Hammer price vs what the buyer paid: four auctions, measured lot by lot

When an auction house publishes a result, the number is almost always what the buyer paid: the hammer price plus the buyer's premium. Comparing those figures across houses, or against a private-sale price, without separating the two quietly mixes different things. We took four sales from the last two weekends, all from the houses' own lot pages, and split every sold lot back into hammer and premium.

What each sale actually charged

How each was checked:

• Broad Arrow: 12% up to $250,000, then 10% above. That formula turns all 63 priced sold lots back into a whole-dollar hammer, with no exceptions.
• RM Sotheby's London: 15% up to £200,000, then 12.5% above, ex-VAT. All 56 sold lots invert to a whole £500. The four non-car lots (a luggage set, two registration numbers and a sign) sat at exactly 20%, so they are left out of the car figure.
• Gooding Christie's Geneva: 15% up to CHF 200,000, then 12.5% above. All 23 lots invert to a whole CHF 1,000, and the published prices add up exactly to the published sale total of CHF 21,073,825.
• Bonhams: the house publishes its own hammer. 19 of 20 sold lots are exactly 1.15× hammer. One lot (205) shows two house figures that disagree, so it is reported as printed.

The effective rate falls on bigger cars because the top tier is cheaper. A 2014 Ferrari LaFerrari hammered at $8,000,000 at Broad Arrow and cost the buyer $8,805,000, which is 10.06%. A 1966 Morris Mini Moke at the same sale hammered at $40,000 and cost $44,800, which is the full 12%.

The same hammer costs different money

Using each sale's own rules (taxes, VAT on premium, transport and import duty not included):

• 100,000 hammer: 112,000 at Broad Arrow; 115,000 at RM London, Gooding Christie's Geneva and the Bonhams sale.
• 1,000,000 hammer: 1,105,000 at Broad Arrow; 1,130,000 at RM London and Gooding Christie's Geneva. (No Bonhams lot in this sale got near that level, so we don't extrapolate one.)

That 25,000 gap on a 1,000,000 car is real money, but it's in local currency and it isn't the whole bill. Your all-in limit should also cover VAT where it applies, sales tax, shipping and anything you'd spend to get the car to resale condition.

Most sold cars hammered below the low estimate

For the three sales that published estimates, we compared each sold lot's hammer with the house's pre-sale range:

That's 76 of 101 sold lots (75%) below the low estimate. These are lots that sold, so the reserve must have been met or waived. Unsold lots aren't counted here: Broad Arrow had 20 that didn't meet reserve and one withdrawal, and Bonhams had 2. RM's London lot pages didn't publish estimates.

Two caveats. First, this is three sales over two weekends, not a market trend. Second, a house estimate is a marketing and consignment figure, not an appraisal. What this sample does show is that an estimate is weak evidence of what a car will hammer for, and that comparing your bid with the estimate is the wrong test.

What to do with this

1. Convert every published result to hammer before you compare it with anything, and use the rules for that specific sale. Houses run different premiums, and RM alone runs more than one.
2. Set your limit as an all-in number and work back to a hammer. Don't add premium on top after the fact.
3. Compare a car with completed sales of the same spec, not with its estimate.

BidBetter's comparable sales are stored with their premium basis, so a with-premium result from one house isn't silently compared with a hammer from another. See how we build a number and which auction house gets more, which compares hammer prices and points here for the premium side.

Source data: each house's own published lot pages for the four sales named above, collected 1–9 Nov 2026. Currency figures are native and not converted.