Combination of Forecasts in the International Context: Predicting Debt Reschedulings

International business is characterized by high uncertainty, making accurate forecasting both important and difficult. Development in forecasting methodology are examined in the prediction of international debt reschedulings. Three basic forecasting approaches are evaluated: single method forecasts, raw data combination forecasts, and forecast-combinations. All three approaches exhibited significant improvement in prediction accuracy over a naive forecasting approach, but only the odds-matrix forecast-combination approach outperformed another non-native approach. The odds-matrix forecast-combination method is less sensitive to data sparsity and unreliability and has the potential to synthesize, effectively and in a theoretically based way, diverse perspectives in the international context.