Are asset owners skilled at selecting fund managers? Research suggests not.
In a new study examining manager selection by pension funds, endowments, foundations, and sovereign wealth funds, finance professors Amit Goyal, Sunil Wahal, and M. Deniz Yavuz determined that the asset managers hired for mandates end up “significantly” underperforming those who were considered, but not chosen.
According to the researchers, these bad hiring decisions tended to be driven by two factors: past performance and personal relationships between the allocators and managers. The conclusions were based on analysis of some $1.6 trillion invested across nearly 7,000 mandates between 2002 and 2017.
First, Goyal, Wahal, and Yavuz concluded that prior returns had an “important influence” on the hiring process: A manager in the 75th percentile of past performance was 30 percent more likely to be chosen than one in the 25th percentile.
“Selection based on prior performance could be rational if there is persistence in performance,” the authors noted. “One explanation is that plan sponsors truly believe that they can select investment managers even when there is no persistence on average.”
Another possible explanation, they said, is that screening managers based on past returns “offers a defense from litigation risk and unpleasant public scrutiny.”
However, these positive returns did not appear to persist after managers were hired: On average, the authors found that chosen managers earned almost 1 percent less over the following three years, compared with fund managers that were not hired. This return gap varied by allocator, with public funds doing worse than non-public funds.
The finance professors also found different results for different types of mandates: While hired equity managers underperformed the non-chosen managers by 1.13 percent, the difference between selected and non-selected fixed-income managers was just 0.02 percent.
“Plan sponsors do not show discernable selection ability,” the authors wrote. “If anything, this first look at post-hiring returns paints a grimmer picture of the outcome of the selection process.”
[II Deep Dive: Manager Research Matters Even More Than You Think]
The other factor driving manager selection, according to the study, was which managers had pre-existing connections to the asset owners — or their consultants. Using proprietary data from Relationship Science, a firm that measures connections between professionals, the authors found that personal ties resulted in a 15 to 30 percent greater likelihood of being hired.
“The post-hiring returns of firms with relationships are, at best, indistinguishable from those without relationships, and often significantly worse,” they wrote.
Given these results, the authors suggested that relationships between managers and allocators were disproportionately beneficial to managers, who got flows — and, therefore, fees. The asset owners, meanwhile, did appear to benefit in any way, as far as the authors could tell. Fees, for instance, were “indistinguishable” between connected and unconnected fund managers.
“It is entirely possible that plans or their trustees receive countervailing benefits that we are unable to measure,” the authors concluded. “What we can say is the magnitude of the implicit price of these potential benefits is meaningful: even a 50 basis point friction implied by $1.6 trillion in mandates implies a dollar cost of $8 billion.”
