【作者】 白寅；

【导师】 刘金兰；

【作者基本信息】 天津大学 ， 管理科学与工程， 2009， 博士

【摘要】 本文在对集团公司预算管理的概念、目标、机制、功能与主要模式进行讨论的基础上,梳理了我国石油集团预算管理的主要内容与环节。结合我国石油集团预算管理的实际情况,针对预算管理部分环节中存在的一些关键问题,本文进行了研究并给出了解决方案。主要研究内容与成果如下:1.在市场预测阶段,对原油价格时间序列建模时,本文针对传统ARIMA-GARCH模型轻描述趋势、重刻画误差因而预测性不强的缺点,结合原油价格在高位时波动较大,低位时波动较小,且上期价格上涨或下跌对本期价格变动的作用不对称的特点,建立了含有非对称成分的ARIMA-GARCH模型。本文对1997.1～2009.7布伦特原油现货价格建立预测模型,显示出较好的预测性。2.在预算目标计划阶段,建立了油价波动条件下,基于石油集团内部供应链价值最大化的油品产、购、销量优化模型,模型可比较各油价下原油产量、销量、进口量、加工量,成品油进口量、销售量以及化工产品产销量,并制定出给定油价下的最优生产计划。此外,模型的对偶模型还可计算出各种产能的边际利润以及市场需求变化的边际利润,为产能调整决策及市场拓展决策提供支持。本文通过案例分析显示出了集团利润最大化与板块利润最大化两种策略存在冲突,并指出了企业内部供应链的风险对冲作用。3.在预算编制与分解阶段,需要对下属单位的盈利能力进行评价。为了构造更为合理的盈利能力评价模型,在SUR模型中加入了时间效应成分,构造了带有时间效应的变系数面板数据模型,并给出了参数估计的方法。该模型用于盈利能力评价与预测时,充分利用公司的财务历史信息,比传统方法更具先进性。通过案例分析,本文比较了某石油集团八个油田的盈利能力,并预测了新增投资的利润。4.在预算执行与控制阶段,将质量控制技术及可视化数据挖掘技术引入财务管理领域。针对财务管理中允许财务数据期望均值发生变化的特点,提出了变均值控制图技术用于预算执行的预警控制;提出新的差异分析技术,即先用“预算树结构”对预算数据进行格式化,在此基础应用“问题发现矩阵”技术查找预算执行异常的原因。更多还原

【Abstract】 In this dissertation, the definition, objectives, mechanisms, functions and main modes of budget management for business group are introduced firstly. Then, the framework of budget management for Chinese petroleum group corporation is discussed and some key issues are raised based on the current situation of management practice. After that, technical solutions are proposed according to the key issues raised before. Main works of this dissertation are as follows:1. A new approach to predict crude oil price named Asymmetric ARIMA-GARCH is proposed. The traditional ARIMA-GARCH model has some weaknesses such as the error part is too large to be used as a predicting model. The oil price series has some characteristics such as the higher price causes larger fluctuation and price rise and drop drives the following price differently. The proposed new model explained the asymmetry of the crude oil price series, which make the prediction more accurate. The Brent oil monthly spot price series during Jan,1997～Jul,2009 is analyzed by the proposed model and the fitness is statistical significant.2. A new idea of inner supply chain value maximization for petroleum group corporation is proposed in the dissertation. Based on this idea, the optimization of inner supply chain planning model is built. Specifically, crude oil production, import, sale, refining quantity, gasoline import, gasoline sale and petrochemicals production are generated by the model given any crude oil price.At the same time, solution for the dual problem give the profit margins created by expand productivity or market share. A case study illustrate that conflicts exist between the overall group value and the branches values. Also, the optimization of supply chain counteracts the risk coursed by crude oil price fluctuation.3. By adding time effects into traditional seemingly related regression (SUR) model, a new variable-coefficient panel data model is proposed in order to evaluate the profitability and to predict the profits of petroleum group corporation’s subdivisions. The parameters’estimation method is given, which fully uses the history of the company’s financial information and shows the advantage of the proposed model. A case study compared eight oil fields’profitability of a Chinese petroleum group corporation. Also, their profits are predicted.4. Quality control techniques and visual data mining techniques are introduced into the area of financial management. Financial targets are allowed to move which is different from quality targets. Thus, the control chart with variable mean is employed as an early warning control tool. Based on that, a new variance analysis method is proposed. Namely, a“budget tree”data structure is designed to format the financial data firstly. After that,“problem detection matrix”technique which is invented by this dissertation is used to compare actual costs and their targets, identify variances, and interpret the source of each variance.更多还原