Tax analysis is a cognitive activity that takes the quantitative aspects of tax economic phenomena as the research object. Deepening the analysis of tax data is helpful to understand the quantitative characteristics of tax, deepen the understanding of tax, infer and predict tax, and thus grasp the initiative of tax management.
Tax analysis index is a calculation formula used to analyze taxpayers' tax risks and its attribute identification, and several index sets are identified through tax analysis and prediction model. Tax risk indicators are divided into tax categories, behavior categories and specific project categories according to the degree of correlation with tax. Tax risk indicators mainly include business tax, enterprise income tax, land value-added tax, personal income tax, property tax, land use tax, stamp duty, deed tax, urban construction tax and surcharges. Behavioral risks include invoice behavior, registration, declaration and tax collection and management identification. Specific risks mainly include non-monetary transactions, debt restructuring, demolition, relocation, bankruptcy, merger, division, land and real estate transfer, etc.
By analyzing and comparing the tax contribution rates of single-family enterprises and local enterprises in the same industry, for enterprises with lower than the average contribution rate of the same industry, early warning information will be given, and in-depth analysis will be made to see whether there are problems such as over-counting production costs, over-counting sales costs, over-counting period expenses, expanding the scope of pre-tax deduction or not counting or under-counting sales revenue. Compare the calculated tax rate with the previous year or the local level of the same industry, and analyze and judge whether the enterprise has the above problems.
In the process of tax analysis and management, tax analysis and identification must be carried out by constructing a set of indicators and models. A scientific model can identify taxpayers' tax problems comprehensively, timely and accurately, and the key to building a high-quality model is to collect effective tax data (characteristics), set accurate indicators and establish a scientific model, so as to simulate the real situation of corporate taxation and strengthen tax supervision.
In the process of model construction, different weights should be set according to the importance of indicators to model management, especially the construction of key indicators. By setting multi-level indicators, such as first-level indicators, second-level indicators and third-level indicators, the actual situation can be dynamically managed. In addition, the construction of the index system itself is also very important, such as: index name, index function, value caliber, comparison method, index weight in the model, calculation of early warning value, etc.