How a local housing market actually works

Stock, turnover, reference pricing and the long lag between an agreement and a record.

Stock and turnover are different things

A town's housing stock is every dwelling standing in it. Its turnover is the small fraction of that stock which changes hands in a given year. In most settled New England towns turnover runs at a few percent of stock. That ratio is the single most useful thing to hold in mind, because almost every confusing feature of a local market follows from it: the market you can observe is a thin slice of the housing that exists.

It follows that supply, in the sense that matters day to day, is not the number of houses in a town. It is the number of owners who have decided, this season, that they would rather have money than the house. That decision is driven by life events, by what the owner thinks they could get, and by where they would go next. When moving on looks hard, listings fall even though the stock has not changed by a single roof.

Why a handful of sales sets the reference price

Housing has no ticker because no two units are the same. A buyer and a seller therefore reason by comparison: they look for recent sales of houses that resemble the one in front of them, and adjust. In a small town there may be only three or four such sales in the past six months, and each will differ in condition, lot, road and finish.

This is why a single atypical transaction can dominate a street for a year. If a thoroughly renovated house sells well above what the street has previously seen, that number becomes the anchor for every subsequent negotiation nearby, including for houses that had none of the work done. Buyers argue that the comparison is invalid, sellers argue that the street has re-rated, and both are partly right. The market resolves it slowly, over the next several sales.

Volatility that is really sample size

Compare two consecutive quarters in a town with forty sales a quarter. If the second quarter happens to contain two large lakeside houses and the first contained none, the median will jump, and a summary will report that prices rose sharply. Nothing rose. The mix changed. This is the most common way local housing figures mislead people who are reading them carefully and in good faith.

The same effect works downward. A quarter heavy with small in-town houses looks like a correction. Where a place has genuinely distinct sub-markets, the sensible move is to ignore the town-wide figure altogether and look at the segment you care about, accepting that the sample will be small enough that a single sale matters.

The lag between agreement and evidence

A price is decided when an offer is accepted. It becomes a public record weeks or months later, when the deed is filed. Any summary of recorded sales therefore describes conditions that prevailed before it was written, and in a period when borrowing costs are moving, the gap can be large enough to invert the story. A set of records showing strong prices may be describing agreements made under quite different conditions.

Nothing about this is a defect. It is simply what a market of infrequent, negotiated, individually financed transactions in unique goods looks like. Reading it well means asking three questions of every figure: how many sales is it built from, are they comparable to each other, and when were the prices actually agreed.