A Dynamic Programming Model for Tax Collection and Payment Based on the Time Value of Tax Payment
基于税款时间价值的税企动态规划模型

In view of the current situation for tax payment and audit,we consider in this paper the time value of tax payment.We first prove that the penalty rate for late tax payment should be higher than bank loan rate when tax payment delay for a business follows a Poisson distribution.Subsequently,we set up a dynamic programming model to analyze the decisions of a tax collection agent and a tax paying business.Using backward induction,we obtain the taxpayer's optimal policy for tax filing and payment at different stages.Using analysis of elasticity,we derive that,when the penalty coefficient is fixed,the daily penalty rate for late tax payment should be lowered.Finally,suggestions are given to encourage corporations to pay their taxes in time.