerations and decisions that have been operated. The 14 S-REITs were selected in Datastream in accordance with our time frame and geographical considerations. In order to get the most complete and representative set of observations, a weekly period of seven years is examined from 2008 - 2014. The S-REITs market is only scrutinized in Singapore currency. Based on the Datastream classification, these 14 firms are divided into five categories. The S-REITs dataset is composed of 5 areas: Office, Retail, Industrial, Hospitality and Healthcare. 4.3. Specifications of the Regression Model
- 16. 16 When researchers deal with financial time series, statistics usually appear with their models to help them on financial issues, for example assessing and predicting the performance of assets or portfolios. However before using statistics which are trying to match performance results to the real world researchers have to be aware of the properties of the models and their assumptions. To determine the stakes of S-REITs and to emphasize our practical considerations, I analyse their performance through their respective return and risk. Then, the data used is the adjusted price and the price index. These extracted results are integrated in an Excel spread sheet and Eview software to perform most of the computations and analysis. 4.3.1 Determination of Beta from the CAPM Model The first step in the calculation of the beta consists of computing the returns of S- REITs. The log-return presents better statistical properties than the simple return. For instance, Chen et al. (1986), Campbell and Shiller (1988) actually use the log-return in their research. I computed the log-return by employing the logarithm function on the adjusted prices due to the continuous compounding effect. Formula 5 shows how each log-return is calculated. Formula 5: Calculation of the Log-Return Where, Log-return t = Logarithmic return of the asset at time t Pt = Price of the asset at time t Even if the net return is commonly utilised in finance, researchers prefer the log- return, additive in time, due to its closer link to the reality and as it measures the continuous compound return. However this distinction is not so important as long as the returns are low (Ruppert, 2006, p. 76). As a second step in the performance evaluation of our S-REITs, I calculate precicley the risk sensitivity with the market through Formula 3 of the beta’s calculation derived from CAPM model. Formula 6: Calculation of the Beta (β) adapted to our case analysis
- 17. 17 The risk of each stock and index is calculated through Formula 7 with the variance (VAR) and standard deviation (SD) computed by Excel function; respectively VARA () and STDEVA (). I use the traditional formula named COV () and VARA () functions due to the similar results I get and to the convenience of Excel. The calculation of the beta is reiterated for each S-REITs and is presented on an annual basis in order to catch the variation of sensitivity between the firm and the market each year instead of having it for a seven year period. In addition, the use of the CAPM model implies the acceptation of the related assumptions (Section 2.1.1.). 4.3.2. The Multifactor Model The single factor model CAPM needs to be extended (Section 2.2.1.). Indeed, I replace the latter by a more comprehensive factor model by applying a multifactor model such as APT, based on the fact that economy-wide factors affect the return of S-REITs. I decompose the analysis in two models to capture better the sensitivity of all factors. The model one integrates our main macro- and micro-factors, and is presented in the Formula 7. Formula 7: Model 1 derived from APT Where, Rit = Expected return of the S-REITs i at time t α = Intercept bk = Sensitivity variable between our S-REITs return and the factor k ε = Non-systematic risk or idiosyncratic error term Formula 8: Model 2 derived from APT Where, Rit = Expected return of the S-REITs i at time t α = Intercept bk = Sensitivity variable between our S-REITs return and the factor k ε = Non-systematic risk or idiosyncratic error term Rit represents our dependent variable which is the actual S-REITs’ returns. The classification granted by Datastream enables us to follow the performance of S- REITs according to its activity in the real estate market. Besides to increase the significance of our results I generated the model 1 and 2 with different variables such as GDP (instead of Real GDP) or interest rates. Nevertheless the
- 18. 18 significance obtained was lower, thus I have chosen to scrutinize the influences of the variables. Additionally, to capture a maximum of information about the factors that can influence the S-REITs’ returns I use both systematic and unsystematic explanatory variables. All of these factors are used as possible explanatory factors. Some dummy variables are used in developing the regression model as they are not readily measurable with quantitative values. (Keating and Wilson, 1986, p. 150-151). 4.3.3. Presentation of the Dataset in Descriptive Statistics This last section presents the dataset used after our linear regression model one and two derived from APT that have been applied by the author. Chapter V – Empirical Findings This chapter presents an overview of FTSE ST Real Estate Investment Trusts Index performance through a benchmark as well as including the outcome of the multifactor model one and two. The descriptive statistical results are generated from Eview software and Excel. 5.1 Overview of the performance of FTSE ST Real Estate Investment Trusts Index FTSE ST REITs STI SSE Dow Jones Mean 718.7745526 3136.232711 2546.118827 14286.40187 Standard Error 2.040820033 5.591945731 17.54776413 69.33248141 Median 716.01 3163.409912 2345.1 13593.37 Mode 646.04 3124.379883 2655.66 11478.13 Standard Deviation 69.23765822 189.7145368 595.3323055 2352.200867 Sample Variance 4793.853316 35991.60548 354420.5539 5532848.918 Kurtosis - 0.641801584 -0.127204366 4.524506563 -1.327571124 Skewness - 0.007128069 -0.522782501 2.032649974 0.147693922 Range 321.05 925.5 3216.34 8326.58 Minimum 569.11 2614.449951 1950.01 9985.81 Maximum 890.16 3539.949951 5166.35 18312.39 Sum 827309.51 3609803.851 2930582.77 16443648.55
- 19. 19 Count 1151 1151 1151 1151 The three benchmarks STI (Straits Time Index), SSE compositeindex (Shanghai Securities Composite Index) and Dow Jones Index are shown the brief statistics given by the table above. Total No. of 1151 counts are used. The mean of FTSE ST REITs performance of 718.7745526 is smaller than STI (3136.232711), SSE(2546.118827) and Dow Jones (14286.40187). In the meantime, the standard deviation of the FTSE ST Real Estate Investment Trusts Index is lower than STI, SSE, Dow Jones. In terms of the risk and return, FTSE ST Real Estate Investment Trusts Index doesn’t show a better performance than STI, SSE and Dow Jones as shown in table above. The Figure below presents the three benchmarks in comparison with my sample. Figure: Comparison of STI, SSE, Dow Jones. Sample (as of 22 July, 2015) The performance of FTSE ST REITs remains stable over the past 5 years. The overall trend is positively correlated to the Straits Time Index (STI),and SSE. As we can see the from the gaph, Dow Jones is leading the overall trend and present the best performance over the other stock index in other countries. Table xx: Correlation matrix of FTSE benchmark FTSE STI SSE Dow Jones 0 2000 4000 6000 8000 10000 12000 14000 16000 18000 20000 7/23/2010 7/23/2011 7/23/2012 7/23/2013 7/23/2014 Chart Title FTSE STI SSE Dow Jones
- 20. 20 FTSE 1 0.854794 0.151994 0.699655844 STI 0.854794 1 0.381103 0.711915673 SSE 0.151994 0.381103 1 0.221535089 Dow Jones 0.699656 0.711916 0.221535 1 The correlation matrix table above indicates that FTSE returns are positively correlation to STI (0.854794) and SSE (0.151994) and DownJones(0.699656). As we can see fromthe table above,the correlationbetweenFTSEandSTI have the higherdegree of correlation,thisismainlydue totheyare inthe Singapore stockexchange market,andthey variesaccordingto the national economicandinvestmentenvironment,sothe correlationis verystrong.Additionally,IalsoobservedthatFTSEhas a quite strongercorrelationwithDow JonesIndex (0.699656). This isbecause the overall stockexchange marketfollowsthe trendof the USA markets. Table 11. Betaper SIICsector Healthcare Industrial Office Retail Hospitality STI Beta 1.195504021 -11.43419559 17.91294967 26.3441103 8.658898576 43.06921819 As we know,betaisto measure the stock’sriskrelatingtothe overall market.If betais1, meaningthe level of riskisthe same asthe overall market.Ina bullishmarket,the stock’sprice increases.VersaVice.If betais greaterthan1, thisstockhas more risk andmore volatile than the market.It will move the same trendasthe marketbut will move togreaterrate.In a bullish market,the stock’sprice will goup at a fasterspeedthanthe market.If beta iszero,meaningto say that the stockhas no relationshipwiththe marketatall.If betais negative,meansthe movingdirectorof betawill be the opposite tothe stockmarket. (http://efinancialresourcecenter.com/stocks-negative-beta/) Applyingthe Formula7fromthe firstmodel introducedinthisdissertation,the table11 presentsthe beta’svalue andeachcategoryof Singapore REITs.We noticedthat industrial class has the onlynegative beta(-11.43419559), whichmeansindustrial categorymovesthe opposite direction of the overall stock market in Singapore. While retail has the highest beta (26.3441103) flowedbythe betaof office (26.3441103), as the overall marketgoup,more investorshave the more powerto purchase the industrial andoffice propertywhichreturnismore thanthe industrial property. The betaof healthcare presents the smallestvalueof betawhichisthe similartothe overall marketrisk. Healthcare Industrial Office Retail Hospitality FTSE Beta 10.54936781 23.55652524 22.19219622 59.18683901 63.13298144 -17.85754567
- 21. 21 5.2 Overview of the Hypotheses I will illustrate the empirical findings in this section in order to solve my two hypotheses. 5.2.1 Hypothesis 1: Some categories of FTSE ST REITs Generate Superior performance than others. In order to analyse further in terms of the performance of FTSE ST REITs, the following graphic illustrates the price evolution of fourteen property stocks over the past six year. To better understand the overview of Singapore real estate market, I have chosen fourteen Singapore real estate stocks and divided them into 5 areas. I have implemented the graph above to demonstrate the price evolution of FTSE ST REITs’ stock over the past six-years to encourage the thoughts of FTSE ST REITs’ properties. I divided sixteen Singapore property shares into 5 classifications. From the graph above, we can see some stocks are perform much better than other classification stock. Investors can forecast a stock’s performance in the future according to the current performance of the stock in the 0 0.5 1 1.5 2 2.5 3 11/1/2008 2/1/2009 5/1/2009 8/1/2009 11/1/2009 2/1/2010 5/1/2010 8/1/2010 11/1/2010 2/1/2011 5/1/2011 8/1/2011 11/1/2011 2/1/2012 5/1/2012 8/1/2012 11/1/2012 2/1/2013 5/1/2013 8/1/2013 11/1/2013 2/1/2014 5/1/2014 8/1/2014 11/1/2014 2/1/2015 5/1/2015 Chart Title Hospitality Healthcare Industrial Office Retail
- 22. 22 market. It is important for an investor to choose which category of the stock to invest. From the graph above, we can see that the healthcare category performed better than other category shares on the equity market from the financial data from 2008 to 2015. While office and industrial category has been seen a steady growth over the period of 2008 – 2015. Hospitality and retails performed better than office and industrial categories from 2008 to 2013, but since Feb 2014, these four categories presented the similar performance from Feb 2014 onwards. Overall, healthcare category dominated have the highest price on the equity market. A quick benchmark from mid of July 2009 to mid of July 2015 is provided by FTSE ST REITs to get the general trend of Singapore REITs using the daily financial data in Singapore. Table 12: Descriptive statistics for FTSE index Healthcare Industrial Office Retail Hospitality FTSE ST REITs Mean 2.103041567 1.24855778 1.323758913 1.38111644 1.501840144 239.6765761 Standard Error 0.008992997 0.006366845 0.005778579 0.003917314 0.003215831 0.659889053 Median 2.21 1.33751 1.3625 1.375 1.4975 240.0333333 Mode 2.35 1.42625 1.52625 1.2725 1.4825 259.8333333 Standard Deviation 0.318077603 0.225191995 0.204385321 0.13855336 0.113742257 23.33993159 Sample Variance 0.101173361 0.050711435 0.041773359 0.019197034 0.012937301 544.7524067 Kurtosis -1.274570644 -1.123334908 -1.14902605 -1.047995738 -0.224877707 -0.658525471 Skewness -0.167499541 -0.414875933 -0.264722411 0.169123526 -0.303829101 -0.027128876 Range 1.37 1.286995 0.7825 0.58 0.641085 107.0166667 Minimum 1.42 0.398235 0.915 1.1025 1.127 189.7033333 Maximum 2.79 1.68523 1.6975 1.6825 1.768085 296.72 Sum 2630.905 1561.945783 1656.0224 1727.776667 1878.80202 299835.3967 Count 1251 1251 1251 1251 1251 1251 The average return for all the five categories are positive. The highest mean return is healthcare (2.103041567) followed by hospitality industry category (1.501840144), the average return of Healthcare is much higher than the other 4
- 23. 23 categories. However, none of the classes’ return exceeds the return of FTSE ST REITs (239.6765761). The coefficient(beta) istoevaluate the performance of the Singapore REITseachsectorand assistus to answerthe firsthypothesis.Accordingtothe table 13 below, Iwill illustrate the coefficientof 14 Singapore REITsbasedon eachclassification. Table 17: Coefficient, classification and FTSE ST Reits (monthly data) Coefficient β Standard Error t Sig Intercept -807.258 210.3437 -3.8378 0.000287 Healthcare 459.8939 55.17455.17496 8.335193 8.28E-12 Hospitality -457.103 135.9031 -3.36345 0.001304 Industrial 48.83247 111.6759 0.43727 0.663388 Office 204.3861 193.9502 1.053807 0.295934 Retail 814.8362 209.2814 3.893494 0.000238 According the table above, at the level of 5%, hospitality and retail classification are statistically signification which present 0.001304 and 0.00238 respectively. In other words, hospitality and retail classifications have real influence on the dependent variable as the coefficient are inferior to 5%. The other three classifications’ β are less than 5% which means they don’t have real influence on the dependent variables. Therefore, investing in the retail area will cause to a rise by 814.8362 units which is the most return of FTSE ST Reits. On the other hand, the performance of the industrial area shows the slowest compare to the other categories, with only an increase of 48.83247. Retail independent variable contribute the most in the explanation of FTSE ST Reits return. Hospitality contribute the opposite to the return of FTSE ST Reits. Table 18: Coefficient, Retail and ST and LT interest rate β Standard Error t Sig Intercept 1.105917 0.056707 19.5023 2.5E-29 ST InterestRate 0.181285 0.038503 4.708394 1.3E-05 LT InterestRate 0.076674 0.027369 2.801488 0.006643 As we can see from the above table, long – term interest rate presents the Beta of 0.006643 in absolute value which is less than 5% level, therefore, it’s the statistically
- 24. 24 significant variable. Meaning to say that long term interest rate has the influence on the dependent variable. Table 19: Coefficient, Healthcare and ST and LT interest rate β Standard Error t Sig Intercept 2.5949 0.198431 13.07707 4.11E-20 ST InterestRate -0.25279 0.134729 -1.8763 0.064971 LT InterestRate -0.42175 0.095771 -4.40378 3.92E-05 From the table above, the Beta for both ST and LT interest rate show more than 5%, indicating that none of them have a real influence on the dependent variable which is 0.064971 and 3.92 respectively. Table 20: Coefficient, Hospitality and ST and LT interest rate Coefficients Standard Error t Stat P-value Intercept 1.542718 0.055401 27.84651 1.49E-38 ST InterestRate 0.017451 0.037616 0.463929 0.644202 LT InterestRate -0.0269 0.026739 -1.00587 0.318097 Table 20 shows that interest rate has no influence on the return of hospitality classification Reits’ return. Table 21: Coefficient, Industrial and ST and LT interest rate Coefficients Standard Error t Stat P-value Intercept 0.559396 0.085993 6.505123 1.15E-08 ST InterestRate 0.10929 0.058387 1.871828 0.065596 LT InterestRate 0.269323 0.041504 6.489148 1.23E-08 Table 20 shows that interest rate has no influence on the return of industrial classification Reits’ return. Table 22: Coefficient, Office and ST and LT interest rate
- 25. 25 Coefficients Standard Error t Stat P-value Intercept 0.759239 0.093141 8.151521 1.28E-11 ST InterestRate 0.16811 0.06324 2.658282 0.009813 LT InterestRate 0.196968 0.044953 4.381615 4.25E-05 Table 20 shows that interest rate has no influence on the return of office classification Reits’ return. 5.2.2 Hypothesis 2: Some Economic Factors Affect the FTSE ST REITs’ Performance. Coefficients Standard Error t Stat P-value Intercept -1801.15 2146.554 -0.83909 0.416584 REALGDP 0.03773 0.023884 1.579722 0.138185 CPI 1.242175 12.87332 0.096492 0.924601 ST 233.5148 239.4796 0.975092 0.347311 LT 7.541249 44.75369 0.168506 0.868779 According to the table of coefficient table, none of the factors have a real influence on the dependent variable because all of them are more than 5% significant level. The parameter under coefficient presents different value in terms of unstandardized coefficients which means the contribution on the FTSE ST REITs’ return of different independent variable are varied. The statistically significant variable is ST interest rate which presents 233.5148 while REALGDP (0.03773) is the lowest Beta. CPI and LT interest rate present 1.242175 and 7.541249 respectively.
- 26. 26 I found the ST and LT interest have more relationship with the return of FTSE ST REITs, in the following sections, I will focus on analyse the coefficient with ST and LT interest with FTSE ST REITs. Table 16: Coefficients,interestratesandFTSTST REITs (Dailydata) Coefficients Standard Error t Stat P-value Intercept 858.1259129 15.53711253 55.23072009 2.5504E-273 LT interestrate -5.090971751 5.857424999 - 0.869148432 0.385030261 ST interestrate -556.0371731 57.21576022 - 9.718251946 3.65986E-21 Table 20 shows that interest rate has no influence on the return of office classification Reits’ return. Table 14 : Coefficients,L-Tinterestrate andFTSTST REITs (monthlydata) Coefficients Standard Error t Stat P-value Intercept 939.287023 45.5214015 20.63396539 2.13E-50 LT interestrate - 122.0978359 15.3477886 -7.955402506 1.57E-13 (Data fromAug 1999 to July2015) Table 1 : Coefficients,S-Tinterestrate andFTST ST REITs (monthlydata) Coefficients Standard Error t Stat P-value Intercept 521.9161 25.012 20.86663 2.16E-48 ST 40.1737* 16.92736 2.373299 0.018767 (Data fromAug 1999 to Aug2013) Table 20 shows that short term interest rate (0.018767) has influence on the return of FTSE ST RETIs. At 5% level, ST interest rate is less than 5% so become significant, therefore, we should reject the hypothesis 2.
- 27. 27 Chapter VI – Analysis This chapter targets to achieve the research purpose of the performance of FTSE ST REITs in six year time series and analyse the economic factors which influence the performance of S-REITs. In order to better understand the outcomes, 2 hypotheses are applied. 6.1 Hypothesis 1: Certain classification in FTSE ST REIT’s performance is superior to others From the table 17 in chapter V, the FTSE ST REITs’ returns are influenced by the hospitality and retail classification. From Appendix 1 the correlation matrix, I observed that five different classification are daily correlated to the S-REITs’ returns. Therefore, the classification can be considered as a representative indicator in the choice of strategy and evaluation of S-REITs performance. All five categories are positively and significantly correlated to the S-REITs, healthcare (0.83643076), industry (0.84792949), office (0.94524881), retail (0.94503900), hospitality (0.51206844), no negative correlation is observed. Hospitality and retail classifications have the strong influence on the performance of FTSE ST REITs as the presentation of table 17 with an advantage of 0.001304 and 0.00238 respectively. As office and retail classifications have the best positive coefficient with FTSE ST REITs, an opportunistic speculator in real estate would invest in office and retail area. In addition, these two categories can be regarded as more competitive than other classes according to my samples. Natale (2000) stated that investors and institutions expect their stocks would increase depending on which subgroup is currently in favour. Applying to my dissertation, real estate investors expect the performance of REITs depending on the FTSE ST REITs classification. According to Ducoulombier (2007), every investment has its own age, structure, localisation, architecture, context, etc. which affect each asset individually. Nevertheless, due to the evaluation of the model is weak, the interpretation need to be more cautious. From financial results in the table 11, the industrial classification is the FTSE ST REITs Real GDP CPI Short term interest rate Long term interest rate FTSE ST REITs 1 0.853634936 0.82603119 -0.28619579 -0.310435072 Real GDP 0.853634936 1 0.960420315 -0.494849766 -0.300249326 CPI 0.82603119 0.960420315 1 -0.522271686 -0.44887364 Short term interestrate -0.28619579 -0.494849766 -0.522271686 1 0.392401628 Long term interestrate -0.310435072 -0.300249326 -0.44887364 0.392401628 1
- 28. 28 only one category to have a negative beta (-11.43419559) accordingto the CAPMtheory.In contrast,the multifactormodel observesthatthe performance of industrial isnotthe worst whichispositive 48.83247. Hypothesis 2: ST REITs’ Performance is affected by Certain Economic Factors Affect the FTSE Take both marco- and micro factors into consideration from the economic perspective, I found the results of multifactor model are interesting. As the studying in Chapter V, the real GDP, CPI, the short-term and long-term interest rate have an influence on the sample FTSE ST REITs returns in terms of the significant level, The empirical findings as the table shown above presents a positive correlation and a significant null sensitivity between the return of FTSE ST REITs and real GDP which is 0.853634936. The beta coefficient of real GDP 0.03773 tends to express that FTSE ST REITs are correlated or extremely highly correlated to the evolution of real GDP. This fact is acceptable and reasonable as GDP plays an influencing role in the real estate market. It is the reflection of the favourable and unfavourable economic climate. Additionally, GDP represents one of the most relevant macro-economics aggregate in the sense that it depicts the level of wealth in a nation (Liow et al.,2006). Therefore, a high real GDP results in a positive reaction from the investors whereas a low real GDP leads to a negative reaction from them. This means the return of FTSE ST REITs and Real GDP move together in a positively and completely linear manner. As for the CPI inflation variable, the examine result is that CPI inflation contributes heavily to the FTSE ST REITs returns. Its correlation and beta coefficient are positively significant which is 0.82603119 and 1.242175 respectively. The values are large, the result corroborate with the hypothesis that an increase in inflation implies an increase of FTSE ST REITs’ return. So when the inflation rises, investors can expect a higher return form FTSE ST REITs. Concerning the interest rate: it is important to distinct between short – term and long – term interest rate. 3 months T-bill yield and 10 year bond yield present the biggest parameter and both are negative correlated, -0.28619579 and -0.310435072 respectively. The intrinsic nature of FTSE ST REITs is to explain the influence between the interest rate and this kind of investment vehicle. The level of significance 10 year bond yield makes the determination of its real influence quite hard. Nevertheless, statistics 3 months T-bill yield makes a high contribution on the return of FTSE ST REITs. The difference in contribution absolutely comes from the interest in FTSE ST REITs investment. As
- 29. 29 investment in real estate is considered as the prudent long-run investment in the view of the conventional sight, as a result, the long-term interest rate prevails on the short – term interest rate by determining a less volatile fluctuation during transaction processes. For example, when the interest rate rises, it will impact directly real estate market due to business rate such as credit rate are mostly based on these reference rates. Therefore, an increase of these rates leads to higher interest cost resulting in the less return. Another impact is that an increase in interest rate resulting in a rise in FTST ST REITs return because it will lead to economic growth and more demand. From the correlation table, which shows that a negative correlation, ST interest rate (-0.28619579), LT interestrate (-0.310435072), from the monthlydataon the coefficient table,showingthatthere issignificantnull sensitivitybetweenthe shortterminterest rate (0.018767) and the returnof FTSE ST REITs. The beta coefficientof shortterminterestrate (40.1737) tendstopresentthatthe return of FTSE ST REITs are uncorrelatedorverylowly correlatedtothe varyingof shortterminterestrate.