본 연구는 한국거래소에 상장된 유로화 통화선물(Won/Euro currency futures)시장의 현물시장(spot market)에 대한 헤지성과를 분석하였다. 유로화 통화선물시장의 적정헤지비율 및 헤지성과분석을 위하여 단순헤지모형(naive hedge model), 최소분산헤지모형, 벡터오차수정모형(VECM) 뿐만 아니라 Bollerslev(1986)의 GARCH 모형을 확장한 동태적인 ECT-GARCH(1,1)모형을 추정하였으며, 주요 실증분석결과는 다음과 같다.
첫째, 유로화 현물가격과 유로화 통화선물가격 수준변수(level variable)간에 공적분관계가 존재하는 것으로 나타났다.
둘째, 내표본(within sample)기간 동안 각 모형별 헤지성과 분석결과에 의하면 시간변동 이변량 ECT-GARCH(1,1)모형의 헤지성과가 정태적인(static) 헤지모형인 최소분산헤지모형, VECM의 헤지성과보다 상대적으로 낮은 것으로 나타났다.
셋째, 외표본(out-of-sample)기간 동안에서도 시간변동 동태적인 ECT-GARCH(1,1) 모형의 헤지성과가 전반적으로 정태적인 헤지모형들의 헤지효과보다 상대적으로 낮은 것으로 나타났다.
이러한 실증분석결과는 Wang and Low(2003), Kroner and Sultan(1993), Lin et al.(1994), Choudhry(2004) 등 해외 주가지수 및 금리선물의 헤지성과에 대한 실증분석결과와 상반되나, Cotter & Hanly(2006), 홍정효, 문규현(2004) 등의 연구와는 일맥상통하는 결과를 보여주고 있다.
This paper studies the hedging performance of Won/Euro futures contracts against the downside risk of Won/Euro cash market. For this purpose we introduce four hedging models such as naive hedge model, conventional minimum variance hedge model, VECM(3) and time-varying bivariate ECT-GARCH(1,1) model. We employ the daily Won/Euro cash market and nearby Won/Euro futures contracts covering from May 26, 2006 to November 8, 2007. The major empirical results are as follows; First, there is a long-term relationship between the level variables of Won/Euro spot market and Won/Euro futures contracts. Considering this co-integration relation, we incorporate an error correction term to estimate several parameter with time varying GARCH model as well as VECM. Second, according the empirical results on hedge ratio with the four models, the ECT-GARCH(1,1) model's hedge ratio is relatively lower than those of naive, VECM and conventional static hedge models within sample period. Third, during the out-of-sample period, we find a similar empirical results that of within sample period.
This paper studies the hedging performance of Won/Euro futures contracts against the downside risk of Won/Euro cash market. For this purpose we introduce four hedging models such as naive hedge model, conventional minimum variance hedge model, VECM(3) and time-varying bivariate ECT-GARCH(1,1) model. We employ the daily Won/Euro cash market and nearby Won/Euro futures contracts covering from May 26, 2006 to November 8, 2007. The major empirical results are as follows; First, there is a long-term relationship between the level variables of Won/Euro spot market and Won/Euro futures contracts. Considering this co-integration relation, we incorporate an error correction term to estimate several parameter with time varying GARCH model as well as VECM. Second, according the empirical results on hedge ratio with the four models, the ECT-GARCH(1,1) model's hedge ratio is relatively lower than those of naive, VECM and conventional static hedge models within sample period. Third, during the out-of-sample period, we find a similar empirical results that of within sample period.
