AllocationADVISOR users are frequently puzzled by the results of their analysis if they select managers rather than market indices at the inputs to a mean variance optimization (MVO). The goal of optimizing active managers is the same as it is for optimizing asset classes, that is, to maximize return and minimize risks. However, the results tend to lead to recommendations that no one in their right mind would follow.
There many reasons why standard mean-variance optimization doesn’t work well for individual managers. Most mutual funds and ETFs don’t have a long enough return history to be useful within a MVO analysis. There are often too many managers in the mix. Quite often the correlations of absolute returns within an asset class are high. However, there are still ways one can optimize active managers by making a few tweaks to the inputs of the MVO.
Over the last decade Barton Waring has been developing a framework to translate Markowitz’s original ideas to optimize manages. The key idea behind this process is that the appropriate way to optimize managers is upon active or relative returns and risk. In other words, rather than looking at the absolute return, standard deviation, and correlations of the line-up of funds, one should look at the excess returns of a manager within an asset class.
|Traditional Model||New Model|
|Goal||Optimize Asset Classes||Optimize Active Managers|
|Return||Absolute Return||Return In Excess of a Benchmark|
|Risk||Standard Deviation||Tracking Error|
|Interaction||Correlation of Absolute Returns||Correlation of Excess Returns|
Logically, this makes sense. If one is going to have a portfolio with multiple managers within an asset class, one hopes that those managers behave differently than one another. If all the managers acted exactly the same, then there would be no diversification benefits. The proper approach would be to calculate the excess returns of managers against their benchmarks, and then use that data series as the inputs for the optimization.
This can be done in AllocationADVISOR and the results are frequently quite good. This process is spelled out in detail in a guide called “Optimizing Manager for Active Risk” . If this process sounds interesting to you, I’d encourage you to find out more by looking to this resource.
Informa Investment Solutions is part of the Business Intelligence Division of Informa PLC
This site is operated by a business or businesses owned by Informa PLC and all copyright resides with them. Informa PLC’s registered office is 5 Howick Place, London SW1P 1WG. Registered in England and Wales. Number 8860726.