The fiduciary duty to monitor – non-fiduciary service providers
In our prior articles we have reviewed the basic principles of an ERISA fiduciary’s duty to monitor persons it appoints to carry out plan-related functions and the application of those principles to the selection of plan investments and investment managers. In this article we discuss the duty to monitor non-fiduciary service providers.
As with our prior articles, we focus on duty-to-monitor issues in DC plans. In this article we focus on two aspects of plan service provider arrangements that can make the monitoring process especially challenging: fees and revenue sharing arrangements.
We begin with some remarks about how monitoring service providers is different from monitoring investment managers.
Monitoring service providers vs. monitoring investment managers
Even though the service provider appointed by the plan fiduciary may not itself be a fiduciary, the appointment of the service provider is (typically) a fiduciary act by the plan fiduciary. In one way, this situation is ‘worse’ for the appointing fiduciary than the situation in which the fiduciary appoints another fiduciary (e.g., an investment manager). In the latter situation, there are two fiduciaries that can be sued: the appointing fiduciary (the plan fiduciary) and the appointed fiduciary (e.g., the investment manager). In the former situation, the only fiduciary who can be sued is the appointing fiduciary; the service provider is typically not a fiduciary and is thus less of a target for lawsuits under ERISA.
Further, at least with respect to active managers, there’s something ‘mysterious’ about monitoring investment managers. Performance is not just a function of meeting a generic standard. As a result, fees cannot generally be addressed as a free standing issue. They are always connected with the issue of performance: ‘excess’ fees may be justified by ‘excess’ performance. Indeed, some active manager fees may in part be a function of performance. Finally, active investment performance itself is a little mysterious. A relevant and useful benchmark is not always available (or constructible). And ‘underperformance’ may, for instance, be explained by the fact that a particular active manager’s style is ‘out of favor’ (consider the problem of evaluating value managers during the tech bubble).
Generally, none of these considerations apply to the evaluation (and review) of service providers. While evaluating the performance, competence and suitability of a service provider may be ‘hard,’ in the sense that it may require a lot of work (e.g., the gathering of a lot of data), it generally won’t be ‘mysterious’ in the way that monitoring active investment management services may be. It is generally possible to define the required service and measure the service provider’s performance. With respect to any given service (e.g., a call center), there will be differences in the quality of services offered by different providers, e.g., high touch versus a more DIY approach. But those differences themselves can generally be defined and measured.
In this context, differences in fees come to the foreground.
Vulnerability of service provider fee decisions to litigation
As we said in our article on monitoring investments, in our view, ERISA 401(k) fee litigation has generally been driven by issues of proof. And, in that regard, the most vulnerable fiduciary decisions are those where a plaintiff can make an A vs. B argument.
Active investment management fees can be defended on the grounds that the active investment manager’s services (e.g., her stock picking or market timing skills) are unique. And because they are unique, they cannot be ‘proven’ to be over-priced: there is no identical but lower-priced service they can be compared to. Thus, the ‘mystery’ of evaluating investment performance can be used as a defense.
Commodity services – such as record keeping – are not unique. Record keeping fees paid by the plan can be compared to the fees paid by other plans. This is exactly what the court did in Tussey v. ABB – the major win, thus far, for the plaintiffs in 401(k) fee litigation. In that case, the lower court compared the annual record keeping fees paid by the ABB plan to those paid by the Texa$aver Plan, the employee retirement system for employees of the State of Texas. While, on appeal, the lower court’s holdings with respect to investment fees were reversed and remanded, the holding with respect to record keeping fees was upheld.
Thus, because service providers typically are providing a ‘commodity service’ (as to which there is no ‘mystery’), service provider fees are a more vulnerable litigation target than investment manager fees. That vulnerability is exacerbated by the practice of paying those fees out of revenue sharing.
Difficulties presented by revenue sharing arrangements
Oversimplifying somewhat, in a revenue sharing arrangement some portion of the investment management fees on one or more funds in the fund menu are used to pay for, e.g., trust and record keeping services. A critical feature of such an arrangement is that the fees paid to the service provider are generally a function of assets-under-management. In a straight fee-for-services arrangement, however, record keeping fees are a function of (i) certain fixed costs plus (ii) a per capita cost, and the reasonableness of record keeping fees are generally analyzed on that basis. There is no necessary relationship between those two approaches (assets-under-management vs. fixed-cost-plus-per-capita).
Because revenue sharing fees are commonly paid on the basis of the assets-under-management in some (but not all) funds in the plan’s fund menu, revenue sharing arrangements create three problems for plan fiduciaries:
A ‘discovery’ problem: the plan fiduciary has to find out (each year/quarter/month) exactly how much is being paid in revenue sharing. Since the finalization of new ERISA section 408(b)(2) regulations (see our article DOL finalizes provider-to-sponsor fee disclosure rules — detailed review), providers are required to provide this information directly to plan fiduciaries.
An ‘analysis’ problem: the fiduciary will generally want to translate the gross amount of revenue sharing paid for, e.g., record keeping into a fixed-plus-per -capita amount, so that the fees paid for record keeping can be compared with the current market cost.
A ‘reasonableness’ problem: because revenue sharing fluctuates with assets-under-management but the reasonableness of record keeping fees are generally analyzed on a fixed-plus-per-capita cost basis, the amount of revenue sharing paid can ‘get out of whack.’ For instance, in a bull market, asset values, assets-under-management and thus the revenue sharing fees paid for record keeping services may increase dramatically even though there has been no change in the plan’s fixed costs and the number of participants.
The Tussey v. ABB litigation vividly illustrates these problems. In Tussey, the lower court found that the plan fiduciaries did not know “the amount of income generated … by revenue sharing” and paid to the record keeper (the ‘discovery’ problem). They did not know “the market costs for comparable services,” which the court found should have been evaluated on a dollars-per-head-per-year basis (the ‘analysis’ problem). And the court found that the revenue sharing-based fees that the plan did pay “far exceeded the market value for record keeping and other administrative services” (the ‘reasonableness’ problem).
Monitoring service providers: bottom line
To repeat what we said above, monitoring service providers may be ‘hard’ (in that it may require a lot of work) but it is not a ‘mystery.’ Plan fiduciaries will want to keep in mind that, because plan services are (in many cases) ‘commodities,’ making possible A vs. B comparisons, fiduciary decisions with respect to them (e.g., the selection of a record keeper) are especially vulnerable to fee-based fiduciary claims. And that revenue sharing arrangements may make monitoring more difficult (although not by any means impossible).
In this context, fiduciaries will want to engage in a regular review of the performance, competence, suitability, and, especially the cost of plan service providers. Assuming the initial decision to retain the service provider was prudent, the critical question for review will be: have there been any changes in any of the relevant circumstances – changes in plan needs, changes in the provider or provider performance or changes in the market? Revenue sharing arrangements will introduce another issue: has there been an assets-under-management driven change in the amount of fees being paid, and are those fees still reasonable?
As we said in our prior articles, ERISA prudence is, critically, about process and documentation. With respect to plan service providers, plan fiduciaries should have a manageable process that deals with all the issues; they should then follow that process; and documentation (e.g., committee minutes) should reflect that process.