Second-hand going rate

Why does a used item settle near a familiar price when nobody anywhere publishes one?

Sell a five-year-old bicycle and something odd happens. You have never sold a bicycle before. No authority publishes what a five-year-old bicycle is worth. Yet within a few minutes of looking around you form a firm view — "about 120" — and so, independently, does every stranger who messages you. There is a going rate, and everybody seems to know it.

So who set it? That is the question worth sitting with, because the honest answer appears to be nobody, and yet "nobody" is not usually a satisfying explanation for a number that people agree on to within ten per cent.

Start with what you actually did before naming your price. You looked at what other people were currently asking for similar bicycles. Not at a valuation, not at a rule — at other people's guesses.

Now ask the awkward follow-up: where did they get theirs? From looking at listings too. Nobody in the chain consulted an authority. Everyone consulted everyone else.

That should feel circular, and it is. So why does a circle of mutual copying settle anywhere in particular rather than wandering off? Because copying is not the only thing happening. Two filters run underneath it.

The first filter is time on the market. Price your bicycle at 400 and nothing happens. Weeks pass. Eventually you cut it, and cut it again, and it sells or it disappears. Price it at 40 and it is gone within the hour. Notice what this means for what the next seller sees: cheap listings vanish quickly and expensive ones linger. What is visible at any moment is systematically skewed toward the too-expensive, which is why the going rate you infer from asking prices is usually a little above what things actually fetch.

The second filter is the outside option. A buyer always has an alternative — a different seller, waiting, or buying new. That puts a ceiling on the band: nobody pays much more than the next listing along. And sellers have a floor: keeping the thing, or scrapping it. The band is squeezed from both directions.

Put the copying and the filters together and you have the whole mechanism. Each participant does something purely local and rather unimpressive — glance at a few listings, guess, adjust if it does not sell. No one has the aggregate in view. But because every guess is fed back into the pool the next guess is drawn from, the errors cancel and the agreements accumulate. A number emerges that no individual computed and no individual could have computed alone.

Two honest qualifications. First, this is a band, not a point. Price dispersion for identical goods is one of the most stubborn findings in empirical economics — it persists even in online markets with near-zero search costs — so "the going rate" is really a fuzzy range that people report as though it were a figure. Second, an emergent number is not necessarily a correct one. The pool has no independent check on itself, which is why a whole category can sit for years at a price nobody would defend from first principles.

The misconception worth correcting is that the going rate reflects the item's value, arrived at by many people independently estimating the same underlying quantity. It does not, quite. It reflects the recent history of what similar items were asked and paid for, plus whatever anchor the category happens to carry — usually the original retail price, which is why depreciation folklore ("half of new") is so sticky even when it fits nothing.

That also explains the sharp edges people notice. A category with lots of turnover has a tight band, because the feedback loop runs fast and errors are corrected quickly. A category with almost no turnover — an unusual instrument, an obscure tool — has no going rate at all, and sellers there flounder, because the pool they would normally copy from is empty. The rate is not a property of the object; it is a property of how much recent trading there has been.

And the loop is exactly what a fresh market lacks. If you have watched a new second-hand platform open, you will have seen the first weeks of chaotic pricing settle into consistency within a few months. Nothing about the goods changed. What changed was that there were now enough completed transactions for everyone to be copying from a shared pool.

• Why asking prices systematically overstate sale prices, and what "sold" filters actually fix.
• Thin markets: how a category with few trades never develops a rate at all.
• The lemons problem — how uncertainty about quality, not just price, shapes second-hand markets.

Every term the collection defines is gathered in the glossary.