Use of Survival Analysis for Shell Companies Analyses in a Brazilian State

Abstract

One of the current problems in tax evasion is the creation of shell companies. Such companies operate in the provision of spurious credits, harming the market and the State. The Tax Authority needs to act as soon as possible to prevent such companies from generating the credit, as well as within the constitutional deadline for recovery of the amounts. The present work aims to analyze how the SEFAZ of the State of São Paulo has acted on this issue. Through the use of survival analysis in Python, it is sought to verify how the performance was in the period from 2015 to 2023. The time to detect shell companies as well as the time required for administrative processes are analyzed. First, an analysis is made through Kaplan Meier. Subsequently, using computational tools, the possibility of parameterizing the survival curve is verified. Through Cox regression, it is verified if there are relevant covariates that can be studied. The results are expected to provide subsidies for the development of lines of action within the SEFAZ to minimize the survival time of such companies.