<img src="../../cfs-file.ashx/__key/CommunityServer.Components.SiteFiles/2102_2E00_/FHFA_2D00_Home_2D00_Prices_2D00_Q4_2D00_2011_2D00_2.png" /</p

Every census division also posted arnpositive change from November to December except the West North Centralrndivision which was down 0.9 percent.  Therngreatest appreciation was in the Mountain division where the December HPI wasrnup 2.5 percent.</p

The HPI is based on the combinedrnmortgage records of Fannie Mae and Freddie Mac and tracks repeat sales of thernsame property.  Last August FHFA beganrnpublishing an “expanded-data” HPI which adds transaction information fromrncountry recorder officers and the Federal Housing Administration’s (FHA)rndatabase.  This index fell 0.8 percentrnover the latest quarter and 2.9 percent over the four quarters it has been inrnuse.</p

FHFA notes that the trend of the tworndata-bases are generally the same, but the purchase-only index has exhibitedrnmore modest price declines over the last four quarters, having dropped 2.4rnpercent vs. 2.9 percent in the expanded series. rnIt should be noted, however, that many of the differences between therntwo series on the state level were quite significant.</p

</p

The current report includes informationrnon a study done by FHFA to investigate an anomaly in the data over time.  Each monthly and quarterly report is given asrnan estimate subject to revision and it has been noted that such revisions haverntended to be negative.  This has beenrnviewed as reasonable since data from Freddie and Fannie comes in on a rollingrnbasis and hence the estimate has a bias toward the earlier part of thernreporting period while revisions pick up more data from the end of the period.  In a period of declining prices, data at thernend of the period would tend to reflect lower prices than data from thernbeginning.   However over the last year, even as pricesrnhave flattened, these negative revisions have persisted.</p

Statisticians began to wonder if therernmight be a feature of distressed sales that were contributing to the phenomena.  Was there, for some reason, a delay inrnreporting these transactions which do tend to be discounted and thus couldrnimpact the revisions?  There was no wayrnto identify private sales where the seller was distressed or bank sales of REO,rnbut FHFA could flag sales of REO sold and also financed by  Freddie Mac and Fannie Mae.  </p

Estimated data for November resulted inrnan HPI that reflected an increase of 1.0 percent.  When that data was revised the followingrnmonth there was a downward revision to +0.7 percent.  When the distressed property theory was testedrnagainst the data it was found that “new data” for November did, in fact,rninclude a relatively large share of Enterprise-Financed Enterprise-REOrn(EFER).    When the November data wasrnfirst estimated EFER transactions accounted for roughly 6.6 percent of thernsample.  The November originatedrnmortgages that subsequently became available included roughly 9 percent EFERrnsales.  Working backward through April itrnwas found that old and new samples contained EFER sales in roughly those samernamounts.</p

FHFA said that while these two or threernpercent changes in the sample seem small, given the REO discounts involved, anrnaverage of 10 percent, they could be having a material impact on indexrnestimates, especially if other distressed sales have a similar lag before enteringrnthe system.  </p

</p

All Content Copyright © 2003 – 2009 Brown House Media, Inc. All Rights Reserved.nReproduction in any form without permission of MortgageNewsDaily.com is prohibited.