What actually moves house prices

Borrowing capacity, not price, is what people shop for

Almost nobody buys a house with the money they have. They buy it with the money a lender will advance plus whatever deposit they have assembled, and the size of that advance is decided by an affordability test applied to their income. This single fact reorganises the whole subject: what buyers are competing over is not really houses, it is credit. When the maximum advance against a given income rises, the amount that can be bid rises with it almost immediately, and prices follow. When it falls, bids fall, though prices are stickier on the way down because sellers withdraw rather than accept less.

Because the constraint is a multiple of income tested against a monthly payment, two things move it. One is income itself, which changes slowly. The other is the interest rate used in the test, which can change in a single afternoon. That asymmetry is why housing markets can be transformed by a monetary decision that has nothing to do with houses.

Interest rates work through the payment, not the price

Consider a buyer who can commit a fixed amount each month. The principal that amount will support depends on the rate. A rise of a couple of percentage points can reduce the supportable loan by a fifth or more, without the buyer's circumstances changing in any way. They have not become poorer; the same money now buys less house. That is the whole transmission mechanism, and it explains why rate changes show up first in the parts of the market most dependent on maximum borrowing - small flats and first purchases - and only later, if at all, in parts of the market where buyers have large deposits or are not borrowing much.

It also explains a pattern that looks paradoxical from outside: falling rates can raise prices without making housing more affordable. If everyone's supportable loan rises at once, the extra capacity is competed away into the price. The monthly payment ends up roughly where it was, the debt is larger, and the deposit needed is larger too. Cheaper credit makes housing easier to buy on a monthly basis and harder to buy on a deposit basis at the same time.

Supply moves slowly and in the wrong places

New building is the obvious lever and the weakest one in the short run. Even an unusually strong building year adds a small fraction to the stock, and the additions rarely land where the pressure is. New dwellings appear where land is available and permitted, which is generally at the edge of a settlement or on redeveloped sites, not in the established streets where demand is concentrated. A large development can therefore change the character and price of one submarket dramatically while leaving the town's average almost untouched.

The more powerful short-run supply variable is second-hand listings. When people believe prices are rising, they list, because they expect to sell well; when they believe prices are falling, they withdraw. Supply and sentiment therefore move together rather than opposing each other, which amplifies swings in both directions. This is the opposite of how supply behaves in most markets, and it is a large part of why housing overshoots.

Constraint on land is a price, not a shortage

Where building is tightly constrained, whether by geography, by protected land or by the planning system, the effect is not simply that fewer houses exist. It is that the value of permission to build becomes part of the price of every existing home. A dwelling in a place where nothing more can be built carries an embedded scarcity value that has nothing to do with its bricks. This is why two physically identical houses a few miles apart can differ enormously in price, and why the difference persists across decades rather than being competed away.

Transaction costs suppress moving

Moving is expensive independently of the price of the house. Transfer taxes, legal work, searches, a survey, lender fees, an agent's commission and the removal itself add up to a substantial sum that is destroyed on each move. That cost has a quiet but powerful effect: it raises the threshold at which moving is worthwhile, which reduces the flow, which thins the market, which makes prices noisier. It also locks households into homes that no longer suit them, because the gain from a better-fitting house has to exceed the cost of getting there. When transaction taxes change, the volume of moves usually reacts more sharply than the price does.

Sentiment is not irrational

It is tempting to treat expectation as noise on top of fundamentals. In housing it is closer to a fundamental. Buying a home involves committing to a large, illiquid, leveraged position that will take weeks to arrange and years to unwind. A buyer who believes prices will be lower in six months is rational to wait, and a seller who believes the same is rational to accept less now. Because both sides are reasoning about the same expectation, belief becomes partly self-fulfilling, and it moves faster than any of the underlying quantities.

The observable trace of sentiment is not price. It is the gap between asking and achieved, the number of price reductions, and the time properties spend listed. Those three move first.

Seasonality and the shape of a year

Housing has a genuine annual rhythm, driven mostly by school years and daylight. Listings and viewings rise in spring, hold through early summer, fall away in high summer, revive briefly in early autumn and go quiet from late autumn. This affects volume much more than price, but it affects reported price too, because the mix of what sells changes: family houses concentrate in spring, smaller units are proportionally more common in winter. A year-on-year comparison is meaningful; a month-on-month one usually is not.

What matters least

Two things are routinely given more weight than they deserve. The first is cosmetic condition, which affects how quickly something sells far more than what it sells for; buyers discount visible work at roughly what it costs plus an inconvenience premium, and no more. The second is the headline national index, which is an average of averages across markets that have nothing to do with each other. Neither is useless. Both are weaker signals than the borrowing arithmetic, the listing count and the reduction rate in the specific submarket concerned.