Estimate upon request

Six lots in this archive were offered without a printed estimate

Six of the 265 auction results recorded in this archive carry an estimate that is not a range of numbers. Instead the catalogue prints a sentence: Estimate Upon Request, or in the region of a figure, or in excess of a figure. They are the six most expensive lots we hold, which is the first thing to say about the practice, and they run from the Chenghua chicken cup at Sotheby's Hong Kong in 2014 to the Klimt portrait of Elisabeth Lederer at Sotheby's New York in November 2025.

The usual reading of a withheld estimate is that the house does not want to anchor the bidding, or that the consignor does not want a public number attached to the property if it fails. Both are probably true and neither is checkable from outside. What is checkable, and what this archive can show because it records the pre-sale reporting alongside the result, is something narrower and more useful: in every one of the six, a number reached the public before the sale anyway.

That is the finding. The withheld estimate is not a missing number. It is a number that has been moved out of the catalogue, which is a document with contractual weight, and into the press, which is a document with none.

The number leaves the catalogue and arrives in the newspaper

In each of the six sales the figure that the catalogue declined to print was briefed to reporters, and in each case it was published days or weeks before the lot came up. Sotheby's printed Estimate Upon Request against the Klimt and the press carried a figure in the region of 150 million dollars; the auctioneer then opened the bidding at 130 million, which is itself a disclosure of where the house believed the floor was. Christie's catalogued the Salvator Mundi at an estimate in the region of 100 million dollars, a form of words that is a number in a sentence rather than a range. Sotheby's Hong Kong offered the Ru guanyao brush washer with the estimate on request and the pre-sale reporting put the expectation in the region of 100 million Hong Kong dollars, quoted at about 13 million US.

The practical effect for a bidder is not that information is withheld. It is that the information arrives through a channel that carries no obligation. A printed estimate sits inside a catalogue whose conditions of sale the buyer contracts on, and houses are careful about what they print there. A figure given to a journalist is subject to no such care, is frequently rounded, is sometimes a range collapsed to its midpoint by the reporter, and can be corrected without anyone being told. When you read the briefed figure you are reading marketing, and it should be weighted accordingly.

The briefed figure is not a better estimate

If withholding the estimate produced a more accurate private number, you would expect the briefed figure to sit closer to the hammer than an ordinary printed estimate does. It does not. Across the six it misses badly, and it misses in both directions, which is the detail that undercuts the idea that these figures are quietly conservative.

Read the last two columns together and the spread is enormous. The Salvator Mundi made four times the figure in its own catalogue. The chicken cup made less at the hammer than the top number the press had been given, and only cleared it once the premium was added, which is a different thing and is exactly the kind of conflation this publication separates by habit. A practice that produces a four-times miss in one direction and a shortfall in the other is not a forecasting method.

In excess of is a floor, and once it was met exactly

The phrase in excess of does something different from estimate upon request, and the difference matters. It publishes a number and declares it a minimum. In doing so it tells the room roughly where the reserve sits, since a house will not normally advertise a floor above the level at which the consignor has agreed to sell.

The Botticelli is the case that makes this visible. Sotheby's catalogued Portrait of a Young Man Holding a Roundel with an estimate in excess of 80 million dollars for the Master Paintings and Sculpture sale of 28 January 2021, and the hammer fell at exactly 80,000,000 dollars. Our record of the sale notes that bidding was thin and slow from 78 to 80 million and closed between two of the house's own specialists on the telephone. The published minimum was met and not exceeded by a single bid.

We would not build a theory on one lot, and one lot is what we have. But it is a clean illustration of a mechanism that is easy to state and hard to see: an advertised floor is a target as well as a threshold, and a bidder who reaches it has done everything the catalogue asked. The Canaletto, offered in excess of 20 million pounds and hammered at 27.5 million, shows the same form of words behaving quite differently in a contested room. The floor is not a prediction. It is a starting position that the room may or may not have a reason to leave.

The same object, catalogued both ways

The Salvator Mundi has been offered twice within living memory, and the two catalogue entries are a controlled experiment that nobody designed. In 2005 the panel came up at the St Charles Gallery in New Orleans as lot 664, catalogued as after Leonardo da Vinci, four lines of description, one photograph of the unrestored surface, and an estimate printed in the trade's bracket form as 1200 to 1800 dollars. In November 2017 the same panel came up at Christie's in New York catalogued as Leonardo, in a contemporary art evening sale rather than an Old Master sale, with no printed range and the words in the region of 100 million dollars.

Everything about the object was the same physical thing. What changed was the attribution, the restoration, the sale category and, at the level this piece is about, the form of the estimate. The 2005 entry told a bidder precisely what the house thought, in a number, in the binding document, and that number turned out to be wrong by five orders of magnitude. The 2017 entry told a bidder a sentence, and that sentence turned out to be wrong by a factor of four.

So the honest conclusion is not that printed estimates are reliable and withheld ones are not. Both were badly wrong about the same panel. The difference is what a reader can do with them. A printed range is a commitment you can hold the house to and check afterwards. A sentence in a press briefing is not, and the archive of these six sales is a record of numbers nobody ever has to answer for.

How to read a lot with no estimate

Treat the briefed figure as the house's opening marketing position and not as its valuation. It is unbound, unattributable and revisable, and this archive holds one case where the eventual hammer was four times it and one where the hammer came in under it. If you are building a comparable set, record the briefed figure with its source and the date it appeared, and never let it into a price series alongside hammer prices, because it is not the same kind of number.

Then look for the disclosures the catalogue does still have to make. The Klimt carried both of Sotheby's symbols for guaranteed property and an irrevocable bid, and neither the guarantee level nor the identity of the irrevocable bidder is published. That combination tells you more about the floor under the lot than any briefed estimate does, because a third party has committed real money at an undisclosed number. The Botticelli, by contrast, carried no guarantee and no irrevocable bid, which is part of why its thin close at exactly the advertised minimum is legible at all.

And record the opening bid. The auctioneer's first number is chosen with knowledge of the reserve and of any guarantee, and it is stated aloud in front of the room. Sotheby's opened the Klimt at 130 million against a briefed 150. That is the house's own view, made in public, on the day, at no cost to anybody. Of the several numbers attached to a lot with no printed estimate, it may be the most honest one available.

Why this archive is free to read

We do this for everyone who loves art. The people who have studied it for years, and the people who just stopped to look. Iconotheca exists so that what we learn belongs to all of them.

Nobody pays us for this. No ads, no sponsors, no commercial interest. These works belong to all of us, and we would rather share them with as many people as we can than keep them to ourselves.

If it gave you something today, tell us to keep going. Follow us, leave a like, or write a positive comment. We read every one, and they are what keeps us going.