When you read a hammer price at Christie’s or Sotheby’s in 2026, you are not reading a market price. You are reading the outcome of a structured product, underwritten in advance by a guarantor whose incentives are not aligned with yours, priced against a reference that the auction house itself helped to set. The hammer is the wrapper. What is inside the wrapper is a derivative.
This is not a conspiracy. It is how the business works now. The question is whether anyone pricing art as an asset has actually internalized it.
A brief recap of the mechanics
A house guarantee is a commitment by the auction house to the consignor that the work will sell at a minimum agreed price. If no bidder reaches that price, the house takes the work. The house is on the hook.
A third-party guarantee, or irrevocable bid, transfers that risk. A third party (often a collector, dealer, or fund) commits to bid at the reserve. If no one bids above, the third party owns the work at the reserve. If someone bids above, the third party walks away with a fee, typically a percentage of the upside above the reserve. Either way, the work sells and the house carries no inventory risk.
Put cleanly: an irrevocable bid is a written put option on the work, purchased by the consignor, sold by the third party, sometimes brokered by the house for a cut.
Why this matters for how you read a hammer
If a lot has a third-party irrevocable bid, and the hammer lands at the reserve, all that tells you is that one person was willing to commit to buy it at that price, and nobody else wanted to bid them up. That is a single data point with a single known counterparty. It is not a market clearing price.
If the hammer lands ten percent above the reserve, all you know is that one other bidder was willing to chase the third-party up by one increment. Still not a market.
Only when a lot clears well above the reserve, with multiple bidders, with no third-party support, do you get something that looks like a price discovery event. The houses do not publish how much of any given evening sale falls into which bucket. You have to infer it from the catalog symbols and the room.
“The hammer is the wrapper. What is inside the wrapper is a derivative.”
What to discount a guaranteed hammer by
Here is the working rule we use, and we would like to be argued out of it if anyone has better data.
- House-guaranteed lot, no third-party, clears at reserve: discount the printed hammer by 10 to 20 percent when you use it as a comparable transaction.
- Third-party irrevocable bid, clears at or near reserve: discount the printed hammer by 15 to 25 percent.
- Third-party irrevocable bid, clears well above reserve with competitive bidding: you can probably take the number closer to face value, but still haircut 5 to 10 percent.
- No guarantee, no irrevocable bid, competitive bidding, clears above high estimate: this is the clean print. This is a market price.
Almost nothing at the top of an evening sale in the current structure is in that last category. Most flagship lots are guaranteed. Many have third-party bids. The hammer numbers you are reading in the Financial Times and the New York Times, and across Artnet and Artprice and the rest of the data services, are numbers that need that discount applied before they are usable as comparables.
Historical examples worth revisiting
The Bouvier era was instructive, and the lesson was mostly buried. Yves Bouvier’s reference prices, brokered privately for Rybolovlev, were inflated by what amounted to informal guarantees against future resale. The market took those prices as comps for longer than it should have. When the reference set collapsed, it took specific Picasso and Modigliani reference points down with it.
The 2015 Picasso “Les Femmes d’Alger” sale is another case worth reading carefully. The hammer came in at a number that printed as the highest-ever auction result at the time. The lot was guaranteed. The guarantor, per trade reports, was a party with a strategic interest in setting a public reference. Was that a market-clearing price or a deliberate print? The answer is both, but the “both” matters when you treat it as a comparable.
Q4 2022 is the most instructive recent example. After a run of softer sales, the houses went aggressive on guarantees through the November New York week, structurally underwriting the top of the sales with third-party bids to ensure sell-through. The resulting totals, widely reported as a sign that the market had stabilized, were in fact a sign that the houses had chosen to spend balance sheet to stabilize the narrative. The footnote read differently from the headline.
The press release and the footnote
The houses are obliged, under their own disclosure rules, to mark lots in the catalog with symbols indicating house guarantee, third-party guarantee, or both. The footnotes are there. They are also, to be polite, not exactly prominent.
A reader of the FT or the Times auction recap is not going to see those symbols. The trade press is more careful but still, on aggregate reporting, lets the hammer numbers carry more weight than they deserve. This is a structural information asymmetry, and it works in favor of the houses and the consignors who use these instruments well.
Lucian Poe has been arguing for two years that every published hammer at evening sale ought to come with a machine-readable disclosure of guarantee structure, third-party presence, and any buyer’s premium rebate. It is the right argument. It is not going to happen without pressure.
Why a guaranteed lot is a structured product
Here is the frame worth carrying with you. A guaranteed evening-sale lot is a multi-party structured transaction:
The consignor has an option to sell at a floor price (the reserve) with upside participation.
The third-party guarantor has written a put (locked at the reserve) and earns a premium in the form of the upside-share fee.
The house has a priced the product, taken a fee, and stripped the inventory risk.
The underbidder, if any, has provided price discovery (thin).
The winning bidder, if different from the third party, has paid above the reserve and the price contains real information.
The market, as an abstract concept, has learned very little. A single transaction with multiple known counterparties is not how you price a broad asset class.
Position implications
For collectors underwriting art as a financial position, the implications are direct.
When you build a comparables set for a work you own or want to buy, audit every comp for guarantee structure. Apply the discount. Your fair value estimate will, on average, come in ten to twenty percent below the naive comp-based estimate. That is where your mark ought to be, not the higher number.
When you consign, understand that the terms a house offers you are pricing you against the third-party market, not against the retail bidder pool. The guarantee you are offered reflects what the house thinks it can syndicate. That is useful information about your work’s true clearing price.
When you read a new auction record in an artist’s market, pause before you extrapolate. Ask: was it guaranteed? By whom? Did anyone bid against the guarantor?
The forward read
We expect the guarantee share of evening sales to increase further, not decrease, over the next four quarters. The houses are economically incentivized to continue expanding this instrument. We also expect, within twenty-four months, either a regulatory or industry-driven disclosure upgrade that makes guarantee structure more legible at the lot level. If that disclosure comes and forces cleaner reporting, expect the printed comp base for the top of the market to reset noticeably downward as the discount we have been recommending becomes industry-standard. If it does not come, the public tape continues to drift further from the real market, and the information asymmetry gets worse. Be careful what you call a comp.
Nothing in this article is investment advice. CreativeSlop is an independent publication. Figures rounded for readability. Names of market participants referenced in good faith from on-the-record and on-background conversations.