题名: |
Rate Setting Analysis: A Statistical Approach to Outlier Analysis in the Rate Setting Process Within the United States Transportation Command. |
作者: |
McGriff, W. B. |
关键词: |
United states transportation command, Statistical processes, Costs, Errors, Probability density functions, Spc (statistical process control), Rate setting analysis, Outlier rates, Ustc (united states transportation command), Transportation rates, Demand forecasting, Twcf (transportation working capital fund) |
摘要: |
This research sought to identify areas in which the current rate setting methodology can be improved. We initially examined the use of six months of historical cost data versus a full year of data to set rates, concluding that there is not a statistically significant difference with respect to their relative effect on the NOR; USTC should proceed with their current practice. The research also identified outliers, first with regard to likelihood of historical rates not being set by the prescribed process and second with regard to whether the rates set by the prescribed process would be an outlier in terms of the marginal contribution to the net operating result. We found that approximately 8%, 10%, and 4% of the rates in FY14-FY16 were likely set using budget analyst experience in lieu of the prescribed method, for the most part imposing a reduction in the prescribed rates. Adapting classical Statistical Process Control (SPC) methods, we found that the prescribed rate setting method does work in aggregate but can induce recurrent outlier rates. However, a pattern in these outlier rates remains elusive some are self-correcting but the demonstrated methodology is shown to be useful for identifying outlier rates that do merit budget analyst experience-informed judgment for rate setting. The final component of this research examined the combination of two factors used in the current methodology to adjust current average weighted costs to set future rates: the Accumulated Operating Result and Composite Rate Adjustment factors. Using historical data from FY08-FY15, we calculate the optimal combined factor values for each respective fiscal year to achieve an NOR equal to $0. In doing so, we concluded that the combination of these two factors contributed to approximately 25% of the induced error in NOR. We suggest a more detailed examination of these rate computations for additional analysis. |
报告类型: |
科技报告 |