Comparison of Different Methods for Measuring Hedonic Price Indexes

This study develops a new method for constructing a price index for Russian art as a whole and different Russian artists separately. With our special approach researchers can construct only one hedonic regression model for all artists without including dummy variables thus increasing the sample size and producing a higher R-squared value. In this paper, we use characteristics such as the year the painting was sold, the canvas size and the type of substrate, the medium of execution, the lot number and the name of auction house. Using a unique sample of 13308 auction records for Russian artworks created by 30 artists sold over the period 1986 to 2018 we will compare our indices with financial indices such as MOEX-Russia index and global art index ARTIMX. The results will show whether art market is more or less risky than financial market and whether these paintings can potentially be used for portfolio diversification.