Of Granularity and Cocktails Methodological Issues in the Tracking of Commercial Property Values: Some Observations Based on the U.S.

Transcription

1 Of Granularity and Cocktails Methodological Issues in the Tracking of Commercial Property Values: Some Observations Based on the U.S. Experience David Geltner, PhD, F.RICS Massachusetts Institute of Technology Remarks for the ECB/Eurostat Conference on Commercial Property Price Indicators May 10-11, 2012 Frankfurt

2 Outline: 1. Evolution of commercial property price indexing in the U.S. From appraisal-based; To transactions price-based; And stock market-based. 2. What have we learned from the new transactionsbased indices: Need to track dynamic segmentation in markets Importance of index granularity Complementarity of different types of indices 3. Recommendation: Triangulation (or use of cocktails of indices). 2 2

3 1999 Value = 1000 First: NCREIF Property Index (NPI), since 1982, appraisal-based 12/31/1999 9/8/2000 5/18/2001 1/25/ /4/2002 6/13/2003 2/20/ /29/2004 7/8/2005 3/17/ /24/2006 8/3/2007 4/11/ /19/2008 8/28/2009 5/7/2010 1/14/2011 9/23/ U.S. Institutional Commercial Property Composite Capital Value, : Various types of indices... NCREIF NPI (appraisal-based) (X-axis is all weekdays; gaps in data are stock market holidays.) Composites are value-weighted. Quarterly Based on small (but important) population (pension funds), law requires regular re-appraisal. (U.S. GAAP does not require appraisals.) Tracks same-property valuations & total returns. Appraisals lag & smooth market values, but widely accepted & used 3 by industry.

4 1999 Value = /31/1999 9/8/2000 5/18/ : NCREIF-based TBI, transaction based 1/25/ /4/2002 6/13/2003 2/20/ /29/2004 7/8/2005 3/17/ /24/2006 8/3/2007 4/11/ /19/2008 8/28/2009 5/7/2010 1/14/2011 9/23/ U.S. Institutional Commercial Property Composite Capital Value, : Various types of indices... NCREIF NPI (appraisal-based) NCREIF TBI (transaction based) (X-axis is all weekdays; gaps in data are stock market holidays.) Composites are value-weighted. Quarterly Originally developed at MIT Center for Real Estate, now produced by NCREIF, based on same population as NPI, but transaction prices of sold properties. Hedonic price model approximated by ratio: TBI(t) = (P(t)/A(t))NPI(t). 4

5 1999 Value = : Moody s/real CPPI, repeat-sales index (transactions based) 12/31/1999 9/8/2000 5/18/2001 1/25/ /4/2002 6/13/2003 2/20/ /29/2004 7/8/2005 3/17/ /24/2006 8/3/2007 4/11/ /19/2008 8/28/2009 5/7/2010 1/14/2011 9/23/ U.S. Institutional Commercial Property Composite Capital Value, : Various types of indices... NCREIF NPI (appraisal-based) NCREIF TBI (transaction based) Moodys/REAL CPPI (X-axis is all weekdays; gaps in data are stock market holidays.) Composites are value-weighted. Quarterly 1 st CRE index using repeat-sales methodology. Based on larger population of properties than NCREIF: Real Capital Analytics database of all sales > $2,500,000. Tracks same-property prices (not total returns). Similar method 5 to Case-Shiller.

6 1999 Value = : FTSE/NAREIT PureProperty Indices, stock mkt based 12/31/1999 9/8/2000 5/18/2001 1/25/ /4/2002 6/13/2003 2/20/ /29/2004 7/8/2005 3/17/ /24/2006 8/3/2007 4/11/ /19/2008 8/28/2009 5/7/2010 1/14/2011 9/23/ U.S. Institutional Commercial Property Composite Capital Value, : Various types of indices... NCREIF NPI (appraisal-based) NCREIF TBI (transaction based) Moodys/REAL CPPI FTSE/NAREIT PureProperty (REIT-based) (X-axis is all weekdays; gaps in data are stock market holidays.) Composites are value-weighted. PureProperty Daily, Others Quarterly Daily-updated index, based on de-geared REIT share prices. REITs are pure plays in commercial property. Index statistically infers movements in underlying property assets implied by movements in REIT share prices. Uses information efficiency and liquidity 6 of stock market.

7 Example of Market Segmentation: Large, non-distressed properties in major mkts, vs distressed properties Pegged to CPPI (based on Dec 2000 = 100) Oct-07 Jan-08 Apr-08 Jul-08 Oct-08 Jan-09 Apr-09 Jul-09 Oct-09 Jan-10 Apr-10 Jul-10 Oct-10 Jan-11 Apr-11 Jul-11 Oct Oct07-09Bottom: 6-city Trophy -42% Distressed -58% Since 09Bottom: 6-city Trophy +57% Distressed +2% "Trophies" vs "Trash", Since the Crash... Since Oct 2007: 6-city Trophy -9% Distressed -57% RCA 6-city Trophy Trophy = >$10M, Non-Distressed, NY,DC,SF,LA,Chi, Bos only. Distressed Source: GA LLC, Based on RCA repeat-sales database, Feb Monthly same-property repeat-sales transactions price based indices based on the Real Capital Analytics database of all sales > $2,500,000. 7

8 Market segmentation Need for index granularity Example of Transaction Price Index Granularity: Moody's/RCA CPPI 10 National Segments All pegged to Dec 2000 = Data source: RCA repeat-sales Mar2012 database MajorMktsApts OtherMktsApts MajorMktsIndust OtherMktsIndust MajorMktsCBDoffice OtherMktsCBDoffice MajorMktsSuburbanOffice OtherMktsSuburbanOffice MajorMktsRetail OtherMktsRetail

9 Market segmentation Need for index granularity Example of Transaction Price Index Granularity: CPPI Major Mkts vs Other Mkts Apts All pegged to Dec 2000 = MajorMktsApts OtherMktsApts Data source: RCA repeat-sales Mar2012 database. 9

10 Market segmentation Need for index granularity Example of Transaction Price Index Granularity: CPPI CBD vs Suburban Major Mkts Office All pegged to Dec 2000 = MajorMktsCBDoffice MajorMktsSuburbanOffice Data source: RCA repeat-sales Mar2012 database. 10

11 Market segmentation Need for index granularity 3 Chicago-APT DC-APT NYC-APT DALLAS-APT SouthFL-APT '01 '02 '03 '04 '05 '06 '07 '08 '09 '10 '11 '12 Metro-level Apartment Properties Price Indices Monthly same-property repeat-sales transactions price based indices based on the Real Capital Analytics database of all sales > $2,500,

12 Different types of indices have different strengths & weaknesses (& this balance varies by country) Index Type: Strengths Weaknesses Appraisal-based Transactions-based Stock Mkt-based Can be available when others not Strong profession & tradition in some countries Actual prices directly reflect mkt equilibrium Objective info, less susceptible to manipulation REITs (or property pure plays ) traded in many countries Uses information efficiency & liquidity of stock mkt Leading indicator, daily updates Therefore Opinions not actual prices Tend to lag & smooth market values Can be subject to influence Requires large historical database Statistical models Some countries have few REITs, or short history, or thin market Information only indirect about actual property mkt 12

13 Suggested method for tracking commercial property prices: Triangulation Stock Mkt Appraisals Transaction Prices 13

14 Or, if you prefer A cocktail Stock Mkt Based Index Transaction Prices Based Index Appraisal Based Index 14

15 (We can debate which ingredient corresponds to which index ) GIN Stock Mkt Based Index VERMOUTH Transaction Prices Based Index The OLIVE Appraisal Based Index 15

16 Similar to the approach in political analysis, where meta-data (averages across different polls) provides better predictions than any one poll 16

17 Summary: 1. Evolution of commercial property price indexing in the U.S. From appraisal-based; To transactions price-based; And stock market-based. 2. What have we learned from the new transactionsbased indices: Need to track dynamic segmentation in markets Importance of index granularity Complementarity of different types of indices 3. Recommendation: Triangulation (or use of cocktails of indices)