Looking Beyond Past Performance in Manager Selection

August 25, 2014 (PLANSPONSOR.com) – A new analysis from Segal Rogerscasey argues past performance is only one of many factors that should be evaluated when assessing active investment manager performance.

Many retirement plan sponsors and other institutional investors feel they are being thorough in looking back as far as five years both in selecting new investment managers and making termination decisions (see “Process for Terminating Investment Managers is Important”). However, even five-year performance can be a poor indicator of future success, according to Segal Rogerscasey, mainly due to a decided lack of long-term performance persistence within nearly all investment classes.

To determine whether past performance is indicative of future results, the firm examined 12 peer groups of investment managers across a variety of active equity and fixed-income styles. Managers were ranked into quartiles for each peer group based on investing performance from January 1, 2004, to December 31, 2008. Next, the performance of top- and bottom-quartile managers from this first period was analyzed for the subsequent five-year period running January 1, 2009, to December 31, 2013.

As explained by Segal Rogerscasey, performance persistence in this analysis was “generously defined” as when more than 50% of a peer group’s top-quartile managers from the five-year period starting in 2004 also finished in the top half of active managers for the five-year period starting in 2008. Drawing this hard line gives “somewhat more significance than is merited” to the peer groups that finished close to the 50% threshold,” researchers explain, but the overall results still hold.

The results of this analysis were particularly striking among the three fixed-income peer groups, Segal Rogerscasey says. The vast majority of fixed-income managers that led their peer groups during the 2004 to 2008 period fell into the third or fourth performance quartile during the following five years. Managers of high-yield fixed income had the greatest level of performance persistence among the three fixed-income peer groups, at 32%. This means just 32% of the high-yield managers who finished in the top quartile for 2004 to 2008 finished in either of the top two performance quartiles from 2009 to 2013.

The performance inconsistency was the most extreme for U.S. core plus fixed-income managers, who invest in a core bond portfolio with additional allocations to non-core assets, such as high-yield or emerging market debt. Nearly eight in 10 (77%) members of this peer group who finished in the first quartile between 2004 and 2008 fell into the bottom two quartiles for the period 2009 to 2013, according to Segal Rogerscasey. Conversely, 91% of the fourth-quartile managers in this peer group in period one (2004 to 2008) rose to the first or second quartiles in the following five-year period.

Another interesting upshot of the research is that this year’s findings about fixed-income peer group performance persistence are significantly different from figures covering five-year periods from 2003 to 2007 and 2008 to 2012—largely because of the lasting impact of the 2008 financial crisis on five-year investment manager return measures. Massive market corrections often have a confounding impact on medium-term performance review efforts, researchers explain, further challenging retirement plan fiduciaries to make sound backward-based predictions of future performance.

Results among the nine equity manager peer groups analyzed by Segal Rogerscasey also support the thesis that outperformance is rarely persistent. In fact, emerging market equity was the only equity peer group analyzed to show performance persistence during the most recent 10-year period, something researchers explain as an essentially random result. Strikingly, among U.S. small cap equity managers of every style, an investor had a much better chance of ending up with a first- or second-quartile manager in the 2009 to 2013 period by selecting from the fourth-quartile managers versus those who finished in the top three quartiles from 2004 to 2008.

In an attempt to identify any patterns in the performance data, researchers looked at the behaviors of the 12% of U.S. large cap core equity managers who finished in the first quartile for both 2004 to 2008 and 2009 to 2013. The only pattern identified was that managers that protected assets better in the down market of 2008 (i.e., whose portfolios fell less than the Russell 1000 in that year) had a higher likelihood of showing performance persistence in the 2009 to 2013 period of the study. Other than that, there was no discernible trend or investing behavior, such as style tilts, that might help one predict which top-quartile managers would manage to outperform in both five-year periods.

One of the more counterintuitive results of the analysis, according to Segal Rogerscasey, is the lack of performance persistence in capacity-constrained sub-asset classes, such as U.S. small-cap core, growth and value, or emerging markets equity. Some may assume that stocks in these areas are less closely followed by sell-side analysts and thus perhaps offer more scope for an informational or tactical advantage among skilled active asset managers.

This may be true to some extent, researchers say, as about 77% of emerging market equity managers have outperformed the MSCI Emerging Markets Index during the 10-year period ending December 31, 2013. However, this insight does little to help one to identify which active manager among a group of peers will perform best, or indeed which ones will even be able to beat their benchmarks.

One explanation for the inability of first-quartile active managers to maintain outperformance relevant to peers may be that a manager with a strong five-year track record is likely to attract large amounts of new investments. A major inflow in strategy assets may limit the manager’s ability to nimbly trade less-liquid assets, which may dampen performance. As stressed by Segal Rogerscasey, this type of research is highly end-point dependent. So in last year’s version of the study, which reviewed different five-year periods, results for each peer group differed substantially.

Given all this, Segal Rogerscasey recommends that investors (and retirement plan fiduciaries) should rely on a forward-looking approach based on fundamental research when selecting active managers. Such fundamental research involves plan fiduciaries closely examining the investment manager’s stated strategies and processes, Segal Rogerscasey says. It’s also important to review credentials and regulatory background for any issues. 

In this scenario, past performance is only used as a validation of a manager’s capabilities, researchers explain. When the plan fiduciaries do look at past performance, they should be sure to look across various market cycles to see how the manager performs in different macroeconomic conditions.

The full Segal Rogerscasey analysis is available here