A Model of Daily Predictions for Sugar Futures Contracts

Abstract

Using monthly data we estimate the best linear model to predict international sugar prices in the New York and London markets. Although we try to use the WTI price and the relationship between the two prices, we found no cointegration between daily wti prices and sugar prices. To identify the short-run relationship between the three series, a VAR model was estimated for the first difference of the series. Our results show that a univariate ARIMA model predicts best the sugar prices in New York and London.