Revealing Asymmetric Spillover Effects in Hazelnut, Gasoline, and Exchange Rate Markets in Turkey: The VECM-BEKK-MGARCH Approach

Abstract

The study used the VECM-BEKK-MGARCH method to model the volatility transmission between the markets of gasoline, exchange rates, and the hazelnut market for the period of 21.07.2005-20.3.2018. The suitability of the VECM-BEKK-MGARCH method was confirmed by statistical testing. The changes in hazelnut prices were not affected by the changes in the prices or final values in the other two sectors (Granger causality). Moreover, the Granger causality tests revealed that, while the change in the gasoline market was not affected by the other two markets, the change in the exchange rates market was affected by the other two markets. Furthermore, especially the volatilities (long-term uncertainties) of the markets were affected by both their own short-and long-term volatilities and other sectors’ short-and long-term volatilities. It was shown that the long-term swings in these three markets were affected by the cross-interaction in the markets. Additionally, as opposed to the case in the pos-itive news, it was observed that pieces of negative news affected the markets.