To help the StreetEasy community make sense of the New York City market, the StreetEasy Research team tracks how sale prices and asking rents change over time in addition to evaluating the state of the market at a given point in time in our monthly and quarterly market reports. To estimate the change in sale price or asking rent for a given home over time, we developed the StreetEasy Price and Rent Indices. We updated the methodology and coverage of these indices as of July 2017, though the current methodology remains similar to how we previously produced these metrics.
The StreetEasy Price and Rent Indices have a key advantage over other popular market metrics such as median sale price and median asking rent: the ability for our indices to control for unit quality over time. The composition of the New York City sales or rental market may vary substantially month to month, with differing proportions of units of a given size, location, or bedroom count available at any specific time. Unlike median sale price or median asking rent, our indices control for this variability by looking at the change in sale price or asking rent for each individual unit over time. Combining the change in prices of several units over time gives us a more complete picture of how prices for a given area or segment of the market change.
The StreetEasy Price Indices cover home appreciation in Manhattan from 1995 to present, and home appreciation in Brooklyn and Queens from 2005 to present. We estimate indices at the borough-wide level as well as for geographic submarkets and price tiers within each borough. Index calculations include data on condos, co-ops, townhouses and single-family homes, but exclude sales for a nominal fee that do not reflect arms-length transactions. We source the underlying data for our price indices from the New York City Department of Finance, which publicly releases data on all closings in New York City. As a result, the population of transactions on which we base our indices does not depend on the population of listings on StreetEasy.
The StreetEasy Rent Indices cover changes in asking rents in Manhattan from 2007 to present, and changes in asking rents in Brooklyn and Queens from 2010 to present. As in the case of our price indices, we estimate rent indices at the borough-wide level as well as for geographic submarkets and price tiers within each borough.[1] All data included in the calculation of our rent indices comes from the full history of listings submitted to StreetEasy by brokerages, property managers, and individual landlords. We calculate the rent indices based upon the last advertised asking rent posted on StreetEasy for a given listing; these rents typically reflect the monthly rental rate net effective of concessions included in the lease.
We set the dollar levels of the StreetEasy Price and Rent Indices such that December 2016 values equal the 2016 median recorded sales price and 2016 median asking rent, respectively. Dollar denomination allows for direct and interpretable comparisons of the indices across time, market segment, and geography. We plan to re-index all indices as needed at future dates to maintain relevant benchmark price levels.
We calculate our Price Indices using a repeat sales methodology, a commonly used methodology to estimate changes in home prices over time. The repeat sales methodology models the price change for all units with more than one recorded sale by regressing changes in home prices on the change in time between two transactions. The coefficients resulting from this regression form the basis for month-on-month changes reflected by the index values. The methodology was introduced in 1963 by Martin Bailey, Richard Muth, and Hugh Nourse, and further developed by Karl Case and Robert Shiller in the late 1980s. Our approach is similar to the methodology underlying both the Federal Housing Finance Agency (FHFA) House Price Index and the S&P CoreLogic Case-Shiller Home Price Indices. To ensure the interpretability of our index, we use an ARIMA model to smooth the result of our index time series.[2]
To calculate the StreetEasy Rent Indices, we adapt the repeat sales methodology to the rental market by substituting changes in asking rents for changes in sale prices. In all other aspects, the methodology we employ for our rent indices mirrors the repeat sales methodology used in our price indices. Though developed independently, this methodology mirrors the approach described by Ambrose, Coulson and Yoshida.
M. J. Bailey, R. F. Muth and H. O. Nourse, “A Regression Method for Real Estate Price Index Construction,” Journal of the American Statistical Association, vol. 58, pp. 933-942, 1963.
K. E. Case and R. J. Shiller, “The Efficiency of the Market for Single Family Homes,” American Economic Review, vol. 79, pp. 125-137, 1989.
B. W. Ambrose, N. E. Coulson and J. Yoshida, “The Repeat Rent Index,” Review of Economics and Statistics, vol. 91, no. 5, pp. 939-950.
[1] For rent indices, we exclude areas which do not have a substantial population of listings on StreetEasy. As of July 2017, this applied to the Queens submarkets of South Queens and the Rockaways.
[2] Indices published before July 2017 used a different smoothing algorithm and may differ substantially from those published since. For more details on our methodologies used prior to July 2017, see streeteasy.com/blog/methodology-streeteasy-price-indices and streeteasy.com/blog/methodology-streeteasy-rent-indices.