This paper brings to the forefront the assumptions that we make when focussing on a particular type of explanation for bank profitability. We evaluate a broad field of research by introducing a general framework for a profit maximizing bank and demonstrate how different types of models can be fitted into this framework. Next, we present an overview of the current major trends in European banking and relate them to each model's assumptions, thereby shedding light on the relevance, timeliness and shelf life of the different models. This way, we arrive at a set of recommendations for a future research agenda. We advocate a more prominent role for output prices, and suggest a modification of the intermediation approach. We also suggest ways to more clearly distinguish between market power and effciency, and explain why we need time-dependent models. Finally, we propose the application of existing models to different size classes and sub-markets. Throughout we emphasize the benefits from applying several, complementary models to overcome the identification problems that we observe in individual models.