Developing a model that enables the analysis of a Common Consolidated Corporate Tax Base
Danuše Nerudová, associate professor at the Department of Accounting and Taxes at Mendel University in Brno, is the leader of FairTax WP5. Together with her colleagues in WP5, she is analysing what impact implementing a Common Consolidated Corporate Tax Base (CCCTB) in EU member states might have.
Photo left: Danuše Nerudová, associate professor at the University in Brno
In their latest deliverable (5.3), the researchers in WP5 have developed a model to research the impacts of the CCCTB system on the budget revenues of individual EU Member States. The model is based on datasets from the Amadeus and Bankscope databases. Amadeus is a database with comparable financial and business information on Europe's largest 510,000 public and private companies in 43 countries. The database includes standardised annual accounts (consolidated and unconsolidated), financial ratios, sectoral activities, and ownership data. It is suitable for research on competitiveness, economic integration, applied microeconomics, business cycles, economic geography, and corporate finance. The Bankscope database is an identical database comprising information on the banking sector only.
“The model we have developed covers groups of companies operating in the EU that meet the criteria for consolidation and group taxation under the CCCTB system. The criterion for the selection of entities is based on a two-layer approach – we selected entities with more than 50% of the voting rights in the controlled company and more than 75% of ownership rights. Further, information on assets, employment, sales, payroll, income (profit or loss before taxation), and long-term tangible assets has to be known for all of the selected groups of companies so that the proper share of the consolidated tax base under the CCCTB system can be determined through formulary apportionment,” said Nerudová.
Method to handle missing data
“However, some required information is often missing in the Amadeus and Bankscope databases. In order to eliminate the negative impacts due to reductions of the dataset, we decided to apply three method of missing data imputation – regression, Bayesian model using an adaptive Metropolis-Hastings algorithm and single imputation method. We performed a sensitivity analysis, and based on the results, we have decided to impute the missing data through the regression model in case all of individual Member States with the exception of United Kingdom, for this method does not allow to impute the data with the smallest deviation from the real data. In case of the United Kingdom Bayesian model using an adaptive Metropolis-Hastings algorithm was applied as it turned up as the method imputing the missing data with the smallest deviation from the real data” Nerudová continued.
Calculating tax bases in each EU member state
In the next step of this research, the model will be used to identify the shares of individual EU Member States in the consolidated tax bases of the groups of companies.
“The dataset we created of companies with imputed data will be separated into partial datasets according to the residency of the parent companies. This means that groups of parent companies with their subsidiaries acting in the Internal Market of the EU for each EU Member State will be created,” Nerudová explained. “The partial datasets will be analysed in detail with respect to group taxation schemes and consolidation regimes currently applied in each EU Member State. Based on these results, it will be possible to determine the total tax base of the subsidiaries of parent companies that are resident in each EU Member State according to the current rules. Subsequently, the consolidated tax bases under the CCCTB system in partial datasets will be determined with respect to individual EU Member States. For this purpose, the formulary apportionment under the CCCTB directive proposal will be used.”
What do you think are the biggest challenges when researching CCCTB?
“We encountered two main challenges. First, the information needed in order to be able to do correct calculations was often missing in the databases. Therefore, we had to search for a method to impute the missing information without shrinking the dataset of companies. The second challenge will be how to reflect changes in the behavior of the companies, in other words, how to create a dynamic research model.”
Merged data create a unique study
Danuše Nerudová and her research team in FairTax WP5 will use the model to conduct a unique study on the impact of a CCCTB.
“The uniqueness lies in the fact that we are going to have a very large dataset of companies. This is possible for two reasons. First, we are using the Amadeus database for all sectors of the economy except NACE K, which includes financial institutions. For this sector, we are merging data with the Bankscope database, which is aimed at the financial sector only. Second, we are also simulating a variant of temporary cross-border loss offsetting (instead of consolidation) as suggested in the new proposal of the European Commission. On the topic of impacts of cross-border loss offsetting on budget revenues, I mean CCCTB without consolidation regime but enabling cross-border loss offsetting, there are no other studies at the moment.”