[Skip to content]

Sign up for our daily newsletter
The Actuary The magazine of the Institute & Faculty of Actuaries
.

Investment: The right balance

Investors often classify fundamental active managers by expected tracking error. Low tracking error managers are expected to provide excess returns while sticking closely to an index, while other managers — unconstrained by an index — are required to provide high excess returns offsetting tracking errors.

But for investors who seek the middle ground — tracking error that is neither too large nor too small — an interesting dilemma occurs. The pursuit of modest tracking error can act as a straitjacket on an active manager’s ability to use his stock-picking skills and may lead to sub-optimal portfolios with inefficient trade-offs between risk and return.

Our analysis shows that how one constrains tracking error is crucial. We believe it is far more effective to control tracking error as the outcome of a process than as an input to the process. Otherwise, a number of risk-and return-related consequences arise, including:
>> Investors may forego risk-adjusted excess returns, as the fund manager must keep an eye on the index and make decisions on stock picks that they would not make under unconstrained conditions.
>> Investors may take too much risk for the expected return, since portfolios managed to minimise tracking error may not be mean-variance-efficient — in other words, they do not maximise the potential returns at a given level of risk.

It is also the case that a fundamental active manager who explicitly targets tracking error must always keep an eye on the index and inevitably ends up making decisions on stock picks that they would not make under unconstrained conditions.

This situation has given rise to the widely held belief that many active managers are increasingly little more than ‘closet indexers’. Measured by assets under management, portfolios that differ significantly from their benchmarks fell from over 98% in 1980 to about 60% in 2003, according to a recent study by Martijn Cremers and Antti Petajisto1. The same study found that managers with a high non-index component in their portfolios tended to create value, while those making sector bets and tracking closely to an index tended to destroy value.

Modest tracking error, or active risk, is an important criterion for many pension funds. As tracking error becomes a more important metric of performance, managers are likely to structure their portfolios with strict benchmark constraints. The active component of the portfolio will decrease and the overall portfolio will look more like the index.

The greater the amount of assets under management with one manager, the more acute this problem becomes. Increasing cash flows into a portfolio tends to eliminate less liquid stocks that offer the opportunity for excess returns, channelling money into the higher capitalization names that tend to be components of the index. Moreover, at a certain point — once a portfolio reaches its capacity — the manager has more to lose by being wrong than to gain by being right.

So how can an investor solve the dilemma of controlling tracking error without limiting the excess return potential of a portfolio? Simply increasing the active portion of the portfolio would just lead to greater tracking error. One way to approach the problem is to build a portfolio in such a way that there is a low correlation of excess returns between each active segment and the benchmark portfolio.

An efficient and effective solution is to segregate the portfolio into segments run by individual portfolio managers or investment teams. Each portfolio manager or team is constrained to one segment whereby overlap is avoided through clear demarcation of each segment through one of two ways:
1 Distinct definition of investment approach or style, such as growth versus value versus special situations, or
2 By dividing the investment universe into separate segments, the most obvious one being a separation of sectors. This approach divides the investment universe among several high-alpha-seeking managers in such a way that the excess returns of one manager have a low correlation to those of other managers. When combined, the portfolios offer an improved risk-return trade-off and lower overall tracking error while maintaining expected excess return. Several investment managers offer investment team portfolios that apply the separated sector approach. One way to do this is to construct a portfolio that is sector-neutral relative to its benchmark but is unconstrained within each sector. When each sector portfolio is integrated into the overall strategy, the expected excess returns generated by the individual sector teams have a correlation of close to zero. While each sector component has a high to very high tracking error in relation to the sector benchmark, the overall tracking error of the strategy tends to be low.

Investors have a variety of tracking error requirements depending on their investment policies, constraints and portfolio structures. However, problems arise when controlling tracking error becomes an explicit input to the investment process rather than an outcome of the process. This needn’t be the case if one finds a portfolio that stresses the excess return generation potential of several managers in a structure that minimises the correlations of these excess returns among managers.

A segment- or sector-segregation approach allows active investment managers to make full use of their stock-picking skills — potentially improving the risk-return trade-off without leading to high tracking error. It meets the challenge for active managers of treating tracking error as the outcome of an investment process rather than as the overriding input to the process. It is a way of taking the asset manager out of an unduly constricting straitjacket, providing an effective solution for pension funds seeking to make the most of the middle ground.

______________________________________________________________

1 ‘How Active is Your Fund Manager? A New Measure that Predicts Performance’ Yale International Centre for Finance Working Paper

______________________________________________________________

Jim Goff is director of research at Janus Capital Group