Revenue Sharing - What's the right way to apply it in a 401(k) Plan?

In order to understand the question of how to apply Revenue Sharing in a 401(k) plan, it begs explanation as to what Revenue Sharing really is, what form it takes and how it's commonly applied.  In short, revenue sharing typically takes the form of basis points (bps) built into the underlying expense ratio of certain mutual funds, collective investment funds or annuity sub-accounts (referred to herein as "funds").  Not every fund has revenue sharing, but many do.  

Ex.) The Vanguard S&P 500 Index Admiral Share Class (VFIAX) costs 0.05%.  There isn't any revenue sharing in this fund.  Whereas, the Dodge & Cox Income fund (DODIX) costs 0.43%, but contains 0.08% (8 bps) of revenue sharing, making the underlying net expenses, 0.35%. 

Both of these funds are considered very good or excellent by most methods of evaluation, yet one has excess revenue built into it and one does not.  So what is this revenue?  Why is it there? What should a plan trustee do with it since it is a plan asset?

Revenue sharing came about as an industry concern as technology advanced in the late 1990's.  It comes in various forms or names.  Some of these names are Sub Transfer Agency (Sub-TA) fees, shareholder servicing fees, finder's fees or even 12(b)-1 fees which are typically used to pay brokerage commissions.  

If one were to look back at what the 401(k) product landscape looked like in the mid and late 1990's, it would be observed that there were two dominant types of products.  The first, is the Group Variable Annuity, where the investments were sub-advised funds, and the second was a mutual fund product offered by a mutual fund company containing only their mutual funds.  These "proprietary products" were often recordkept by that insurance company or that mutual fund company's transfer agency.  Transfer agencies are required by the investment industry and act as the record keeper for such things as share-lot accounting, tax basis, customer information and the like.  Transfer agencies are not revenue generating centers, rather they are a cost center to the investment houses.  They are financed by a ledger transaction on the fund company balance sheet called a 'Transfer Agency Fee' or a TA Fee.

During this period of time, technological systems and customer demand necessitated the creation of 'Multiple Fund Family' products, a predecessor to true open architecture.  These products still exist today.  Typically, this type of product would be distributed by a group of fund families partnered together.  Ex.) 8 fund families, 400 funds. Arguably, this is better than one fund family's 40 fund product on its own.  Often these products would only include those funds typically distributed by brokers for a commission (so no Vanguard, no DFA, etc.), and it is here that we find the birth of Revenue Sharing as we see it today.

Technology gave rise to "independent" record keeper service providers.  These record keepers were independent of the fund families, and often had superior technology and the ability to link together investment trading for multiple fund houses.  They took over the fund companies role as transfer agent for the 401(k) clients.  This freed up the aforementioned TA Fee for these types of accounts.  It didn't take the fund companies long to realize that this TA Fee could become an incentive to the record keepers for priority shelf space, similar to having a sugary cereal at eye level for a seven year old to see.  This incentive became the SUB-Transfer Agency fee (Sub-TA).  This was/is a legal practice, but it did have some systemic abuses.  In the following 5-15 years it was mainly a hidden secret of the retirement industry.  It was really in the mid-late 2000's when it gained notoriety, and then with the advent of the fee disclosure rules, primarily 408(b)-2 it became more widely known.

All that said, revenue sharing can be a good or bad thing.  It has its place in the fiduciary decision tree on the mechanics of paying for necessary services.  The question now shifts away from discovery of the existence of revenue sharing in the funds in a plan to how to properly allocate this revenue sharing.  

The Department of Labor (DOL) has been somewhat vague on this.  The 'Frost Letter', DOL Advisory Opinion 97-15A is the best guidance the industry has and essentially says that if a provider is a plan fiduciary, that it cannot retain revenue sharing received from funds and that it must credit it back to the plan.  This avoids any potential Conflict of Interest in the form of Self Dealing.  Okay, fair enough.  We, the fiduciary, pick a fund, it has revenue sharing, we collect it and give it back to the plan.  That makes sense.  Hey DOL, so how do we give it back?  In other words, what is the appropriate method?  The DOL is silent on this issue.  

Last year, in DOL Advisory Opinion 2013-03A, the DOL had a chance to illuminate the industry on what is the appropriate allocation method and, specifically, it opted not to; saying "this letter also does not address any fiduciary issues that may arise from the allocation of revenue sharing among plan expenses or individual participant accounts . . .”   Leading ERISA Attorney, Fred Reish reminds us that the method of allocation is a fiduciary decision and must be prudently considered by the responsible plan fiduciary (http://fredreish.com/advisory-opinion-2013-03a/).

The argument some in the industry make is this. Since only some funds have revenue sharing, while others do not, and revenue sharing is experienced as a cost by those who invest in those funds, but not experienced by participants who don't invest in those funds, shouldn't the revenue sharing be rebated back ONLY to those participants who experienced it as a cost?  Sounds right.  That's how other rebates work in other industries.....however, common industry practice is for revenue sharing to offset provider costs. If these costs are not offset, they would've been passed to ALL of the participants prorata.  Some providers don't give the plan trustees the option.  Some do.  Some providers provide a gross invoice that is then offset by revenue sharing and then give the client the choice to write a check on the net invoice or pass it to the participants.  

Some providers have shifted to the "participant-level" revenue sharing rebate process.  This then begs the question of when?  When is it applied, daily, monthly quarterly, annually?  What if the amount expected to receive differs materially from what is actually received?  How does one collect and then apply the difference?  Are their earnings adjustments required, if so who pays for those?  Could plan discrimination issues arise, benefits, rights and features issues?  All these are valid questions.  The issue of application of revenue sharing has now become a differentiating issue for providers, i.e. a product feature issue that can be sold or sold against.

After reading Fred's blog post linked above, he mentions forewarned is forearmed.  Rather than this becoming a potential fight one needs to be armed for, how about this instead?  As a challenge to the DOL, PLEASE COME UP WITH A 'SAFE HARBOR' METHOD ON HOW TO APPROPRIATELY ALLOCATE REVENUE SHARING WHEN IT EXISTS WITHIN A 401(K) PLAN!!!!  This would sure solve a lot of unnecessary problems or potential future problems.