A Study on Statistical Forecasting Methods of Foreign Patent Cost

Abstract

Filing a foreign patent is extremely expensive. The indirect costs have a major impact on corporate earnings. Therefore, firms must manage the cost of filing for a foreign patent by forecasting such costs in the forthcoming fiscal years and including them into their management strategy. However, costs vary with the timing and number of inventions, thus complicating cost forecasting. Thus, firms need to use simple methods to predict foreign patent costs for the fiscal year in question by referencing the previous year's actual numbers. The present study compares multiple statistical forecasting methods to find a more accurate method of estimating foreign patent costs, particularly for the United States, and to apply this method to corporate practice. The study found a precise method of forecasting future overseas patent costs by selecting the most appropriate methodology for the existing circumstances from among multiple forecasting methodologies by using a certain algorithm and developing a forecasting system.