How the numbers are made.
Every figure on Oras — market price, deal classification, alerts — is computed transparently. Here is exactly what we do, and what we do not do, with the data behind your wishlist.
The honest median.
The price you see for each reference is the rolling 30-day median of completed secondary-market transactions across our tracked sources. Asking prices and unsold listings are excluded — they tell you what the market hopes for, not what the market clears.
For pieces with thin liquidity, we widen the window to 90 days and surface a low-confidence flag. We never extrapolate from a single data point.
Three labels, no opinions.
A watch is labelled good when its market price sits at or below MSRP, fair within the +0% to +15% band, and stretch beyond that. Bands are reference-class adjusted; what counts as fair for a GMT-Master is not what counts as fair for a Nomos.
Labels are descriptive, never prescriptive. Whether a stretch piece is worth its premium is a question for you, not for us.
Movement, not noise.
An alert fires when one of four things happens to a piece on your wishlist: a price drop greater than 1% week-over-week, an MSRP change announced by the manufacturer, a target you set being reached, or a return to stock at an authorised retailer.
We rate-limit per-watch to one alert per kind per day. The alerts feed is meant to be checked weekly, not refreshed.
No predictions.
Oras does not forecast prices, score watches as investments, or rank pieces by expected return. Watches are not assets and we will not pretend otherwise. What you see is what is, with the dates clearly marked.