Monte Carlo methods for business management : applications in hi-tech and financial companies

• Main Authors: Chung, Ming Tao, 鍾明道
• Other Authors: Chi, Yen Ping
• Format: Others
• Language: en_US
• Published: 2017
• Online Access: http://ndltd.ncl.edu.tw/handle/dp8d9w

博士 === 國立政治大學 === 資訊管理學系 === 106 === This dissertation utilizes Monte Carlo methods to solve business problems in hi-tech and financial companies. There are two essays: The first one is titled “Optimal Outsourcing Strategy: A Stochastic Optimization Approach”： As the production capacity of a company over a certain period of time is limited, enterprises must carefully consider product line development or outsourcing options. Unlike traditional studies that use static or comparative static analyses to determine optimal production strategies, essay 1 proposes a stochastic optimization model that can be used to determine optimum quantities of multi-period production/outsourcing plans. Based on the proposed approach and utilizing the real demand and production capacity data of a high-tech production company in Taiwan, we can quantify the expected financial benefit of an optimal outsourcing strategy. In addition, we consider 3 types of outsourcing partners. The type of outsourcing partners is based on their flexibility to accept outsourcing requests. Therefore, the proposed approach can be applied to a broad range of possible outsourcing partners and can quantify the benefits of flexibility in outsourcing requests. The second essay is titled “Fast Simulation of Operational Risk for Financial Institutions”： Quantification of operational risk has led to significant concern regarding regulation in the financial industry. Basel Accord II and III for banks and Solvency II for insurers require insurance companies and banks to allocate capital for operation risk. Because the risk measure used for Basel regulatory capital purposes reflects a confidence level of 99.9% during one year and the loss distribution of operational risk has high skewness and kurtosis, it is almost infeasible to get an accurate estimate of such a risk measure if a crude Monte Carlo approach is used. Therefore, we develop a novel importance sampling method for estimating such a risk measure. Numerical results demonstrate that the proposed method is very efficient and robust. The main contribution of this method is to provide a feasible and flexible numerical approach that delivers highly accurate estimates of operational risk with a high confidence level and meets the high international regulatory standard for quantification of operational risk.