The most common measure of a property market is the house price index, which used to track the price of real estate over time.
The three most common types of real estate price indices are the House Price Index (HPI), the Land Price Index (LPI) and Commercial Property Price Index (CPPI). The HPI measures the price of residential property, while the LPI and CPPI measure the price of land and commercial property.
In this article we will explore the HPI.
Table of Contents
Unlike other goods and services that have a fixed price per unit, each sold property is always priced differently. Furthermore, there are no transaction prices every month/quarter on the same property. Instead, there are infrequent transactions on diverse properties.
Therefore, simplistic price indices such as average or median price will provide a noisy estimate of price levels and changes. The set of houses sold in every period is small and not necessarily representative of the total stock of housing. Changes in the mix of properties bought and sold each period will therefore affect the sample median price.
An even bigger problem is systematic bias. A median price index can be biased if the quality of the housing stock changes over time. Furthermore, median price index can be upward biased if the average quality improves over the years. Bias can also arise if certain types of houses sell more frequently than others and at the same time exhibit different price changes.
Some of the variables that might impact the price of a property include:
- the location of the property,
- the number of bedrooms and bathrooms,
- the size of the property,
- the age of the property,
- the presence of features like a swimming pool or a fireplace,
- the crime rate in the area
Consequently, statisticians and data scientists have devised several approaches which take the property features into account when modelling real estate prices. The image below shows House Price Indices created by modelling prices based on features of properties sold.
However, feature-driven models is not the only approach. There are several property indices that utilise other methodologies. Here is a summary of most popular techniques statisticians use to create a House Price Indices:
- Stratification / Mix-adjustment – Separate the total sample of houses into a number of sub-samples or strata. For example, the sub samples can be houses of the same age, number of bedrooms, location etc. The overall price is calculated as an weighted average of sub-samples.
- Repeat Sales – Utilises information on properties sold multiple times. This approach requires that there are multiple sales for the same property over a reporting period. As the resulting price change is for the same property, there is no need to rely on other features to model the overall price for a particular market.
- Repeat Sales with Appraisals – Uses assessed values as proxies for selling prices in a repeat sales methodology. This is a variant of the repeat sales approach, where artificial data is created based on assumed value of existing stock in the market. In other words, we pretend that a property sale took place, estimate it’s price and add it to the sample.
- Hedonic Regression – Models heterogeneous properties by their characteristics. In layman terms, hedonic regression aims to establish a typical price of properties by removing the added hedonic value that the features of the property and the surrounding environment create. For example, a swimming pool, neighbourhood vibe, schools in the area etc.
Hedonic Regression: The Queen of all House Price Indices
To re-iterate, the number one problem for real estate price indices is that changes in the average price of properties reflect in part changes in the quality-mix of properties transacted. For example, there are more 2-bedroom apartments sold in the current period than in some reference period.
Hedonic regression tackles this problem by determining the (marginal) value of an additional unit of each feature, such as the number of bedrooms, bathrooms, square footage of property, parking, postcode, proximity to a train station, quality indicators of local school, and so forth. However the features do not have a market price, only the property as a whole.
In comparison to other indices, hedonic regression is the most advanced and also most accurate approach in estimating real estate prices. It solves the underlying problem using a statistical method that takes into account not just properties sold but also all of the possible features of these properties.
Hedonic price indices answer questions like: “What is the average price increase for a property that has an additional bedroom?” or “What is the average price difference between properties in a high-crime and a low-crime neighbourhood?”
Unfortunately, hedonic regression requires a large enough dataset with a good mix of sold properties (and features of these properties). Because of this, hedonic price indices commonly only exist for large property markets i.e. capital cities, regions, states and countries.
How are the House Price Indices Used?
Price indices can be useful for a variety of reasons. For example, they can help identify trends in the real estate market, and it can help to identify which areas of the market are growing or shrinking.
Policymakers often use house price indices when making decisions about things like housing subsidies or zoning regulations.
Additionally, economists use price indices to measure the rate of inflation in the real estate market. This is important because it can help to identify when the market is becoming more or less expensive.
In summary, here are possible uses for a house price index:
- as a macro-economic indicator of economic growth;
- for use in monetary policy and inflation targeting;
- as an input into estimating the value of housing as a component of wealth;
- stability or soundness indicator to measure risk exposure;
- support individual’s decision making on whether to buy (or sell) a residential property;
- as an input into the consumer price index, which in turn is used for wage bargaining and indexation purposes;
- for use in making inter-area and international comparisons.
Most Well-known House Price Indices
The S&P Case-Shiller Index is probably the most well-known among other HPIs around the world. It is the leading measure of U.S. residential real estate prices. The indices are published on a monthly basis and are constructed to accurately track the price path of typical single-family homes located in each metropolitan area.
The Case-Shiller indices are not calculated using the sale prices of all homes, but rather the sale prices of homes that have been sold two or more times. This methodology is used to ensure that the indices reflect only price changes for typical homes and are not affected by changes in the composition of the sample of homes sold.
The UK House Price Index is an index compiled by the UK Office for National Statistics (ONS) that measures the average change in the prices of residential properties sold in the UK. It contrast to the U.S. Case-Shiller Index, it uses a hedonic regression model to produce the indices.
The Australian house price price index is published by the Australian Bureau of Statistics (ABS) every quarter. The index measures the movement in prices of residential dwellings in Australia’s eight capital cities. ABS calculates the index using a hedonic regression model, which takes into account the characteristics of dwellings that influence prices, such as location, size, age and type.
|Residential property prices||Jun Qtr 21 to Sep Qtr 21 % change||Sep Qtr 20 to Sep Qtr 21% change|
|Weighted average of eight capital cities||5||21.7|
The most recent release of the ABS’ real estate price index showed that prices increased by 5.0% in the September quarter of 2021, following a 6.7% increase in the June quarter. Hobart and Sydney were the two cities with the biggest price increases in the September quarter, with prices in Hobart up by 8.2% and prices in Sydney up by 6.2%.
Typical Price by HtAG Analytics
Although house price indices are available for major capital cities and some regions in Australia, prices for smaller property markets continue to be reported using Median Price. As a consequence, the problems we highlighted at the start of the article become highly relevant for reported LGA or suburb prices.
Suburb median prices are frequently “noisy” and biased, which makes them unsuitable to support decision making. We visualise the problem in this article about Typical Price and also explain the solution.
Typical Price represents values for houses or units pertinent in Australian LGAs and suburbs. HtAG calculates the prices using the stratification methodology coupled with a data fitting technique. Prices are first calculated for number of bedrooms, type and suburb strata using a data fit method. We then calculate the overall Typical Price using a weighted average of the sub-stratum.
Typical Price for Gold Coast LGA. Source: HtAG Analytics
The resulting metric, called Typical Price is available for over 5,000 suburbs and 400 LGAs country-wide and more correctly represents house prices and their changes over time in council area and suburb property markets. It is widely used by real estate professionals and property investors to support decision making when assessing property markets for investment potential.
Typical price is a revisionary metric and is recalculated every month for the current month and all previous month. Despite its revisionary nature, the Typical Price offers a more accurate representation of the current market and long-term trends, when compared to median prices.
There are several ways to calculate a House Price Index (HPI). All methodologies have both advantages and drawbacks. Notably, all HPIs only highlight prices for larger real estate markets such as cities, states and countries.
In lieu of an HPI, simplistic median price indices are used to report prices for smaller property markets, which results in noisy and biased data. HtAG’s Typical Price metric addresses this problem using an innovative methodology that combines stratification and data fitting techniques to produce HPI for suburbs and LGAs Australia-wide.
Typical Price for LGAs and suburbs as well as more than 20 other real estate metrics are available for download from HtAG Digital Store.