Freight rate volatility spillover analysis on primary capesize spot routes

Abstract

This thesis focuses on the examination of dynamic volatility spillovers between five primary spot Baltic Capesize routes. The time series sample consists of daily prices of the Baltic Capesize spot freight rate indices C2 (Tubarao – Rotterdam), C3 (Tubarao – Qingdao), C5 (West Australia – Qingdao), C7 (Bolivar – Rotterdam) and C17 (Saldanha Bay – Qingdao) that cover the period from 2014 to 2021. The logarithmic returns of the individual Capesize routes freight rates were fitted in a Vector Error Correction Model (VECM) which provided information on the lead-lag relationship between the routes, as well as the basis for the subsequent Impulse Response Function Analysis (IRF). The resulting IRF graphs provided great insight into the way a shock in one endogenous variable affects the rest of the endogenous variables.In the next step, a multivariate VECM-DCC-GARCH(1,1) model was estimated in order to accurately explain the time-varying conditional volatility and simultaneously capture the dynamics between the variables over time. The conditional variance estimates predicted by the DCC-GARCH model were then fitted on a VAR Model of order 3, so that the Diebold-Yilmaz Spillover Index approach could be applied. Based on the Generalized Factor Error Variance Decomposition, the Diebold-Yilmaz Spillover Indices confirmed the existence of dynamic volatility spillovers between the primary spot Capesize routes.