The Mnuchin Sale Post-Mortem: What the Market Got Right and Wrong

Every major sale is a test. The auction house sets the estimates; the room decides whether they're right.

The recent Mnuchin Gallery evening sale was a particularly clean case study: a focused, high-quality consignment with coherent estimate ranges across the board. No obvious loss-leaders, no vanity lots. Just a house betting that its client relationships and scholarly positioning would translate to competitive bidding.

Here is what happened — and what the Paddle Supernova prediction market saw coming before the hammer fell.

The headline number, and why it doesn't tell the story

The sale totaled within its pre-sale estimate range, which headline writers will call a success. It is not the interesting number.

The interesting numbers are lot-level pass rates and premium-to-estimate ratios. A sale that aggregates to its total estimate while half the lots sell below the low and the other half dramatically exceed the high is a very different market than one where every lot lands quietly in range.

For this sale, the distribution was bimodal. Lots with strong exhibition histories and clean provenance flew — often 30–60% above the high estimate. Lots where the story required more faith sat on the low end or bought in entirely.

What the crowd knew

Paddle Supernova had live prediction markets on six lots from the sale. Before the auction opened, the aggregate signal was striking:

• The three lots that ultimately exceeded estimate all had YES probabilities above 68% on Paddle Supernova at lock. The crowd had already priced in the upside.
• Of the two bought-in lots, one had a NO probability of 74% — the clearest signal in the set. The estimate had been set aggressively and the market knew it.
• The sixth lot, a smaller work with contested attribution, was nearly 50/50 at lock and settled just above the low estimate. The uncertainty was correctly priced.

This is the core value proposition: not that a prediction market is always right, but that it aggregates the distributed knowledge of people who have looked closely at the work, the comparables, and the room dynamics.

Three lessons for reading the next sale

1. The estimate range shape matters more than the midpoint.
A range of $800K–$1.2M is a very different signal than $900K–$950K even if the midpoints are close. Wide ranges signal the house has less conviction, or is trying to accommodate a cautious seller's reserve. Narrow ranges signal the house has priced it precisely and the room will either affirm or reject.

2. Bought-ins cluster by category within a sale.
When an evening sale buys in two lots, they are rarely from the same aesthetic period by coincidence. Weak demand in one segment tends to compound: phone bidders who were on the fence for the first lot of a period they follow watch the buy-in and pull back their bids on the next.

3. Exhibition history is still the most reliable predictor.
Lots with major institutional exhibition histories — MoMA, Tate, Guggenheim, Serpentine — have significantly higher hammer-to-estimate ratios in our data than comparable lots with only gallery or private-collection provenance. The signal is durable and shows no sign of decay.

Join the market for the next sale

If you have a view on where the hammer falls, prove it. Paddle Supernova opens markets on lots from Christie's, Sotheby's, Phillips, and other major houses weeks before each sale. No real money, no registration fee — just skill on the line.

Your predictions are public. The leaderboard keeps score.