Value Premium in International REITs ERES Conference 2014 Ytzen van der Werf and Fred Huibers

  • Published on
    22-Feb-2016

  • View
    36

  • Download
    0

DESCRIPTION

Value Premium in International REITs ERES Conference 2014 Ytzen van der Werf and Fred Huibers 27 June 2014 y.vanderwerf@asre.nl. Outline. Introduction Aim Literature review Methodology Data Results Conclusion. Introduction . Value investing attractive for common stocks - PowerPoint PPT Presentation

Transcript

<p>PowerPoint-presentatie</p> <p>Value Premium in International REITs</p> <p>ERES Conference 2014</p> <p>Ytzen van der Werf and Fred Huibers27 June 2014y.vanderwerf@asre.nlGood afternoon Ladies and Gentlemen,</p> <p>Today I will be presenting the results of our study into the Value premium within International REITs. Fred Huibers and I are both employed by the Amsterdam School of Real Estate and have conducted this study to the existence of the so called Value premium in REITs within International Developed markets.1OutlineIntroductionAimLiterature reviewMethodologyDataResultsConclusionMy outline for today will be:</p> <p>I will start with a short introduction followed by the Aim of the study and the Literature review. Subsequently the methodology and the data will be discussed and finally I will present Results and Conclusions.</p> <p>2Introduction Value investing attractive for common stocksValue premium internationally accepted phenomenonValue premium = excess return from buying cheap and selling expensive stocksExtensively researched within common equitiesLatest research into question whether it is a reward for risk or rather a result of behavioural inaccuracy Value investing is attractive investment strategy for Equity investors. If this study to the value premium in International developed REITs turns out to be as successful this might be very beneficial to investment managers who are able to structurally outperform the market benchmarks.</p> <p>The value premium is an internationally accepted phenomenon. Founding fathers of this phenomenon were Graham and Dodd with their classic book Security Analysis in 1934 claiming that the most effective way to invest in equities was to buy low and sell high. But how do we define low and high!</p> <p>The Value premium is defined as the excess return of buying cheap stocks while (at the same time) selling expensive stocks. Cheap stocks are defined as stocks that have relatively low market prices versus a fundamental value. This fundamental value could be Book Value, earnings (per share) or dividend (per share). </p> <p>The value premium has been studied extensively within common stocks. In the Journal of Finance alone nearly 10 authors have shown that for common stocks the value premium exists and is statistically significant. Recent studies have predominantly examined the cause of the Value premium in equities. </p> <p>Two different schools of thought with respect to the cause have emerged. On the one hand there is the risk based school of thought led by Fama and French who argue that the excess return on the value stocks can only be explained by a higher risk profile of the portfolios (Markowitz / CAPM). On the other hand there is a behavioural school of thought led by Lakonishok, Schleifer and Vishny who posit the value premium could be a result of over optimism and nave extrapolation of past results by investors.</p> <p>3Aim of this studyTo find out whether: the value premium exists for REITs in international developed markets and whether it is a reward for risk or a behavioural phenomenonThe aim of our study is to examine whether: Is there a significant value premium for International developed REITs and if so can this excess return be explained by a higher risk profile or is it a result of behavioural inaccuracy.</p> <p>Our contribution to the literature lies within the extended period of time (including the GFC) and broadening the geographical scope outside the US and therefore including currency risk in the results. </p> <p>4Literature review REVast number of studies into US REITsSome find consistent value premium in US REITs for the 90s. Gentry et al. (2004) 11-22%. Ooi et al. (2007) 8.5%.One study into direct real estate (Addae et al. 2013). Assume properties with low initial yield are growth investments. Find value premium of 6% p.a. for offices and 8% for retail.No study so far into non-US REITsThere is an emerging body of evidence that the value premium exists for US REITs. Gentry, Jones and Mayer (2004) in a working paper for the NBER (National Bureau Economic Research) found significant alphas of 11 22% and Ooi, Webb and Zhou (2007) in a comprehensive study on US REITs for 1993-2001 found a value premium of 8.5%. Both using the Book to Market multiple as ranking variable to establish whether stocks where either cheap or expensive. Ooi, et al. (2007) also examined whether the US value premium was a reward for higher riskiness of the value portfolio or nave extrapolation of past results.</p> <p>There has only been one study into direct real estate. Within Direct Real Estate it is not possible to compare the Book value of a property with the Market value. Therefore Addae et al. compared the Net Operating Income with the Market value. Essentially they were assessing the yield of the property. They define properties with a low NOI relative to value as expensive (or growth) properties and high NOI/Value as cheap (or value) properties. In other words, low yielding properties are growth properties since they are expected to increase in value and therefore delivering additional (indirect) return. Addae et al found an average Value premium of 6% per annum for US and Asian offices and 8% for retail units. </p> <p>None of the studies we have encountered investigated the REITs outside the US so far. Most studies use the Book to Market multiple as their ranking variable and group stocks (REITs) into 5 Quintile portfolios. Our contribution to the literature will be an extension of the period of research and a broadening of the geographical scope outside of the US.</p> <p>5MethodologyExclude REITs with negative Book to Market and daily trading volume &lt; 0.5 million EURRank stocks at the 30th of June each year on book to market (B/M)After ranking the REITs, we group them into Quintiles (20% highest B/M = Value REITs). Calculate the equally weighted average return of the different (quintile) portfolios, assuming an equal amount is invested in each REIT</p> <p>Book to Market represents the book value of the Equity versus the Market Value of the Equity. Remember this takes leverage into account. Highly leveraged REITs could have negative book value of Equity and will be excluded from our analysis since distressed (Real Estate) equity is said to behave differently.</p> <p>Every 30th of June we will rank REITs on their Book to Market multiple and group them into 5 quintiles. REITs with a high B/M are said to be cheap (Value) REITs and labelled Q1. Stocks with a low B/M are relatively expensive and therefore Growth REITs and labelled Q5. To ensure liquidity we exclude REITs with a daily trading volume in June less than .5 million.</p> <p>Subsequently we calculate equally weighted mean total returns of the different Quintile portfolios for post formation year. First formation date was 30 June 1993 and total return of July 93-June 94 has been calculated. </p> <p>6Methodology Book to Market multiple1.00HighestLowest0.00Discount to NAVPremium to NAV30 JuneBook to Market represents the book value of the Equity versus the Market Value of the Equity. Remember this takes leverage into account. Highly leveraged REITs could have negative book value of Equity and will be excluded from our analysis since distressed (Real Estate) equity is not necessarily equivalent to growth Equity.</p> <p>Every 30th of June we will rank REITs on their Book to Market multiple and group them into 5 quintiles. REITs with a high B/M are said to be cheap (Value) REITs and labelled Q1. Stocks with a low B/M are relatively expensive and therefore Growth REITs and labelled Q5. To ensure liquidity we exclude REITs with a daily trading volume in June less than .5 million.</p> <p>Subsequently we calculate equally weighted mean total returns of the different Quintile portfolios for post formation year. First formation date was 30 June 1993 and total return of July 93-June 94 has been calculated. </p> <p>7MethodologyBook to Market multiple1.00HighestLowest0.00Discount to NAVPremium to NAVQ1Q5Q2Q3Q430 JuneBook to Market represents the book value of the Equity versus the Market Value of the Equity. Remember this takes leverage into account. Highly leveraged REITs could have negative book value of Equity and will be excluded from our analysis since distressed (Real Estate) equity is said to behave differently.</p> <p>Every 30th of June we will rank REITs on their Book to Market multiple and group them into 5 quintiles. REITs with a high B/M are said to be cheap (Value) REITs and labelled Q1. Stocks with a low B/M are relatively expensive and therefore Growth REITs and labelled Q5. To ensure liquidity we exclude REITs with a daily trading volume in June less than .5 million.</p> <p>Subsequently we calculate equally weighted mean total returns of the different Quintile portfolios for post formation year. First formation date was 30 June 1993 and total return of July 93-June 94 has been calculated. </p> <p>8Methodology1.00HighestLowestDiscount to NAVPremium to NAVQ1Q5Value REITsGrowth REITsannual portfolio return1 July 30 June30 Juneannual portfolio return1 July 30 JuneBook to Market represents the book value of the Equity versus the Market Value of the Equity. Remember this takes leverage into account. Highly leveraged REITs could have negative book value of Equity and will be excluded from our analysis since distressed (Real Estate) equity is said to behave differently.</p> <p>Every 30th of June we will rank REITs on their Book to Market multiple and group them into 5 quintiles. REITs with a high B/M are said to be cheap (Value) REITs and labelled Q1. Stocks with a low B/M are relatively expensive and therefore Growth REITs and labelled Q5. To ensure liquidity we exclude REITs with a daily trading volume in June less than .5 million.</p> <p>Subsequently we calculate equally weighted mean total returns of the different Quintile portfolios for post formation year. First formation date was 30 June 1993 and total return of July 93-June 94 has been calculated. </p> <p>9DataInternational Developed REITs with viewpoint of European Investor (returns in Euro)23 countries with a total of 1,152 REITsUse minimum daily trading volume of 0.5 m EUR to ensure liquidity and positive B/M</p> <p>We have collected all REITs from Thomson Reuters Datastream for 23 developed countries. A total of 1,152 REITs were collected and REITs with a negative Book to Market Multiple at 30 June of each year are excluded, since negative B/M multiples imply distressed assets. </p> <p>Furthermore we required a minimum trading volume of 0.5 million euro per day to ensure liquidity. This gave us the following number of REITs to choose from each year. 100 REITs in 1993 and over 400 in 2012. </p> <p>When forming quintile portfolios this leads to approximately 20 stocks in a portfolio in 1993 and around 80 in 2012. In 2012 these REITS have a total market cap of 700 bn EURO of which 40 bn are Value stocks (6%). </p> <p>10ResultsCharacteristics (Q1 = Value REITs)</p> <p>*** significant at 1% level</p> <p>Cumulative Total Returns (1,2,3 year holding)</p> <p>** significant at 5% level; *** significant at 1% levelTR1 is the average yearly return of holding a value (growth) portfolio for 1 year and then rebalance the portfolio with (possible) new REITs with high (low) book to market multiples. TR02/03 cumulative total return of 2/3 years after portfolio formation</p> <p>Q1Q2Q3Q4Q5Diff. Q1-Q5B/M2.231.06.84.65.361.86***MV (m EUR)4455987491,0791,5431,098***Q1Q2Q3Q4Q5Diff. Q1-Q5TR1.218.145.130.100.115.103 **TR02.439.276.276.236.266.173 ** TR03.712.464.435.378.422.290***The first table shows the descriptive stats of all quintile portfolios. Firstly the mean Book to Market multiple over the 20 year period and secondly the mean Market value. It is clear that Q1 (Value) has a higher B/M multiple and a lower average market size.</p> <p>The lower table shows that cumulative return on a 1, 2 and 3 year basis is 21.8, 43.9 and 71.2% respectively for Value REITs and only 11.5, 26.6 and 42.2% respectively for Growth REITs</p> <p>75% of the years deliver higher returns for Q1 compared to Q5. Including the GFC in the study period does not alter results substantially, 2/10 years negative performance of Q1-Q5 before 2003 and 3 years after 2003. </p> <p>11Explanation (I)Risk based school of thought (Fama and French)Higher return is a reward for higher riskTest whether volatility of value REITs is higher or whether beta within CAPM framework is higher for value stocks</p> <p>* significant at 10% level** significant at 5% level *** significant at 1% level</p> <p>The risk based school of thought led by Fama and French argued that the Value premium should be a result of higher risk taken when investing in the value portfolio. We have tested this first by examining the standard deviation, Sharpe and Treynor ratio of the portfolio returns and subsequently by establishing whether value portfolios have a higher risk loading (beta) within the well known CAPM model and/or significant excess return over the risk adjusted return (alpha). In the second case we regressed portfolio returns (minus risk free rate) on the returns of the market portfolio (a weighted portfolio of common stocks in developed markets) obtained from Kenneth French website.</p> <p>Sharpe ratio = (Return portfolio risk free return) / [standard deviation excess return] Higher Sharpe ratio better risk adjusted performance</p> <p>The Treynor ratio relates excess return over the risk-free rate to the additional risk taken. The higher the Treynor ratio, the better the (risk adjusted) performance of the portfolio under analysis</p> <p>Treynor = (Ri Rf) / beta </p> <p>Q1 (Value REITs) clearly shows a higher Sharp and Treynor ratio, although their standard deviation is a little higher as well.</p> <p>Within the CAPM model Q1 exhibits even lower beta compared to the Growth portfolio (Q5) .499 versus .649</p> <p>12Explanation (II)Behavioural school of thought (Lakonishok)Higher return in value REITs is a result of naive extrapolation of results from the past to the future</p> <p>Test whether growth REITs indeed show higher past performance and whether this changes after portfolio formation and vice versaPerformance measured as Total Return </p> <p>The behavioural school of thought argues that value stocks (or REITs) do not have to be riskier at all. The excess return could be explained by investors that are overly excited about stocks that performed well in the recent past and expect their more than average performance to hold for more years.</p> <p>This hypothesis can be tested by examining pre and post formation returns or fundamentals such as growth of earnings and dividends. 13Explanation (III)Pre and post-formation total return performance</p> <p>** significant at 5% level; *** significant at 1% level</p> <p>This sheet clearly shows the turn in pre and post formation returns for Value as well as Growth REITs. Value REITs exhibit a significantly lower total return than Growth REITs prior to formation date (13.7% with p-value &lt; 0.01 in the last year) and a significantly higher Total Return in the first year after formation (10.3% (5%...</p>

Recommended

View more >