Good Artwork Always Sells at the Right Price.
by Michael McCullough
A dozen years ago anyone who foretold the size and importance of today’s global auction market would have been looked on as a lunatic. And yet the truth is that the present market ought to have been predictable even during the Great Recession. Something like it was bound to happen as soon as the financial benefits of globalization started to kick in.
Of course, much of the success of the auction market rests upon the mechanics of the auction process itself. Take, for example, how auction estimates are set: every auctioneer will tell you that low estimates bring high prices. This sounds counter-intuitive, but it’s a fact that competitive bidding is never as robust as when the bidding starts at a relatively low level and works up toward the object’s market value. In consequence of this, auctioneers will always advise a seller to accept estimates lower than his or her expected return on an object in the hope that the bidding goes above the high estimate (auction estimates are always a range, such as $5,000-$10,000, with the former being the low estimate and the latter being the high estimate).
The dilemma for the seller is that the “reserve price,” the confidential price at which the auctioneer will sell the object, cannot exceed the low estimate (at least in honest auctions), so the seller takes the risk of the object selling to a single bidder at the reserve price in the absence of competitive bidding. A seller who wants the certainty of obtaining a high price will insist on higher estimates, and many of these lots fail to sell because bidders are reluctant to start bidding competitively at a high level. This week at Christie’s, Modigliani’s “Monsieur Baranowski” failed to sell because the estimates were set too high. The picture, estimated at $25–$35 million, would have had a much better chance of selling at the $25 million mark had the estimates been set lower at $20-$25 million, or at least would have sold within those estimates. In contradistinction, Kandinsky’s “Schwarz und Violett,” that had been estimated conservatively at $4.5 million to $7.5 million, sold for $12.5 million, still a good buy. All of this is to say that competitive bidding is as much an exercise in psychology as it is in economics.
Traditionally, about a third of the objects offered at the major auction houses do not find a buyer. Of those that do sell, about half sell within the estimates and half sell above the high estimate. This means that global auctioneers have a .667 success rate, which is also an outstanding batting average in the major leagues. It also means that the auction houses are very good at convincing their clients to accept reasonable estimates.
So, why is this important? Because auction houses make their money from charging the buyer’s premium, which is a percentage of the sale price of an object, calculated on a sliding scale that averages about 20%. While they also charge the seller a commission, referred to as the “vendor’s commission,” it’s usually negotiated away by sellers of high-priced objects and only ranges between 6-10% for lower priced objects. But the buyer’s premium is sacrosanct, never waived and is where most of an auction house’s revenue is gained. As a matter of fact, the major auction houses make no profit from the vendor’s commission at all, as it just pays for the lights. All of the year’s profits are made from collecting the buyer’s premium, a majority of which comes from only five collecting categories: Impressionist and Modern, Post-War and Contemporary, Old Master Paintings, Jewelry and Special Collections (single-owner sales such as Jacqueline Kennedy-Onassis or Liz Taylor).
All of this is why the “White Glove” sale, where 100% of the lots are sold in an auction, is so coveted among auctioneers. It means that the most profit possible was achieved from the sale, and results in the popping of champagne corks and cartwheels in the street…..literally.
Note: An earlier version of this article appeared on Gallery Intell.
Michael McCullough's Blog 