The FTSE NAREIT PureProperty Index Series including comparison with the NPI, NFI-ODCE, NTBI, and MSCI US Liquid Real Estate Index

Transcription

1 The FTSE NAREIT PureProperty Index Series including comparison with the NPI, NFI-ODCE, NTBI, and MSCI US Liquid Real Estate Index Presentation by Brad Case, PhD, CFA, CAIA Senior Vice President, Research & Industry Information to the Research and Portfolio Management Committees, National Council of Real Estate Investment Fiduciaries June 25, 2015 National Association of Real Estate Investment Trusts REITs: Building Dividends & Diversification 1875 I St, NW Suite 600, Washington, DC

2 Presentation Outline Background: Measuring Returns for Targeted Real Estate Portfolios Conceptual Approaches and Difficulties The FTSE NAREIT PureProperty Index Methodology Conceptual Approach to Measuring Pure-Play Returns Liquidity-Optimization Deleveraging Comparing the PureProperty Index with Other Real Estate Indices NPI, NFI-ODCE, and NTBI MSCI US Liquid Real Estate Index Portfolio and Risk Management Uses of the PureProperty Index Series Real Estate Portfolio Benchmarking / Monitoring Strategic Real Estate Investment Rapid Capital Deployment Tactical Portfolio Allocation Reduce / Hedge Concentration Risk Swap Real Estate Returns for Non-RE Returns Target Leverage Portable Alpha 2

3 Introduction to the PureProperty Index Series Investment-oriented product introduced June 1, 2015 Initial information-oriented product introduced June 2012 Not marketed significantly, pending investment-oriented modifications Daily historical data from December 31, 1999 Series of 88 indices Six property types: Apartment, Industrial, Office, Retail, Health Care, Hotel Four regions: East, Midwest, South, West (NCREIF definitions) Eleven region/type combinations E-Apt, M-Apt, S-Apt, W-Apt, E-Off, M-Off, S-Off, W-Off, E-Ret, M-Ret, S-Ret Nationwide all-type aggregate Total return and price return Equity and Property versions Daily historical data from September 23, constituents, aggregate market cap $630 billion 38,000 properties, aggregate market value $1 trillion 3

4 Measuring Returns on a Targeted Portfolio Valuation Target Property (and aggregate up) Appraisal NCREIF Property Index IPD Property Index NFI-ODCE 1 PREA/IPD Fund Index 1 Green Street CPPI 2 (REIT equity analysts) Valuation Approach Transaction Moody s/rca CPPI CoStar CRSI MIT TBI NCREIF TBI 3 Portfolio (and dis-aggregate down) (possible, but none in use) REIT indices 4 PureProperty 1 Fund managers value individual properties by appraisal and aggregate up to fund level, but NFI-ODCE and PREA/IPD measure returns on diversified portfolios (funds) rather than targeted portfolios. 2 Combines completed transactions, pending transactions, and (informal) appraisals; not published for targeted portfolios. 3 Uses transactions to compute a correction factor that is then applied to appraisals, with the same factor used for all targeted portfolios. 4 Disaggregated only by property type based on dominant property-type holdings of each REIT. 4

5 Problems in Measuring Returns on a Targeted Portfolio Valuation Target Property (and aggregate up) Portfolio (and dis-aggregate down) Appraisal Appraisal inaccuracy 1,2 Appraisal infrequency Appraisal smoothing 3 Non-appraisal smoothing Appraisal lag Appraisal inaccuracy 1,2 Appraisal infrequency Appraisal smoothing 3 Non-appraisal smoothing Appraisal lag Disaggregation method Valuation Approach Transaction Transaction infrequency Transaction smoothing 4 Transaction lag Disaggregation method 1 Cannon & Cole [2011]: On average, appraisals are more than 12% above or below subsequent sales prices... Even in a portfolio context where errors can cancel out each other, results are not appreciably better: appraisals are off by an average of 4%-5% of value because under- and overvaluations are highly correlated across properties at the same points in time. 2 Appraisals are also subject to the possibility of strategic behavior, including selectivity and intentional bias (the Libor scandal problem). 3 Appraisal smoothing arises from the cognitive problem of anchoring. 4 Transaction smoothing arises from aggregation of all transactions occurring within a given month or quarter, which is necessitated by transaction infrequency. 5

6 A Simplified Example of the PureProperty Methodology 1. Let s say there are just two REITs (REIT 1 and REIT 2). We observe their portfolio holdings: REIT 1 is classified as an apartment REIT, but has some office properties: its portfolio is 80% Apartment and 20% Office. REIT 2 is classified as an office REIT, but has some apartment properties: its portfolio is 25% Apartment and 75% Office. 2. Let s say that we observe these stock price changes: REIT %, REIT 2-0.5%. +1.4% is NOT the change in apartment values, because REIT 1 s stock price was also affected by the change in value of its office holdings. -0.5% is NOT the change in office values, because REIT 2 s stock price was also affected by the change in value of its apartment holdings. 3. Mathematically, there is only one combination of property value changes that work exactly, given what we observed: Algebra, with two equations in two unknowns, gives us a way of computing the answer: Apartment values must have gone up by 2% and office values must have gone down by 1%. The value of REIT 1 s portfolio changed by 80% * +2% (Apartment) plus 20% * -1% (Office) = +1.4%. The value of REIT 2 s portfolio changed by 25% * +2% (Apartment) plus 75% * -1% (Office) = -0.5% 4. In the real world things the numbers don t work out exactly, so we use regression: Which combination of property value changes BEST explains what we observed? 6

7 This document was presented during the 2015 NCREIF The PureProperty Summer Conference. Equity Index Methodology (still simplified) Observe the property holdings of all stock exchange-traded U.S. REITs: With holdings primarily in the six property types (Apartment, Health Care, Hotel, Industrial, Office, and Retail) Compute portfolio share by market segment Property Type, Region, and Type/Region Combination) Observe the returns for all REITs on each day: Stock price change (price return) Plus income distributions (total return) Do a cross-sectional regression of returns on portfolio shares: r it = β it E x it E + β it M x it M + β it S x it S + β it W x it W + ε it where x it E + x it M + x it S + x it W = 1 The slopes of the regression equation are the changes in the PureProperty indices To compute the index weights, stop just short of running the regression: β = (X X) 1 X r w = (X X) 1 X The weights identify a portfolio of positions in investable REIT constituents Geek alert! The returns to the portfolio are the pure returns to the target market segment 7

8 Example: Purifying the Apartment Index In December 2011, only seven REITs had pure apartment portfolios Equity Residential, Apt Investment & Management, Camden Property Trust, Home Properties, Mid-America Apartment Communities, American Campus Communities, and Post Properties Most apartment REITs also had non-apartment holdings AvalonBay Communities, Essex Property Trust, and BRE Properties (office holdings) UDR (retail holdings) Colonial Properties Trust (substantial office and retail holdings) 11 non-apartment REITs received small negative index weights Office: SL Green (-0.148%), Alexandria (-0.091%), Mack-Cali (-0.059%), BioMed (-0.056%), and Boston Properties (-0.045%) Retail: GGP (-0.059%), Kimco (-0.013%), Macerich (-0.010%), and Simon (-0.003%) Diversified: Highwoods Properties (-0.021%) and Liberty Property Trust (-0.010%) Negative weights (short positions) totaled just 0.51% Small negative weights on several companies with holdings mostly of nonapartment properties counteracted the effects of non-apartment holdings on the stock price movements of apartment REITs Thereby purifying the apartment index 8

9 Example: Weights for the Apartment Equity Index Town & Country Trust AMLI Residential Properties Gables Resident GMH Communities Trust Highwoods Prop Other Home Properties BRE Properties Associated Estates Realty Campus Crest Communities Education Realty Trust Colonial Prop American Campus Communities Mid-America Apartment Comm Post Properties Other Associated Estates Realty GMH Communities Trust Highwoods Prop Campus Crest Communities Education Realty Trust Town & Country Trust Colonial Prop AMLI Residential Properties 0.5 Essex Prop Trust American Campus Communities Mid-America Apartment Comm 0.4 Camden Property Gables Resident Post Properties UDR Home Properties 0.3 BRE Properties Apartment Inv Management Essex Prop Trust 0.2 Camden Property Archstone-Smith Trust Avalonbay Communities Equity Residential UDR Apartment Inv Management Avalonbay Communities Archstone-Smith Trust Equity Residential 9

10 The PureProperty Property Index Methodology REITs use leverage, so REIT returns are the returns of equity investments made in properties, but through the balance sheets of the REITs that own those properties The PureProperty Property indices measure unlevered (property) returns Using leverage means borrowing capital (and making payments on it) De-levering means lending capital (and receiving payments on it) with yields matched as closely as possible to REITs weighted average cost of debt (WACD) For the Property indices, two constituents are added: FTSE US Government Bond Index FINRA-Bloomberg Active High Yield U.S. Corporate Bonds Index Both fixed-income indices are chosen because investors can almost perfectly replicate them using liquid fixed-income ETFs that is, there are liquid fixed-income ETFs that explicitly track very similar indices The aggregate weights on fixed-income indices depend on aggregate use of debt by the constituents of each PureProperty index The specific weight on each fixed-income index is derived from a regression of REIT WACD on fixed-income index yields 10

11 The MSCI US Liquid Real Estate Index Objective: use liquid securities to create an investable index matching, as closely as possible, the return characteristics of the IPD US Quarterly Property Index One aggregate property-level (delevered) total return index only Does not disaggregate by property type or geography; reflects mix in MSCI Core Real Estate Index No corresponding equity version, or price return version Portfolio uses: Strategic real estate investment Rapid capital deployment Swap returns on a broadly diversified real estate portfolio for non-re returns Constituents: same as MSCI Core Real Estate Indexes; includes non-reits Weights: computed using MSCI s Volatility Tilt Factor Index methodology Over-weights constituents with lower volatility, relative to market cap weights Deleveraging methodology: similar to PureProperty Fixed-income component measured by Markit iboxx short-duration inflation-linked index Very slightly greater volatility than PureProperty Perhaps because constituents include non-reits such as developers Perhaps because of differences in deleveraging methodology ishares MSCI Target US Real Estate UCITS ETF in market (USRE LN) 11

12 What do the PureProperty Indices Look Like? 12½+ years (50 quarters) of data, 9/23/02 to present Frequency: daily not monthly, not quarterly no transaction smoothing Timing: same day not four weeks after the end of the quarter (or month) no stale data Average Gross Total Returns: slightly better than NCREIF measures in aggregate (9.98% for PP vs 8.88% for NPI, 8.86% for NTBI, and 8.01% for ODCE) note: investment costs should be much less than the costs of private real estate investment Volatility: comparable to NCREIF measures annual frequency: 12.2% vs 10.5% for NPI, 12.5% for NTBI, and 14.4% for ODCE) quarterly frequency: 11.4% vs 10.9% for NTBI Risk-Adjusted Gross Returns (Sharpe Ratio): comparable to NCREIF measures annual frequency: 0.73 vs 0.77 for NPI, 0.65 for NTBI, and 0.52 for ODCE quarterly frequency: 0.88 vs 0.77 for NTBI note: risk-adjusted net total returns should be better than for NCREIF measures 12

13 What do the PureProperty Indices Look Like?

14 Comparing Returns: PureProperty Property $40,000 $35,000 $30,000 $25,000 Gain from: Income on Initial Investment = $7,020 Income on Reinvested Income = $2,381 Cap Apprec on Initial Investment = $7,487 Cap Apprec on Reinvested Income = $2,934 TOTAL = $19,822 Capital Appreciation on Reinvested Income $20,000 $15,000 Capital Appreciation on Initial Investment Income on Reinvested Income Income on Initial Investment $10,000 $5,000 Initial Investment $0 Sep-02 Sep-03 Sep-04 Sep-05 Sep-06 Sep-07 Sep-08 Sep-09 Sep-10 Sep-11 Sep-12 Sep-13 Sep-14 14

15 Comparing Returns: MSCI US Liquid Real Estate Index $40,000 $35,000 $30,000 Gains TOTAL = $15,675 $25,000 $20,000 $15,000 Total Gains on Initial Investment $10,000 $5,000 Initial Investment $0 Sep-02 Sep-03 Sep-04 Sep-05 Sep-06 Sep-07 Sep-08 Sep-09 Sep-10 Sep-11 Sep-12 Sep-13 Sep-14 15

16 Comparing Returns: NTBI $40,000 $35,000 $30,000 Gain from: Income on Initial Investment = $4,823 Income on Reinvested Income = $1,074 Cap Apprec on Initial Investment = $6,948 Cap Apprec on Reinvested Income = $1,792 TOTAL = $14,637 $25,000 $20,000 $15,000 Capital Appreciation on Reinvested Income Capital Appreciation on Initial Investment Income on Reinvested Income $10,000 Income on Initial Investment $5,000 Initial Investment $

17 Comparing Returns: NCREIF Property Index $40,000 $35,000 $30,000 Gain from: Income on Initial Investment = $7,443 Income on Reinvested Income = $2,958 Cap Apprec on Initial Investment = $3,835 Cap Apprec on Reinvested Income = $1,764 TOTAL = $16,000 $25,000 $20,000 Capital Appreciation on Reinvested Income Capital Appreciation on Initial Investment Income on Reinvested Income $15,000 Income on Initial Investment $10,000 $5,000 Initial Investment $

18 Comparing Returns: NFI-ODCE $40,000 $35,000 $30,000 Gain from: Income on Initial Investment = $6,711 Income on Reinvested Income = $2,418 Cap Apprec on Initial Investment = $2,643 Cap Apprec on Reinvested Income = $1,262 TOTAL = $13,035 $25,000 $20,000 Capital Appreciation on Reinvested Income Capital Appreciation on Initial Investment Income on Reinvested Income $15,000 Income on Initial Investment $10,000 $5,000 Initial Investment $

19 Comparing Returns: PureProperty Equity $40,000 $35,000 $30,000 Gain from: Income on Initial Investment = $8,341 Income on Reinvested Income = $3,069 Cap Apprec on Initial Investment = $11,896 Cap Apprec on Reinvested Income = $5,412 TOTAL = $28,718 Capital Appreciation on Reinvested Income $25,000 Capital Appreciation on Initial Investment $20,000 Income on Reinvested Income $15,000 Income on Initial Investment $10,000 $5,000 Initial Investment $0 Sep-02 Sep-03 Sep-04 Sep-05 Sep-06 Sep-07 Sep-08 Sep-09 Sep-10 Sep-11 Sep-12 Sep-13 Sep-14 19

20 Volatility Comparison: PureProperty and NTBI Market Segment Price Appreciation Total Return PureProperty NTBI PureProperty NTBI Apartment 11.1% 10.9% 11.1% 11.0% Industrial 11.1% 10.6% 11.2% 10.7% Office 11.6% 10.9% 11.6% 10.9% Retail 12.1% 11.3% 12.1% 11.5% East 12.3% 11.3% 12.3% 10.2% Midwest 10.9% 10.3% 10.9% 10.4% South 10.3% 10.6% 10.3% 10.7% West 12.2% 10.9% 12.3% 10.9% Aggregate 11.3% 10.9% 11.4% 10.9% 20

21 Sharpe Ratio Comparison: PureProperty and NTBI Market Segment Average Net Total Return Sharpe Ratio PureProperty NTBI PureProperty NTBI Apartment 9.60% 7.78% Industrial 8.76% 7.40% Office 8.53% 6.98% Retail 10.05% 10.22% East 9.92% 8.63% Midwest 8.78% 6.43% South 8.85% 7.74% West 10.33% 8.31% Aggregate 9.48% 7.86%

22 How can the PureProperty Indices be Used? Real Estate Portfolio Benchmarking / Monitoring Targeted diversified portfolios, no measurement/reporting lag, no smoothing, no error Strategic Real Estate Investment Stronger (risk-adjusted) net total returns without sacrificing liquidity, low transaction costs Rapid Capital Deployment Daily liquidity in targeted diversified portfolios Tactical Portfolio Allocation Long or short positions, index replication, ETFs, derivative contracts Reduce / Hedge Concentration Risk Bring portfolio exposures to target ranges without selling high-quality assets Swap Real Estate Returns for Non-RE Returns Lock in risk-free return, variable interest rate, inflation rate, etc. Target Leverage Combine equity and property indices to adjust leverage exposure synthetically Portable Alpha Maximize allocations to superior asset manager while maintaining real estate beta 22

23 Challenges in Direct Real Estate Investment Illiquidity Large transaction costs Large unit size (indivisibility / non-granularity) Difficulty of constructing a diversified property portfolio Cheng & Roulac [2007]: [T]o reduce the diversifiable risk to about 10% of a portfolio s total risk, investors would have to invest in at least sixty-six MSAs, an impossible task for virtually any institution.... [G]eographically concentrated portfolios may be severely under-diversified in the sense that they contain high diversifiable risks, and it is questionable whether the returns of those portfolios sufficiently compensate investors for taking the risks. Large capital outlay Especially for a diversified portfolio Difficulty of measuring asset values and monitoring risks Especially using appraisal-based indices Infrequent and lagged performance measurement Impossibility of short positions 23

24 Challenges in Synthetic Real Estate Investment Difficulty of measuring asset values / returns Appraisal-based indices Appraisal error Appraisal lag Appraisal smoothing Transaction-based indices Transaction lag Transaction smoothing Possibility of inadequate transaction volume Backward revisions (historical restatement) All are variable over the market cycle and difficult to predict Reporting lag Possibility of strategic behavior (LIBOR scandal) Infrequent return measurement Large margin requirements (infrequent settlement dates) Largest inter-date movement in NPI was 8.29% compared to 5.25% for the PureProperty index Lack of investable underlying Impossible to exploit arbitrage opportunities Requires identification of a counterparty to every trade Counterparty risk 24

25 The Promise of Synthetic PureProperty Investment Index is Based Entirely on Liquid Securities Daily liquidity Minimal transaction costs Granularity (small unit size) Daily Index Computation No measurement lag No reporting lag Tiny margin requirements No Realistic Possibility of Strategic Behavior Reduces reputation risk / litigation risk Increases willingness to enter contracts on both sides Underlying Index is Entirely Investable Arbitrage opportunities, including shorting Exposures can be hedged, so no need to find counterparties Index Can Be Tracked by ETFs Intra-Day Liquidity 25

26 An Important Advantage of Private Real Estate Investment Lag and Smoothing Are Beneficial Reduces perceived volatility, increases perceived risk-adjusted returns Increases perceived diversification benefits These benefits are false, but reportable Liquid securities do not benefit from lag and smoothing BUT lag and smoothing can be replicated using liquid securities! Backward-Forward Trading Contract Derivative whose value in the future ( forward ) depends on the lagged value ( backward ) of liquid securities Example: Backward-Forward 7/7 pays off 7 quarters from now based on the average daily value of the PureProperty index over the preceding 7 quarters B/F 7/7 NPI ODCE Gross TR 8.48% 8.83% 7.78% Volatility 4.7% 6.1% 8.7% Sharpe Ratio Jun-04 Correlation between B/F 7/7 and: NPI: 91.5% ODCE: 90.9% Jun-05 Jun-06 Jun-07 Jun-08 Jun-09 Backward-Forward 7/7 contracts NCREIF Property Index NFI-ODCE Jun-10 Jun-11 Jun-12 Jun-13 Jun-14 26

27 Final Thoughts about the PureProperty Methodology The methodology can easily be applied to different market segments Gateway MSAs CBD vs Suburban markets Industry-related real estate exposures (e.g., oil patch markets) Application of the methodology to non-u.s. markets is straightforward, depending only on data and securities Listed property companies (LPCs) such as listed equity REITs Data on property holdings by market segment For property-level indices: Balance sheet data (use of debt) Data on weighted average cost of debt (WACD) Liquid fixed-income securities with yields approximating WACD For more information to to:

28 Conclusion The PureProperty Index Series provides, for the first time, an entire suite of risk management tools for private real estate portfolios. I don t see much from our real estate teams from a risk perspective, and feel the risk team didn t know much about portfolio risk except for leverage. U.S. pension fund quoted in MSCI, Long-Run Investment Ambitions and Short-Run Investment Processes: Findings of the 2013 Global Asset Owner Survey (February 2014) True risk management in real estate portfolios may lead to dramatically increased strategic allocations to the real estate asset class. Although many investors have significant exposure to real estate, the asset class is, as described by Roger Urwin, dripping with issues that represent major concerns for the boards and CIOs of many asset owners. IPD, The Asset Owner Real Estate Investment Process: Risk Management Insights from the MSCI/IPD Survey (February 2014) 28

29 Technical Appendix: Details of the Investable PureProperty Equity Index Methodology Problem: less-liquid REITs with portfolios concentrated in under-represented market segments could have large weights relative to their liquidity Index weights are determined by an optimization that is not quite regression explicitly to avoid liquidity problems Larger REITs (in enterprise value) get larger weights because stock price movements for larger REITs are higher-quality signals of changes in portfolio values No REIT gets a weight larger than 10% in any index No REIT gets a weight that would cause liquidity problems, defined as: that REIT s weight times $500 million > 20% of average daily trading volume Specifically, subject to Max W = min W X Ω 1 X W X Ω , 0.2ADTV Vw k w i,k Min 0.1, 0.2ADTV and N i=1 w ik = 1 Vw k Really BIG g 29

30 Disclaimer NAREIT is the worldwide representative voice for REITs and listed real estate companies with an interest in U.S. real estate and capital markets. Members are REITs and other businesses that own, operate and manage income-producing real estate, as well as those firms and individuals who advise, study and service those businesses. NAREIT is the exclusive registered trademark of the National Association of Real Estate Investment Trusts, Inc., 1875 I St., NW, Suite 600, Washington, DC Follow us on REIT.com. Copyright 2015 by the National Association of Real Estate Investment Trusts, Inc. All rights reserved. This information is solely educational in nature and is not intended by NAREIT to serve as the primary basis for any investment decision. NAREIT is not acting as an investment adviser, investment fiduciary, broker, dealer or other market participant, and no offer or solicitation to buy or sell any security or real estate investment is being made. Investments and solicitations for investment must be made directly through an agent, employee or representative of a particular investment or fund and cannot be made through NAREIT. NAREIT does not allow any agent, employee or representative to personally solicit any investment or accept any monies to be invested in a particular security or real estate investment. All REIT data are derived from, and apply only to, publicly traded securities. While such data are believed to be reliable when prepared or provided, such data are subject to change or restatement. NAREIT does not warrant or guarantee such data for accuracy or completeness, and shall not be liable under any legal theory for such data or any errors or omissions therein. See for important information regarding this data, the underlying assumptions and the limitations of NAREIT s liability therefore, all of which are incorporated by reference herein. Performance results are provided only as a barometer or measure of past performance, and future values will fluctuate from those used in the underlying data. Any investment returns or performance data (past, hypothetical or otherwise) shown herein or in such data are not necessarily indicative of future returns or performance. Before an investment is made in any security, fund or investment, investors are strongly advised to request a copy of the prospectus or other disclosure or investment documentation and read it carefully. Such prospectus or other information contains important information about a security s, fund s or other investment s objectives and strategies, risks and expenses. Investors should read all such information carefully before making an investment decision or investing any funds. Investors should consult with their investment fiduciary or other market professional before making any investment in any security, fund or other investment. For more information please visit: 30