Fractionally Cointegrated Vector Autoregression Model of Spread Estimation for Metals

Abstract

The predictability of asset prices on the exchanges is the most relevant topic of financial research. High/low prices help to analyze the volatility of the commodity price in the current auctions at the stock exchange. In the study we investigate the impact of daily high/low spreads on close/open spreads for non-ferrous and precious metals using the multifractal analysis and the Fractionally Cointegrated Vector Autoregressive (FCVAR) models. For fractional data testing, it is proposed to use an indicator based on the width of multifractal spectrum. Results of full series and positive spreads models for all metals except zinc and lead indicated that any increase in high/low spreads leads to a decrease in close/open spreads. For all negative spreads models this relationship has positive trend.