The Future(s) of House Prices

The UK has a new tradeable derivatives contract on house prices, which shows house prices falling by 7% next year. That has Tim Harford worried: futures markets are "better than cheap-talk forecasts" in terms of being right, he says.

But is that true? It's certainly not true in terms of currency futures, which are atrocious as predictions of future currency moves at all – they're governed overwhelmingly by differences in interest rates. I think that Harford might be confusing futures markets with prediction markets, which can be very accurate. But even prediction markets wouldn't do so well when it comes to housing, as Harford's commenter "Joe Bloggs" explains:

Property investors (whether professionals or just Joe Bloggs with a mortgage) are structurally long a somewhat illiquid asset: it's impossible to sell houses short. Consequently, everyone has the same risk, and it follows that hedging that risk is expensive. It could be that at least some of that perceived 7% is a premium that investors are prepared to pay just on grounds of general nervousness. Markets of this sort have been available in spread betting markets and have often been oversold in times of uncertainty.

Interestingly, the UK has had a prediction market in housing prices since 2002, when Nick Denton recommended going short as prices were "teetering on the edge of an ugly correction". (How did that trade work out, Nick?)

It's also worth noting that the US version of this market, the futures contracts traded on the Chicago Mercantile Exchange, are so illiquid as to be utterly useless. And it's also worth noting that all the cognoscenti were spectacularly wrong about the future direction of house prices for most of the past decade. Show me a housing futures market with any kind of track record of being right, and I might pay attention to it. For the time being, I'll stick to my conviction that nobody knows anything.