Repeat-Sales Methodology

Repeat-sales methodology estimates property-price change from multiple transactions on the same assets, reducing sales-mix bias while introducing distinct sample risks.

Repeat-sales methodology estimates property-price change by comparing two or more eligible transactions involving the same property and combining many such pairs in a statistical model. Matching a property with itself reduces distortion from changes in the mix of homes sold, but it does not automatically adjust for renovations, deterioration, or the fact that frequently traded properties may differ from the wider housing stock.

The method is best known for residential House Price Index (HPI) construction. It can also be applied to other assets with stable identifiers and repeated observations, although the data requirements and biases may differ.

Key Takeaways

  • A repeat-sales model uses price relatives for the same property rather than comparing different homes across periods.
  • Matching reduces sales-mix bias, but “same property” does not guarantee constant quality between transactions.
  • The classic model estimates period effects from a network of overlapping sale pairs, not by averaging individual appreciation rates.
  • Weighted versions give less influence to observations expected to contain more measurement error, often including long holding-period pairs.
  • Filters for non-market transfers, duplicate records, implausible prices, very short holding periods, and property changes are central to index quality.
  • Repeat-sales indexes exclude properties without a second eligible observation, creating possible sample-selection bias and sparse local samples.
  • New sales can create new pairs that affect earlier periods, so historical index values may be revised.
  • FHFA and Case-Shiller both use repeat transactions, but their samples, filters, weights, aggregation, and release processes are not identical.

Methodology vs. Index

The repeat-sales methodology is the procedure used to select pairs, estimate period effects, weight observations, aggregate results, and revise the series. A repeat-sales index is the numerical output of that procedure.

That distinction matters because two providers can both use repeat sales and still publish different results. Their methodologies may differ in:

  • eligible property and transaction types
  • source records and geographic coverage
  • treatment of appraisals and refinance observations
  • holding-period requirements
  • filters for distressed, non-arm’s-length, or extreme transactions
  • weighting and regression specifications
  • treatment of renovations or property-type changes
  • seasonal adjustment and aggregation
  • publication lag and revision policy

Calling an index “repeat sales” therefore describes its broad design, not every rule needed to reproduce it.

Required Data

A compiler generally needs:

Data elementWhy it matters
Stable property identifierMatches the same parcel or dwelling across records
Transaction or valuation dateAssigns each observation to a model period
Sale price or eligible valuationProvides the price relative for a matched pair
Transaction typeHelps distinguish market sales from transfers, refinances, or other events
Property type and geographySupports eligibility filters and local aggregation
Record-quality fieldsHelp identify duplicates, incomplete addresses, implausible values, or coding errors
Property-change information, when availableCan flag major renovations, demolition, subdivision, or conversion

The exact fields depend on the index. A public-record index may use deed and assessor data. A mortgage-based index may use purchase prices and refinance appraisals tied to loan records. Those sources observe different parts of the market.

How the Method Works

1. Select Eligible Observations

The compiler defines the target market before matching properties. Rules may specify:

  • residential or commercial property
  • single-family homes, apartments, or another type
  • existing homes, new construction, or both
  • arm’s-length purchases only or purchases plus appraisals
  • minimum data quality and acceptable price ranges
  • geographic and time coverage

Transfers without a meaningful market price, duplicate records, incomplete identifiers, and implausible observations are normally excluded. Provider rules can also remove or separately classify foreclosures, short sales, rapid resales, and properties whose type changed.

2. Match the Same Property

Eligible records are grouped by a stable property identifier. A home sold in two periods produces one sales pair. If it sells three times, the methodology may form multiple pairs depending on the provider’s rules.

Accurate matching is not trivial. Parcel identifiers can change after subdivision or consolidation, addresses can be recorded inconsistently, and condominium unit numbers can be missing. A false match combines different properties; a missed match discards usable information.

3. Calculate the Log Price Relative

For property (i), observed in earlier period (s) and later period (t), a simplified repeat-sales return is:

$$ r_i = \ln\!\left(\frac{P_{i,t}}{P_{i,s}}\right) $$

where:

  • (P_{i,s}) is the earlier price or eligible valuation
  • (P_{i,t}) is the later price or eligible valuation
  • (r_i) is the continuously compounded price change across the full interval

The logarithm turns a price ratio across several periods into an additive change. It does not reveal how much of that change occurred in each intervening month or quarter; the regression estimates those period effects using all overlapping pairs.

4. Estimate Period Effects

A simplified model is:

$$ r_i = \sum_{k=1}^{T} x_{i,k}\beta_k + \varepsilon_i $$

where:

  • (x_{i,k}) identifies the earlier and later periods for pair (i), commonly with (-1) at the purchase period, (+1) at the resale period, and 0 elsewhere
  • (\beta_k) represents cumulative log price movement relative to the base period
  • (\varepsilon_i) captures property-specific change, data error, and other unexplained effects

The model uses many pairs with different start and end dates. Together, they form a network that lets the regression infer market-level movement across periods.

If the base index is 100, a simplified index level can be recovered as:

$$ I_k = 100e^{\beta_k} $$

Production methodologies can use different parameterizations, weighting stages, aggregation rules, and robustness procedures. These equations explain the core logic; they are not a substitute for a specific provider’s methodology document.

5. Weight Noisy Pairs

An unweighted model gives each eligible pair the same regression weight. A weighted repeat-sales model recognizes that error variance may differ across observations.

Longer holding periods often create more opportunity for unobserved renovations, deterioration, or other property-specific changes. A common three-stage approach is:

  1. estimate an initial repeat-sales regression
  2. model the squared residuals as a function of the time between observations
  3. re-estimate the index using weights based on the predicted error variance

Pairs expected to be noisier receive less influence. Some production indexes add robust weights to limit extreme observations or use value weights when aggregating market segments. The phrase weighted repeat sales does not identify one universal weighting formula.

6. Aggregate and Normalize

Local estimates may be combined into regional or national indexes. Providers can weight component areas by housing stock, transaction value, market value, or another measure. The aggregation choice determines how much influence each market has on the published total.

The final index is normalized to a base such as 100. Changing the base rescales index levels without changing period-to-period growth rates.

A Simplified Example

Consider three eligible sales pairs:

PropertyEarlier observationLater observationRaw pair change
AQ1: $300,000Q2: $315,0005.0%
BQ1: $400,000Q3: $436,0009.0%
CQ2: $250,000Q3: $260,0004.0%

Property B spans both intervals. The regression uses all three pairs to estimate a Q1-to-Q2 market effect and a Q2-to-Q3 market effect that best fit the observed log price relatives. It does not simply average 5%, 9%, and 4%, because those changes cover different periods.

For illustration, if the estimated effects were 5% in the first interval and 4% in the second, an index beginning at 100 would chain as follows:

1Q1 index = 100.0
2Q2 index = 100.0 x 1.05 = 105.0
3Q3 index = 105.0 x 1.04 = 109.2

That chain is illustrative rather than an estimate from the small table. A production index uses far more observations, statistical weights, quality filters, and explicit aggregation rules.

Data Filters and Quality Controls

Repeat matching alone does not create a defensible index. Common controls include:

  • Arm’s-length screening: removing family transfers, gifts, and transactions that may not reflect market value.
  • Property-type consistency: excluding pairs where the recorded type changes between observations.
  • Price validation: removing zero, extremely low, duplicate, or implausible values.
  • Holding-period rules: excluding same-period or very rapid resales that may reflect recording problems, distressed activity, or non-market circumstances.
  • Extreme-change treatment: excluding or down-weighting price relatives more likely to reflect bad records or major property changes.
  • Address and identifier checks: ensuring that both records refer to the same asset.
  • Geographic consistency: assigning each property to the correct and current market boundary.
  • Minimum sample requirements: withholding granular estimates when too few pairs support a stable result.

Filters improve consistency but can also change the represented population. Removing distressed sales, for example, may answer a useful question while producing an index that no longer describes the full transaction market. Users need the exact rule set, not only the index label.

Why Repeat-Sales Indexes Are Revised

The method naturally uses information that arrives over time. A property’s first transaction does not become a repeat pair until another eligible sale or valuation occurs. When the later observation arrives, the new pair adds information about price movement since the earlier date.

Revisions can also result from:

  • delayed deed, mortgage, or appraisal records
  • corrected prices, dates, addresses, or property identifiers
  • seasoned loans or older transactions entering a mortgage dataset
  • updated seasonal factors or geographic weights
  • revised filters or methodology

For historical research, the current series may differ from the data available to an analyst on the original decision date. Retaining release vintages is therefore important when testing forecasts or policies in real time.

Repeat Sales vs. Other Methods

QuestionRepeat salesHedonic regressionStratification or mix adjustment
How is quality controlled?Match each property with itselfModel observed property characteristicsCompare prices within defined groups
Must a property sell more than once?YesNoNo
Characteristic data requiredRelatively limited, beyond matching and filtersOften extensiveEnough to define useful groups
Main coverage riskRepeat-sale properties may not represent all housingMissing or mismeasured attributesQuality can still vary within groups
Renovation treatmentUsually imperfect unless changes are observed or filteredCan include observed improvementsUsually indirect
Revision tendencyNew pairs can revise earlier estimatesRe-estimation and new data can revise resultsWeight and source revisions can change history
Small-area performanceOften constrained by too few repeat pairsDepends on transaction count and attribute qualityDepends on sufficient observations in each group

No method is automatically best for every market. Repeat sales is attractive when reliable property matching and long transaction histories exist. Hedonic methods may use more observations but require richer attribute data and model choices. Stratification is transparent but may leave material differences within each group.

Major Limitations

Renovation and Depreciation

The method treats a matched property as the same asset, but the structure can change. Remodeling, additions, damage, deferred maintenance, and normal aging can make the price relative differ from pure market appreciation.

Some providers remove suspected changes or reduce their influence. Complete renovation data are rarely available, so constant quality remains an approximation.

Repeat-Sale Sample Bias

Only properties with multiple eligible observations enter the model. Frequently traded homes may differ from long-held homes in age, location, price, condition, investor ownership, or borrower characteristics.

The estimated trend therefore describes the repeat-observation sample most directly. It may not represent newly built homes, never-resold properties, or segments with low turnover.

Sparse and Uneven Data

Small areas and quiet markets may produce too few pairs for stable monthly or quarterly estimates. Transaction volume can also fall sharply during periods of financial stress, increasing uncertainty when users most want timely local evidence.

Transaction-Source Bias

A deed-record index, mortgage index, and appraisal-based index observe different events. Cash purchases, jumbo mortgages, government-backed loans, refinances, and non-market transfers may enter one dataset but not another.

Revisions and Model Dependence

Weights, filters, aggregation, and seasonal procedures are modeling choices. New records can change previous estimates, and results can differ across providers even when both use valid repeat-sales methods.

Common Mistakes

  • Using a single property’s percentage change as the index formula: An index estimates time effects from many overlapping pairs.
  • Assuming identical property means identical quality: Renovation, aging, damage, and neighborhood change can affect the second price.
  • Averaging raw pair returns: Pairs cover different periods and may require different statistical weights.
  • Treating the sample as all housing: Properties without a second eligible observation do not enter the repeat-sales regression.
  • Ignoring source coverage: Mortgage-based and public-record samples can represent different transactions.
  • Comparing index levels as currency: A level of 250 is a normalized number, not a $250,000 price.
  • Ignoring revisions: Current historical data may include information unavailable in an earlier release.
  • Using a regional index as an appraisal: Market movement cannot capture a specific property’s condition, features, and exact location.
  • Calling one implementation universal: FHFA, Case-Shiller, and other compilers apply different eligibility, weighting, and aggregation rules.

Uses in Finance

Repeat-sales indexes can support:

  • mortgage collateral and Loan-to-Value Ratio monitoring
  • regional mortgage default and loss-severity scenarios
  • housing-cycle and financial-stability analysis
  • mortgage prepayment and home-equity modeling
  • comparison of housing-price movement across regions
  • index-based updates to portfolio property values, subject to model controls

The output should be combined with property-level evidence, sales volume, inventory, income, mortgage rates, delinquency, and other relevant data. A broad repeat-sales index is not an appraisal, an affordability measure, or a forecast.

Methodology Review Checklist

Before relying on a repeat-sales index, verify:

  1. the eligible property and transaction types
  2. the source records and property-matching method
  3. the treatment of appraisals, refinances, distressed sales, and non-market transfers
  4. minimum and maximum holding-period rules
  5. filters for property changes, outliers, and duplicate records
  6. the regression specification and weighting procedure
  7. geographic aggregation and component weights
  8. index frequency, base period, and seasonal treatment
  9. publication lag and revision policy
  10. whether the sample represents the exposure or decision being analyzed

Authoritative Sources

  • Repeat-Sales Index: The numerical price index produced from matched property observations and a repeat-sales estimation method.
  • House Price Index: The broader category of residential property-price measures, which can use repeat-sales, hedonic, stratified, or other methods.
  • Case-Shiller Home Price Index: A U.S. index family constructed from eligible repeat sales recorded in public property data.
  • FHFA House Price Index: A U.S. index family using repeat mortgage transactions, primarily from Fannie Mae and Freddie Mac.
  • Appraisal: A property-specific valuation process that can supply observations to some index variants but serves a different purpose.
  • Real Estate Index: A broader index category covering property prices, rents, market activity, or investment performance.

Check Your Understanding

Loading quiz…

FAQs

What is repeat-sales methodology?

It is a statistical method that estimates property-price movement from multiple eligible transactions involving the same properties. Many overlapping pairs allow the model to estimate price changes for individual periods.

Does repeat sales fully control for housing quality?

No. It holds the property’s identity constant, which reduces differences between sampled homes. Renovations, deterioration, additions, damage, and other unobserved changes can still affect the price relative.

Why use logarithms in a repeat-sales model?

The logarithm of the later-to-earlier price ratio expresses the multi-period change in an additive form. That lets a regression estimate the period effects that best fit many pairs with different start and end dates.

What is weighted repeat sales?

It is a repeat-sales approach that gives observations different influence based on estimated measurement-error variance or other methodology rules. Long holding periods and extreme residuals are examples of factors some providers address through weights.

Why are repeat-sales indexes revised?

New sales can create new pairs linked to earlier observations. Delayed or corrected records, seasoned mortgage acquisitions, updated weights, and methodology changes can also revise history.

Can a repeat-sales index value a particular home?

Not by itself. It measures market-level movement for a defined sample. A property-specific value requires current comparable evidence and consideration of the home’s precise location, condition, features, and changes since its earlier transaction.

This article is educational and does not provide an appraisal, lending decision, investment recommendation, or individualized financial advice.

Browse Mortgages and Real Estate Finance