The article introduces new STATA commands for estimating different types of models for predicting outcomes. These models can handle situations where there are a lot of zero outcomes and can predict positive, neutral, and negative outcomes separately. The researchers tested the accuracy of these models using simulations and found that they work well. They also compared the models to see which one is the best fit for the data. The new commands were applied to predict the U.S. federal funds rate target in a real-world example.