(April 9, 2012) — Investor evaluation of a fund manager is marred by a lack of certainty, and a Bayesian approach can help provide greater clarity, according to consulting firm Towers Watson.
A Bayesian approach involves a standard set of procedures and formula to perform a specific calculation. “In the real world…investors often react to experienced underperformance by firing their fund manager. To some consultants and commentators, these actions are evidence of investor irrationality. They talk of ‘chasing returns’ and claim that much value is lost,” a newly released paper by Towers Watson asserts.
According to the paper, investors could gravitate to Bayesian statistics to better evaluate underperfomance of fund managers. “The essence of the Bayesian approach is to provide a mathematical rule explaining how you should change your existing beliefs in the light of new evidence. In other words, it allows scientists to combine new data with their existing knowledge or expertise,” the paper by Towers Watson says, citing The Economist.
The paper considers three parties: an investment board (‘the investor’), a fund manager, and an investment consultant. According to Towers Watson, an investor can use a Bayesian approach to update its belief about whether a consultant picked a star manager. “This answer will then update the investor’s view about the manager’s expected return, and guide it on whether it should retain the manager.”
The paper continues: “Studies show that investors tend to lose value when they hire and fire managers for performance reasons. One popular reason for this loss is that investors tend to fire manager that have underperformed, only to see the performance from these managers rebound in the following years.”
In the same vein, however, the paper cautions that models simplify the real world but can also lead to investors missing the complexity of real life. Therefore, investors should use models — such as the Bayesian approach — alongside existing qualitative ways of evaluating managers.
The paper by Towers Watson follows similar themes expressed by consulting firm Rogerscasey, which asserted earlier this year that recent investment performance should be only one of several criteria used to determine when to hire or fire a manager, although longer track records should not be ignored. According to the report, if a manager is having a lull, it will show up in the periodic performance numbers. However, it is imperative to realize that these statistics represent only single snapshots in time. “As a result, recent underperformance can affect the periodic returns dramatically. More often than not, these relative returns can and will change directionally from quarter to quarter. In many cases, we see the poor performance turn positive in less than four quarters if the manager begins to perform well,” asserted the paper’s authors Dave O’Donovan, Jason Bailin, and Nick Catanese. “…We advise clients to make all of their hiring and firing decisions based on a diverse set of qualitative and quantitative factors with an eye on prospects for future performance,” the authors concluded.
While Rogerscasey concluded that its clients should not retain the entirely of its underperforming managing, the decision to terminate should not be made out of frustration, as “it is the job of the consultant to remove emotion from the decision and to present the facts as to why the managers may or may not outperform once again.”