原油期貨基差之實證研究

An Empirical Study on oil futures basis

張甄庭(Chen-Ting Chang)

指導教授 ： 郭文忠

國立臺北大學/社會科學學院/經濟學系/碩士(2014年)

摘要

基差為期貨價格減去現貨價格，在實務操作中常作為短期期貨價格變動之重要參考指標。本論文以西德州原油期貨基差為研究對象，蒐集1993/1/3-2013/11/22期間日資料，探討影響原油基差的重要解釋變數，藉由估計不同的計量模型，包括GLS模型、GARCH模型、EGARCH模型及EGARCH-M模型，進行模型估計與預測。
結果顯示延遲一期的基差、美國信貸危機、原油期貨成交量及原油現貨報酬率於各計量模型估計結果皆顯著且穩定，延遲一期的基差與基差為正相關，係數估計值介於0至0.6之間，表示基差變數有mean reverting現象。此外，信貸危機與基差為正相關，表示信貸危機發生，各家銀行將利率調升，使民眾持有現貨成本增加，故有較高意願購買原油期貨。原油現貨報酬率與原油現貨報酬率為負相關，根據期貨反轉效應，石油現貨價格漲，但期貨價格的成長幅度不如現貨，表示期貨價格有反轉趨勢，使基差值轉小。本論文最後並以樣本外預測結果，比較各實證模型之優劣。

關鍵字

未平倉合約 ； 金融危機 ； 樣本外預測 ； EGARCH-M 模型 ； EGARCH 模型 ； GARCH 模型 ； GLS 模型 ； 期貨基差 ； 成交量 ； VIX 指數

並列摘要

Basis is the difference between oil futures price and oil spot price. In practice basis is an important reference index as the movements of short-term futures prices. In this paper, we collect the daily data of West Texas Intermediate crude oil futures basis over the period January 1993 to November 2013 for research to observe which explanatory variables are impact significance in different time series models such as GLS model、GARCH model、EGARCH model、EGARCH-M model. And observe predicted results of these models. Finally, we compare the result of out-of-sample forecasting.
The investigation results indicate that one lag oil basis, U.S. crisis, oil futures volume and oil spot price returns in every econometric model are significant and stable. Lag one oil basis and basis have a significantly positive correlation; the coefficient is from 0 to 0.6, which means that the basis has a mean reverting phenomenon. Besides, U.S. crisis has a positive correlation with basis, which means that banks raise interest rates when crisis happen, so that investors prefer to buy oil futures because of the cost of spot crude oil is more expensive than before. Oil spot price returns and basis have a significantly negative correlation. Refers to futures reversal effect, oil spot and futures price rising, but the growth rate of futures is less than spot price. It indicates that futures have reversal effect that makes basis become smaller.

並列關鍵字

EGARCH - M model ； EGARCH model ； GARCH model ； AR model ； oil basis ； out-of-sample forecasting

參考文獻

Bessembinder H., Chan, K. and P. J. Seguin (1996) “An empirical examination of information, difference of opinion, and trading activity,” Journal of Financial Economics, 40, 105-134.

Bollerslev, T. (1986) “Generalized autoregressive conditional heteroskedasticity,” Journal of Econometrics, 31: 34-105.

Engle, R. F. (1982) “Autoregressive conditional heteroskedasticity with estimates of the variance of UK Inflation,” Econometrica, 50, 987-1008.

Fleming, J. and B. Ostdiek (1999) “The impact of energy derivatives on the crude oil market,” Energy Economics, 21, 135-167.

Garbade, K. D. and W. L. Silber (1983) “Price movement and price discovery in futures and cash markets,” Review of Economics and Statistics, 65, 289-297.