Experimental New Condominium Apartment Price Index (NCAPI)

Detailed information for fourth quarter 2018

The Experimental New Condominium Apartment Price Index (NCAPI) is a quarterly series that measures changes over time in the contractors' selling prices of units in new condominium apartment buildings, where detailed specifications pertaining to each unit remain comparable between two consecutive periods.

Data release - May 2, 2019

• Questionnaire(s) and reporting guide(s)
• Description
• Data sources and methodology
• Data accuracy

Description

The Experimental New Condominium Apartment Price Index (NCAPI) is a quarterly series that measures changes over time in the contractors' selling prices of units in new condominium apartment buildings, where detailed specifications pertaining to each unit remain comparable between two consecutive periods. The survey is conducted in nine census metropolitan areas (CMAs) and six of these CMAs, as well as a six CMA composite are published as experimental price indexes.

Measuring inflation in the new condominium apartment market has inherent challenges in collection and estimation that makes this index unique in Canada and one of the few found among national statistical agencies. One defining attribute of the condominium apartment market is the high heterogeneity of the units in terms of their quality characteristics. Units within a building may differ in size and amenities, the floor they occupy or the direction they face. Each of these characteristics is relevant to the price. The statistical challenge is to extract a pure price signal amidst the high heterogeneity in characteristics, which requires identification of correct and, usually implicit, price premiums and discounts for the characteristics.

Currently, the collected data do not always or consistently include complete information on the dollar values of these adjustments. In addition, price differences of otherwise identical units, which are however in different buildings, carry yet a different set of building- and location-related premiums and discounts. Price movements from a sold-out development to a new one that is for sale cannot currently be captured given the difficulty in obtaining information that would allow the accurate estimation of these premiums and discounts. As a consequence of high heterogeneity and incomplete data collection, the accuracy of the pure price signal is unknown.

The index methods and data sources continue to be explored and tested in order to increase the sample size, particularly in Ottawa, Vancouver and Victoria; to enhance the questionnaire to more accurately capture premiums and discounts for various quality characteristics; and to refine the editing, imputation and hedonic modelling processes. For these reasons, the indexes are considered experimental at this time.

In addition to general feedback about this experimental data product, users are encouraged to provide information on potential data sources or data collection methods in order to increase the sample, particularly in Ottawa, Vancouver and Victoria, and to provide more accurate pricing for the purposes of quality adjustment. All feedback is appreciated through the feedback page (https://www.statcan.gc.ca/eng/contact/feedback?pid=1810016201).

The NCAPI is designed to be used by economists, academics and the general public to monitor trends in this important component of the construction sector. Within Statistics Canada, components of these series are used in the calculation of some elements of the Consumer Price Index. In addition, the series are used by Canada's System of Macroeconomic Accounts for deflating the value of the national housing stock.

Collection period: The collection process occurs over a two-week period beginning on the 15th of the reference month or the next business day if the 15th is not a business day.

Subjects

• Construction
• Construction price indexes
• Prices and price indexes
• Residential construction

Data sources and methodology

Target population

The target population consists of all units in new condominium apartment buildings which are listed for sale or have been sold.

The observed population includes a sub-set of all unit types sold in new condominium apartment buildings. The sub-set includes: one bedroom, one bedroom plus den, two bedroom, two bedroom plus den, and three bedroom units. Based on feedback from respondent testing and industry analysis, these unit types are representative of the unit types commonly available for sale.

In addition, the observed population is limited to units in new condominium apartment buildings in nine CMAs: Halifax, Québec City, Montréal, Ottawa, Toronto, Calgary, Edmonton, Vancouver and Victoria. Six of these CMAs (Montréal, Ottawa, Toronto, Calgary, Vancouver and Victoria) were assessed to have enough ongoing new condominium construction and sales activity to support the estimation of quarterly index movements. Based on data from the Canada Mortgage and Housing Corporation (CMHC), these six CMAs accounted for 86.7% of absorbed units in new condominium apartment buildings in 2017. An absorbed unit is a housing unit that has been completed and sold.

Instrument design

The Experimental New Condominium Apartment Price Index (NCAPI) is based on an electronic questionnaire. The questionnaire has been field-tested with potential respondents and their comments on the design and content have been taken into consideration.

The new condominium apartment price report (NCAPR) collects information on contractors' selling prices for five types of units that are commonly offered for sale in new condominium apartment buildings. The prices used in the index calculation exclude value added taxes such as the goods and services tax, the harmonized sales tax and the Quebec sales tax.

Developers are asked to report for up to four developments where they have any of the five targeted units for sale in new condominium apartment buildings. For each development, they select a building in which less than 70% of the targeted units have been sold. This ensures that selling prices can be collected repeatedly for the same development and building combination over time. For each building, developers are asked for the number of each targeted unit type that they sold in the reference month and for detailed information on up to ten units with a maximum for each unit type. They are asked to report for units which were sold closest to the 15th of the reference month. If no units were sold, list prices are collected for the targeted unit types that were available in the reference month.

In addition to prices, the questionnaire collects detailed information on the characteristics of the condominium development, the building and the units for which the respondent has agreed to report. Information on the characteristics of the condominium units is used to evaluate the dollar value of the difference in the "quality" between the units for which prices have been reported and a base unit. Reported selling prices are adjusted for these quality differences so that data collected with the NCAPR can be used to measure "pure" price change over time for a base unit. A base unit is one with no floor or orientation premium and no purchased options.

Sampling

This is a sample survey with a cross-sectional design and a longitudinal follow-up.

The initial source for the survey frame is building permits. The monthly Building Permits Survey collects data on the value of permits issued by Canadian municipalities for both residential and non-residential buildings, and the number of residential dwellings authorised (see IMDB Record Number 2802).

Building permits by themselves are not sufficient as a survey frame. Frequently, builders start selling condominium apartments in the planning phase before building permits can be acquired. Other sources used to complete the survey frame include information received from municipalities such as zoning and planning applications, builder associations, new home buyer insurance companies, and governmental/non-profit home buyer protection services, Internet research, as well as advertisements.

The NCAPI has a multi-stage sample design. Representative units are selected into the sample at each stage, based on judgement. There are unique sampling units associated with each stage in NCAPI sample selection. These include condominium developers, as well as their developments, buildings and units. In the first stage of sampling for NCAPI, a census is taken of the condominium developers identified in the survey frames for each CMA covered by the survey.

The remaining sampling stages occur when in-sample developers are contacted for the first time and when data are collected. During the initial contact, developers are asked to confirm that they are actively selling apartment condominiums units in the CMA for which they have been selected. If they are in-scope, they are requested to identify up to four developments in any of the nine surveyed CMAs. Ideally, each development must have at least one condominium apartment building which has any of the five targeted unit types and these units are less than 70% sold.

The NCAPI collects information for one building per development. In cases when more than 90% of the units are sold in the same building, developers are asked to choose another building with the most unsold units.

For each building, developers are asked to provide characteristics and pricing information for up to ten representative units. These can include up to 3 one bedroom units, 2 one bedroom plus den units, 2 two bedroom units, 2 two bedroom plus den units and 1 three bedroom unit. The exact number of units and their characteristics will vary each month depending on which units were sold or listed (if no units were sold) during the reference month.

Over time, condominium developers will enter and exit the market. In addition, new condominium developments will be launched when sales of existing ones are completed. As a result, the sample of developers, developments, buildings and units must be refreshed over time. Statistics Canada subject matter staff continuously review and add to the NCAPI sample of developers. Questions in the NCAPI help developers identify when and what criteria to use in choosing new condominium developments and buildings for price collection so that their reported selling prices continue to represent their current new condominium construction profile.

Data sources

Data collection for this reference period: 2018-01-15 to 2019-01-03

Responding to this survey is mandatory.

Data are collected directly from survey respondents and derived from other Statistics Canada surveys.

All new condominium developers in the sample are contacted to confirm that they are in-scope for the survey, to identify the condominium developments for which they will report, as well as to collect a respondent name and contact information. This pre-contact is done by computer-assisted telephone interview.

Computer-assisted telephone interviews are also used to collect the first month of data from new respondents. Interviewers request detailed information from developers for the condominium development, building and unit combinations for which they will report. They also collect prices for the targeted condominium units. Finally, they will ask for permission to pre-fill future electronic questionnaire(s) with information provided by the respondent for a previous period.

After this initial collection, respondents will be sent an electronic questionnaire to self-complete each month for each new condominium development for which they are reporting.

Telephone follow-up is done for failed edits and non-response.

View the Questionnaire(s) and reporting guide(s) .

Error detection

Error detection is an integral part of both collection and data processing activities.

During collection, edits are applied to data records to identify missing data, as well as capture and reporting errors. Respondents are asked to validate their data when these collection edits fail.

During data processing, subject matter specialists review the collected data to identify and resolve errors, inconsistencies and outliers. This review focuses on the behaviour of the reported price changes, both to validate them directly and to ensure they are representative of new price movements generally.

Imputation

Missing data are generally estimated by a systematized imputation process. In any given period, price data may not be available for estimation. In such cases, missing data are imputed using the average price movement of remaining units within the same CMA (overall mean or targeted mean imputation method).

Estimation

The NCAPI is comprised of both elemental and aggregate calculations. The elemental index is calculated using the Jevons formula (equally weighted geometric mean) combining price relatives across a CMA. The aggregate NCAPI is the weighted combination of city estimates to a six census metropolitan area composite total. The index used for this calculation is a fixed weighted Laspeyres type (Lowe) formula.

Sources of weights
Price relatives for all units in all buildings in a CMA are given equal weight in the calculation of CMA-level indices. Weights to aggregate the CMA-level indices are determined by economic activity, which is derived from the value of sales of new condos in each CMA in the preceding three years (i.e., sales weights). These weights are updated annually, and introduced with the release of the second quarter indices. As an example, weights for the 2017 indices come from the value of new condos sold in each CMA during the 2014, 2015, and 2016 calendar years, with weights for the 2018 indices covering sales in each CMA from 2015 to 2017.

The weight assigned to each CMA is the ratio of the total value of sales in that CMA over the preceding three years to the value of sales in all CMAs in the NCAPI over the preceding three years.

Data for the weights come from CMHC's Market Absorption Survey, which gives a census of newly built residential properties for each CMA in the NCAPI. Sales data for condos in the Market Absorption Survey (sales date and price) are used to calculate the weights for each CMA.

Quality evaluation

Prior to publication, the NCAPI is analyzed for comparability with historical trends, as well as for coherence with results from related economic indicators, known current events and information from other external price index sources.

Disclosure control

Statistics Canada is prohibited by law from releasing any information it collects which could identify any person, business, or organization, unless consent has been given by the respondent or as permitted by the Statistics Act. Various confidentiality rules are applied to all data that are released or published to prevent the publication or disclosure of any information deemed confidential. If necessary, data are suppressed to prevent direct or residual disclosure of identifiable data.

Collected data are converted to price indexes and data are released as such, so that it is not possible to identify the business supplying pricing information.

Revisions and seasonal adjustment

With each release, data for the previous quarters may have been revised. The indexes are not seasonally adjusted.

Data accuracy

The statistical accuracy of this index depends on price and value data. Price data are obtained from a sample survey. The accuracy of the value data depends on the quality of CMHC's Market Absorption Survey. Both data sources are subject to their own errors.

Though the NCAPI uses a sample survey methodology to obtain the necessary information, confidence intervals are not currently estimated, due to the longitudinal nature of price index series.

Indexes for all levels of aggregation are considered to be statistically reliable unless otherwise denoted with a quality indicator value.

Response rate
The average monthly response rate for builders in the survey for 2018 was 77%.

Non-sampling error
When the detailed specifications of a condo unit changes between two consecutive periods, Statistics Canada uses values provided by the builder to render the prices for the unit comparable over time. When inaccuracies in these dollar values occur and are not detected, they are a source of non-sampling error.