The Shanghai Containerized Freight Index (SCFI) is a highly cited metric used to discuss the health of global trade. In the past 30 days alone, news outlets and industry journals have published over 100 articles referencing the SCFI.

The SCFI is of enormous importance for ocean freight companies, forwarders, traders and consumers alike. 

Despite all the uncertainty, developing a sense of how the SCFI will behave in the short, medium or even long term is of great value to all these market participants.

With the help of AI, a dynamic stochastic model can be created to predict SCFI:

  • Multivariate and non-stationary model
  • Assessment of seasonal effects
  • ARMA model (autoregressive moving average)
  • Update with publication of the SCFI

The following graphics and tables show the relevant stochastic key figures for the evaluation of the different forecast periods.

  • "reference": defines the forecast period
  • "obs_count": the number of observations
  • "sdev_average": Standard deviation in relation to the mean value
  • "sdev_median": standard deviation in relation to the median
  • "rmse": Root Mean Squared Error
  • "sdev_adjusted": adjusted standard deviation

Almost all forecasts achieve deviations or correlation coefficients of around 0.9, even for full-year forecasts. See the years 2018 and 2019. 

These values are therefore very well suited for data-driven decision systems that want to automatically incorporate the influencing factor "market", in this case SCFI.