Between 1996 and 2006 the U.S. has experienced an unprecedented boom in house prices. As it has proven to be difficult to explain the large price increase by observable fundamentals, many observers have emphasized the role of speculation, i.e. expectations about future price developments. The argument is, however, often indirect: speculation is treated as a deviation from a benchmark. The present paper aims to identify house price expectation shocks directly. To that purpose, we estimate a VAR model for the U.S. and use sign restrictions to identify house price expectation, housing supply, housing demand, and mortgage rate shocks. House price expectation shocks are the most important driver of the boom and account for about 30 percent of the real house price increase. We also construct a model-based measure of exogenous changes in price expectations and show that this measure leads a survey-based measure of changes in house price expectations. Our main identification scheme leaves open whether expectation shifts are realistic or unrealistic. In extensions, we provide evidence that price expectation shifts during the boom were primarily unrealistic and were only marginally affected by realistic expectations about future fundamentals.

A number of observers have suggested that shifts in house price expectations have been an important driver of the US house price boom that preceded the financial crisis. Arguments in favor of this hypothesis have either been indirect by using deviations from benchmark models or have relied on survey measures of house price expectations. The present paper proposes an empirical strategy to quantify the contribution of price expectation shocks directly using observed housing variables. It uses a structural VAR framework and identifies price expectation shocks with sign restrictions. We argue that if there is an exogenous increase in expectations about future house prices, economic theory delivers a number of predictions on the behavior of variables, such as current house prices, residential investment and vacancy rates that can be used to distinguish the price expectation shock from traditional shocks such as mortgage rate, demand for housing services, and housing supply shocks.