Home / Time to get serious about manager selection

Comment by John Schaffer*

On October 14, Investor Strategy News published an article entitled: “What value is manager research and selection?”  Whilst the article purported to have an even-handed expose of views on the subject, it could easily have left one with the impression that manager research doesn’t add value, that we are unable to select good managers more often than not and that even if we could we focus too much on manager selection relative to asset allocation.

With the greatest respect to both the author, Greg Bright (one of our industry’s most experienced journalists and a close friend to boot) as well as other quoted luminaries (including another close friend, Ron Liling, Asia Pacific’s first Pension & Investment’s asset consultant of the year and my co-founder at InTech) I think the subject deserves better than what we received in the article mentioned.

On the subject of whether manager selection is effective in adding value, it must be said there are few, if any, sophisticated analyses on the subject.  It’s a bit like the debate about stock selection. The analyses tend to be simplistic and unhelpful.  Those arguing in favour of passive stock selection cherry pick performances which show the average active manager under-performs the index.  Active managers include all those who are not index or enhanced indexed.  If one manages with a tracking error constraint, that manager is assumed to be in the same category as an index agnostic manager.  The analysis immediately becomes a nonsense. Similarly, an analysis on manager selection needs to consider the type of managers included in the universe under consideration.

I also believe that if historically we haven’t been able to select managers well – and I doubt that’s universally true – that doesn’t mean we cannot do so effectively going forward.

I believe that there are significant opportunities to enhance the quality of manager selection outcomes.  Currently most analytical approaches tend to use the same template for every manager.  For example, applying traditional performance attribution to index agnostic managers potentially has significant limitations, particularly when conducted with a limited grasp of each managers’ idiosyncrasies of style, process and strategy.

Finally, on the subject of whether investors need to spend more time on asset allocation versus manager selection, again one is left scratching one’s head.  Of course asset allocation can have profound impacts, but that goes across the spectrums of both risk and return, and remember we are talking about predicting movements of markets, which are highly influenced by unpredictable elements of human behaviour and so called “black swans”.

Good manager selection can provide strong excess returns over time, if effectively executed, arguably with lower risk than asset allocation because of the diversification implied in most multi-manager processes.

For me, the problem is that we do not devote the time and intellectual resource required to thoroughly analyse managers individually as effectively as we could and, if we’re serious, should.  The potential rewards would justify it.  So let’s get serious about manager selection.

*John Schaffer, the co-founder of InTech, is now a third-party marketer representing Johnston Asset Management, a US-based active global equity manager.