Efficient Frontier Analysis

Efficient frontier analysis is used when there are two competing goals. One goal can be chosen as the target for an optimization and the other can be constrained to be no “worse” than a specified limit. Then a sequence of optimizations is performed, each time changing the limit value for the constraint.

Although efficient frontier analysis is generally used for financial portfolio optimization, it can be applied to a variety of problems where there are two competing goals. As one example, one might be trying to determine a waste management system that minimizes cost and minimizes the pollution level. Because these goals are pulling in opposite directions, the solution might be to minimize cost while putting various upper limits on the allowable pollution level. Alternatively, a solution could be to minimize the pollution level by putting various upper limits on the allowable cost.

Evolver can perform efficient portfolio analysis for any model with two competing goals. One goal can be chosen as the target to optimize, and an “efficient frontier” type of constraint can be added to the other goal, with the list the bounds to try, and the optimization is specifically set to perform an efficient frontier analysis. Then Evolver solves a sequence of optimizations, one for each bound specified, and it presents the results in graphical and tabular form.