Blessing of dimensionality in forecasting real house price growth in the nine census divisions of the US: 57th International Statistical Institute Conference 2009

Abstract

This presentation deals with forecasting real house price growth in the nine census divisions of the US. Data rich environment of the FAVAR models that include a wide range of macroeconomic series of the US economy, besides the house price growth rates of the census divisions, are more informative when forecasting the house price growth rate of the nine census regions. The role of fundamentals in affecting the housing market cannot be underestimated. What is needed first is a detailed look at the factors to determine the dominant macroeconomic variables that comprise these factors, and secondly to incorporate the role of the house price growth rate of the neighboring division(s) in the forecasting process of a particular census division, by developing spatial versions of the various models.