Browsing by Subject "GARCH"

The volatility of commodity prices, such as agricultural products, metals and energy, have been fluctuating from year to year. Studies show that there are many fundamentals affecting the volatility of commodity prices. The increasing financialisation of commodity markets might affect the price volatility among other fundamentals. One way to invest widely in commodity markets is to invest in sector indices which consist of numerous underlying securities. The objective of this research is to measure the price volatility of commodity sector indices during economic fluctuations. This study also tries to answer whether volatilities of selected sector indices correlate to each other. In this research four sector indices, S&P GSCI Agriculture Official Close Index ER, S&P GSCI All Cattle Official Close Index, S&P GSCI Energy Official Close Index and S&P GSCI Industrial Metals Official Close Index, were selected to measure the volatility and correlation. Fifth variable, S&P 500 Index was used to identify the periods of up and down markets and measure the correlation of commodity sector indices against that. A covariance-variance matrix was applied to measure the correlation. Econometrical approach to measure volatility was achieved by applying seven different univariate and multivariate GARCH (Generalised Auto Regressive Conditional Heteroscedasticity) models. The results suggest that the volatility and correlation in volatility models have been highest during the period of when the prices of S&P 500 Index were lowest. The volatility and correlation coefficient mainly follow the market fluctuation of S&P 500 Index. The results also indicate the pattern of volatility of the selected indices seem to be mainly similar except for one (S&P GSCI All Cattle Official Close Index). The mean values of coefficient correlation in multivariate models were mainly positive, fluctuating between -0.0000459 and 0.003934. This study supports the theory that volatility increases when market prices decrease. This study also shows evidence that when the volatility is high, also the correlation is high. This means that in the bear markets when correlation is higher, the benefit of diversification in these selected indices does not, in general, minimise the risk.