Daily appraisal of commercial real estate a new mixed frequency approach

Abstract

We present a mixed frequency repeat sales model for commercial real estate, taking into account changes in net operating income between the date of buying and selling the property. Moreover, we relate monthly private market index asset returns to lags, up to 1 year, of daily (REIT) index returns. The underlying daily model enables us to interpolate the monthly private market index returns on a daily basis, and to predict the private market index asset returns going forward. The mixed frequency repeat sales model facilitates daily appraisal of commercial real estate portfolios. We apply the model on sale prices (all property types, and apartments only) in the period January 2006-July 2020. We find that the mixed frequency repeat sales model reduces 1-monthahead forecasts errors and index revisions, compared to a benchmark model without daily returns.