農產品期貨價格波動性的到期效應與交易量效應

Maturity and Volume Effects of Agricultural Futures Price Volatility

農產品期貨可以規避農產品價格波動所造成的風險，然而農產品期貨的避險績效決定於避險者對期貨價格波動性的精確估計與掌握；關於農產品期貨波動率的分析，過去的研究皆以每日收盤價資料來進行，忽略交易日內價格變動的資訊。本文則採用農產品期貨的日內價格資訊計算期貨價格上漲變幅與下跌變幅，作為期貨價格波動性的代理變數，藉以分別估計農產品期貨短部位與長部位的價格波動性。同時本文將價格變幅資料配合Chou（2006）的不對稱條件自我相關變幅（ACARR）模型，檢測農產品期貨價格波動性的到期效應與交易量效應。本文以玉米、棉花、黃豆、肉牛期貨作為實證樣本，資料期間從1991 年到2003 年，檢測結果發現這四種農產品期貨的價格上漲變幅與下跌變幅皆具有顯著的交易量效應，包括成交量效應與未平倉量效應；但是這四種農產品期貨的價格上漲變幅並不具到期效應， 僅有價格下跌變幅呈現顯著的正向到期效應， 拒絕Samuelson（1965）的負向到期效果假說，而且此一結果也顯示下跌變幅蘊含較多期貨價格資訊。最後本文發現農產品期貨下跌變幅的到期效應受到交易量效應的支配，換言之，期貨交易量變數是影響期貨波動率的主要因素，這個結果支持Grammatikos 與 Saunders（1986）的發現。

Agricultural futures can help hedge the price risk of agriculture, butthe hedge performance of agricultural futures is dependent on theaccurate estimation and the understanding of futures price volatility. Mostvolatility models are based on the information of closing prices of thereference period, failing to use the information contents inside thereference. In the article we adopt the intraday price information tocalculate both the upward price range and the downward price range ofthe futures, in order to estimate the price volatility of short position andlong position, respectively, of futures trading. Meanwhile, we use Chou(2006) asymmetric conditioned autoregressive range (ACARR) model tofit the futures price range data, so as to examine the maturity, volumeand open interest effects for the agricultural futures contracts. Fouragricultural futures contracts including corn, cotton, soy bean, and feedercattle futures are tested in the article. The whole sample period rangesfrom the year 1991 through 2003. Our results show that very strongvolume effects, including trading volume effect and open interest effect,exist in both the upward price range and the downward price range.However, the upward price range does not demonstrate the maturityeffect; while only the downward price range demonstrates the positivematurity effect. The results reject Samuelson (1986) negative maturityeffect hypothesis, and also show that the downward range contains moreprice information of futures than upward range. Finally, we find that thematurity effect of futures volatility is significantly dominated by the volumeeffects; in other words, the volume of the futures contract is the keyvariable for the agricultural futures price volatility. The result supports thefinding of Grammatikos and Saunders (1986).