A Shortrun Price Forecast for Coffee

摘要

Six econometric models were used to estimate the expected value of the monthly consumer price index for ground coffee. The hypothesis was based on the observation that effects of changes in the coffee price at import level are spread over 4 to 6 months at retail. Estimation models were used to make forecasts 6 months ahead. Several statistical criteria were applied to evaluate the forecasting performances. In addition, a linear correction function using proportional error estimates was applied to estimates from the best fitting relationship to improve the forecasting results.