This paper presents an example of an approach to teaching financial theory at the college and post-graduate levels that I call "teaching backwards". In the more traditional approach, instructors begin by explaining financial theory, then proceed to give examples of the way this theory can be applied to a business problem, structuring data around the predetermined theory. When teaching backwards we reverse the process, suggesting various ways to organize and report data related to a business problem and then inviting students to analyze these data to identify relationships between the variables that enable them to see potential applications of financial theory, or even to discover the theory itself. The example presented employs computer modeling and computer simulation as essential enabling tools for self-directed learning.