| Summary: | 博士 === 國立中正大學 === 財務金融所 === 95 === This thesis consists of two essays dealing with issues related to the futures hedging effectiveness. There are two different aspects applied to observe the futures hedge to aim the same purpose. The requirement of “perfect hedge” is that the spot and futures prices are complete positive correlated. However, in reality the underlying assets returns and the futures returns is not completely positively correlated. As a result, hedge ratios calculated based on different models are always used for hedging purpose.
The first essay compares the hedging effectiveness of eleven models under different price trends. The models employed to evaluate optimal hedge ratio include: Regression model, VECH model, EWMA model, Matrix-Diagonal model (MD), ECM-MD, BEKK model, ECM-BEKK model, constant conditional correlation GARCH model (CCC), ECM-CCC, Kalman-filter model, and Kalman smoother model. All the models point out the same fact: the hedging effectiveness is different under upward or downward price trends regardless of the models we employed.
The results from the different evaluation methods indicate that investors should take account of the upward or downward price trends when they decide on their hedge portfolios.
The second essay tries to find a proxy for the unknown factor which affects the linkage between cash price and futures price. Because the futures price is the prediction value of the expected future cash price, the spread between the cash price and futures price, i.e., the basis, might contain important information. This paper incorporates the basis into the calculating of the optimal hedge ratio in an attempt for reducing the variance of portfolio formed by cash and futures.
Therefore, the second essay focuses on a new empirical model, STFB with TAR-GARCH, to improve futures hedging effectiveness. We suspect that the spreads between spot and futures prices are a proxy of unknown factors to adjust the optimal hedge ratio, and suggest a smooth transition function of basis (STFB) model to control the varying hedge ratio. In this essay, the STFB model outperforms the conventional model including linear regression, BEKK, DVECH, CCC GARCH, and DCC GARCH models.
|
|---|
