The statistics include the level and daily evolution of the sales made by companies included in the Immediate Supply of Information (SII) system. The daily sales series comes from the VAT management system based on the SII, implemented since July 2017 (RD 596/2016, of December 2). This system allows the exchange of tax information practically in real time between the Tax Agency and the taxpayers obliged to the SII by sending the details of the billing records within a period of four days through the electronic headquarters of the Tax Agency.
The daily data in this publication can be considered an early indicator of the sales published in the monthly Sales, Employment and Salaries of Large Companies report and the quarterly Sales, Employment and Salaries in Large Companies and SMEs report. Both offer information from VAT declarations and withholdings for work income; The first includes exclusively taxpayers considered as Large Companies for tax purposes, while the second adds information on SMEs with the form of a public limited company and limited liability company. In addition to being a complement to these two publications, the rapid availability and broad coverage of the information represent added value in the short-term forecasting and monitoring of both tax collection and macroeconomic variables.
Those obliged to comply with their tax obligations through the SII are Large Companies (those with a volume of operations in the previous year greater than 6 million euros), VAT groups and companies covered by the Monthly Refund Registry (REDEME). . Given the size and profile of these taxpayers, the daily sales available represent around 70% of the total domestic sales of all VAT taxpayers (the percentage is higher if the total declared volume of operations is considered), with a great diversity of coverage by activities (the SII companies are very representative of some sectors, while in others their number and weight are reduced, as is the case in part of the construction, in the hotel and catering industry or in some modalities Of transport).
The geographical scope is the so-called Common Fiscal Regime Territory; That is, companies that operate exclusively in the territories managed by the estates of the Basque Country and Navarra, and companies that do so in the territories that are outside the scope of application of VAT (Canary Islands, Ceuta and Melilla) are excluded.
Taxpayers are territorially assigned to the place where they have their tax domicile. Information on the tax domicile of the seller and buyer can be obtained from the SII system, but the place where the sales were made cannot be determined unequivocally.
The companies included in the Immediate Information Supply system (SII) are classified according to the activity declared by the companies themselves. Activities are classified according to the headings of the Economic Activities Tax (IAE). Its regulation is found in Royal Legislative Decree 1175/1990, which approves the rates and instructions for the Economic Activities Tax, and its successive updates. For the purposes of publishing the report, companies are classified at four digits of the CNAE-2009 and are grouped based on it.
The variable presented in this statistic is obtained from the aggregation of the non-exempt subject bases available in the record book of invoices issued. This tax base includes both the part corresponding to the sale of goods and the provision of services and is comparable to the concept of domestic sales (excluding exempt sales) available in the Sales, Employment and Salaries statistics in tax returns.
Daily economic series pose a series of problems that are not observed as clearly in lower frequency series (monthly or quarterly). These problems arise, above all, from the complex and unstable structure of the calendar (different length and composition of the months, mobile seasonal elements such as Easter, leap years and a mobile work calendar that also interacts with the weekly composition of the months. ) and the existence of seasonal elements that overlap cycles of different frequencies.
To this we must add the intensity that its irregular component usually presents (smoothed when it comes to monthly data) and the impact of exogenous elements that distort the usual behavior of the systematic components of these series (in the case of the SII these elements can be , for example, different billing dates in different companies, which cause abnormally high values to appear that interact with other components of the series).
For all these reasons, the estimation of a robust economic situation signal is especially difficult and requires a treatment based on econometric models that jointly and statistically satisfactorily represent the aforementioned phenomena. To do this, a structural time series model is used that flexibly includes all these elements and that includes, in order to preliminarily correct a part of these effects, a treatment through exogenous variables of a deterministic nature. These last variables are designed to control the presence of anomalous observations and specific calendar effects that, due to their aperiodic nature or mobile periodicity, do not fit into the structural representation considered.
Estimating a robust economic situation signal in daily series is especially difficult and for this purpose an approach based on structural econometric models of time series is used. In this case, the TBATS modeling approach is applied to the corrected series of deterministic effects.1.2, which is based on the representation of unobserved components (trend, seasonality, irregularity) through explicit dynamic models. This approach is very flexible since it allows the simultaneous consideration of several seasonal components (weekly, monthly and annual) and, thanks to its trigonometric representation, allows the treatment of fractional periodicities (for example, those induced by leap years), offering a representation very compact of these components. From this modeling, the trend series and the corrected series of seasonal and calendar variations (CVEC) can be obtained, as well as the corresponding seasonal factors at their different frequencies (weekly, monthly and annual). The average of said factors is zero at each of said frequencies.
1 TBATS: Trigonometric seasonality, Box-Cox transformation, ARMA innovations, Trend and Seasonality (De Livera et al., 2011).
2 A specific application of these series can be found in Cuevas, Ledo y Quilis, SERIEs, 2021, 10.1007/s13209-021-00251-7. https://link.springer.com/content/pdf/10.1007/s13209-021-00251-7.pdf