Investment Returns and Risk for Art: Evidence from Auctions of American Paintings

Numerous economic studies have attempted to empirically document the profitability of investment in various collectibles including coins, violins, wine, antique furniture, and paintings. With profitability information, the collector will have some idea of the expected consumption cost of collecting (the residual between alternative financial returns and those of the collectible), and thus be in a better position to make a purchase decision. This study contributes to this economic literature by focusing on a rather expensive collectible, paintings by well-known U.S. artists. This paper extends previous analyses of returns and adds results on risk, a topic receiving less attention in earlier studies. The empirical analysis is based on data from over 25, 000 paintings by U.S. artists sold at auction from 1971 to 1996. Following Chanel, et al. [1996] I use a hedonic log price model. Dummy variables that reflect temporal, spatial, and characteristic variations allow the simultaneous estimation of price indices for paintings and shadow values for spatial and other characteristics. The analysis is disaggregated and focuses on individual artists, genres and quality levels. The findings show significant sensitivity of both returns and risk to the particular segment of the painting market. When and what one invests in matters a great deal just as it does in traditional financial markets. Overall returns are lower than equity markets and risk is greater although high-end paintings do well. The early 1990s were quite bearish although some painting sectors recovered significantly by 1996. Art may play a potential diversification role for the collector/investor even though returns and risk alone are not generally appealing.