The diversification of currency loans: A comparison between safety-first and mean–variance criteria

This paper presents a model and a method for solving currency loan problems under stochastic and deterministic constraints. The currency loan problem is a mathematical programming problem, i.e. how to diversify borrower's loans in order to reduce currency and interest rate risks for the borrower. The second objective is to illustrate similarities and differences among alternative portfolio selection criteria. Applying a chance-constrained programming algorithm to the above currency loan problem we can solve it by using three different criteria, Markowitz's Mean–Variance and Safety-First criteria by Telser and Kataoka. This allows us to compare these criteria under both equality and inequality constraints.