Change the way you look at style

This essay appeared in Citywire Selector on November 13, 2009.

As fund strategies become increasingly complex, selection processes must continuously innovate to stay ahead of the curve. Traditional methods of style analysis are helpful in understanding the drivers of performance and the historic positioning of funds. They can provide an efficient way to check a manager’s story, but they have clear limitations and are prone to misuse.

The negative consequences of overly ‘style-centric’ investing are often highlighted and are very real. Being too focused on style reduces a manager’s ability to make the most of his talent – the potential for alpha is squeezed out of the investment process. To own a stock because it is in your benchmark index but against your better judgement runs contrary to the basic tenets of active asset management.

Selectors can fall into the trap of concentrating on funds that are a good ‘fit’ but do not have the potential to produce performance above passive options because of restrictions on how the manager is able to invest. Finding a balance between fit and alpha is a serious challenge facing selectors. Selectors set themselves up for disappointment by expecting the manager to continue along the same path of performance while adhering to their historically-defined style.

As style analysis grew in popularity through the 1990s, more asset managers – predominantly in the US – launched investment products to satisfy the need created by new ‘styles’ entering increasingly complicated asset allocation models. When managers launch funds based on filling style boxes and not on investment expertise, the tail is wagging the dog.

To take style analysis to its useful limits and maintain control of processes in the hands of selectors, I propose the application of an ‘expectation-based’ approach. The approach recognises the importance of a qualitative assessment of a manager. It is done in the context of quantitative information gathered from returns and holdings-based style analysis, along with other quantitative techniques. It directly confronts the limitations of relying too heavily on historic information and the potential for slipping into ‘style traps’.

Using this approach, selectors develop a range of expectations about a manager’s potential positioning and shifts in style as market opportunities and risks change. The approach recognises that the investment environment is constantly changing and evolving. It assumes that fund management, properly executed, should have the capacity to act according to this changing information.

An approach based on developing expectations provides selectors with the ability to not only understand the past and current positioning and style of a fund, but to set ranges for acceptable future positioning and style. It puts selectors in control of determining the importance of style for their own needs and under what conditions it matters.

Further, it is customisable to moderate the level of allowable style drift relative to what a selector knows is a reasonable and calculated use of the manager’s talents and the constraints of the fund being evaluated. This qualitative information is often gathered through conversations with the manager.

Such an approach may lead to a more robust selection process and more predictive power for the selector. It will become particularly important when combining multiple unconstrained strategies that, by definition, are seeking returns from an increasingly broad array of investment opportunities and don’t fit neatly into a box.