Premium
Multivariate GARCH hedge ratios and hedging effectiveness in Australian futures markets
Author(s) -
Yang Wenling,
Allen David E.
Publication year - 2005
Publication title -
accounting and finance
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.645
H-Index - 49
eISSN - 1467-629X
pISSN - 0810-5391
DOI - 10.1111/j.1467-629x.2004.00119.x
Subject(s) - heteroscedasticity , autoregressive model , autoregressive conditional heteroskedasticity , econometrics , futures contract , ordinary least squares , mathematics , hedge , multivariate statistics , economics , statistics , financial economics , volatility (finance) , ecology , biology
We use the All Ordinaries Index and the corresponding Share Price Index futures contract written against the All Ordinaries Index to estimate optimal hedge ratios, adopting several specifications: an ordinary least squares‐based model, a vector autoregression, a vector error‐correction model and a diagonal‐vec multivariate generalized autoregressive conditional heteroscedasticity model. Hedging effectiveness is measured using a risk‐return comparison and a utility maximization method. We find that time‐varying generalized autoregressive conditional heteroscedasticity hedge ratios perform better than constant hedge ratios in terms of minimizing risks, but when return effects are also considered, the utility‐based measure prefers the ordinary least squares method in the in‐sample hedge, whilst both approaches favour the conditional time‐varying multivariate generalized autoregressive conditional heteroscedasticity hedge ratio estimates in out‐of‐sample analyses.