Statistical Shenanigans: Misrepresenting Auction-Sale Results

Bill Ruprecht, Sotheby's president

I've said it before in the Wall Street Journal (and numerous times on CultureGrrl). But it looks like I'm going have to say it again. Even Bill Ruprecht, the financially savvy president and chief executive of Sotheby's, hasn't yet gotten the message. So let's review:

If you're going to compare sale totals to presale estimates, you've got to use the hammer-price total for that comparison, NOT the total of hammer price-plus-buyer's premium. Presale estimates are predictions of hammer total, plain and simple. You have to compare them to the achieved HAMMER total (not hammer-plus-buyer's premium) to do a meaningful analysis. If you do the apples-to-oranges comparison, inflating the total hammer price with the buyer's premium, you are making the auction look stronger than it actually was. Apples-to-apples is a comparison of hammer price total to the presale estimate of the hammer price total.

Have I said this enough times yet for it to sink in? It's really not that difficult to understand.

It's bad enough when the auction-house PR people, whose job it is to make their firm look good (but also to do it accurately), make to apples-to-oranges comparison. It's a serious professional lapse when so many of the leading members of the art press, whose job it is to separate the hype from the facts, parrot the misleading comparisons in the press releases (as has occurred again with this round of sales).

But its even worse when the head of the company does this. Here's Sotheby's Ruprecht, analyzing this sale in his written comments for his auction house's just-released announcement of first-quarter financial results:

Last night's Impressionist sale brought outstanding results, ...coming in near the high end of the presale estimate.

As Bill surely knows, the presale estimate of the sale's hammer total was $141.35 million to $204.25 million. The total hammer price achieved was $171.27 million, slightly below the midpoint between the low and high estimates, not "near the high end." Only the apples-to-oranges comparison ($195.7 million, with buyer's premium) makes it appear to be "near the high end."

In its postsale press release, Christie's, which had its record Picasso to crow about, didn't make any claims about how its sale total compared with the presale estimate. But it did falsely claim that one of its Picassos, "Le peintre et son modèle," 1964, "sold above expectations" at $10.72 million (hammer price: $9.5 million). That painting actually fetched a hammer price BELOW the presale estimate of its hammer price---$10 million to $14 million. Even when inflated by the buyer's premium, the final price was near the low end of the estimate, not "above expectations."

Doesn't truthful accounting matter? And why do so many art scribes keep letting them get away with it?