本研究將以紐約商品交易所(NYMEX)交易的西德州原油為標的，其2000年1月4日至2004年8月18日之期貨及現貨價格資料進行研究實證。先檢驗原油報酬統計分配特性，再次檢驗出資產報酬的厚尾及聚集現象。此外，由於風險報酬隨機模型與GARCH模型在決定避險比例與避險時機不盡相同。所以，本研究接著評估一般化及總合誤差分配的隨機模型與GARCH兩種方法的動態避險效果，利用移動視窗法作為動態避險之程序後，發現GARCH模型動態避險效果優於隨機模型。 Crude oil products play an important role in economic development and industrialization. Not only for energy purpose, has the crude oil also supplied important raw material for the chemical industry. The price volatility and hedging strategy is emphasis by oil and chemical companies.
The characteristic of oil price volatility is the best target to test volatility measurement model. Engle (1982) proposes the ARCH model and Bollerslev (1986) revises the model (GARCH) to estimate and predict price volatility. More and more researches are suggested to discuss and describe the price volatility.
The paper use the price of WTI traded on the NYMEX from January 4, 2000 to August 18, 2004 to test the hedging efficiency of different model. The empirical shows that GARCH can catch the price volatility of WTI oil price. The explanation power of GARCH model is superior to the OLS model. Overall the GARCH-GED has the best performance of hedging efficiency among all the model.
