It is essential to take the inter-temporal structure of commercial real estate returns into account in the asset allocation decision. The trends in real estate are significant, and they affect not only forecasts of future risk and return, but the estimation of covariance between real estate, financial assets, and institutional liabilities.

We use a Factor Augmented Vector Autoregression (FAVAR) to capture the relationship between property returns, asset returns, and macroeconomic variables including inflation and unemployment. The general idea of the model is to augment the VAR structure with additional economic variables that are important covariates of the series of primary interest. We use the augmented VAR estimation models to forecast distributions of assets and liabilities. These in turn can be used to quantify the distributions of portfolio returns given various percentages in real estate, stocks, and bonds.

The ultimate intent of the project is to understand the role of real estate in the investment portfolio. We wish to answer a few basic questions such as whether real estate serves as a hedge against inflation and wage increases, and whether it is useful as a diversifying asset.