Financial Forecasting Based On Time Series Analysis

Financial forecasting is one of the most important methods in the process of business administration. It is especially important to make an accurate prediction when it comes to making big plans. Though traditional ways of prediction like percentage of sales approach and linear regression analysis have been carried out over the years, their accuracy need to be improved. Time series forecasting is the use of a model to predict future values based on previously observed values. It has been experimented in various industries such as econometric, mathematical finance and weather forecasting, earthquake prediction and so on. This research aims to find out whether time series will do a good job in financial forecasting.In this research, we built a model which includes three main parts: the data processing part for collecting data from various balance sheets and arrange them in the series of time; the data predicting part for analyzing all the data collected and make accurate predictions of them; the result analyzing part which compares the results we got to the real data and find out whether its errors are acceptable.In this paper, we introduced the traditional ways of financial prediction and discuss the flaws they have. Then we explain why we want to try time series and built a model to test whether time series can work well in prediction. After trying out the model on a real company, we came to the conclusion that time series can make much more accurate forecasting when the national politic policies remain the same.