It has long been understood that ownership of works of art isn’t just a collection or a hobby, or a way to brighten up the walls of one’s home. It’s a portfolio decision.
But it is a portfolio decision that involves non-fungible and very infrequently traded assets. So: how does one estimate values without actually liquidating, and how does one measure the performance of different classifications of art (impressionist versus contemporary) against one another over time? How exactly does one go about treating art as an investment decision?
Those are the questions behind a new paper from Fabian Y.P. Bocart, Eric Ghysels, and Christian M. Hafner. Bocart is with Artnet, the art market website run out of New York. Ghysels is with the Department of Economics, University of North Caroilina. Hafner is with the Universite Catholique de Louvain, in Belgium.
They use a new econometric approach to the issue of the broad movement of the art market, invoking a dynamic state space model. In statistics, such a model is a graphic representation of variables as a set of first-order ordinary differential equations. It turns out that the mathematical/ statistical trickery involved allows for updates to the art index even in periods of no transactions “via observed transactions in other markets that have a link with the art market.”
This author is a hopeless amateur on the mathematics, and can only recommend a careful reading of the paper by those qualified to understand its modeling/indexing approach. The authors make quite general claims for their method. They believe it can be employed in a number of other markets where, as with paintings, “trading occurs infrequently during periodically scheduled auctions,” as with rare coins, stamps, wines, etc.
Bocart et al compute five price indexes. Four of them speak to classes of art by period: impressionist, modern, post-war, and contemporary. The fifth merges these into a global art market index.
Their indexes involve the work of a total of 200 artists (50 associated with each of the four periods just listed) whose works have been sold at auction between January 2002 and September 2017. The 200 artists represent 3,059 works of art that were sold at auction during those years for a total consideration of $43.9 billion.
Having these indexes “at the monthly frequency” as these authors say, “brings us in line with the more traditional asset pricing literature.”
For purposes of comparisons, Bocart et al also constructed an equally weighted basket of art related companies (Sotheby’s, Artnet, Artprice, and Collector Universe) and an equally weighted basket of luxury companies (Dior, Moët, etc.). Both of these baskets were constructed with the data available from Yahoo!Finance.
As one might expect, “all art movements positively correlate with the monthly stock returns of art companies.” But the art index “seems to correlate negatively with those of luxury companies,” which, these authors say, seems a bit counter intuitive.
The art indexes show no correlation, negative or positive, to the West Texas Intermediate oil price, U.S. Treasury bonds, or the returns of the iShares U.S. consumer goods ETF.
The indexes thus compiled indicate that throughout the period 2002 to 2017, the contemporary art index and the post war art index, which stayed close to one another, easily outperformed the other art indexes. The impressionist index was the least impressive. It merely held its value through this period, making some gains in the period 2002-2007 but then losing them in the global financial crisis and remaining flat since 2010.
Modern art did somewhat better than impressionism, but not much. It rose further in the pre-crisis period and then gave up some though not all of its gains – and it, too, has been flat since 2010.
Perhaps more important for non-specialists: art works do not appear attractive as an investment when compared to stocks and bonds. Contemporary art has worked best as a portfolio decision, it “appears to perform almost as well as the S&P 500” [emphasis added] with regard to volatility and the Sharpe ratio. Contemporary and Post-War art performed better than real estate, again with regard to the Sharpe ratio. The Art and Luxury companies did better than any of the art indexes, though.