|
In 2003, the father of classical optimization, Nobel prize winner Harry Markowitz tested Resampled Efficiency. He, and colleague Nilufer Usmen, hypothesized that classical optimization would work better than resampled efficiency as long as better inputs were used. They pitted two fictitious players against each other—one using classical optimization and the other using resampled efficiency. Resampled Efficiency won all thirty tests, even those where the classical optimizer had better inputs! A better optimizer is more important than improved inputs.
|
"Resampled Frontiers vs. Diffuse Bayes: An Experiment"
Journal of Investment Management Q4 2003
|