We present an analytical framework for active foreign exchange reserves management that integrates risk-return objectives with macroeconomic, macro-prudential and sovereign debt management concerns. Our framework allows for very general objective functions, does not restrict the class of eligible stochastic processes or limit the investment universe, and can incorporate many types of macroeconomic concerns. It incorporates several kinds of risk constraints in order to obtain benchmarks satisfying possible central bank requirements of safety, liquidity, returns, and stability. Feedback between outcomes and decisions is easy using tools that reshape distributions and functions of the outcomes. And the model can be run on a PC-based platform. We apply the framework to several common reserves management problems focusing especially on the formulation of model equations, generation of trees and estimation of density functions of outcomes. We compare our approach to those used by many central banks and discuss advantages to our approach.