學術產出-NSC Projects

Article View/Open

Publication Export

Google ScholarTM

政大圖書館

Citation Infomation

  • No doi shows Citation Infomation
題名 損失分配法下作業風險值快速蒙地卡羅法的設計
其他題名 An Efficient Monte Carlo Method for Computation of Operational Var under Loss Distribution Approach
作者 謝明華
貢獻者 風險管理與保險學系
日期 2013
上傳時間 20-Apr-2016 15:18:59 (UTC+8)
摘要 近年來,作業風險的量化已經成為金融機構監理的一個重要議題。例如,保險監理 的 Solvency II 與銀行監理的巴塞爾協定都要求保險公司與銀行需要計提作業風險資 本。在巴塞爾協定的進階測量方法 (Advanced Measurement Approaches) 下,金融機構 有自由去選擇使用的隨機模型。損失分配法 (Loss distribution approach) 是一個符合這 個目的的標準隨機模型。在損失分配法下,事業單位與損失形態的組合組成一個矩陣; 而矩陣中的每一個元素有自己的損失分配。這些損失分配的相關性通常是透過 copulas 來做連結。金融監理上對作業風險資本計提的需求, 通常是需要金融機構計算一年內, 在九十九點九的信賴度下,作業風險可能帶來的最大損失。在這樣的高標準要求下, 傳統的蒙地卡羅法無法提供一個準確的估計值。因此,本計畫的主要目的是希望設計 一個有效率的蒙地卡羅演算法,以達成快速且正確計算作業風險值的目標。
In recent years, quantification of operational risk becomes an important issue for regulation in financial industry. For example, Solvency II for insurers and Basel Accord for banks are required insurance companies and banks to allocate capital for operation risk. Under the Advanced Measurement Approaches (AMA) within Basel Accord, financial institutions are given freedom concerning the stochastic models used. Loss distribution approach (LDA) is a standard approach. LDA approach concerns the measurement of risk for random losses generated from an m by d matrix whose element corresponds to a combination of business line and event type. The dependence structure of these random losses is usually modeled through copulas. The risk measure used for regulatory capital purposes reflects a holding period of one-year and a confidence level of 99.9%. It is almost infeasible to get an accurate estimate of such risk measure if naïve Monte Carlo approach is used. Therefore, in this project, we wish to propose an efficient Monte Carlo simulation algorithm for computing such risk measure.
關聯 計畫編號 NSC 102-2410-H004-062
資料類型 report
dc.contributor 風險管理與保險學系
dc.creator (作者) 謝明華zh_TW
dc.date (日期) 2013
dc.date.accessioned 20-Apr-2016 15:18:59 (UTC+8)-
dc.date.available 20-Apr-2016 15:18:59 (UTC+8)-
dc.date.issued (上傳時間) 20-Apr-2016 15:18:59 (UTC+8)-
dc.identifier.uri (URI) http://nccur.lib.nccu.edu.tw/handle/140.119/85689-
dc.description.abstract (摘要) 近年來,作業風險的量化已經成為金融機構監理的一個重要議題。例如,保險監理 的 Solvency II 與銀行監理的巴塞爾協定都要求保險公司與銀行需要計提作業風險資 本。在巴塞爾協定的進階測量方法 (Advanced Measurement Approaches) 下,金融機構 有自由去選擇使用的隨機模型。損失分配法 (Loss distribution approach) 是一個符合這 個目的的標準隨機模型。在損失分配法下,事業單位與損失形態的組合組成一個矩陣; 而矩陣中的每一個元素有自己的損失分配。這些損失分配的相關性通常是透過 copulas 來做連結。金融監理上對作業風險資本計提的需求, 通常是需要金融機構計算一年內, 在九十九點九的信賴度下,作業風險可能帶來的最大損失。在這樣的高標準要求下, 傳統的蒙地卡羅法無法提供一個準確的估計值。因此,本計畫的主要目的是希望設計 一個有效率的蒙地卡羅演算法,以達成快速且正確計算作業風險值的目標。
dc.description.abstract (摘要) In recent years, quantification of operational risk becomes an important issue for regulation in financial industry. For example, Solvency II for insurers and Basel Accord for banks are required insurance companies and banks to allocate capital for operation risk. Under the Advanced Measurement Approaches (AMA) within Basel Accord, financial institutions are given freedom concerning the stochastic models used. Loss distribution approach (LDA) is a standard approach. LDA approach concerns the measurement of risk for random losses generated from an m by d matrix whose element corresponds to a combination of business line and event type. The dependence structure of these random losses is usually modeled through copulas. The risk measure used for regulatory capital purposes reflects a holding period of one-year and a confidence level of 99.9%. It is almost infeasible to get an accurate estimate of such risk measure if naïve Monte Carlo approach is used. Therefore, in this project, we wish to propose an efficient Monte Carlo simulation algorithm for computing such risk measure.
dc.format.extent 1940691 bytes-
dc.format.mimetype application/pdf-
dc.relation (關聯) 計畫編號 NSC 102-2410-H004-062
dc.title (題名) 損失分配法下作業風險值快速蒙地卡羅法的設計zh_TW
dc.title.alternative (其他題名) An Efficient Monte Carlo Method for Computation of Operational Var under Loss Distribution Approach
dc.type (資料類型) report