ASYMMETRIC VOLATILITY AND EXOGENOUS FORECASTING OF CONTAINER FREIGHT RATES IN THE SHANGHAI-INDONESIA (CN.SHA-IDTPP) TRADE LANE

Abstract

Container shipping rates are a critical cost signal for exporters in Shanghai and importers in Indonesia, but much of the forecasting literature treats South-East Asia as an aggregate market and relies on univariate models. This study consolidates a route-level econometric framework combining unit-root and seasonality diagnostics, ARCH-LM testing, symmetric and asymmetric GARCH models, and ARIMAX/SARIMAX-type forecasting with lagged exchange-rate, bunker-price, and port-dwell-time regressors. Because a verified weekly Shanghai Indonesia spot-rate series has not yet been assembled, the framework is evaluated using 300 weeks of synthetic data covering January 2021 to October 2026. The  path represents a pandemic-era peak, a post-pandemic correction, and a later recovery, while the data-generating process embeds annual seasonality, exogenous effects, and EGARCH asymmetry. Results show mixed unit-root evidence, no statistically significant Fourier seasonality, weak-to-borderline ARCH evidence, and no detectable asymmetry despite a positive asymmetry parameter in the generating process. Exogenous regressors reduce one-step-ahead RMSE by 28.3% 31.9% relative to ARIMA, with Diebold Mariano p-values of 0.004 and 0.002; improvements at four weeks are smaller and statistically insignificant.