Focus on U.S. Real Estate Benchmarks: Moody’s/RCA CPPI and CoStar CCRSI

Brad Case smallBy Brad Case, Ph.D., CFA, CAIA

A variety of indices are published regularly and may be appropriate for benchmarking, risk assessment, and other real estate investment purposes. I plan to focus on the strengths of each index series, starting with two that share similar approaches: the Moody’s/RCA Commercial Property Price Index (CPPI) and the CoStar Commercial Repeat-Sales Index (CCRSI). Both are published monthly based on actual property transactions; both measure average price appreciation at the property level; and both are available for the aggregate U.S. commercial property market as well as for important market segments.

The CPPI is published jointly by Moody’s Investors Service and Real Capital Analytics (RCA), which tracks all transactions of properties valued at more than $2.5 million, while the CCRSI published by CoStar is based on an even broader set of properties. Both index series are based on a version of the repeat-transactions methodology popularized by Case and Shiller: basically, two transactions of the same property show how much the value of that property changed during the time between the two transaction dates, and “averaging” the changes across a large number of transaction-pairs (with a variety of first-sale dates and second-sale dates) enables us to estimate how much of the change in overall property values happened during each month. The CPPI is based on a methodology originally described in a paper available here, while the CCRSI methodology is detailed here.

What They Measure: The CPPI and CCRSI measure the implied monthly change in market value for a property that has had “normal” levels of both economic depreciation and capital reinvestment. They do not measure income or total returns at the property level, nor do they measure returns to investors on property investments in which any debt is used.

Reporting Schedule and Access: Both indices are published roughly 45 days after the end of each month, and both can be accessed freely: the CPPI from here, and the CCRSI from here.

Geography:

  • The publicly accessible CPPI is available for the entire U.S. and separately for “major markets” (Boston, Chicago, Los Angeles, New York, San Francisco, and Washington D.C.) and “non-major markets” (all others). In addition, office property indexes are available for the entire U.S. and separately for central business district (CBD) and suburban areas.
  • Real Capital Analytics separately publishes indices for five regions (Northeast/Mid-Atlantic, Midwest, Southeast, Southwest, and West), 20 states, and 34 metro areas. The RCA CPPI index series, which is available only to subscribers, is described here.
  • The publicly accessible CCRSI is available for the entire U.S.; separately by region (Midwest, Northeast, South, and West); and separately for “prime market” and “non-prime markets” for the multi-family, office, retail, and industrial property types. (Prime markets are defined in each CCRSI release.)

Property Types:

  • The publicly accessible CPPI is available for Apartment, Industrial, Office, and Retail property types, as well as a “core commercial” aggregate of industrial, office, and retail properties. The Hotel property type is available to subscribers through the separate RCA CPPI index series.
  • The publicly accessible CCRSI is available for Hospitality, Industrial, Land, Multifamily, Office, and Retail property types, as well as a “composite excluding multifamily” aggregate.

Geography/Type Combinations:

  • The publicly accessible CCRSI is available for 16 combinations of region (Midwest, Northeast, South, and West) with property type (Industrial, Multifamily, Office, and Retail).
  • The publicly accessible CPPI is available on request for 10 “building block” combinations of market category (“major” or “non-major”) with property type (Apartment, Industrial, CBD Office, Suburban Office, Retail), while the subscription-based RCA CPPI index series includes multiple additional combinations.

Weighting:

  • All properties are weighted equally in computing the 10 “building block” indices for the CPPI; the national aggregate index, as well as the region and property type indices, are then constructed from the “building block” indices using dollar value of transaction volumes as weights.
  • Both value-weighted and equal-weighted versions of the CCRSI are available, except for the Hospitality and Land property types which are available only in equal-weighted versions.

Advantages: Both the CPPI and the CCRSI are published monthly rather than quarterly, and both are based on actual transaction prices rather than appraisals. This means that both indices produce substantially more accurate measurements of movements in property values over time compared to appraisal-based indices, whether at the property level (such as the NCREIF Property Index) or at the fund level (such as the NCREIF ODCE Index, the PREA|IPD U.S. Quarterly Property Fund Index, or the Cambridge Associates Real Estate Index). In addition, both cover substantially the entire commercial property market in the U.S.

Disadvantages:

  • Both the CPPI and the CCRSI are significantly delayed relative to actual movements in property values, with both published at least six weeks following the end of the month being measured.
  • Both the CPPI and the CCRSI use only property transactions that have already been completed. Because it takes a long time to complete a transaction even after the buyer and seller have agreed on the transaction price, both indices suffer from illiquidity lag, which means that movements in property values tend to be measured roughly six months (on average) after they actually occurred.
  • Because of the small number of properties that transact in any given month—and the even smaller number of transacting properties for which RCA and CoStar also have data on a previous transaction—both the CPPI and the CCRSI are based on relatively small data sets and neither measures movements for any period less than a month.
  • Because transactions completed at any time during a given month must aggregated together to have enough data to compute the indices, both the CPPI and the CCRSI suffer from illiquidity smoothing (also called “aggregation bias”), which makes property values appear less volatile than they truly are.
  • Repeat-transacting properties may not be representative of properties that did not transact, or that transacted less frequently, so both indices may be biased although the direction of the bias is not clear.
  • Both the CPPI and the CCRSI cover substantially the entire commercial property market in the U.S., but many institutional investors own portfolios that are concentrated in the higher-quality and more-expensive segments of the market; therefore, neither may be appropriate for benchmarking a higher-quality real estate portfolio.

Notwithstanding these disadvantages, both the CPPI and the CCRSI can be very valuable for benchmarking, risk assessment, and other investment purposes.

Brad Case is senior vice president, research & industry information for the National Association of Real Estate Investment Trusts (NAREIT).  Dr. Case has researched residential and commercial real estate markets, domestically and globally, for more than 25 years.  His research encompasses investment return characteristics including returns, volatilities, and correlations with other assets; measuring appreciation in property values; inflation protection; use of DCC-GARCH and Markov regime switching models to measure and predict investment characteristics; the length of the real estate market cycle; and the role of the investment horizon.  He holds patents as the co-inventor of the FTSE NAREIT PureProperty(r) index methodology and the backward-forward trading contract.  Dr. Case earned his Ph.D. in Economics at Yale University, where he worked with Robert Shiller and William Goetzmann, and holds the Certified Alternative Investment Analyst (CAIA) designation.

Be Sociable, Share!

5 Comments

  1. Bill Conerly
    December 2, 2014 at 3:17 pm

    Nice explanation. Could you comment on how these indices compare to NCREIF and NAREIT?
    thanks,
    Bill


  2. Arthur
    December 15, 2014 at 11:32 am

    Great post! I’d also like to hear your comments on how these indices compare to NCREIF and NAREIT?

    Arthur.


  3. Brad Case, Ph.D., CFA, CAIA
    January 25, 2015 at 10:03 am

    (Sigh…) I wrote a long and detailed answer to your questions, Bill and Arthur, but for some reason it didn’t publish. I’ll do it again, but only if you’re still interested. Should I?


  4. Brad Case
    January 27, 2015 at 2:32 pm

    Regarding the NCREIF Property Index (NPI), please check out my separate article on it. Here are the main comparisons between the NPI and both the CPPI and the CCRSI:
    (1) The NPI measures (gross) income and (gross) total return in addition to capital appreciation.
    (2) As detailed in my article on the NPI, the NPI measures capital appreciation–meaning that the NPI subtracts out capital expenditures–whereas the CPPI and CCRSI measure changes in property values (without subtracting capital expenditures). NCREIF publishes a separate index of price change, but you have to dig a little bit deeper for it; moreover, NCREIF recently announced a new Market Value Index (http://ncreif.org/research-single.aspx?post=30291) that will be more comparable to the CPPI and CCRSI.
    (3) The NPI is available for many more geographic areas than the CPPI or the CCRSI.
    (4) The NPI is quarterly, whereas the CPPI and CCRSI are monthly.
    (5) The NPI measures returns for a sample of roughly 7,000 properties owned (or partially owned) by institutional investors with fiduciary responsibility, whereas the CPPI and CCRSI attempt to measure essentially all properties. The NPI data base tends to be dominated by high-quality, large, high-value properties, whereas the CPPI and CCRSI data bases include a large number of lower-quality, smaller, lower-value properties.
    (6) The CPPI and CCRSI are based on transaction prices, whereas the NPI is based on appraisals. I can’t emphasize strongly enough the importance of this distinction, because any index based on appraisals will be much, much less accurate than an index based on actual transactions. In particular, while an index based on transactions of illiquid assets will be smoothed relative to actual market values of those assets, an appraisal-based index will be much, much more smoothed–which means that estimates of asset volatility based on appraisal-based indices will be much more severely biased downward relative to actual volatility. Similarly, an index based on transactions of illiquid assets will be lagged relative to movements in actual market values, but an appraisal-based index will be much more lagged: by 4 to 5 quarters for the NPI, compared to maybe 6 months for the CPPI and CCRSI. Combining smoothing and lag means that estimates of correlations based on appraisal-based indices will be much more severely biased downward relative to actual correlations. In short, appraisal-based indices will cause assets to look as though they have much better risk-adjusted returns, and much greater diversification benefits, than they really do.


  5. Brad Case
    January 27, 2015 at 2:32 pm

    With respect to the FTSE NAREIT index series, here are the main comparisons with the CPPI and CCRSI:
    (1) The FTSE NAREIT measures returns on the common stock of the REITs that own the properties, whereas the CPPI and CCRSI measure the properties themselves. That’s not as big a difference as it sounds, because in both cases we’re talking about a diversified portfolio, and the predominate factor that drives changes in the REIT stock index is the same as what drives in a property index such as the CPPI or CCRSI: changes in the aggregate value of a diversified portfolio of properties.
    (2) The FTSE NAREIT measures income and total return (of stock in the REITs that own the properties) in addition to price change, whereas the CPPI and CCRSI measure only price change (of the properties themselves). Keep in mind that REIT stock returns are net of general and administrative (G&A) expenses, so while both the FTSE NAREIT index and the NPI measure gross income and gross total return, in the case of REITs gross return is the same as net return except for investment management fees (such as mutual fund fees) whereas for the NPI net returns are about 1% per year less than gross return.
    (3) Because almost all REITs have geographically diversified property holdings, the FTSE NAREIT index series is not available for any geographic areas within the U.S.
    (4) The FTSE NAREIT index is available daily, and published immediately, whereas the CPPI and CCRSI are available only monthly and (like the NPI) published several weeks after the end of the month.
    (5) The properties owned by REITs (like those in the NPI data base) tend to be higher-quality, larger, and higher-value than those in the CPPI and CCRSI data bases.
    (6) The FTSE NAREIT is based on actual transactions in a fully liquid market, whereas the CPPI and CCRSI are based on actual transactions but in a very illiquid market. That means there is no smoothing or lag at all in the REIT index, whereas both the CPPI and the CCRSI suffer from lag and smoothing–just not nearly as much as in an appraisal-based index such as the NPI.
    I hope that’s helpful!


Leave A Reply

← Wars and Failed Mergers Make for a Tricky October The Economics of Happiness →