How watch market values are calculated

Every pre-owned watch has three prices at once: what sellers ask, what dealers sell for, and what it actually changes hands for between private parties. Price guides that blur these together are often 10% or more off in either direction. A useful market value starts by deciding which of those prices it is measuring.

Ours measures the last one: the price a typical private sale of the watch would reach today, for a pre-owned example with its box and papers, averaged across regions. Everything else on a watch page (dealer price, asking price, new or unworn, without box and papers, a specific region) is derived from that baseline.

Asking prices are evidence, but biased

Most of the visible data in the watch market is asking prices on listings. They matter: a listed price tells you where sellers think the market is. But asks sit above what watches sell for, and the gap is not constant. A listing that has been up for months was probably overpriced from the start, which is why it hasn't sold.

So a good valuation counts asks, discounts them toward sale prices, and discounts stale listings more than fresh ones. Completed sales and auction results count most, because they are what buyers actually paid.

Comparing like with like

Two copies of the same reference can be worth noticeably different amounts. Condition (new or unworn versus worn), whether the box and papers are present, whether the seller is a dealer or a private individual, and where in the world the watch is sold all move the price.

Statisticians call adjusting for these differences hedonic pricing: compare copies of the same reference sold around the same time, and measure what each difference is worth. Once every price is adjusted to the same baseline, they can be combined into one estimate instead of being averaged blindly.

Following the value over time

Most references trade a handful of times a month, not hundreds. A simple 90-day average reacts slowly when the market moves and jumps around when a single odd sale comes in. A better approach treats the true value as something that drifts day to day and updates it with each new price, weighting reliable prices (sales) more than noisy ones (asks), and growing more uncertain when no new prices arrive.

That is how we do it, and it is why every value on this site comes with a range: 80% of private sales are expected to land inside it.

How to judge any price guide

Ask three questions. Which price is it measuring: asks, dealer sales or private sales? Does it adjust for condition and box and papers? And does it publish its accuracy against real sales it hasn't seen? Our methodology page states the publication standard. Accuracy figures are published only when a sufficient real-sales holdout has been evaluated; simulated backtests are not live-market accuracy.