ESS Standard for Quality Reports Structure (ESQRS)
National Statistical Institute of Bulgaria
|Contact organisation unit|
Consumer Prices, Housing prices and PPP Department
|Contact person function|
|Contact mail address|
2 Panayot Volov Str., 1038 Sofia
|Contact email address|
|Contact phone number|
+359 2 9857 181
|Contact fax number|
The House Price Index (HPI) is a quarterly indicator that measures the changes in the dwellings transaction prices that households acquire on the market (both newly built and already existing). It is calculated according to a harmonized approach and a single set of definitions.
The HPI is an indicator comprising two sub-indices:
• HPI of new dwellings (newly built and existing but new to the households’ sector);
• HPI of existing dwellings (trade between households).
The HPI emphasis on the market price of dwellings, so non-marketed prices are ruled out from the scope of the HPI (self-builders are excluded). All transactions are included (both cash and mortgage) acquired by households regardless of its final use, so dwellings bought by households for uses other than owner- occupancy are included (for investment, e.g. to rent it out). Price include land value.
Due to the low share of house transactions at the moment HPI is calculated only for apartments.
Luxury properties are out of the scope of HPI.
HPI is constructed and produceed according to Regulation (EU) 2016/792 on harmonised indices of consumer prices and the house price index and the Commission Regulation (EU) No 93/2013 concerning harmonised indices of consumer prices, as regards establishing owner-occupied housing price indices.
National HPI series are published according to the following categories:
H.1. Purchases of dwellings
H.1.1. Purchases of new dwellings
H.1.2. Purchases of existing dwellings
Regional HPI series are published for the following levels:
- statistical zones according Classification of Territorial Units for Statistics in Bulgaria (NUTS1)
- statistical regions according Classification of Territorial Units for Statistics in Bulgaria (NUTS2)
- for the six cities with population more than 120 000 inhabitants
ESA 95 sector S. 14 (Household sector). HPls cover residential properties purchased by households in the reference period irrespective of the type of area in which they live, their position in the income distribution and their nationality status.
|Statistical concepts and definitions|
The published data are as follows:
1) House price indices (HPI) by type of purchased dwellings (new/existing/total HPI), national level
• HPI, national level, 2015 = 100
• HPI, national level, previous quarter = 100
• HPI, national level, corresponding quarter of the previous year = 100
2) House price indices (HPI), by type of purchased dwellings (new/existing/total HPI), regional level
• HPI, by statistical regions, 2015 = 100
• HPI, by statistical regions, previous quarter = 100
• HPI, by statistical regions, corresponding quarter of the previous year = 100
• HPI, for the six cities with population more than 120 000 inhabitants, 2015 = 100
• HPI, for the six cities with population more than 120 000 inhabitants, previous quarter = 100
• HPI, for the six cities with population more than 120 000 inhabitants, corresponding quarter of the previous year = 100
Market price of a dwelling purchased by a private household.
Any published index or rate of change refers to the households' expenditures for the acquisition of residential property on the territory of Bulgaria.
The statistical population consists of households' expenditures spent for the acquisition of residential property on the economic territory of the country. Household sector covered in HPI includes all persons or groups of persons irrespective of the type of area in which they live, their position in the income distribution and their nationality status.
HPI covers all monetary transactions of residential properties (new and existing apartments) on the territory of the country, purchased by private domestic and nondomestic households (i.e.'domestic concept).
HPI covers the territory of the Republic of Bulgaria
Н.1. Total HPI
• data are available since the Q1 2005;
• data for Q1 2005 - Q4 2008 are based on estimates;
• data for Q1 2009 - Q4 2014 are provisional indices of dwellings prices. The data source are: a quarterly pilot survey of newly built dwellings' prices, and a quarterly survey "Market price of dwellings" carried out by BNSI between 1993-2014;
• since 2015 BNSI has started regular production of HPI for new and for existing dwellings according to a harmonised methodology and coverage in compliance with Commission Regulation (EU) No 93/2013, therefore breaks in time series must be considered in Q1 2015.
Н.1.1. New dwellings
• data are available since Q1 2005;
• data for Q1 2005 - Q42008 are based on estimates;
• data for Q1 2009 - Q4 2014 are provisional indices of new dwellings prices. The data source is a quarterly survey of newly built dwellings' prices;
• since 2015 BNSI has started regular production of HPI for new dwellings according to a harmonised methodology and coverage in compliance with Commission Regulation (EU) No 93/2013, therefore breaks in series must be considered in Q1 2015.
Н.1.2. Existing dwellings
• data available since Q1 2005;
• data for Q1 2005 - Q4 2014 are indices of existing dwellings prices. The data source is a quarterly survey "Market price of dwellings" carried out by BNSI between 1993-2014.
• since 2015 BNSI has started regular production of HPI for existing dwellings according to a harmonised methodology and coverage in compliance with Commission Regulation (EU) No 93/2013, therefore breaks in time series must be considered in Q1 2015.
Regional HPI series are available since Q1 2015.
HPI are based on a quarterly survey - “House price indices”, that gives information on the market prices of the dwellings and on the major price determining characteristics.
HPI are based on a sample of enterprises with main economic activity in the field of real estates: real estate agencies, construction enterprises and investors. The sampling is done by using information from the Business Register to BNSI, the annual reporting of enterprises and the Register of the National Real Estate Association.
The sampled units are selected using non-probability sampling techniques: so called ‘purposive’ or ‘cut-off’ sampling. In the sample are included:
• real estate agencies (68.31 NACE 2008) - in the big cities the agencies with turnover exceeding a certain threshold are included; in the rest of the towns all active real estate agencies are selected;
• residential construction companies - includes enterprises with a turnover exceeding a certain threshold;
• residential construction investors - includes enterprises with a turnover exceeding a certain threshold.
Sampling is done annually. The initial sample for 2018 consists of 3168 sampling units. Experts from regional structures of NSI may add respondents to this sample through the whole reporting year.
The Property Register of the Registry Agency is the main source for the calculation of HPIs weights.
|Frequency of data collection|
Since 2016 data collection is based on a new information system on the site of NSI - IS Prices of dwellings, where firms can report online and in real time. Remains the possibility for the respondents to provide information through questionnaires on paper or in electronic format (in excel), sent by post or by email to the responsible experts from department "Statistical surveys" in RSO.
Depending on the type of transaction data are recorded using two questionnaires:
• Market prices of new and existing dwellings, municipal and state private property purchased by the households;
• Market prices of exsisting dwellings (trade between households).
The following information on the market prices of the dwellings and on the major price determining characteristics is collected:
- month of the transaction,
- locality of the dwelling,
- closed type complex ('yes'/'no'),
- type of dwelling ('apartment', 'house', 'other'),
- stage of construction works (for the newly built dwellings),
- age of the building (for the existing dwellings),
- type of construction (for the existing dwellings),
- dwelling’s condition (for the existing dwellings),
- development of district's infrastructure/public utilities (for the new dwellings),
- level of construction execution – standard/extra/luxury (for the new dwellings),
- total area in square meters,
- total area of the yard (for the houses),
- availability of garage or parking space,
- number of rooms,
- furnished dwelling or not,
- service tax for the transaction,
- date of the announcement for sale (new, since 2016),
- initial price (new, since 2016).
Data quality checks and validation work is carried out on both regional (in the regional structures of NSI) and national level. The experts in the regional structures validate the information from the respondents and perform initial control and checks for coding and typing errors. The main part of the work is performed by the "Consumer Prices, Housing prices and PPP" department in the central office of NSI. Data quality checks focusses on big deviations in the average price for the strata (extreme and unusual price levels/changes); missing prices; missing characteristics of the dwelling; changes in dwelling’s specifications, etc. Consultations with data providers are made and corrections are done where needed. There is no automatic rejection of observed prices in this validation process. Each problematic price is considered individually and any necessary modifications are made only on the basis of relevant information. Additional information is required from the sampled units without transactions within the observed quarter. The reasons for none-response must be filled in by the respondent: “no sales within the quarter”, “other type of activity”, “sales of land only”, etc., or by the RSO’s expert: “not find on the address”, “refuse to give information”, etc. None-response rates by regions and total non-response rate are calculated and the reasons are analysed.
Bulgarian HPI cover only purchases of apartments. The purchases of houses are excluded due to their insignificant number of transactions.
HPI are calculated as a chain-linked Laspeyres type price index. Indices are constructed both on national and on regional level using stratification combined with hedonic method to take into account the quality changes.
Major stages of HPI construction:
• calculation of the price per square meter for each dwelling;
• calculation of the average price and index for the strata at the lowest level of stratification by Dutot formula (missing prices are imputed);
• aggregation of indices from the lower levels of stratification to the regional level and high levels to produce an index of the prices of new dwellings, price index of existing dwellings and total HPI are done using Laspeyres formula.
The stratification method is appled using the following stratification variables and levels of aggregation:
• newly built and existing dwellings - level 1;
• locality 1: statistical region (NUTS II) - level 2;
• locality 2: district (NUTS III) - level 3;
• other: detailed stratification by micro location (city centre/broad centre/neighbourhood), stage of completeness, type of construction and etc. - level 4.
HPIs for 6 districts (NUTS III) with sufficient number of transactions namely: Burgas, Varna, Plovdiv, Sofia city and Stara Zagora (for newly built dwellings) and Sofia city (for existing dwellings) are calculated using Hedonic regression (characteristics method). The dependant variable is a logarithm of the price, the regression factors (locality, total area, micro location and etc.) are defined in the beginning of each year based on the purchases made in the previous year. The coefficients for year t are calculated in year t-1 and are constant during the whole year t.
The use of Property Register maintained by RA in combination with data from the former stages of OOH/HPI survey are considered as the most adequate approach for derivation of weights for the HPI and its sub-indices.
The weights at the lowest level of aggregation (level 4) are calculated using the transaction information from the quarterly surveys from the previous three years. The weights for level 3 are derived from the Property Register data in previous year (t-1). Detailed weights for these two levels of aggregation are updated every three years.
The weights for level 2 (the 28 districts - NUTS III) and for the level 1 (type of purchase: of newly built and existing dwellings) for year t are calculated using the information from the Property Register for the preceding three years (t-3, t-2, t-1), and each year is price updated to fourth quarter of t-1.
Each year HPI is calculated using fourth quarter of the previous year=100 as base reference period and are rebased to 2015=100 using the chain linked method.
No seasonal or calendar adjustments are made.
A combination of detailed stratification and hedonics is applied to adjust the quality changes in the prices of the dwellings, as described in point Data compilation.
All data are checked for completeness and consistency with statistical concepts and definitions. HPI are produced in compliance with HPI methodological requirements of the Owner-Occupied Technical Manual.
The quality of the HPI index is routinely reviewed using a framework that is based on the OOH technical manual and the European Statistical System (ESS) definition of quality.
Data on the house price indices are used for analysis, prognoses and management purposes by the government institutions, agencies and other organizations.
The main national users are: BNB, Ministries, National and State Agencies; information media;
Major international user are : Eurostat, ECB, IMF, UN ECE.
No missing data. BNSI produces and delivers the full set of HPI indices and weights.
Due to the low share of house transactions at the moment HPI is calculated only for apartments.
|Data completeness - rate|
|Accuracy and reliability|
The HPI data collection is based on a sampled survey. The type of survey, sampling and the price collection methods are designed to ensure sufficient coverage and timeliness. The HPI are with good accuracy assured by strictly following Eurostat's methodological recommendations and regulations.
HPI are produced using non-probability (purposive) sample which makes it difficult to assess the sample error. Therefore and due to the complexity of price index structures NSI does not produce estimates on sampling errors. Nevertheless, NSI aims to avoid possible bias due to sample misrepresentation by using a sample of respondents (real estate agencies, construction firms and investors) including the ones with the bigger share of revenues according the annual reporting of enterprises.
|Sampling errors - indicators|
Non-sampling errors are not quantified for the HPl. NSI tries to reduce non-sampling errors through continuous methodological improvements, survey process improvements and validation methods which help us to avoid coding and typing errors.
HPI are based on a sample of enterprises with main economic activity in the field of real estates: real estate agencies, constructuion enterprises and investors. All sample units from the target population are included in the frame population of the survey.
|Over-coverage - rate|
|Common units - proportion|
Information system Prices of dwellings (IS Prices of dwellings) is used to enter data on prices and characteristics of the transacted dwellings. Some technical controls are implemented to prevent against typo errors, wrong input of prices, measure units and characteristics. Nevertheless additional data quality checks and validation work is done in regional statistical offices and central office as described in point Data validation.
|Non response error|
|Unit non-response - rate|
|Item non-response - rate|
Data on the dwellings transactions are stratified according pre-defined criteria using IS Prices of dwellings. An IT system is developed and performs indices calculations. Logical controls are implemented in the software to minimise/eliminate the processing errors. Experts from the central office of NSI calculate the imputation prices and indices in excel.Validations and the stages of HPI constuctions are done as described in p.Data validation.
|Imputation - rate|
|Model assumption error|
|Data revision - policy|
HPI are revisable under the terms of Commission Regulation (EC) No 1921/2001 of 28 September 2001. Published HPI may be revised for mistakes, new or improved information and methodological improvements.
|Data revision - practice|
HPI are published as preliminary and are usually confirmed or revised in the following quarter.
|Data revision - average size|
|Timeliness and punctuality|
Data are published according to NSI Release calendar - 90 days after the end of the reference quarter.
|Time lag - first results|
HPI are final when first published - 90 days after the end of the reference quarter.
|Time lag - final results|
HPI are published 90 days after the end of the reference quarter.
Data are always delivered and published on the pre-announced release dates.
|Punctuality - delivery and publication|
|Coherence and comparability|
|Comparability - geographical|
HPI data are calculated on both national and regional levels. Regional stratification is performed in the construction of the indices and regional HPIs are also available.
HPI is based on harmonised definitions and classifications set by series of legal acts. Therefore, the index can be considered to be comparable to the other EU Member States HPIs.
|Asymmetry for mirror flows statistics - coefficient|
|Comparability - over time|
In principle HPI are comparable over time. Nevertheless, from the first construction of HPI till now, the methodology has constantly developed and improved and therefore some breaks in time series must be considered:
• data for 2005 - 2008 is based on estimates;
• data for 2009 - 2014 are preliminary HPI and the data sources for indices construction are: (1) quarterly pilot survey of newly built dwellings' prices and (2) quarterly survey "Market prices of dwellings" carried out by NSI between 1993 - 2014;
• since 2015 NSI has started regular production of HPI for new and existing dwellings according to a harmonised methodology and coverage in compliance with Commission Regulation (EU) No 93/2013. Since Q1 2015 HPI are comparable over time.
In compliance with Regulation (EC) No 2015/2010 since 2017 the base year for HPI has been changed and the all indices have been calculated and published at 2015 as a base year.
|Length of comparable time series|
Since 2015'Q1 until 2017'Q4 - 12 quarters.
|Coherence - cross domain|
Only one HPI set is available.
|Coherence - sub annual and annual statistics|
Annual HPI are derived from the quarterly indices and are fully reconcilable.
|Coherence - National Accounts|
|Coherence - internal|
HPI are internally coherent. Higher level aggregations are derived from detailed indices in accordance with strictly-defined procedures.
|Accessibility and clarity|
Electronic publication 'Key indicators for Bulgaria'
HPI are available to all users on the NSI website under the heading Housing price statistics: http://www.nsi.bg/en/node/13023
|Data tables - consultations|
|Metadata - consultations|
|Documentation on methodology|
EU and EC regulations:
|Metadata completeness – rate|
Quality report in ESS Standard for Quality Reports Structure (ESQRS) - http://www.nsi.bg/en/content/15357/метаданни/house-price-index-hpi
|Cost and burden|
|Confidentiality - policy|
• Law on Statistics;
• Regulation (EC) No 223/2009 on European statistics (recital 24 and Article 20(4)) of 11 March 2009 (OJ L 87, p. 164), stipulates the need to establish common principles and guidelines ensuring the confidentiality of data used for the production of European statistics and the access to those confidential data with due account for technical developments and the requirements of users in a democratic society.
|Confidentiality – data treatment|
lndividual data are not published in accordance with article 25 of the Law on Statistics. The publishing of individual data can be performed only in accordance with article 26 of the same law.