Two finance professors claim to have finally shown how selection bias skews the supposed returns of art through indices.

The selection bias arises when returns are based on indices built on repeat sales of fairly illiquid assets that are not sold at random. Many of the returns based on those kinds of indices—including the S&P/Case-Shiller Home Price Indices—may be biased upwards.

Not only are the returns of art lower than investors think, but also the risk is higher. Our analysis, of 20,538 paintings repeatedly sold between 1972 and 2010, found the Sharpe Ratio for art is 0.04, rather than the 0.24 that has been previously found.