This note outlines the interest of Revenue Administrations (RAs) and National Statistical Offices (NSOs) in the quality of data at their disposal, and how collaboration between these organizations can contribute to improving data quality. The similarities between the data collection and processing steps in revenue administration and in the production of economic statistics underlie meaningful information and data sharing. Mutually beneficial collaboration between RAs and NSOs can be achieved, particularly in efforts to improve the coverage of registers and to update register information; classify economic activity; and analyze joint data to address data shortcomings. Since there are differences in concepts and definitions used in revenue administration and official statistics, dialogue is necessary to ensure the effective use of data from the partner organization. Collaboration can improve the quality of data available to both institutions: for RAs, this can assist in realizing improved taxpayer compliance and revenue mobilization, and for NSOs, tax-administrative data sources may enable expanded coverage of the economy in official statistics and reduce timeframes required for publishing economic time series and national accounts. Together, these outcomes can enhance the policy formulation, planning, and service delivery capability of governments. To that end, this note delineates concrete steps to engender sustainable and meaningful interchange of information and data between the RA and NSO.
We introduce unFEAR, Unsupervised Feature Extraction Clustering, to identify economic crisis regimes. Given labeled crisis and non-crisis episodes and the corresponding features values, unFEAR uses unsupervised representation learning and a novel mode contrastive autoencoder to group episodes into time-invariant non-overlapping clusters, each of which could be identified with a different regime. The likelihood that a country may experience an econmic crisis could be set equal to its cluster crisis frequency. Moreover, unFEAR could serve as a first step towards developing cluster-specific crisis prediction models tailored to each crisis regime.
The ever-increasing digitalization of businesses has accelerated the need to address the many shortcomings and unresolved issues within the international corporate income tax system. In particular, the customer or “user”—through their online activities—is now considered by many as being a critical driving force behind the value of digital services. Furthermore, the rapid growth of digital service providers over the last decade has made them an increasingly popular target for special taxes—similar to wealth and solidarity taxes—which can also help mobilize much-needed revenues in the wake of a crisis. This paper argues that a plausible conceptual case can be made to tax the value generated by users under the corporate income tax. However, a number of issues need to be tackled for user-based tax measures to become a reality, which include agreement among countries on whether user value justifies a reallocation of taxing rights, establishing the legal right to tax income derived from user value, as well as an appropriate metric for valuing user-generated data if it is ever to be used as a tax base. Furthermore, attempting to tax only certain types of business is ill-advised, especially as user data is now being exploited widely enough for it to be recognized as an input for almost all businesses. Several options present themselves for consideration—from a modified permanent establishment definition combined with taxation by formulary apportionment, to user-based royalty-type taxes—each with their own merits and misdemeanors.
Tamas Gaidosch, Frank Adelmann, Anastasiia Morozova, and Christopher Wilson
This paper highlights the emerging supervisory practices that contribute to
effective cybersecurity risk supervision, with an emphasis on how these practices
can be adopted by those agencies that are at an early stage of developing a
supervisory approach to strengthen cyber resilience. Financial sector supervisory
authorities the world over are working to establish and implement a framework
for cyber risk supervision. Cyber risk often stems from malicious intent, and a
successful cyber attack—unlike most other sources of risk—can shut down a
supervised firm immediately and lead to systemwide disruptions and failures.
The probability of attack has increased as financial systems have become more
reliant on information and communication technologies and as threats have
continued to evolve.
Jin-Kyu Jung, Manasa Patnam, and Anna Ter-Martirosyan
Forecasting macroeconomic variables is key to developing a view on a country's economic outlook.
Most traditional forecasting models rely on fitting data to a pre-specified relationship between input
and output variables, thereby assuming a specific functional and stochastic process underlying that
process. We pursue a new approach to forecasting by employing a number of machine learning
algorithms, a method that is data driven, and imposing limited restrictions on the nature of the true
relationship between input and output variables. We apply the Elastic Net, SuperLearner, and
Recurring Neural Network algorithms on macro data of seven, broadly representative, advanced and
emerging economies and find that these algorithms can outperform traditional statistical models,
thereby offering a relevant addition to the field of economic forecasting.
International Monetary Fund. Communications Department
Address at the Bank of England Twentieth Anniversary Conference
September 29, 2017
International Monetary Fund Managing Director Christine Lagarde delivered this address at the Bank of England conference, “Independence—20 Years On” in London, U.K., on September 29, 2017.