The Efficiency of the Market for Single-Family Homes: A Critique on Case and Shiller (1989)

Abstract

This paper comments on the Weighted Repeated Sales (WRS) method in Case and Shiller (1989). We find that Case-Shiller’s model for step-two of WRS is conceptually mis-specified and empirically inaccurate, which are likely to cause the S&P/Case-Shiller Home Price Indices to be biased for the most critical housing markets (i.e. nine of the ten cities with housing related derivatives based on the indices). Based on our examination of real estate market risks, we propose a quadratic model for the second step of WRS and show that our specification is actually more consistent with the findings by Case and Shiller (1989).