The Case-Shiller Index measures changes in U.S. existing single-family home prices using matched repeat sales and value-weighted market indexes.
The Case-Shiller Home Price Index is a family of U.S. housing-price indexes that measures changes in the value of existing single-family homes by comparing sales of the same properties over time. It is designed to track market-level price movement at a constant level of housing quality, not to estimate the current price of a particular home.
The current official name is the S&P Cotality Case-Shiller Home Price Indices. The series was previously branded S&P CoreLogic Case-Shiller and is still commonly called the Case-Shiller Index. The family includes a U.S. national index, 10-city and 20-city composites, and individual metropolitan-area indexes.
$6,000 gain.The phrase Case-Shiller Index can refer to several related series. Always identify the geography and seasonal treatment before quoting a change.
| Series | What it measures | Best use |
|---|---|---|
| U.S. National Home Price Index | A value-weighted composite of single-family home-price indexes for the nine U.S. Census divisions | Broad national housing-price trend |
| 10-City Composite | The 10 original metropolitan-area indexes combined using housing-market value weights | Long-history comparison across major metros |
| 20-City Composite | The 20 covered metropolitan-area indexes combined using housing-market value weights | Large-metro housing-price trend |
| Individual metro indexes | Average price change for existing single-family homes in each covered market | Comparing the 20 represented metro markets |
| Price-tier indexes | Low-, middle-, and high-price segments within available metro data | Testing whether price changes differ across market tiers |
The national index is not simply an average of the 20 cities. It covers all nine Census divisions and uses a separate national aggregation method. Likewise, the 20-City Composite is not a proxy for every U.S. city, rural area, or property type.
The methodology starts with public transaction records. It searches for an earlier eligible sale of the same property and forms a sales pair. The percentage change between those two arm’s-length transactions contributes evidence about market movement while holding the property’s identity constant.
flowchart TB
A["1. Match an earlier eligible sale of the same home"]
B["2. Exclude non-market or unreliable sales pairs"]
C["3. Apply interval, value, and robust weights"]
D["4. Aggregate by market in a three-month window"]
E["5. Publish the index and revise as records arrive"]
A --> B --> C --> D --> E
This process is more sophisticated than averaging individual appreciation rates. The methodology applies interval, value, and robust weights to reduce the influence of observations that are less representative of broad market movement.
The methodology considers available arm’s-length sales of identifiable existing single-family homes. It excludes or down-weights observations that may not reflect ordinary market price change, including:
This treatment improves consistency, but it cannot make every home perfectly constant in quality. Renovations, deferred maintenance, neighborhood change, and incomplete records can still affect observed sale prices.
Suppose two eligible homes in the same market have the following repeat transactions:
| Home | Earlier sale | Later sale | Time between sales | Unannualized change |
|---|---|---|---|---|
| A | $300,000 | $345,000 | 3 years | 15.0% |
| B | $600,000 | $630,000 | 1 year | 5.0% |
For Home A, the simple annualized growth rate is:
For Home B, the one-year change is 5.0%. The two observations suggest similar annual price movement, even though the unannualized gains are 15% and 5%.
The published index is not the simple average of 15% and 5%, or even the average of the two annualized rates. Case-Shiller combines many sales pairs in a repeat-sales model and applies weights for the time interval, initial home value, and whether a price change appears atypical for the market. The example only shows why transaction interval and weighting matter.
If Home A was substantially renovated between sales, part of its 15% gain could reflect a quality change rather than market appreciation. The methodology can reduce the influence of an unusual observation, but public records may not capture every physical change.
Each monthly index point uses sales-pair data for that month and the two preceding months. For example, a June index point incorporates pairs associated with April, May, and June.
The moving window increases the usable sample and helps manage delays in county recording data. It also means the index is deliberately smoother and less immediate than a measure based only on the latest month’s transactions.
Available metro data can include low-, middle-, and high-tier indexes. These are relative segments, not permanent dollar bands shared by every city.
The methodology calculates price breakpoints for each period so that, after exclusions, the eligible sales are divided into three groups of similar count. The breakpoints are smoothed over time. A repeat-sales pair is assigned using the first sale price relative to the breakpoints at that earlier date, not the later sale price.
This has several implications:
Price-tier indexes are useful for testing whether broad price movement is uneven across a metro. They are not direct measures of affordability, borrower credit quality, housing condition, or expected investment return.
The Case-Shiller indexes use January 2000 = 100 as their base. The index level shows cumulative price change relative to that base; it does not show the average sale price.
If a series reaches 250, its measured price level is approximately 2.5 times the January 2000 base level:
1Cumulative change since base = (250 / 100) - 1 = 150%
This does not mean the average home costs $250,000, that every covered home gained 150%, or that a particular property followed the index.
For a period-to-period change, compare two index levels:
1Index change = (Current level / Earlier level) - 1
Suppose a metro index rises from 300 to 306:
1(306 / 300) - 1 = 2.0%
The measured market index increased 2% over that interval. The result says nothing by itself about rent, mortgage payments, transaction costs, inflation-adjusted purchasing power, or the return earned by a leveraged homeowner.
Case-Shiller values are nominal. To estimate an inflation-adjusted change, use compatible dates and an explicitly selected price measure.
Suppose a Case-Shiller index rises from 300 to 315 while the selected inflation index rises 4% over the same period. Nominal home-price growth is 5%:
The approximate inflation-adjusted change is not simply assumed to be the nominal rate. Using the ratio of growth factors:
The result depends on the chosen inflation measure and matching period. A broad consumer-price index may not represent construction cost, owner expenses, or the purchasing power relevant to every analysis.
Two percentage changes commonly appear in housing reports:
| Measure | Calculation | Interpretation |
|---|---|---|
| Month-over-month | Current month compared with the preceding month | More responsive, but more affected by seasonality and short-term noise |
| Year-over-year | Current month compared with the same month one year earlier | Reduces normal seasonal comparison problems but reacts more slowly to turning points |
A positive year-over-year rate can coexist with a negative month-over-month rate. That combination means prices remain above their level one year earlier even though the latest monthly comparison declined.
Do not describe slower positive growth as a price decline. If the annual rate falls from 6% to 3%, prices are still rising year over year, but at a slower measured pace. A negative rate is required to describe a decline over the stated comparison period.
The underlying indexes are calculated on a non-seasonally adjusted (NSA) basis, which includes recurring seasonal patterns. Seasonally adjusted (SA) versions apply a statistical adjustment intended to remove those patterns.
Seasonal adjustment is an estimate and can be revised. The choice does not make one series universally better; it changes the analytical question.
The indexes are published monthly, generally on the last Tuesday, with index levels released about two months after the measured month. A release published in August may therefore report June data.
The delay reflects the time required for local sale records to become available and for properties to be matched and screened. Recent values may be restated when additional transactions arrive. Under the current methodology, revisions for newly received transaction data are generally limited to the latest 24 months, while seasonal factors and national-index inputs can affect a longer history.
For analysis, record three separate dates:
This prevents look-ahead bias in historical models. A value that is available today may differ from what an analyst actually knew on the original release date.
Both Case-Shiller and the FHFA House Price Index use repeat transactions, but their samples and weighting differ.
| Feature | Case-Shiller | FHFA HPI |
|---|---|---|
| Primary source data | Sales information obtained mainly from public recorder and assessor records | Mortgage transactions associated mainly with loans purchased or securitized by Fannie Mae and Freddie Mac |
| Core transaction type | Purchase prices for eligible repeat sales | Purchase-only and other index variants; some FHFA series also use refinance appraisal data |
| Weighting | Value-weighted, giving higher-value housing more influence on aggregate movement | FHFA describes its standard repeat-sales approach as weighting property price trends equally rather than by home value |
| Geographic products | National, 10-city, 20-city, and 20 individual metro series, plus additional products | National, division, state, metro, county, ZIP-code, and census-tract products |
| Coverage constraint | Existing single-family homes with eligible public sales pairs | Enterprise mortgage data create a different financing and transaction sample |
The indexes often move in the same broad direction, but short-run growth rates can differ. Neither is automatically wrong. The difference can result from geography, loan coverage, transaction eligibility, weighting, revisions, or the chosen series variant.
Housing prices affect household balance sheets, mortgage credit, construction, consumption, and financial-system risk. Analysts use Case-Shiller data to help evaluate:
The index should be combined with sales volume, inventory, days on market, rents, mortgage rates, household income, delinquency, construction, and local employment data. Price movement alone cannot explain whether a market is affordable, liquid, overvalued, or financially stable.
An analyst may use a relevant local index to estimate broad collateral movement when current appraisals are unavailable for every loan. Suppose:
$400,000$320,000A simple index-updated value is:
The estimated current loan-to-value ratio becomes:
The original LTV using $400,000 was 80%. The example shows how a market-level decline can reduce an estimated equity cushion even when the loan balance also amortizes slowly or remains unchanged.
This is a portfolio approximation, not a property appraisal. The result can be misleading if the index geography, property type, original valuation date, home condition, or price tier does not match the loan. Credit risk also depends on payment capacity, loan terms, borrower behavior, insurance, local foreclosure conditions, and other factors.
Case-Shiller is not:
A property can gain value while its metro index falls, or lose value while the index rises. Location within the metro, structure, lot, condition, renovations, property type, price tier, and transaction circumstances all matter.
$250,000 price.Before using a Case-Shiller figure, verify:
This article is educational and does not provide an appraisal, lending decision, investment recommendation, or individualized financial advice.