Appraisals, transaction incentives, and smoothing

This article is structured around three principal objectives. The first is to determine whether any incentives for appraisals support an underlying purchase offer, which may be termed a transaction bias. Appraisals that are lower than purchase prices could involve additional cost for justification and thus undermine the transaction. The second objective is to test whether appraisal data are smoothed or exhibit less volatility than purchase data. The article compares the volatility of separate appraisal and purchase data. Given separate appraisal and purchase time series, the third objective is to derive the implied optimal appraisal updating rule. The model is applied to appraisal and purchase price indices for 3.7 million repeat transactions on mortgages bought by Fannie Mae and Freddie Mac by using monthly data from January 1975 to December 1993. The estimation procedure uses generalized autoregressive conditioned heteroskedastic (GARCH) analysis to take account of persistence in means and volatility in the house price time series. The article draws three principal conclusions. First, appraisals are systematically higher than purchase data, a first-moment differential. Second, appraisal smoothing does not occur generally. Third, the appraisal updating rule for the United States appears to involve error correction whereby underappraisals from pervious periods are eventually adjusted.