Research projects

Project CEMAPRE internal

TitleForecasting Under Real Estate Bubbles
ParticipantsDaniel Esteves dos Reis, Nuno Sobreira (Principal Investigator)
SummaryThe objective is to evaluate the predictive accuracy of linear and
nonlinear time series models in the real estate market context, a segment of the
economy that witnessed periods of sharp increases and falls in prices, that is, bubbles. In our
understanding, a bubble period characterizes a significant change in
the behavior of the economic agents and can be interpreted as a structural break in
the path of an observed time series. We believe that nonlinear models can provide
more flexible characterizations of the DGP to incorporate such observed changes
in series of the real estate market which have experienced periods of bubbles. In
order to date-stamping the beginning and the end of the real estate bubble, we
use PWY and PSY procedures (Phillips et al., 2009 and Phillips et al., 2015b) on
the monthly time series of the price-rent ratio for the United States and for seven
Metropolitan Statistical Areas (MSAs) during the period between 1987 and 2019.
Our expectation is that, in this context, the forecasting accuracy statistics favor
the nonlinear models under competition.