FORECASTING MODELS FOR INLAND WATERWAY GRAIN TRAFFIC

Water carriers, Class I railroads, grain shippers, port authorities, and government policy makers need forecasts of water carrier grain tonnage. Although forecasting studies have been conducted for almost every mode, no forecasting studies of quarterly water carrier grain transportation have been published. The objectives of this paper are (1) specify quarterly water carrier grain transportation forecasting models for the Illinois Waterway, the Ohio River, and the Mississippi River, and (2) empirically estimate the models. Given the lack of potential explanatory variables with data published in quarterly frequency and the focus on short run forecasting, time series models were employed to forecast water carrier grain tonnage. The models were estimated using the Maximum Likelihood estimation procedure for the 1989:1 to 1994:4 period. The actual water carrier grain tonnage for this period was compared to the forecast tonnage generated by the time series models. For the three models as a group, only 54 percent of the quarterly grain tonnage forecasts resulted in a forecast error of 10 percent or less. For the three models as a group, they predicted the direction of change (increase or decrease) in actual quarterly grain tonnage in 88 percent of the cases. The models were employed to generate quarterly grain tonnage forecasts for each of the three rivers for the years 2000 and 2001.

Language

  • English

Media Info

  • Features: Figures; References; Tables;
  • Pagination: p. 88-128

Subject/Index Terms

Filing Info

  • Accession Number: 00804773
  • Record Type: Publication
  • Files: TRIS
  • Created Date: Jan 10 2001 12:00AM