Abstract
We use prices realized for Picasso prints at auctions worldwide, as well as the 100 prints that comprise his Vollard Suite, to test the law of one price: the proposition that identical art objects sold contemporaneously should command the same price regardless of the auction house or geographic region where the sale takes place. Picasso is the most prolific printmaker of the twentieth century and, from 1977 to 2004, his prints appreciated in price significantly faster than the prints of modern masters as a whole. We find that Picasso prints sold in the United States command higher prices than in Europe. However, prices realized at Sotheby’s in New York are no longer higher than at Christie’s in New York, nor at Kornfeld than at other auction houses. We find evidence of “irrational exuberance” in the transitory nature of the extraordinary prices realized for the Picasso prints included in the 1997 sale of the collection of Victor and Sally Ganz at Christie’s in New York. More generally, we find substantial noise in auction outcomes, a result well known to savvy auction goers.
This is a preview of subscription content, log in via an institution to check access.
Similar content being viewed by others
Notes
The set is named after the art publisher Ambroise Vollard, to whom Picasso traded the 100 copper plates in 1937.
A repeat sale occurs whenever the “identical” print (i.e. artist, catalogue, raisonné number, signed or not) is sold on two different occasions. For each pair of sales, the log-price relative is calculated: the log of the price on the later sales date less the log of the price on the earlier date. The log-price relatives are then regressed on a set of dummy variables, one for each observation of the log-price index. For each observation of the dependent variable, the dummy is set equal to +1 at the time of the second sale, −1 at the time of the initial sale and 0 at all other times. If the initial sale is in the first time period, there is no dummy variable corresponding to the initial sale. To address the fact that the errors in the repeat sales regression may increase with the time between repeat sales, Case-Shiller propose a three-stage procedure to lessen the weight assigned to repeat sales that are widely separated in time. First, the index is estimated using Ordinary Least Squares (OLS). Second, the squared residuals are regressed on a constant term and the time between sales. Third, the index is reestimated by generalized least squares (GLS) with each observation divided by the square root of the fitted value in the second stage. For a discussion of the issues related to the repeat-sale regression methodology, see Goetzmann and Peng (2002).
Due to the large number of repeat sales of prints, as well as the relatively short sample period, the results that we obtain with OLS and with the Case-Shiller method are very similar.
Biey and Zanola (2005) also investigate the returns to Picasso prints, using data for the period 1988–1995. The primary interest of Biey and Zanola is a comparison of the repeat sales, hedonic price model and a hybrid model as strategies for calculating a price index for art. However, it merits note that their results do mirror those obtained in Pesando and Shum (1999) and in our current article, but with a different data set. For example, using repeat sales, Biey and Zanola also find a sharp correction in 1990 to the bull market of the late 1980s.
These artists are: Chagall, Miro, Matisse, Whistler, Nolde, Heckel, Schmidt-Rottluf, Kirchner, Kandinsky, Klee, Mueller, Bellows, Benton, Hopper, Munch, Renoir, Sloan, Vuillard, Braque, Bonnard, Cassatt, Kollwitz, Laurencin, Leger, Rouault, Toulouse-Lautrec and Villon.
Our programme identifies the first sale of a particular print and then compares its price with the price of all sales in the “other” market within the next 29 days. If there are no such sales, the programme identifies the second sale of the print and repeats the procedure. After a comparison has been made, the programme identifies the first sale after this window has elapsed and then continues. If there are two or more sales of the print in the same market within the “window,” the average price is used in the price comparison.
Contrary to the “folklore” of the art trade, there was no evidence that the prices realized by certain artists (such as the German Expressionists) were systematically higher in certain geographic markets (such as Germany) than elsewhere (see Donson 1977).
In addition to the large difference in realized prices that may occur for different impressions of the same print, there are occasional errors or nuances in the data set itself. For example, in 1998, the price for Picasso (Bloch Catalogue No. 1635, from the 347 Series) was $6,325 at Christie’s in New York on April 28th and $145,500 at Sotheby’s in New York on April 30th. In fact, the price at Sotheby’s was for 66 different prints from the 347 Series, starting with Bloch Catalogue No. 1635. In 1999, a Picasso print (Bloch No. 859) sold for $266,500 at Christie’s in New York on May 3rd, but for only $134,500 at Sotheby’s in New York on April 29th. In fact, the impression sold at Sotheby’s, although having the same catalogue raisoneé number, was an edition proof that was signed on the reverse and not an “identical” print.
In order to eliminate catalogue errors and sharp differences in price due to condition problems, we eliminate those observations in which the higher-priced print sold for more than double the price of the paired print. In so doing, we reduce the total number of paired sales. For example, the number of paired sales between the United States and London markets declines by a relatively large amount, from 244 to 218, when the filter is imposed. The number of paired sales between Sotheby’s and Christie’s decline by a relatively small amount, from 140 to 137 in New York and from 58 to 56 in London, when the filter is applied. The relatively large number of deleted observations when sales in the United States are compared to those in London is due, presumably, to the presence of relatively minor auction houses which are more likely to offer damaged prints for sale.
See William Landes (2004) for an excellent, and detailed, review of the Ganz sale.
One referee finds unpersuasive our suggestion that “irrational exuberance” is evident in the exceptional prices realized for Picasso prints sold at the Ganz auction.
To our knowledge, none of the impressions (which are numbered and thus identifiable) sold at the Ganz sale have as yet been re-offered at auction. As a result, we cannot state that any of the impressions sold at the Ganz sale has been subsequently sold at a substantially lower price.
Subject to the above caveat, we believe that the case is very strong that—when offered for resale—these impressions are likely to sell for dramatically lower prices than those realized in the Ganz sale.
We have now reviewed sales of (different impressions of) these prints subsequent to the period (1998–2000) cited in Table 4. Auction prices continue to be dramatically lower than the prices realized at the Ganz sale. For example, Tete de femme (B.1065) sold for $32,250 in 2002 (well beneath $90,500) and Deux femmes sur la plage (B. 789) sold for $12,189 in 2002 (well beneath $34,500). In the print market, unlike some other segments of the art market, provenance is not a marker for authenticity and is not likely to enhance market/resale value on this account. Finally, three of the five Picasso prints sold at the Ganz sale were published in 1962. Picasso prints published at this relatively late date, when his art was already so expensive, tend to be of relatively uniform condition as well as of the same quality.
Most auction houses prefer not to offer more than one impression of a print at auction. If two or more impressions are available from consigners, most auction houses will defer the sale of one impression to their next sale.
This measure of “noise” is calculated after the filter has been applied to eliminate cases where the higher-priced print sold for more than twice the price of the lower-priced print.
We again note that differences in taxes may add to the noise in these pairwise comparisons of auction outcomes.
References
Biey, M. L., & Zanzola, R. (2005). The market for Picasso prints: A hybrid model approach. Journal of Cultural Economics, 29, 127–136.
Case, K. E., & Shiller, R. J. (1987). Prices of single family homes since 1970: New indexes for four cities. New England Economic Review, (September/October) 45–56.
Donson, T. B. (1977). Prints and the print market. New York: Thomas Y. Crowell.
Goetzmann, W., & Peng, L. (2002). The bias of the RSR estimator and the accuracy of some alternatives. Real Estate Economics, 30(1), 13–39.
Landes, W. M. (2004). The test of time: Does 20th century American Art Survive? In V. A. Ginsburgh (Ed.), Contributions to economic analysis: the economics of art and culture. Elsevier Science.
Mei, J., & Moses, M. (2002). Art as an investment and the underperformance of masterpieces. American Economic Review, 92(5), 1656–1668.
Mei, J., & Moses, M. (2005). Beautiful assets: Art as an investment. Antiques and Fine Art, V(5), (Anniversary), 34–35.
Pesando, J. E. (1993). Art as an investment: The market for modern prints. American Economic Review 83(5), 1075–1089.
Pesando, J. E., & Shum P. M. (1996). Price anomalies at auction: the market for modern prints. In V. A. Ginsburgh & P.-M. Menger (Eds.), Economics of the Arts. Elsevier.
Acknowledgements
We thank Simiao Zhou and Lindsay Eng for excellent research assistance and the three referees for very useful comments on an earlier draft of this article.
Author information
Authors and Affiliations
Corresponding author
Rights and permissions
About this article
Cite this article
Pesando, J.E., Shum, P.M. The law of one price, noise and “irrational exuberance”: the auction market for Picasso prints. J Cult Econ 31, 263–277 (2007). https://doi.org/10.1007/s10824-007-9046-7
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s10824-007-9046-7