How Often Should You Benchmark Your 401(k) Plan?
Every excessive-fee headline prompts the same question inside a committee: when did we last check the plan against the market — and would that check hold up if we had to produce it? ERISA names no interval. It imposes an ongoing duty of prudence, which courts read as monitoring on a cadence a careful professional would keep.
The cadence that has become the standard
Review fees and services annually as part of normal committee work, and run a full market benchmark roughly every three years, or sooner on a trigger event. The annual review confirms nothing has drifted; the periodic benchmark tests the plan against what the market would actually offer today. For sponsors who would rather hold that cadence on a set schedule than track it themselves, Regency's annual Tune-Up benchmarking subscription runs the review on a fixed interval.
Triggers that shouldn't wait for the calendar
- Assets crossing a meaningful threshold — a grown plan has usually earned better economics than it's getting.
- A recordkeeper merger or platform migration affecting your provider.
- Service breakdowns — missed deposits, compliance errors, participant complaints.
- Corporate events: acquisitions, divestitures, sharp headcount change.
- A long gap since the last live check — if no one remembers it, that is the trigger.
A database printout is not a benchmark
Many advisors close the conversation with a survey report showing the plan near some median. Those reflect lagging, self-reported data — not what providers would bid for your plan now. A live-bid benchmark routinely surfaces pricing below database medians, and it is far stronger evidence of prudence: it shows the committee tested the real market, not read about an old one.
The documentation is the point
As Hughes v. Northwestern made clear, a benchmark the committee cannot produce might as well not exist. Each review should leave a record: what was examined, what the market showed, what was decided, and why. Regency structures every engagement to produce exactly that record. Our open-architecture screening covers 118 vendors, and no provider pays us to bend the result. We adapt the scope to the plan in front of us. The standard the evidence has to meet does not move.
Common questions
Is annual benchmarking legally required?
ERISA does not specify an interval; it imposes an ongoing duty of prudence that courts read as regular monitoring. The widely adopted standard — annual fee reviews plus a full market benchmark every three years or on a trigger — exists because it is the cadence sponsors can defend.
What is the difference between a fee study and a live-bid benchmark?
A fee study compares your costs against survey data; a live-bid benchmark solicits actual pricing from providers for your specific plan. Live bids are current, plan-specific, and far stronger evidence of prudence — and they often surface savings a database study never would.
Does benchmarking mean we have to change providers?
No, and most benchmarks don't end in a transition. A common outcome is the incumbent repricing once it sees live competitive data. The duty is to test the market and act on the result, not to switch vendors for its own sake.
What should a benchmark leave behind?
A written record: the plan's restated all-in costs, the alternatives examined, the committee's deliberation, and the decision. That documentation — not the benchmark itself — is what protects fiduciaries if fees are ever challenged.
Speak with Regency.
Questions about this topic as it relates to your plan? We welcome the conversation.