No abstract

This article presents a methodology for producing a quarterly transactions-based index (TBI) of property-level investment performance for U.S. institutional real estate. Indices are presented for investment periodic total returns and capital appreciation (or price-changes) for the major property types included in the NCREIF Property Index. These indices are based on transaction prices to avoid appraisal-based sources of index “smoothing” and lagging bias. In addition to producing variable-liquidity indices, this approach employs the Fisher-Gatzlaff-Geltner-Haurin (Real Estate Econ., 31: 269–303, 2003) methodology to produce separate indices tracking movements on the demand and supply sides of the investment market, including a “constant-liquidity” (demand side) index. Extensions of Bayesian noise filtering techniques developed by Gatzlaff and Geltner (Real Estate Finance, 15: 7–22, 1998) and Geltner and Goetzmann (J. Real Estate Finance Econ., 21: 5–21, 2000) are employed to allow development of quarterly frequency, market segment specific indices. The hedonic price model used in the indices is based on an extension of the Clapp and Giacotto (J. Am. Stat. Assoc., 87: 300–306, 1992) “assessed value method,” using a NCREIF-reported recent appraised value of each transacting property as the composite “hedonic” variable, thus allowing time-dummy coefficients to represent the difference each period between the (lagged) appraisals and the transaction prices. The index could also be used to produce a mass appraisal of the NCREIF property database each quarter, a byproduct of which would be the ability to provide transactions price based “automated valuation model” estimates of property value for each NCREIF property each quarter. Detailed results are available at http://web.mit.edu/cre/research/credl/tbi.html . Copyright Springer Science+Business Media, LLC 2007Transaction-based index, Commercial real estate investment performance, NCREIF database, Commercial property type returns, Constant liquidity index, Ridge regression techniques,

No abstract

This paper compares housing price indices estimated using three models with several sets of property transaction data. The commonly used hedonic price model suffers from potential specification bias and inefficiency, while the weighted repeat-sales model presents potentially more serious bias and inefficiency problems. A hybrid model combining hedonic and repeat-sales equations avoids most of these sources of bias and inefficiency. This paper evaluates the performance of each type of model using a particularly rich local housing market database. The results, though ambiguous, appear to confirm the problems with the repeat sales model but suggest that systematic differences between repeat-transacting and single-transacting properties lead to bias in the hedonic and hybrid models as well. Copyright American Real Estate and Urban Economics Association.

No abstract