Prediction of Baltic Capesize Freight Index Based on GARCH Model

Abstract

This paper analyzed the statistical data of the Baltic Capesize Freight Index (BCI) and the daily return rate sequences to improve forecast reliability of the international dry bulk shipping market. It used the first-order logarithmic difference method to get the BCI daily return rate sequence, showing that the BCI daily return sequence had a leptokurtosis and fat-tail fluctuation feature and other features such as integration. To analyze volatility persistence, sensitivity and hysteresis of the sequence, a GARCH (1, 1) model was introduced. The GARCH (1, 1) model constructed the forecast method of BCI sequence, and it predicted the BCI daily return rate by optimizing lag phases. The logarithm sequences of BCI daily return rate finally reverted to BCI, which was forecasted at last. The conclusion is supposed to improve the international dry bulk shipping market forecasting method.