July 2026 Home Price Index Gap: What Differs

The July 2026 home price index gap was 0.7 percentage point between two rounded annual headlines released September 29: FHFA’s 2.6% seasonally adjusted gain and Case-Shiller’s 1.9% unadjusted national gain. Their samples and methods also differ. The releases do not establish how much each difference contributed to the gap.

What the September 29 releases reported

Both releases concern July 2026, although their treatment of that month differs.

FHFA’s September report, page 3 identifies its figures as seasonally adjusted and nominal. S&P’s release reports its annual headline on a non-seasonally-adjusted basis; its tables show monthly national gains of 0.12% unadjusted and 0.29% adjusted.

The monthly readings therefore agree at one decimal place when adjustment status matches. The annual headline comparison does not hold seasonal adjustment constant. Neither set of figures is adjusted for inflation.

Use the national Case-Shiller series for this comparison. Its 20-city composite is a separate geographic measure. Related releases are collected in our home-price data and reports.

Which transactions enter each index?

Both producers follow repeat transactions on individual homes. Comparing the same property over time helps limit changes caused merely by a different mix of homes selling. FHFA’s March 1996 technical paper describes this mechanism, but it does not make the two samples identical.

FHFA’s July headline is its purchase-only house price index. The September report’s overview specifies conventional, conforming mortgages purchased or securitized by Fannie Mae or Freddie Mac—the Enterprises. That excludes cash purchases, jumbo loans and FHA- or VA-backed purchases. Refinance appraisals belong to FHFA’s all-transactions measure; they cannot explain a difference involving this purchase-only headline.

Case-Shiller selects qualifying recorded sales of existing single-family homes without requiring Enterprise financing, so cash and jumbo-financed transactions can qualify. The national measure excludes new construction and condominiums. Its nine-division construction also incorporates FHFA data. These rules are described in its August 2025 methodology.

“National” therefore does not mean an identical census of transactions. Identify the financing population before applying either result to a market segment.

Matching repeat transactions also leaves a separate limitation: a property’s condition can change between sales. Our discussion of renovation bias in repeat-sales indexes examines that issue.

Weighting changes whose price movement counts most

Coverage determines which transactions can enter an index. Weighting determines how much individual properties influence the estimate and how geographic estimates combine.

Case-Shiller gives higher-priced homes greater influence. FHFA does not add that property-value weighting, although its repeat-sales model does apply statistical weights. FHFA’s May 2015 methodological comparison explains the value-weighting distinction; its technical paper explains the statistical treatment.

For geographic aggregation, FHFA combines state price changes using shares of one-unit detached housing stock, as described in its May 2011 report, FAQs 25–26. Case-Shiller’s national aggregation uses housing-stock values, according to its methodology.

Consider a hypothetical market where expensive homes appreciate more slowly than cheaper homes. Giving expensive homes more influence can lower the aggregate gain. Reverse their relative performance and that influence can raise it. The mechanism alone cannot tell us why FHFA’s July headline was higher.

“July” covers different time windows

FHFA estimates separate monthly values from transactions grouped by month, as its May 2011 explanation, FAQ 24, describes.

Case-Shiller pools sales pairs over three months: July represents May through July, while June represents April through June. The overlapping windows support sample size and accommodate recording delays. They can also spread a market turn across releases. See the August 2025 methodology.

Smoothing and seasonal adjustment address different problems. Pooling combines observations across months; seasonal adjustment estimates recurring calendar effects. Comparing adjusted monthly rates improves consistency, but it does not erase differences in coverage, weighting or time windows. Our guide to understanding home-price measures provides broader context.

Choose the measure that fits the claim

For a claim about appreciation among homes financed through the Enterprises, FHFA’s purchase-only series provides the closer sample match. For a broader financing mix among existing single-family resales, Case-Shiller provides another perspective, with its value weighting and rolling window disclosed.

Neither national rate establishes a particular listing’s market adjustment. FHFA’s own July annual changes ranged from 0.6% in the Mountain division to 6.3% in the Middle Atlantic, according to its September 29 release. Regional variation is already substantial before narrowing to a neighborhood or property type.

The gap itself is not evidence that either producer is wrong. Determining its causes would require recalculating with matched samples and methods. Averaging the two headlines would create an undocumented blend, without evidence that it improves accuracy.

For a presentation based on these releases, retain the September 29 vintage and put the series name, geography, comparison period and adjustment status beside each figure. Keep both readings visible when the claim spans their different populations.