Christie's have introduced the maddest auction house fee yet. They now propose an extra 2% seller's premium if lots exceed the high estimate. So if Christie's misidentifies your consignment and sells a sleeper for a fraction of its true value, but a multiple of its estimate, you'll owe them a 'performance fee'! It creates a perverse incentive for Christie's to push for the lowest possible estimate, and it gives sellers an incentive to insist on the highest possible estimate. The estimate becomes a commercial negotiation rather than a guide to its likely sale price. And prices can be driven up by all sorts of idiosyncratic reasons that have little to do with the auctioneer's actions.
Christie's says the new fee will 'incentivise and reward high performance'. Of course that's just marketing guff, but it's stupid marketing guff because it implies that they need new incentives to perform highly - aren't they trying at the moment? And it's foolish to create an expectation that their performance should be judged on the difference between estimate and sale price. Sometimes they do a superb job, but buyers just aren't interested.
The sellers' premium is mainly a negotiating ploy, only really charged on low-value consignments. It's a suckers' charge for naive sellers who aren't aware that it's negotiable. So I suspect most sellers will avoid this fee, but it will still create unnecessary additional suspicion of estimates, though they are already as much a marketing tool and a negotiation with ambitious sellers as much as a true guess at hammer price. It's a bit ironic because the buyers' premium is in practice a charge on sellers, as I explained in a previous post.
Transaction costs in the art market are notoriously high and notoriously vague. Stepped increments mean that the difference between hammer price and actual cost is hard to work out. It also makes it difficult to assess how lots performed against expectations, because the auctioneers like to compare estimates without premium with sale prices including premium. So something said to have sold around the low estimate might actually have missed it by about a third, if you actually compare like with like. We should therefore be grateful to Felix Salmon, a rare beast who writes intelligently about both art and economics, for he has just published an auction premium calculator online, which lets you work out the premium from the hammer price, and the hammer price from the total.
I also wrote a premium calculator a few months ago, but couldn't embed it on the blog (technological incompetence). Still happy to send out if anyone wants my sheet, which will give you the underlying formulae to play around with. The one thing you can't do with the online version is adjust for tax (charged on the premium and sometimes on the lot itself, at different rates ... taxation is almost as complicated as auction house pricing).