earticle

영어

This study aims to analyze the contemporary and lead-lag relationship between Hanwoo(Korean native cattle) farm price volatility and Hanwoo wholesale price volatility. Due to change from grain to animal food in food consumption pattern, the demand for livestock products has been increased. The stability of Hanwoo price is major concerns for both producers and consumers. Especially, Hanwoo price stabilization is important for stabilizing the income of Hanwoo breeding farmers. Hanwoo price destabilization weakens the domestic base of Hanwoo industry. Prior studies on this subject focus mainly on the relationship between Hanwoo farm price in level and Hanwoo wholesale price in level,otherwise, on the relationship between the observed volatility of Hanwoo farm price and that of Hanwoo wholesale price. On the contrary, this study focuses on the relationship between the permanent volatility of Hanwoo farm price and that of Hanwoo wholesale price and the relationship between transitory volatility of Hanwoo farm price and that of Hanwoo wholesale price. This is the main contribution of this paper. Permanent volatility is the long-run component of observed volatility and transitory volatility is the short-run component of observed volatility. Monthly data of Hanwoo farm price and Hanwoo wholesale price are used in this study. Data for the analysis are obtained from ‘Materials on price, supply and demand of livestock products’ published by National Agricultural Cooperative Federation. The sample period in this paper ranges from January 1994 to June 2009. The summary statistics show the standard deviation of the return series of Hanwoo wholesale price is higher than that of Hanwoo farm price. The Jarque-Bera statistic significantly rejects normal distribution for the return series of Hanwoo farm price and accepts normality of the return series of Hanwoo wholesale price. The Ljung-Box Q-statistics Q(10) and Q2(10) indicate that there exist serial correlations in returns and squared returns for two variables. The procedure and results of this paper are as follows. First, Component GARCH model is used in order to estimate observed volatility, permanent volatility, and transitory volatility. Because each time series is autocorrelated significantly, the mean equation in Component GARCH model contains a MA(1) term in each model. The Component GARCH model estimation results show that persistent rates of long-run components of farm price volatility and wholesale price volatility are nearly one, and persistent rates of short-run components are smaller than that of long-run component. This means that long-run components of farm price volatility and wholesale price volatility have much slower mean-reverting rate than short-run components.Second, ADF test is employed to test for the stationarity of data. The appropriate lag order is selected by the AIC. ADF test results show that all volatilities are stationary series. Therefore, Granger causality in a VAR should be applied. Third, in the correlation analysis, it is found that there are statistically significant contemporaneous correlations at 1% level in all cases of volatilities. Fourth, this study employs the Granger causality test to investigate the lead-lag relationship between Hanwoo farm price volatility and Hanwoo wholesale price volatility. Granger causality test is a useful statistical technique to examine the direction of causality between two variables. Because each volatility has not a unit root, Granger causality test based on VAR is applied. The optimal lag order of the VAR model is determined by the AIC. The Granger causality tests show that, in the case of observed volatility, there are two-way Granger causalities between the observed volatility of Hanwoo farm price and that of Hanwoo wholesale price. In the case of permanent volatility, namely, the trend component of observed volatility,there is one way Granger causality from the permanent volatility of Hanwoo wholesale price to that of Hanwoo farm price. This test result means the permanent volatility of Hanwoo wholesale price precedes that of Hanwoo farm price. Transitory volatility is the cyclical component of observed volatility, namely, it is the short-run portion and passes away with time. The test result also shows that there is one way Granger causality from the transitory volatility of Hanwoo wholesale price and that of Hanwoo farm price. This means that the cyclical portion of Hanwoo farm price volatility is preceded by that of Hanwoo wholesale price volatility. Fifth, the impulse response function analysis is applied to examine the magnitude of Hanwoo farm price volatility response to the impulse on the Hanwoo wholesale price volatility. The impulse response functions show that Hanwoo wholesale price volatility impulse increases the magnitude of Hanwoo farm price volatility in all cases. In short, these results of this paper suggest that Hanwoo wholesale price volatilities help to estimate Hanwoo farm price volatilities and indicate that stabilizing Hanwoo wholesale price volatility is necessary for Hanwoo farm price volatility stabilization. The implication to policy makers from this study is that stabilizing Hanwoo wholesale price volatility is important for the income of Hanwoo breeding farmers and Hanwoo industry. The area for further work is to apply other statistical methods estimating permanent volatility and transitory volatility, and to investigate the relationship between farm price volatility and wholesale price volatility of other livestock.

한국어

본 연구에서는 한우 산지가격 변동성과 한우 도매가격 변동성 간에 어떤 관계가 있는지를 동시기적 관계 분석과 선도-지연 관계 분석으로 구분하여 분석하였다. 한우 산지가격과 한우 도매가격 간의 관계에 대한 기존의 연구들에서는 수준 변수 간의 관계 또는 관측된 변동성 간의 관계에 대해서만 분석이 이루어져 왔으나 본 연구에서는 기존 연구들과는 차별적으로 Component GARCH 모형을 이용하여 관측된변동성, 영속적 변동성 부분, 일시적 변동성 부분을 추정하여 분석에 사용하였다.동시기적 상관관계 분석에서는 산지가격 관측된 변동성과 도매가격 관측된 변동성의 상관관계, 산지가격 영속적 변동성과 도매가격 영속적 변동성의 상관관계, 산지가격 일시적 변동성과 도매가격 일시적 변동성의상관관계 모두에 있어서 상관관계가 유의한 것으로 나타났다. 그랜저 인과관계 분석을 이용한 선도-지연 관계 분석에서는 먼저 관측된 변동성의 경우에는, 한우 도매가격 관측된 변동성과 산지가격 관측된 변동성이양방향의 그랜저 인과관계가 있는 것으로 나타났다. 변동성 중에서 추세부분에 해당되는 영속적 변동성 간의 관계는 도매가격 영속적 변동성이 산지가격 영속적 변동성에 대해 단방향의 그랜저 인과관계를 갖는 것으로 나타났다. 그리고 단기적으로 존재하는 변동성 부분인 일시적 변동성의 경우에도 도매가격 일시적 변동성이 산지가격 일시적 변동성에 대해 선도관계가 있는 것으로 나타났다. 이와 같이 변동성 중에서 추세부분에 해당되는 도매가격 영속적 변동성 뿐만 아니라 일시적 도매가격 변동에 의해 생성되는 변동성 부분도산지가격 변동성에 영향을 미칠 수 있다는 것으로 도매시장 안정의 필요함을 역설하는 실증분석 결과이다