An algorithmic approach to handle circular trading in commercial taxing system

TL;DR

This paper presents algorithms for detecting and analyzing circular trading in commercial tax systems by modeling transactions as directed graphs, demonstrated on real government data from Telangana, India.

Contribution

It introduces a novel graph-based algorithmic approach to identify circular trade patterns in tax systems, addressing a specific form of tax evasion.

Findings

Effective detection of circular trading in real datasets

Identification of key actors involved in circular trade

Potential for improving tax compliance and fraud prevention

Abstract

Tax manipulation comes in a variety of forms with different motivations and of varying complexities. In this paper, we deal with a specific technique used by tax-evaders known as circular trading. In particular, we define algorithms for the detection and analysis of circular trade. To achieve this, we have modelled the whole system as a directed graph with the actors being vertices and the transactions among them as directed edges. We illustrate the results obtained after running the proposed algorithm on the commercial tax dataset of the government of Telangana, India, which contains the transaction details of a set of participants involved in a known circular trade.