- Christophe Spaenjers
Statistical measure showing the development of art prices since a chosen base year. Index series are often represented as graphs, and allow for a comparison with the performance of other assets. An index also enables the measurement of the correlation of art returns with changes in valuations of other investments. Two techniques are commonly used to construct an art price index based on auction transaction data. First, so-called ‘hedonic’ methods use all available sales information to measure changes in quality-adjusted average transaction prices. Second, ‘repeat-sales’ regression models only use price information on artworks for which at least two transactions are observed to estimate the average return in each period.
Measuring the returns to art Investments is not methodologically straightforward. While for a publicly traded financial asset (e.g. a stock in a large company) a price can typically be observed on any given day, in the Art market each item is unique and trades only very infrequently. Ideally, an art index would track the total monetary value of a representative portfolio of objects over time, but this is not possible as we do not observe prices for the same set of artworks in every period. For this reason, even an index based on unadjusted average prices will not accurately capture changes in the willingness to pay for art over time. For example, even if the average price of all transacted artworks is twice as large in one period compared to the previous one, this does not mean that the typical item has doubled in value; it could be that in the second period there were more transactions of relatively more attractive works....