IMPERFECT INFORMATION AND INVESTOR INFERENCES FROM HOUSING PRICE DYNAMICS

We examine characteristics of housing price dynamics that may be consi stent with rational learning and not simply irrational feedback tradin g. We find significant patterns of temporal and spatial diffusion that are more amenable to explanations that allow for rational components. First, we execute our tests not simply on housing price changes, but on town-by-town differentials from regional average price changes. Sec ond, we find significant relationships with own and neighboring town d ifferentials, but not with control groups of non-neighboring towns. Th ird, we find that population density, a proxy for scale economies in i nformation production, accelerates the diffusion process. Test were pe rformed on quarterly data for large samples from Connecticut and the S an Francisco area, employing method of moments estimators.