Quarterly Gross Domestic Product (QGDP)
To access the QGDP press here.
Quarterly Gross Domestic Product at market price (QGDP), the main macro-economic aggregate of national accounting, represents the final result of production activity for resident productive units, for a certain period, a quarter, respectively:
Quarterly Gross Domestic Product at market price is estimated by three approaches:
- output approach:
GVA=gross value added at basic prices;
TP=taxes on products;
SP=subsidies on products.
- expenditure approach:
FC= actual final consumption;
GCF=gross capital formation;
E=exports of goods and services;
I=imports of goods and services.
- income approach:
CO= compensation of employees;
GOS=gross operating surplus;
TPI=taxes on production and imports;
SPI=subsidies on production and imports.
The main data sources used for quarterly Gross Domestic Product estimation are:
- statistical sources: short-term surveys regarding industrial production, construction, services, trade; production account for agriculture; short-term surveys regarding earnings and employment;
- financial-accounting sources: accounting statements of financial institutions;
- administrative sources: execution of state budget and local budgets, and of social security budget; balance of payments.
Quarterly Gross Domestic Product is estimated in current prices, in the prices of the corresponding period of the previous year and in the average prices of the year 2000. The estimates in average prices of 2000 are calculated by chain-linking volume indices.
Besides the gross estimates of quarterly Gross Domestic Product, seasonally adjusted estimates are also compiled, based on the regressive method, this method being recommended by the European regulations.
The seasonal adjustment envisages the removal of seasonal effects from the data series in view to highlight the real economic evolution during consecutive periods.
In order to adjust the main aggregates series, based on which the GDP is estimated by all three methods, JDEMETRA software package is used (TRAMO/SEATS and X-13ARIMA-SEATS methods). This leads to the estimation of seasonal effect (events taking place each year at the same time, with the same amplitude and orientation, such as: seasons, holidays, etc.), of the working days number different from one month to another and the calendar effect (Orthodox Easter, leap year and other national holidays) as well as to the outliers identification and correction of outliers (additive outlier, transitory change, level shift) and to missing data interpolation.
The seasonally adjusted series was obtained by removing this effect from the unadjusted series, by means of correction coefficients, selected depending on the regression model used (additive or multiplicative). The additive or multiplicative model used for regression is automatically identified by the JDEMETRA software, depending on the nature of series that are subject to adjustments.
The seasonally adjusted GDP is obtained through the direct method, thus leading to a statistical discrepancy between the GDP and the sum of its components, which are independently seasonally adjusted.
The selection of regression models used for each series takes place at the beginning of each year and entails a revision of adjusted series compiled during the previous year (revision due to the changes of selected models, of regression factors number and of available observations number).
The revision of the quarterly accounts data is periodically done, when a new version of yearly national accounts is available. The revision of data has as objective to keep the coherence between the quarterly accounts and the yearly accounts. The periodical revision of national accounts unadjusted data series also entails the revision of seasonally adjusted series.
Last update: September 7, 2018