Fantastic article. It always made me nervous while attempting to model an optimal portfolio that the output was so extremely sensitive to the inputs. Even shifting a 15-year input period by a couple months could alter the output dramatically. I am sure we agree that that sort of statistical analysis is best used as one tool in the toolbox and at a broad level. Using an optimization as gospel does seem foolish given the magnitude and sensitivity of assumptions being made.
Fantastic article. It always made me nervous while attempting to model an optimal portfolio that the output was so extremely sensitive to the inputs. Even shifting a 15-year input period by a couple months could alter the output dramatically. I am sure we agree that that sort of statistical analysis is best used as one tool in the toolbox and at a broad level. Using an optimization as gospel does seem foolish given the magnitude and sensitivity of assumptions being made.