01
Start with a defensible exit assumption
Do not begin with the seller’s ask and negotiate downward by habit. First estimate the outcome the watch could reasonably support in the channel you expect to use. Review relevant comparables, distinguish asking prices from completed evidence where possible, and adjust for the specific watch’s condition, completeness, and selling venue. Write the assumption down so it can be challenged later.
Example: suppose your research supports an illustrative exit assumption of 4,800 currency units for a complete example in the inspected condition. That number is not a forecast or a real-time quote. It is simply the top line of a decision model. If the watch lacks accessories or needs work, revise the assumption or the expected costs instead of hiding the difference inside intuition.
02
Subtract every expected transaction cost
List the costs that stand between acquisition and cash received: inspection, service, parts, inbound and outbound shipping, insurance, marketplace or payment deductions, packaging, and any repair you are prepared to fund. Use the costs that fit your actual process. A fee that is unlikely to apply should not be invented, but a recurring cost should not disappear merely because it is inconvenient.
Timing also has an economic effect even when you do not assign it a formal charge. A watch that may take longer to sell ties up capital and consumes follow-up work. Keep that consideration separate from direct cash costs so the model remains legible. You can then decide whether a slower opportunity needs a wider margin or simply does not fit your current inventory position.
03
Reserve room for uncertainty and return
After costs, subtract the return you require for the work and risk involved. Then include a reserve for unresolved condition, sparse market evidence, or a seller claim you have not verified. The reserve is not a prediction that something will fail; it is explicit space for the model to absorb a reasonable adverse outcome without turning an acceptable purchase into a loss.
Example: from the illustrative 4,800-unit exit assumption, you might subtract 420 units of expected transaction costs, 700 units for the return required, and 280 units as an uncertainty reserve. The resulting 3,400-unit maximum is arithmetic based on those assumptions, not an instruction to buy. Change any input and the safe-buy boundary changes with it.
04
Stress-test the number before offering
Run at least one less favorable case. Reduce the exit assumption, increase the likely service cost, or extend the expected holding period, then see whether the deal still meets your requirements. If a small revision erases the return, the original maximum was fragile. Your walk-away number should survive ordinary uncertainty rather than depend on every optimistic input landing correctly.
Market ranges are context, not guaranteed outcomes. They help define the evidence available when you decide, but they cannot promise a future sale price, timing, or buyer. Save the assumptions with the watch in Patina AI, revisit them after inspection, and be willing to decline the deal when new information moves the boundary below the seller’s acceptable price.