Evaluating the effectiveness of the security measures undertaken to protect a distributed system (e.g., protecting privacy of data in a network or in an information system) is a difficult task that, among other things, requires a risk assessment. We introduce a logical framework that allows one to reason about risk by means of operators that formalize causes, effects, preconditions, prevention and mitigation of events that may occur in the system. This is work in progress and we describe a number of interesting variants that could be considered.