The FHFA House Price Index (HPI) is a set of numbers that tracks how prices of typical single‑family homes change over time in the U.S. It is a “weighted, repeat‑sales” index: FHFA finds homes that have sold (or been refinanced) more than once and compares their prices across time, then uses a statistical method to average those changes while controlling for differences in home quality. The flagship index uses data on conforming, conventional mortgages for single‑family homes that Fannie Mae or Freddie Mac have purchased or securitized since 1975, and it reports how much the value of a typical home in that sample has gone up or down over a month, quarter, or year.
“Seasonally adjusted” means the index has been statistically corrected to remove regular, predictable patterns that happen at the same time every year—such as more home buying in spring and summer and less in winter—so that month‑to‑month changes mainly reflect underlying market moves rather than normal seasonal swings. FHFA does this using the U.S. Census Bureau’s X‑13 ARIMA procedure, the same type of method many government agencies use for other key economic data.
“Purchase‑only data” means the index is based only on mortgages used to buy homes (purchase loans), excluding refinances where owners simply replace an existing loan. For the flagship FHFA HPI, the sample is limited to single‑family homes with conforming, conventional mortgages that Fannie Mae or Freddie Mac have bought or securitized. FHFA uses these data because, as the regulator of Fannie Mae and Freddie Mac, it has long‑running, detailed, and consistent nationwide records back to 1975, allowing it to build a large, stable, repeat‑sales index; other FHFA index variants add refinances, FHA loans, or county deed records, but the main “purchase‑only” series stays focused on actual sales.
The September change was revised from 0.0% to a 0.1% decline because FHFA routinely updates past HPI values as more complete data arrive. There is a built‑in delay: new mortgages typically reach Fannie Mae and Freddie Mac 30–45 days after closing, and the Enterprises also buy some older (“seasoned”) loans. Each time a property shows up again in the data (through a sale or refinance), it can slightly change the estimated average price change for earlier periods, so recent months—like September—are often revised as additional transactions are incorporated.
A 1.7% year‑over‑year increase means that, on average, single‑family homes in the conforming‑loan market that FHFA tracks are estimated to be about 1.7% more expensive than they were 12 months earlier. FHFA notes that this “one‑year” change is its best estimate of how much the value of a typical property in the sample has risen over a year, but it is a national average (local markets can differ a lot) and it is not adjusted for inflation—so if consumer prices rose faster than 1.7%, real (inflation‑adjusted) housing values may have been flat or even down.
In the census divisions FHFA cites: • West South Central division: Arkansas, Louisiana, Oklahoma, and Texas. • Middle Atlantic division: New York, New Jersey, and Pennsylvania. These are the standard U.S. Census Bureau groupings that FHFA uses when reporting regional HPI figures.
All three (FHFA HPI, Case‑Shiller, and CoreLogic) are repeat‑sales home price indexes, but they differ mainly in data coverage and methodology: • FHFA HPI uses only conforming, conventional mortgages on single‑family homes that Fannie Mae or Freddie Mac purchased or securitized, so it excludes jumbo, FHA/VA, and most cash sales. It is equal‑weighted and provides very broad geographic coverage (nation, regions, states, 400+ metros, counties, ZIP codes). • S&P CoreLogic Case‑Shiller uses public records and related data on arm’s‑length sales of single‑family homes, including non‑conforming and higher‑priced homes; it value‑weights transactions so expensive homes have more influence and focuses mainly on a national index and large‑metro composites. • CoreLogic’s HPI uses public records plus servicing and securities data, covers a wide range of loan types (including many non‑conforming loans and nondisclosure states), and is also a value‑weighted repeat‑sales index available at national, state, metro, and ZIP‑code levels. Because FHFA’s index is limited to the conforming‑loan segment and equal‑weighted, it can show somewhat different growth rates from Case‑Shiller or CoreLogic, which capture a broader and more expensive slice of the market.