Persistence: Does it exist? Can it be proven?

Professional investment management has come ahead in leaps and bounds over the past decade or so. The latest trend to alternative and bespoke benchmarks has undoubtedly given pension funds more ammunition to test the skill and remuneration of their managers, either external or internal.

Greg Bright*

But the big and seemingly age-old question as to whether active managers can prove that they have skill and that this skill will persist remains unanswered. Academic thought still leans to the negative, but pension funds continue to want to chance their arms with manager selection.

The old BARRA information and research group has had a long history studying performance persistence. In the early 1990s it produced a landmark paper which failed to find substantive evidence of persistence of good performance for equity managers. The paper did find evidence of persistence of bad performance though. And it also showed, somewhat surprisingly, that bond managers could show persistence of good performance.

BARRA merged with the RogersCasey consulting firm for a while and now resides as part of the MSCI index information group. The three remain linked, philosophically at least, in their continued questioning of manager skill in relation to the benchmarks which are used to demonstrate or measure that skill.

RogersCasey has published a paper on low-volatility benchmarks, which represent the latest in a range of new benchmarks for pension funds to challenge their traditional cap-weighted portfolio measurement sticks (see separate research report).

In that paper, the firm questions whether the apparent favourable risk/return characteristics of a portfolio of low-volatility stocks will persist and points out that, almost by definition, there will be difficulties in pension funds benchmarking those portfolios.

One of BARRA’s founders, Barr Rosenberg, went on to form what is now the troubled AXA-Rosenberg global quant manager. To be honest, all quant managers have had their troubles in the past three years, although AXA-Rosenberg’s were exacerbated by its failure to inform clients of a mistake in its model.

Quant managers, though, were the first to show that manager skill takes a very long time to prove, arguably decades, if it can ever be proven. They also show that certain investor behaviours do persist and a clever model has the ability to exploit them.

Where the quants appear to have come unstuck in the most recent period is the weight of money which has gone into their strategies, which has rendered them either less effective or completely ineffective.

So, where does this leave the average pension fund which is looking to put together a group of strategies to deliver on its investment goals, with either an in-house team of portfolio managers or an outsourced team?

Sadly, there is no easy answer to this question. The television commercial for painkillers might provide the best answer: if pain persists, consult your doctor.

Meanwhile, the development of new and alternative benchmarks is at least helping pension funds question the fees being charged by active managers. The costs associated with running their portfolios probably represents the one area over which pension funds have some control.

*Greg Bright is the Beijing-based publisher of