Price-volume Correlation in the Housing Market: Causality and Co-movements (P.19)

Decomposing the Price–volume Correlation
This section analyzes which relations—the Granger causality between prices and
trading volume or the effects of exogenous variables—help explain the price–
volume correlation in housing markets. We first calculate the raw price–volume
correlation for each MSA using the series of home appreciation rates and changes in
trading volume. Then, based on results from estimating the panel VAR model, we
decompose the home appreciation rates and trading volume (both demeaned due to
the within transformation) into fitted components and residuals.

We then calculate the correlation between      and      , as well as the correlation
between ui,t and vi,t, which is the component of the price–volume correlation that
cannot be explained by our model.
Panel A in Table 6 reports the across-MSA averages of the raw price–volume
correlations, the “fitted” correlations, and the correlations between residuals, using

Next Page 20



Leave a Reply

You must be logged in to post a comment.