台灣雨量選擇權之定價與避險

Abstract

近年來全球暖化、極端氣候事件頻繁，導致重大天災事件頻仍， 造成社會、經濟及生命的嚴重損失。台灣地區特殊之地形、地理位置及地質條件，經常受到颱風、地震、土石流、山崩等天然災害之侵襲，不僅造成財產損失，亦威脅民眾生命安全。其中颱風和梅雨帶來的災害最為嚴重，而降雨為主要因子。本研究以台灣地區特有之降雨型態，發展雨量為標的物之氣候衍生性商品，提供遭受雨量影響之企業作為避險工具。 本研究以中央氣象局自2002年至2010年間10雨量資料10個雨量測站之汛期（5月∼11月）時雨量氣候資料，以紐曼-史考特矩形跳動模型進行雨量模擬，模型中假設鋒面的產生服從卜瓦松分配，每個鋒面可產生的雨雹數目及雨雹強度服從指數分配，每個雨雹的延時服從指數分配。在參數估計方面，參考Farve et al.（2004）提供之參數估計方法，使用1小時及24小時歷史累積降雨量作為樣本資料進行估計，即完成雨量模型之設定。在選擇權定價方面，採用蒙地卡羅法模擬10,000次日雨量，最後以2009年莫拉克颱風時雨量來進行實際避險應用。

Due to the special geographical and geological conditions, Taiwan has been threatened for years by typhoon, earthquake, debris flow and landslide. Recently, extreme climate events due to global warming has increased the frequency and intensity of those natural catastrophes which cause more damages than before in Taiwan. As precipitation is the most important factor which causes most of the damages, this study attempts to develop the precipitation option which can be used by individuals and/or enterprises to hedge the rainfall risk. Using historical precipitation data (2002~ 2010) of 10 observation stations from Central Weather Bureau, we simulate the monthly rainfall distribution with Neyman-Scott Rectangular Pulses Model in which storm arrives following Poisson distribution. Parameters are estimated with 1-hour and 24-hour cumulative historical precipitation data. The precipitation model is built and we apply Monte Carlo method to simulate the path of the precipitation. To demonstrate the use of precipitation option in rainfall risk hedging, Typhoon Morakot (2009) is analyzed as an example.