Sale-rooms
Discriminating sellers
Peter Watson
anhattan responded to the great art sales last week as only Manhattan can. In a florist's on Lexington Avenue an arrange- ment of irises in the window was accompa- nied by the legend: 'At a mere $53.9 million, these must now be considered a bargain'. Wall Street whizzes invented their own 'Dr Gachet' highball: Dutch bitters, French polish, American sugar and a secret ingredient that only the Japanese will swallow. And Johnny Carson, inevit- ably, referred to Christie's record-breaking sales — as taking place at Sotheby's.
Now that the dust has settled, however, and we have been put in our place by Mr Ryoei Saito, who said that this $160.6 million purchase of both the Van Gogh and the Renoir was 'no big deal', there is still one contentious issue unresolved. This is the matter of guarantees.
Sotheby's have offered financial guaran- tees to sellers on objects to be sold at auction for longer than is generally recog- nised. According to Mr Frank Herrman, in his official history of the firm, guarantees were first given on Impressionist pictures in the 1950s. Christie's, who introduced them for the first time last week, have traditionally opposed the idea because, they say, it can lead to a conflict of interest. Auction houses will be more apt to 'push' those lots in which they have a stake and other lots, accordingly, will suffer.
Well, last week's sales contained enough guaranteed pictures for this matter to be studied. Sotheby's main auction had 13 lots which were guaranteed and 54 which weren't. If Christie's fears are justified then they ought to show up in one or both of two ways. In the first place, the auc- tioneer should treat guaranteed lots dif- ferently from non-guaranteed lots and in the second place, guaranteed lots should sell better than other pictures.
An analysis of the auctioneering at last week's New York sale suggests that Sotheby's do treat guaranteed lots some- what differently but that, last week any- way, it had no effect on the saleability of the art. First, I studied where auctioneer John Marion opened the bidding on each lot at the main sale. Usually, an auctioneer will start the bidding at 50 per cent of the low estimate. Occasionally he will start lower, more rarely higher. Last week, with guaranteed lots Marion started below the 50 per cent level on only two pictures (15 per cent). With non-guaranteed lots the corresponding figure was 22 per cent, but with such a small sample these two figures must be considered equivalent. Ironically, it was Christie's, and not Sotheby's, who treated their guaranteed lots differently. Of the five lots guaran- teed, Christopher Burge started the bid- ding on four of them above 50 per cent of the low estimate. This 80 per cent of an admittedly low sample was nevertheless in marked contrast to the rest of the sale where the corresponding figure was only 44 per cent. The level of opening bid does matter, as a comparison with actual results shows. Briefly, this may be summarised as follows: lots where the bidding starts under 50 per cent of the low estimate are much less likely to sell, and to sell well, than lots where the bidding starts higher. In one way, this is common sense, of course. If an auctioneer knows there is a lot of genuine interest in a particular painting why waste time at the bottom of the scale? On the other hand, it does mean that an auc- tioneer is to some extent telegraphing his view, and/or his inside knowledge, to the sale-room through the level at which he opens the bidding. In this context it was again Christie's, and not Sotheby's, who discriminated most between guaranteed and non-guaranteed lots. Having started the bidding higher on guaranteed lots, all five items reached or exceeded their estimates, as compared with only 38 per cent of the non- guaranteed lots. In contrast, in Sotheby's sale, 46 per cent of the guaranteed lots reached or exceeded their estimates, com- pared with 60 per cent in the non- guaranteed part of the sale. In this particular case, the quality of Christie's guaranteed lots (the Robert Leh- man collection: Van Gogh's 'Self-portrait, Toulouse-Lautrec's 'Fille a la fourrure', Modigliani's 'Jeanne Hebuterne', etc) probably meant that they would have sold whether the bidding started on the chande- lier or the spittoon. But there seems to be no statistical evidence yet for saying that, on the rostrum, Sotheby's treat guaranteed lots more favourably than non-guaranteed lots.
On the other hand, the fact that a higher proportion of guaranteed lots failed to reach the estimates (54 per cent against 40 per cent in the rest of the sale) suggests that guaranteed lots are given more aggres- sive estimates and in that sense Sothebys could be said to be 'pushing' this merchan- dise.
One sale, even one with a lot of guaran- teed paintings, cannot prove a case one way or the other and we shall have to wait for future sales to provide us with a decently reliable sample. But here at least is a methodology for examining the charge. We can return to it at later date.
Previous page
