ANNEX. Reporting institution: Cyprus

Transcription

1 ANNEX I. Generic template for national reports providing information on quality, sources and methods, together with information on the statistical processes used for the compilation of the statistics underlying the MIP indicators (including concrete specifications for housing price statistics) Reporting institution: Cyprus European Code of Practice principles elected principles Public Commitme nt on European tatistics by the ECB elected elements ome specific examples for indicators and related questions (Response options, unless indicated otherwise: not implemented; partially implemented; fully implemented) tatistical domain specifications Institution al Environme nt Institutiona l Environme nt CoP/P C IMF DQA F IM Question(s) Housing price statistics

2 CoP Professiona l Independen ce PC Professiona l Independen ce./..4/ The independence of [name of reporting institution], from political and other external interference in developing, producing and disseminating official statistics is guaranteed by law. Legal basis: Pls confirm the independence by referring to the relevant national and EU legal act(s) and the formulation used in the legal act... The tatistical ervice of Cyprus (CYTAT) is the national statistical institute responsible for the production and dissemination of official statistics in the country. CYTAT functions under the Ministry of Finance; however, as explicitly stipulated in Article 2(2) of the tatistics Law, No. 5(I) of 2000, the tatistical ervice maintains its autonomy in technical matters and has exclusive responsibility for the choice of methodology, technique, definitions and procedures for the realisation of the annual and five-year programmes of statistical activities, as well as for prescribing the time and manner of dissemination of the statistical output. Furthermore, the tatistical ervice decides about the sources of information it will use, during its surveys or work. The [name of reporting institution] has an annual statistical work..2 In accordance with the provisions of Article 7 of the tatistics Law No.5(I) of programme which is 2000, the tatistical ervice produces an annual programme of statistical activities every made public. year in May preceding the reference year. The programme is published on the website of tatistics work the tatistical ervice. Website link to the Annual Work Programme, 205 : programme: Pls confirm the existence endocument of a statistics work programme and specify where to find it (or similar document), 2

3 if available. CoP2 Mandate for data collection PC2 Mandate for 2./ data collection The mandate(s) of the.2. In total, seven people are involved in the production of the House Price Index (HPI). [name of the Their full time equivalent is estimated to one full time employee. The main tasks institution(s)] to collect undertaken to produce the HPI are: data collection, processing of data, information for the validation/consistency checks, compilation of the index and transmission of the index to development, production Eurostat. The main source of the data is the Department of Lands and urveys. No data is and being purchased. dissemination of European tatistics is specified in law. Allocation 7. responsibilities: explain the allocation of responsibilities in producing the relevant statistics..2.2 of Pls -Commission Regulation on Owner Occupied Housing No 93/203 provides for the quarterly transmission of HPI data by 85 days after the end of the quarter and of the annual weights by 5 June of year following the year to which the weights relate. -The tatistics Law No. 5(I) of 2000 Legal basis: Pls -According to article 0 of the tatistics Law No.5(I) of 2000, the tatistical ervice has confirm the mandate the right of free access to the administrative records of the ministries and services of the by referring to the government of the Republic and of public corporations, to the extent to which the data to relevant legal act(s) be collected is necessary for the production of statistics. and any other kind of formal agreements. CoP6 PC6 Impartiality Impartiality and and 6.6/ Advance notice is given on release calendars and on major revisions or changes in.3. Advance release calendar: HPI is not yet released by CYTAT but it is expected to be released for the first time for the first quarter of 206. The annual release calendar is presented on CYTAT s website at: 3

4 objectivity objectivity tatistical Processes tatistical Processes CoP/P C IMF DQA F.20. IM methodologies, source data and techniques. Advance release calendar: Pls confirm the existence of an advance release calendar and provide a link to it. Revision policy: Pls explain succinctly the revision policy and provide a link to material on the revision policy, if existent. Question(s) pendocument.3.2 Revision policy: The main reasons for revisions are new or revised datasets, calculation mistakes or revisions of the compilation method. All data are clearly identified as provisional, or revised where this is applicable. At first release some data might be provisional. The data become final as soon as all data used in the compilation become final. Housing price statistics CoP7 PC7 ound ound methodolog 7./7. methodolog y y Pls specify the relevant 2.. General remarks: The actual transaction prices are used for the compilation of the statistical standards to House Price Index The methodology used is in compliance with the Methodological which the statistics Manual referred to in the Commission Regulation (EU) No 93/203 and the Residential abide. Property Price Indices Handbook. The Rolling Year Hedonic Regression Model which is Links/references to a Time dummy variable hedonic approach mentioned in paragraph 45 of the OOH/HPI existing comprehensive Manual combined with stratification is used for the production of a chain-linked index. methodological documentation on specific aspects of the national sources and 4

5 2. methods can also be provided CoP8 Appropriat e tatistical procedures PC8 Appropriate 8.2/8.2 tatistical procedures Pls specify adherence or deviations from concepts and rules according to international standards and highlight coverage gaps that are of material relevance for the statistics. Pls indicate whether actions to address the deviations and coverage gaps are envisaged. Pls provide a succinct assessment of the robustness of the statistics by elaborating briefly on the relevant sources of information, statistical methods and procedures used across the various For a number of sales, the square meters are not available. As a result, the accuracy of the index may be affected by the fact that these transactions cannot be used. Additionally, the data obtained from the Department of Lands and urveys is incomplete as regards dwellings characteristics. Therefore, the hedonic model used for the compilation of the Cypriot HPI includes only a few explanatory variables and does not adjust for some important housing standards (e.g. condition, view) and for the age of the building. We are in the process of getting some important explanatory variables for old dwellings (class, age). For new dwellings we will add two important variables, foreign citizenship and luxury (based on criteria like swimming pool, quality of materials, square meters, number of bedrooms e.t.c). ources: The only administrative data source used for both prices and weights is the Department of Lands and urveys, Ministry of Interior. The data cover all areas which are under the control of the government of the Republic of Cyprus. The data sets made available to CYTAT include new and used houses and apartments. The transactions covered are sales and contracts of sales of dwellings (actual price of each dwelling sold). The following characteristics are included in the data sets: district, town or village, quarter, registration number, declared price, accepted price, property type, agreement date, acceptance date, square meters (enclosed, 5

6 6 frequencies. covered and uncovered area of the dwelling). Pls provide links/references to trata existing comprehensive methodological Old and new dwellings documentation on Method of calculation of the index in each stratum specific aspects of the national statistical In order to calculate the sub-indices for old and new dwellings, a hedonic model of procedures. The the following form is used in each stratum: objective should be to get an overview on log p = Xβ + c t I t + c t I t +c t 2 I t 2 +ε, how much the statistical output are where p is is the price, X is the matrix of characteristics, I t, I t-, I t-2 are the time anchored in actual dummy variables for the quarter t and the previous two quarters, respectively and ε collected data rather are the error terms. than on estimation methods or similar The rolling window method uses a rolling window length of M=4 quarters. judgemental inputs. An initial regression model is estimated and the appropriate indices are calculated using data pertaining to the first M=4 quarters of the data set, where M is the window length, i.e. the number of periods used in each regression. In our case, the dataset includes all transactions of dwellings for all quarters from 200 and the initial regression is based on the first 4 quarters of 200. For the next quarter, a new regression model is estimated where the data consist of the initial data less the data for period (quarter) of 200 but adding the data for period (quarter) M+= 5, in our case the st quarter of 20. Appropriate price indices are calculated for this new regression model, but only the rate of increase/decrease of the index from period M to M+ is used to update the previous sequence of M index values. Given the estimated parameters of the 6

7 model β,c M+, c M,c M the index from period M to M+ is calculated as the ratio:. exp(c M+) / exp(c M). This procedure is continued with each successive regression dropping the data of the previous earliest period and adding the data for the next period, with one new update factor being added with each regression. Calculation of weights In order to calculate weights the following method is used: For the st quarter of each year, the annual weights of the previous year are used since the base period is the 4th quarter of the previous year. In order to calculate these weights the following formulae are used: w old, = N i= N0 P i= i N2, w new, = i=, N0 P i= i where p i, is the transaction price for dwelling i, N is the number of transactions of old dwellings for the previous year and N 2 is the number of transactions of new dwellings for the previous year and N 0 =N + N 2 is the number of transactions of all dwellings for the previous year. For the 2nd quarter the weights of the previous year are updated to the st quarter of the year (which is the base period) in the following way: 7

8 w old,2 = I old N i= I old N P i= i +I new N2 P i= i, w new,2 = I new N2 i= I old N P i= i +I new N2 P i= i where I old is the HPI for old dwellings in the first quarter and I new is the HPI for new dwellings in the first quarter (with base period the last quarter of the previous year). For the 3rd quarter the weights of the previous quarter are updated to the 2nd quarter of the year (which is the base period) in the following way: w old,3 = N i= I old2 I old N I old2 I old N i= + I new2 I new 2, i= w new,3 = I new2 I new N N 2 i= N 2 I old2 I old i= + I new2 I new i= where I old2 is the HPI for old dwellings in the 2nd quarter and I new2 is the HPI for new dwellings in the 2nd quarter (with base period the first quarter of the year). For the 4rth quarter the weights of the previous quarter are updated to the 3rd quarter of the year (which is the base period) in the following way: w old,4 = N i= I old3 I old2 I old N I old3 I old2 I old N i= + I new3 I new2 I new 2, i= 8

9 w new,4 = I new3 I new2 I new N N 2 i= I old3 I old2 I old i= + I new3 I new2 I new i= N 2 where I old3 is the HPI for old dwellings in the 3rd quarter and I new3 is the HPI for new dwellings in the 3rd quarter (with base period the second quarter of the year). Calculation of Index When I new,t the HPI for new dwellings at quarter t with base period the previous quarter t-, and I old,t the HPI for old dwellings at quarter t with base period the previous quarter t- are obtained with the method described above, the House Price Index for all dwellings (old and new) at quarter t, I t, with base period the previous quarter t-, is calculated as follows: I t = I old,t w old,t + I new,t w new,t 9

10 tatistical Output tatistical Output CoP/P C IMF DQA F IM Question(s) CoP Relevance PC Relevance.-.3 Pls specify briefly why At national level, the House Price Index can be used for the following purposes: the statistics are - to assess housing affordability over time relevant by referring to - an input for decision making in respect to the housing market the multiple purposes for which the statistics - a measure of specific price trends are used at (i) The Central Government, Banks and Real Estate Agents are some of the users of the international, (ii) EU House Price Index. and (iii) national level References to the statistical legislation and other formal or informal commitments could be sufficient. In addition, if applicable, the outcome of user consultation processes could also be given. 0

11 CoP2 Accuracy and reliability PC2 Accuracy and reliability (including stability) 2./ The analysis should be The data are received quarterly from the Department of Lands and urveys, Ministry of based on the following Interior, as described in Principle 8 Appropriate tatistical Procedures. ince additional indicators (to be characteristics for new and second-hand dwellings are needed in order to build an compiled by the appropriate model, CYTAT has allocated staff to the Department of Lands and urveys ECB/Eurostat on a in order to get additional information for contracts of sales and is currently examining the common data vintage) possibility of receiving additional characteristics for sales. Data are checked for internal on the reliability of consistency and data cleaning is carried out to remove duplicate cases or cases with first releases for the missing data or erroneous data. relevant series (selected for their material impact and to avoid compensatory effects in the case of balancing items): - Directional reliability indicators; - Revisions indicators (e.g. MAPE, MACE or RMRE depending on the indicator). and provide a succinct assessment of the accuracy and reliability of the statistics on the basis of the results, with an attempt to provide a relative objective measure, explaining also the Methodological rules and guidelines prescribed in the official OOH/HPI Manual are closely followed in the processes of modelling, weighting and estimation. The methodology used for the compilation of the House Price Index is in compliance with the Methodological Manual referred to in the Commission Regulation (EU) No 93/203 and the Residential Property Price Indices Handbook. Every effort is made in order to maintain and improve the quality of the House Price Index. ince the data source for HPI is administrative data, there is no sampling error. The main reasons for revisions are new or revised datasets, calculation mistakes or revisions of the compilation method.

12 main reasons for revisions. CoP3 Timeliness and punctuality PC3 Timeliness 3. (including 3.4 punctuality) 4..6 Pls provide a succinct As described in Principle 8 Appropriate tatistical Procedures, in order to calculate the assessment of the level sub-indices for old and new dwellings, a Rolling Year Hedonic Regression Model is used. of the statistical These sub-indices are then aggregated together using transaction-based weights, as discrepancies ( internal described above, in order to produce the HPI. Therefore, the overall aggregate index is fully consistent with its component indices. consistency ) Pls specify the timeliness requirements for international and EU purposes and the national target publication dates. Pls specify punctuality defined as deviations in relation to the timeliness requirements and target dates specified above National requirements: HPI will be published on CYTAT s website in the second quarter of 206 with reference period the first quarter of EU and other international requirements: Data are transmitted to Eurostat according to the OOH Regulation (indices: 85 days after the reference period and weights on the 5th of June of the year following the year to which the weights relate). ome changes/improvements as regards the timeliness of HPI during the last two years have occurred and mainly have to do with the improvement of the data collection and data processing. 2

13 CoP4 PC4 Coherence Consistency 4./4. and and comparabili comparabili ty ty Pls provide a succinct 3.4. assessment of the The OOH may partly overlap the scope of the HPI. The Central Bank of Cyprus (CBC) results of the compiles a Residential Property Price Index based on property valuations. The Royal consistency check of Institution of Chartered urveyors in Cyprus publishes a quarterly price index that tracks the statistics with other property across all districts and across all property types in Cyprus. There are differences related domains or data in the data published by each source, as there are differences in both the data and methodology used. sets with which the statistics must show coherence ( external consistency ). The results should provide a relative objective measure to facilitate cross-country comparability, e.g. be presented as a ratio of the inconsistency between compared statistics, as a percentage of GDP, where appropriate. 3

14 CoP5 Accessibilit y and Clarity PC5 Accessibilit y and Clarity 5./ Pls provide a succinct assessment of the time series consistency, specifying (i) whether the time series presents breaks and if so explaining the reasons For the period , a proxy for the indicator of luxury dwellings was used in the for the breaks; (ii) the regression model for new dwellings. An indicator of luxury dwellings based on real data has been used since the first quarter of 205. Although it is assumed that the proxy is a length of the series for good approximation of the indicator, this change may cause a break in the series. which back data are consistent; and (iii) whether the statistics are consistent across the various frequencies (monthly/quarterly/ann ual) where relevant. Pls provide a brief description of the dissemination policy HPI will be published on CYTAT s website in the second quarter of 206, with and dissemination reference period the first quarter of 206. means, including links to the data and respective metadata. 4