A Simple Test to Identify the Boom and Bust of Housing Bubbles

The recent housing boom and bust in the United States was marked by large differences in the run-up and the subsequent decline of the housing prices both across metro areas and across market segments in the same area. One common observation in many metro areas is that the low-tier S&P Case-Shiller price indices realized the largest gains during the boom and saw the sharpest decline during the bust of the market.In this paper we propose a new time-series empirical test to identify housing bubble periods. Our test determines the dates of origin and burst of bubbles without market fundamentals: bubble periods are identified by the misalignment in the movements of the tiered price indices over time. Our empirical analysis is based on a theoretical model that explains why during a bubble prices of lower-tier houses tend to rise faster than prices of high-tier houses. For quite general specifications of the households’ utility function we show that, when given unrestricted access to credit, low income households take mortgages that constitute a larger share of their future expected income, pay higher interest rates, and are more likely to default given that their future income is uncertain. The main consequence of this household behavior is that the low tier price increases at a faster pace when households borrow funds to finance their home purchases and declines more sharply when lower income households start defaulting. We implement our test on 15 U.S. Metropolitan Statistical Areas during the most recent housing bubble.