Yesterday OFHEO put out it's home price index for January, and it is up 1.9% over December. Moreover, this is the first statistically significant up movement in the index since the housing downturn started.
What's going on? The OFHEO price index is based on houses transacted or assessed under conforming mortgages (or other mortgages eligble for purchase by Agencies):
The HPI is a broad measure of the movement
of single-family house prices. The HPI is a weighted, repeat-sales
index, meaning that it measures average price changes in repeat sales
or refinancings on the same properties. This information is obtained by
reviewing repeat mortgage transactions on single-family properties
whose mortgages have been purchased or securitized by Fannie Mae or
Freddie Mac since January 1975.
So some of this may be sample selection by who comes in for refinancing -- people who are underwater because of price declines aren't going to come in to refinance (and the lesser run-up compared to the Case-Shiller HPI is also due to the exclusion of houses purchased with non-conforming mortgages) and refis are generate appraisal prices, not transaction prices, and lots of the Jan data may be from refis and not transactions, compared to December data. Oops, the data above is purchase only data (there is also an index with refi appraisals included.)
Nevertheless, houses in the conforming mortgage market are exactly where you'd expect the recovery to start, since they appreciated less and are less mispriced, and conforming mortgages are much better deals now than non-conforming mortgages. An interesting data point given how persistent price movements are. It will be interesting if this shows up for a couple more months.
Very little attention to this in the press. Casey Mulligan notices it, that is about it. But it isn't clear to me how much information ought to be in OFHEO price data given other sources the market has.
Update #1: This is weird, from the press release:
Month-to-month changes in the geographic mix of sales activity explain most of the
unexpected rise in prices in January. The January home sales reflected in the FHFA data
disproportionately occurred in areas with the strongest markets. While it is difficult to
perfectly control for changing geographic mix in estimating house price indexes, the data
suggest that if one were to remove those effects, the change in home prices in January,
while still positive, would have been far less dramatic.
This makes no sense: a well constructed HPI isn't going to be driven by these sorts of shifts in sales activity, since the weights ought to be determined by housing stock values, not sales volume. I can see this driving up measurement errors, but that ought not creating a bias. What are they doing? Are they not weighting by location? This is still important even if this is a repeat sales index calculated from appreciation since the last sale for the same house (i.e. it looks at appreciation, not prices).
Skimming the technical details, it does appear that, surprisingly, the OFHEO HPI uses a single factor model, so it treats all homes as priced by one aggregage price index and an idiosycratic factor. Or that is simply the exposition, and they add in more factors when they actually calculate the index (which is implied by some of the phrasing). But the press release implies they don't. It isn't very clear.
In any case, it is important to pay attention to factors that drive prices and transactions and how they interact:
Second, the method does not recognize that houses which are sold more frequently in any
interval may be a non-random sample of the population of owner-occupied dwellings. Empirical
evidence (e.g., Englund, et al., 1999) suggests that houses on the market are a decidedly nonrandom sample of owner-occupied dwellings.
Moreover, what housing transacts changes a lot over time and is related to the factors that drive prices.
Update #2: on the other hand, this would be price stabilization in an environment with zero points 30 year fixed rate conforming mortgages around 5% or a bit lower. Higher rates without immediate offsetting inflation may drive prices down again.
Finally, high debt financed home prices aren't a social goal. Time to start thinking how to phase out the mortgage interest tax deduction.