天然巨災可能造成人生命的危害，也可能造成個人財產的損失，而這些危害輕微則可能造成個別家庭經濟生活的失衡，嚴重則造成國家社會的經濟、政治的不安定，因此如何進行風險管理以避免危險，或減少危險事件發生後所造成損害，便成為一相當重要之研究課題。 台灣地區由於地理環境位置的特性，最常見的天然巨災為地震及颱風，本研究即針對地震及颱風兩類天然巨災，藉由現代之風險管理觀念，考量天然巨災風險的特性及來源，整合國內外相關之基礎研究，例如地震危害度分析、地震損害分析、各種易損性曲線相關研究及淹水潛勢圖等，建立了台灣地區地震及颱風洪水風險評估模型，並引入動態財務分析的概念，將巨災風險評估模型與動態財務分析模型作結合，最後再透過物件導向程式設計理念，開發建立具有親和力使用者介面之天然巨災風險評估系統，並取名為 Risktrace-EQ及RiskTracer-TY/FL。 本研究所研發的天然巨災風險評估系統採用投資組合分析模式，並建立較能反應損失關聯性之累計損失分佈，主要可以提供損失超越機率曲線的分析、了解不同財務觀點的損失分佈情形、藉由不同自負額，限額等的保險條件擬定設計較佳的保險策略、依據不同風險忍受度來訂定最大可能損失以及風險移轉機制之設計及分析等功能。 Natural catastrophe can cause the damage of the human being and loss of the property. General speaking, the damage can hurt the economy of a family and induce the uncertainty of economy environment and politic of the country. Therefore, how to avoid risk by rsik management or reduce the loss after natural catastrophe event is a very important topic of the catastrophe rsik management. The most common natural catastrophe is earthquake and typhoon due to the special geographical environment position of Taiwan area. The purpose of this research is to delvelop the natural catastrophe risk assessment model include earthquake and typhoon. Using the modern concept of the risk management, this research considered the characteristic of the natural catastrophe risk and integrated the national and international relevant primary researches, such as seismic hazard analysis, seismic vulnerability analysis, fragility curve researches regarding catastrophe, flood potential analysis and etc. It also applied the concept of the dynamic financial analysis to the catastrophe rsik assessment model. Finally, we use the Object- Oriented Programming method to build up the natural catastrophe risk assessment software with user-friendly interface. The software names are “Risktrace-EQ” and “RiskTracer-TY/FL”. The natural catastrophe risk assessment software we built uses portfolio analysis method. It uses the cumulative loss distribution which can reflect the correlated relationship between event losses. It can provide the functions such as loss exceeding probability analysis, loss distribution by different financial perspective, design the better risk program by different insurance condition, probale maximum loss by different risk tolerance, design and analysis of the risk sharing scheme.
