The Estimation of Quality-Adjusted Auction Returns with Varying Transaction Intervals

Abstract

Previous research has separately addressed the problem of estimating risk in the presence of infrequent trading and the problem of estimating quality-adjusted returns in markets with quality variation in the observed price series. This paper simultaneously addresses both problems by applying a signal extraction method for unequally spaced data to decompose the observed price series with varying times between transactions into a quality-adjusted, permanent component (which would be observable in the absence of quality variation) plus a stationary, transitory quality variation component. Stamp auction transaction prices provide an application. Auction quality grading is treated in a manner analogous to bond ratings. Almost all of the observed variance is attributed to the auction quality variation. The observed auction returns and stock index returns are not well related.