Fees Significantly Vary Performance for Public DB Private Equity Investors
A report from Stanford University and Harvard Business School researchers provides food for thought about fee negotiations with private equity investment providers.
An analysis from the Center for Retirement Research (CRR) at Boston College revealed that public defined benefit (DB) plans had significantly larger assets in non-traditional asset classes—such as private equity, hedge funds, and real estate—than their corporate DB plan counterparts.
Recently, corporate DB plans have been encouraged to increase assets in private equity as a way to boost returns and update their liability-driven investing (LDI) strategies.
However, a report from Stanford University and Harvard Business School researchers provides food for thought about fee negotiations with private equity investment providers. According to the researchers’ estimates, public DB plans would have earned nearly $45 billion more on their investments—equivalent to $8.50 more per $100 invested—had each received the best fee (based on actual results rather than forecasts) contract in its respective funds.
The researchers found consistent winners and losers in the sense that some pensions systematically pay more fees than others even when investing in the same fund—what they call “pension effects.” They estimate some pensions in their sample could have earned as much as $15 more per $100 on their investments.
“We argue that this empirical finding is due to ex-ante differences in willingness to pay for the same fund,” the researchers state.
The researchers found evidence that larger pensions (as measured by overall assets under management) with stronger ties to the general partners (GP) of a fund tend to outperform other investors in the fund. In addition, pensions that have more member representation on their boards also appear to pay lower fees, “perhaps because more member representation lowers agency frictions at public pensions.”
However, the researchers note that these pension characteristics account for only a modest amount of the total pension effects that they find. “Strikingly, the distribution of pension effects is virtually identically whether we control for observable characteristics or not. In addition, controlling for pension characteristics does little to change which pension funds stand to gain the most and how much they would gain if they paid similar fees as the best performing investors in our sample,” the researchers note. They say this suggests that two seemingly similar pensions choose different fee structures when investing in the same fund.
Pensions may agree to pay different fees because they differ in their information about manager skill, meaning they have expectations about the gross return of the fund that come from outside sources, the researchers say. “A profit-maximizing GP would then optimally elicit these expectational differences and charge different fees accordingly,” they contend. A GP could offer investors a menu of fee structures from which to choose, or GPs might offer fee breaks to more informed investors in order to attract less informed investors into the fund.
The researchers also say that what they call “optimization frictions” might arise from biased beliefs about gross fund returns, a failure to fully internalize the cost structure of private market investments, poor negotiation skill or agency frictions. The suggestive evidence the researchers have for “optimization frictions” include that less than 5% of pension investors in their sample have any mention of performance fees or carry on their annual report, “despite the fact that we find differences in carry to be an important component of price dispersion.” In addition, there is some evidence that frictions in labor markets and political considerations distort public pension investment decisions.
The researchers conclude: “Size, relationships, and governance account for some of the pension effects, but the majority appear orthogonal to these observable characteristics. We argue that these facts are puzzling from the perspective of several rational models of fee determination and consistent with the idea that public pensions fail to optimize when choosing fees.”