Empirical models for forecasting in Finance are predominantly based on linear regression models. A strength of linear regression models is the interpretability of the relationship between the variables in the model. However, a prediction model must fulfill further requirements: it should provide precise estimates, build a stable relationship between dependent variable and independent variables, and be robust towards outliers. Against this background, the development of an appropriate forecasting model is a complex task, and linear regression models may not always provide the best solution to this problem (e.g. in case of non-linearities). Thus, machine learning methods provide a potentially valuable tool to accomplish the above-mentioned modeling challenges and achieving more accurate predictions by allowing a rich set of possible model specifications compared to conventional methods. Therefore, the Institute of Finance compares different machine learning and linear regression models with respect to their forecasting performance to develop the best possible model in the respective application in Finance.