Online Appendix for: Is the Behavior of Sellers with Expected Gains and Losses Relevant to Cycles in House Prices?

Abstract

This is the Online Appendix for: Is the Behavior of Sellers with Expected Gains and Losses Relevant to Cycles in House Prices?

Appendix 1 provides support for using assessed value to mitigate unobserved quality.
Appendix 2 describes the calculation of normalized assessed value (NAV) and summarizes results of leave-one-out (LOO) cross-validation.
Appendix 3 summarizes the analytical framework.
Appendix 4 displays our sample construction process.
Appendix 5 reports the first-stage hedonic regression results.
Appendix 6 shows the comparison between repeat sales index (based on repeat sales pairs), loss index and gain index.
Appendix 7 provides additional summary statistics that supplement Table 2.
Appendix 8 summarizes double results using asking price.
Appendix 9 summarizes results using an alternative house price cycle.
Appendix 10 reports additional robustness tests.
• A10.1 report robustness tests with various combinations of spatial and temporal fixed effects.
• A10.2 report robustness tests by restricting to the repeat sale sample only.
• A10.3 report results using asking price as the dependent variable.
• A10.4 reports robustness tests by adding the loan-to-value ratio.
Appendix 11 summarizes Clapp and Zhou (2019)’s method for correcting unobserved quality and report results using their method.
Appendix 12 presents and discusses contrast-relative analyses holding loss/gain coefficients constant.

Keywords: Housing Cycles, Anchoring, Reference Dependence, Price-volume Relationship, Cyclical Turning Points

JEL Classification: E00, G11, N2, R3