مقاله انگلیسی رایگان در مورد ترکیب متغیر وابسته خود رمگذارها برای انتخاب مورد حسابرسی مالیات بر ارزش افزوده – الزویر 2020

1. Introduction

2. Related work

3. Methodology

4. Method deployment

5. Conclusions and future work

Acknowledgments

References

Abstract

In this work, we address the problem of targeted Value Added Tax (VAT) audit case selection by means of machine learning. This is a challenging problem that has remained rather elusive for EU-based Tax Departments, due to the inadequate quantity of tax audits that can be used for conventional supervised model training. To this end, we devise a novel Gated Mixture Variational Autoencoder deep network, that can be effectively trained with data from a limited number of audited taxpayers, combined with a large corpus of filed VAT returns. This gives rise to a semi-supervised learning framework that leverages the latest advances in deep learning and robust regularization using variational inference. We developed our approach in collaboration with the Cyprus Tax Department and experimentally deployed it to facilitate its audit selection process; to this end, we used actual VAT data from Cyprus-based taxpayers. This way, we obtained strong empirical evidence that our approach can greatly facilitate the VAT audit case selection process. Specifically, we obtained up to 76% out-of-sample accuracy in detecting whether a significant tax yield will be generated from a specific prospective VAT audit.

Introduction

Valued Added Tax (VAT) is a consumption tax charged on the value of almost all the goods and services sold or consumed within the European Union (EU). It constitutes an indirect tax collected by enterprises on behalf of the state, and is ultimately paid by the final consumer. As such, it represents an important source of revenue for all EU Member States; according to the European Commission Taxation Trends Report, 2018 edition [1], indirect taxes comprise more than 30% of the total tax revenue in the EU.

The European commission uses the concept of VAT-gap to estimate the lost revenue of the EU Member States due to taxpayer non-compliance with the VAT legislation. It is defined as the difference between the estimated VAT amount that should have been collected and the actually collected amount. The latest European Commission study [2] estimated the VAT-gap at 12.3% (147 billion Euros) of the total expected VAT revenue in the EU; in Cyprus it stands at around 5% (83 million Euros). This exemplifies how important it is that we come up with effective methods for reducing the VAT-gap.