Files in this item
|(no description provided)|
|Title:||Time-varying hedge ratio estimation for selected agricultural commodities and products|
|Author(s):||Roh, Jae Sun|
|Doctoral Committee Chair(s):||Garcia, Philip|
|Department / Program:||Agricultural and Consumer Economics|
|Degree Granting Institution:||University of Illinois at Urbana-Champaign|
|Abstract:||The use of autoregressive conditional heteroskedasticity (ARCH) models to estimate time-varying hedge ratios suggests that conventional procedures may not provide the optimal hedge ratios. However, the initial results using ARCH models raise several questions regarding time-varying hedge ratios. First, what is the sensitivity of time-varying hedge ratios to alternative specifications of time-varying variances and covariances, and what tests can be used to select the most appropriate model? Second, what is the degree to which the variance of returns is reduced with alternative procedures? Third, do alternative approaches to capture the time-varying nature of hedge ratios exist?
The primary objectives of this study are to identify procedures which can be used to find appropriate model specifications for the improved estimation of time-varying hedge ratios and to develop testing procedures for the constancy of the correlation coefficient in the bivariate generalized ARCH (BGARCH) model. Also, a random coefficient model is applied as an alternative approach to estimate time-varying hedge ratios. This model also permits a direct test of the constancy of hedge ratio.
Using the data on cash and futures price changes of corn and soybeans, the constancy of the correlation coefficient in the BGARCH model is weakly and strongly rejected, respectively. Following model specification tests, a diagonal vech parameterization is found appropriate for both corn and soybeans. The BGARCH diagonal vech parameterization provides the largest reduction in the variance of the portfolio return and is consistent with diagnostic and specification test results. For both corn and soybeans, the constancy hypothesis of the hedge ratio was rejected when tested against the random coefficient autoregressive (RCMAR) model. However, the RCMAR procedure does not lead to large reductions in the variance of returns.
The other objective of this study was to develop a joint time-varying hedging estimation method for multiple risk situations. Here, we examine the effectiveness of a multivariate GARCH (MGARCH) for modeling dynamic behavior in the soybean complex. Empirical results show that a constant correlation MGARCH model adequately represents cash and futures prices in the soybean complex. The results of hedging effectiveness also indicate that MGARCH scheme performs better than one-to-one and the other time-varying separate hedging procedures.
|Rights Information:||Copyright 1992 Roh, Jae Sun|
|Date Available in IDEALS:||2011-05-07|
|Identifier in Online Catalog:||AAI9305668|
This item appears in the following Collection(s)
Graduate Dissertations and Theses at Illinois
Graduate Theses and Dissertations at Illinois
Dissertations - Agricultural and Consumer Economics