| Summary: | 碩士 === 輔仁大學 === 應用統計學研究所 === 96 === The government finance is a foundation stone of economic development. The main income of the fiscal revenues is tax revenues. How to estimate and forecast the budget of the tax revenues, in government's fiscal administration, is a subject that is worth paying attention to. This Research offers references for the government’s enacting financial and economic policies as well as annual revenue budget by getting rid of the non-economic factor that may influence tax change, analysing the actual intension of tax, and predicting the tax in the future. The main purposes as this research are as follows: First, working out the tax seasonal index, and probing into the seasonal variation and light-busy situation, in order to adopt the best levy tactics at the right time. Second, setting up single variable ARIMA models according to the historical monthly variation tendency characteristic of tax, excluding the influence by the non-economic factor (such as time difference, season ), doing a forcast on the tax revenue in the future, as references for budgeting the annual revenue next year.
This research contains five chapters altogether. Chapter one is the introduction, stating the motives, purposes and procedures of this research. Chapter two is documentation discussions, sketching the tax structures at present, introducing the predict theories, and extracting the relevant domestic researches. Chapter three is research approach, exposing the basic conception and setting-up model procedures of time series analysis, and showing the seasonal index method. Chapter four is the analysis results of real examples, doing the trend and seasonality analysis on the tax item characteristics, explaining the setting-up, and diagnosis ARIMA model, and predicting the results. The fifth chapter gives conclusions and suggestions based on my research.
|
|---|
