What Moves the Price-Rent Ratio for Housing? A Modified Present-Value Approach

This paper proposes a modified present-value model that takes into account the fact that movements in the price-rent ratio for housing may not be mean-reverting. Our approach decomposes the price-rent ratio into expected real rent growth, expected housing return and a non-present-value (NPV) component that represents the deviation of the price-rent ratio from its conventional present-value level for the 18 U.S. metropolitan areas and the nation from 1975 through 2012. This NPV component takes into account non-stationarity of the price-rent ratio. To estimate this modified present-value model, we use the unobserved component approach. Our findings suggest that the NPV component is significant and sometimes very large both at the national and the regional level. This is especially true for the MSAs that have experienced frequent booms and busts in the housing market. We also find that the MSAs that display larger deviation from the present-value model are more sensitive to mortgage rate changes. We also compare our results with a recent statistical test for periodically collapsing bubbles. The results from this test and for our model indicate that the MSAs that had large NPV components are also the MSAs that witnessed explosive sub-periods in their price-rent ratios, especially during the 2005-2007 subsample. Our approach also allows us to estimate the correlation between expected rent growth, expected housing return and the NPV component. We find that a shock to expected housing return and a shock to the non-stationary NPV component are highly positively correlated in the pre-2006 sample period, implying that they fed off of each other. This correlation declined significantly in the post-2006 sample period. Our results also show that most of the variation in the present-value component of the price-rent ratio arises due to the variation in expected housing return.