The article explores ways to improve predictions of economic trends using different models. By combining three types of models and using full Bayesian predictive distributions, the accuracy of predictions increased significantly. This method led to a mean improvement in probability of over 50% compared to using a single model. Additionally, combining predictive densities from all three models further enhanced the accuracy of predictions by almost 42%. These findings suggest that a combination of models and full Bayesian predictive distributions can greatly enhance the accuracy of economic predictions.