If your wonkish side was troubling you over the disparity between the Case-Shiller price index and the FHFA price index here’s the reason for the different results.

From the WSJ Real Time Economics blog:

“The FHFA uses the prices of homes backed by mortgages sold or guaranteed by Fannie Mae and Freddie Mac, excluding refinances. That means it’s only looking at homes sold within the conforming loan limits, which are set at $417,000 in most markets and rise to as high as $729,750 in the most expensive housing markets,” Nick Timiraos wrote. “The housing crisis hit the bottom end of the market first, so it makes sense that a recovery would begin there.”

Other issues may be contributing to the differences, primarily due to the index’s reliance on loans backed by federal agencies. Abiel Reinhart of J.P. Morgan Chase notes that the FHFA index is based on loan pairs of repeat transactions through one of the government agencies. “Consider what this means for a home that was bought with an agency mortgage in the late 1990s, bought again with a non-agency mortgage in the midst of the housing boom, and then bought again in the last few months. Because the middle transaction is excluded, the FHFA price index will calculate the change in the home price based on the 1990s and very recent transaction. Even after home price declines, this difference will often be positive (the FHFA index is up 57% over the last decade),” he said.

The article goes on to discuss some other more minor sources of noise in the FHFA numbers. One is pretty interesting. New home sales aren’t counted no matter the source of the mortgage. So a new home that was sold to a buyer who subsequently defaulted and the house is resold after foreclosure it won’t show up in the data. As we all know, a lot of the foreclosure resales are just that — new homes that the buyer lost and first time buyers or investors are snapping up on the cheap.

The moral of this story is that the increase in home prices that the FHFA showed the past two months is questionable. FHFA used to be my favorite index simply because it tracked repeat sales and had tons of data sets. Now it looks like the serious dislocations in the housing market may cause its methodology to produce suspect numbers. Probably better to stick with Case-Shiller as a broad measure of market conditions.