Housing stock value

Where I blithely muttered earlier about a return to [the] 1980s peak values for houses in terms of GDP, commenters both at Felix's Reuters House of Fun and elsewhere have noted that houses are, you know, bigger than they used to be. Better, even. To the extent homes nationally are bigger and better than they were in 1989, they should be worth more. The 1989 housing %GDP peak was 116%. We're at 127% now. Maybe that's fair value.

The million-dollar REO around the corner from me was built in the 1950s, mind you. But maybe.

I did want to nail down another moving part I had ignored: total units. Have we just built up in a way the 1950s never knew? I deflated the flow-of-funds assets data by CPI ex-shelter and divided by total units, and came up with a surprisingly tight fit to my naive %GDP chart. It would seem recent price movements swamp everything else. Intriguingly, this was not true of the 1989 peak: that one is much moderated in unit terms. Per unit, we are 40% above 1989 in real dollars per unit, which makes me wonder a little.

If that REO next door would only drop those 40% tomorrow, I'd be in the market.