The financial crisis of 2007–2008 showed, among other things, that the information technology in the financial sector was inadequate to assist institutions in producing reliable reports and keeping track of the securities issued. This, in turn, compromised the chances of correctly estimating financial risk. In this paper, I explain how a well-designed computational ontology of finance could help improve data management. I also argue that this ontology should be realism-based and conformant to Basic Formal Ontology (BFO). This choice, however, does not come without complications because it requires some preliminary reflections on the nature of the entities involved. Fortunately, some philosophical literature can be a valuable starting point for these reflections; in the paper, this is shown by considering money as a case study.