The high-end market is sewn up by Christie's and Sotheby's, and usually reported as a messianic struggle between them. But reporters are measuring the wrong thing when they compare value of sales, a metric chosen because it's easily available and easily understood rather than because it's actually important or meaningful. If you're selling pictures, you want to know which auctioneer will get the higher price, and it's not clear to me that either has a clear edge. Knowing how much they've sold for other people doesn't answer that question. If you're a shareholder, you want to know about risk-adjusted return. Sale volume doesn't help that analysis: you need to know the commission charged and guarantees granted. And mere observers like me would be happy if either of them could just make navigating their websites easier than bitcoin mining.
On the face of it Christie's sale was a disaster. But it really comes down to a handful of pictures. Their top lot, Hoffmann's Hare didn't sell, despite being one of the best things offered in London in recent years. Maybe private equity magnates think hanging a fluffy bunny on their wall would convey the wrong message. When I first saw it I thought the estimate would be £3m-£5m, but it was actually pitched at a slightly more aggressive £4m-£6m. Given its appeal, I didn't think that unreasonable, and it was quite plausible that a couple of interested collectors could have pushed it well beyond that. A small Claude drawing sold recently for just shy of the Hoffmann low estimate, after all. Sotheby's top lot, a Constable that didn't inspire me, sold to a single bidder at its reserve. If that buyer had gone for the Hoffmann at Christie's, it would have looked very different.
We don't know the economics of either auction, particularly how much commission they earned (assume negative seller's commission). The other expensive lot at Christie's was withdrawn; surprisingly, it was consigned by Christie's itself, having taken ownership when it failed to meet its guaranteed price a few years ago. It was sold privately, I'm told above its high estimate. That would also have boosted their total significantly. It ought also to boost profits too, as it would have been taken to inventory at the lower of estimated realisable value and purchase price. I'd say a prudent accounting treatment would be to discount by at least 20%, causing a big loss then but a decent profit now.
A few other observations:
- Vive la France! The Louvre made a great purchase of the Master of the Dreux Budé Triptych's Arrest of Christ for £965k (pictured above). It's a good and unusual picture that would complement any of the great collections of northern renaissance painting. Well done them for getting it directly from auction. Another fine French museum acquisition was Cabanal's Portrait of an Artist from the Winter Collection sale, for just £6,250. There are plenty of opportunities for good but cheap pictures that museums could be buying, but it's rare for them to do so. Do watch La Tribune de l'Art for the most thorough coverage of museum acquisitions, including the Cabanal.
- Things I got right: of the pictures I thought looked cheap, the Moillon doubled estimate, Flinck, Mignon, Bachiacca and Testa all exceeded estimate (even before premium). Two early Italian pictures I thought seemed cheap sold for approximately double and triple their respective estimates.
- The pictures I thought expensive mostly did poorly, which wasn't my expectation. The Constable scraped its reserve with one bid, view pictures and the Bosschaert still life did poorly, and the Henry VIII sold below low estimate. Tudor portraits have been madly expensive recently; historically interesting, but artistically dull.
- Things I got wrong: the Cariani was a bargain at its low estimate (£100k, £125k with premium), though dark pictures of Christ with the Cross are not the most 'commercial'. Baciccio made its high estimate (£35k with premium), but I still say it was cheap at that price. Mr Market didn't share my fondness for the early Cologne picture that made just £50k. I was surprised the Verspronck didn't sell; I thought it might exceed the estimate. His portraits are not rare, but this was a good one (though not in perfect condition). The Standard Bearer by Verspronck made $1.5m in 1998, and is now in the National Gallery, Washington. This one wasn't nearly as spectacularly swashbuckling, but similar order of quality.
- So I was more right than wrong. I'm a genius, right? Well, alas no. Estimates aren't really predictions. Some pictures are estimated low to encourage interest (and tease us). Some are estimated 'robustly' to signal their importance (e.g. Hoffmann). And others are estimated high because the seller has demanded a high reserve. The auction house has marginal costs that aren't fully recovered from cataloguing fees, and they don't want too many unsold lots. But against that, they want to ensure their sales have a reasonable balance and low marginal costs make it worthwhile to take a punt if there's even a small chance of selling. There might also be opportunities for commission from private sale after the auction, and including some lots with high reserves might facilitate other consignments from the same seller. Sometimes unsold lots are no surprise to anyone, even the auctioneers. But that doesn't mean they're acting irrationally.