The Investment Performance of Classical Swedish Painters

This paper studies the investment properties of paintings. It applies a hedonic regression model on a unique sample of more than 8,000 paintings by classical Swedish painters traded in the period 2010-2023. To address sample selection bias regarding sold and unsold paintings, it uses a Heckman correction. Despite some very high prices for paintings, the price index for the sample paintings underperforms well-used indexes for stocks, bonds, and real estate, while its volatility is much higher. The study shows that paintings of classical Swedish painters have a very limited role when constructing optimal investment portfolios.