'

The seemingly obscure issue of when a payment from an annuity contract purchased under a Defined Contribution plan is considered to be a "payment from an annuity" is actually one of the most pressing tax issues that needs to be resolved in the area of DC annuitization.  Resolution of this issue determines when spousal consent rules apply and notices given.  What makes it so important to have a clear set of spousal rights rules is that the timing of the application some of the actuarial rules used in determining the level of benefit under certain kinds of products can be affected and, more universally, the actual administration of the spousal consent rules can be challenging in the best of worlds.

At issue, again, in the latest PLR from the IRS is how the spousal consent rules apply to a "hybrid" annuity of the sort which is being widely designed in the 401(k) marketplace. These hybrid products guarantee lifetime income while also maintaining a participant account account balance under which the participant remains invested in equities. These types of products tend to have complex designs, but they basically permit participants to withdraw amounts from their account balances throughout their lifetime and provide "longevity insurance" payments at a guaranteed level once the account balances reach zero or the participant reaches a certain age.  

Simply put, the IRS was asked if electing and beginning a flow of withdrawals from an account balance at actuarially determined levels, where the actual "insurance" does not kick in for a number of years, constitutes "payment as an annuity" (I won't go into here the Sect 417 differences between "payments as an annuity" and "payments NOT as an annuity").

The IRS first answered that question late in 2009 in PLR 200951039.  It ruled that the that the initial election from this sort of product was legally two different elections: it was first an election to take “payments as an annuity” at that later date, when "insurance payments" arising from the longevity insurance kicked in, which then triggers application of the 417 rules. Secondly, for any payment prior to that "longevity insurance" payment start date, it was an election to take “payments not as an annuity.” 

In the new PLR, 201048044, the IRS apparently shifted from this initial position.  A vendor introduced a new product for the 401(k) marketplace which is (from a strictly legal view) strikingly similar in key respects to the product described in In the prior PLR.  The taxpayer asked for guidance on whether the initial payments under the scenario described above would be considered "payment as annuity."

Surprisingly, and in contradiction to its earlier ruling, the IRS answered that it would be.The new PLR ruled that the initial distribution election characterizes ALL periodic payments from the contract as “payments as annuity,” beginning with the first payment. This is so even though all the payments are withdrawals from an account balance, reducing that account balance. 

Lets look at the details to see why this appears to be a shift in the IRS position.

In the new PLR, the product is a specially designed “”Guaranteed Lifetime Withdrawal Benefit,” or GLWB with a number of unique design elements, held by a plan. The product in the prior PLR was a unique version of variable annuitization, distributed from a plan. The two products, however, share common design elements  which are determinative of how to apply the "payment as an annuity" rules:

-Participant DC account balances are deposited in equity based, insurance pooled separate accounts, which balances will vary with investment performance.

-The participants elect to begin taking a monthly payment at time of retirement. Positive investment performance serves to increase the monthly benefit over time, investment losses serve to decrease the benefit. Under both products, there is a guarantee base under which the benefit will never go below, in spite of investment losses.

-Payments reduce the account balance.

-The amount of the monthly withdrawals are actuarially determined. Under the product in the prior PLR, the monthly payment is calculated based on life expectancy. In the product in the new PLR, the base payment is calculated using an average calculation from a pool of insurers.

-The death benefit while there is an account balance equals the account balance.

-The participant can withdraw all or any portion of the account balance at any time. Under the product in the prior PLR, this right ceases at the earlier of a stated age or when the account balance reaches zero. Under the product in the new PLR, that right ceases at the time the account balance becomes zero.

-The insurance company continues to pay the elected monthly benefit for the rest of the participant’s life, even after the account balance reaches zero (thought he monthly benefit is ultimately reduced by early withdrawals). This is the true income guarantee.

Of the two different positions taken by the IRS, one position is not inherently better than the other. They both have their strengths and their weaknesses (which I will not go into here). Whether one is better suited than the other is so much dependent on the specific product design and the administration and system capabilities of the insurer. This new PLR is actually more consistent with the tax treatment of similar types of payments of "non-qualified" annuities (that is, annuities not sold in relation to qualified plans). 

The problem is that the differences in these two rulings are difficult to reconcile, and muddies the waters for designing DC annuitization programs. There are a number of important differences between the products described in the two PLRs, but none of them appear to impact the determination of when payments should be characterized as "payment as an annuity." If there is a critical, minute fact difference upon which the the new rule was based, it would be helpful for the IRS to highlight it so we can use it in our design efforts  I suspect, however,  that the difference lies in process: the request for this new PLR was put in much different terms than the first, and the PLR was issued out of a different IRS group than the first.

In any event, we now have contradictory rulings. While the IRS has firmed up the position that payments from these sorts of hybrid products are (at some time or another) "payments as an annuity" subject to 417, it left us with confusion as to when that status actually occurs. An opportunity  to settle the issue has been missed. Instead, we are left with a new, self-constructed bump in the road for DC annuities- without any particular policy consideration serving as its basis. 

BNA's  Tax Management Advisory Board published a Memorandum as an "Advisory Board Analysis" last week, "Income Guarantees in Defined Contribution Plans."   Click on the title to download a copy. It speaks in some detail of the technical issues confronting the provisions of annuities through a defined contribution plan.

I authored this paper, and have also committed to Al Lurie and NYU to do an even more detailed paper in the NYU "Review of Employee Benefits and Executive Compensation: 2010" to be published this summer. This paper will be a bit more extensive, further detailing the ERISA Title 1 issues related to these contracts, as well as a number of other points made in the BNA paper.

I look forward to your comments.

 

___________

 

Any discussion on any tax issue addressed in this blog (including any attachments or links) is not intended to be used, and cannot be used, for purposes of avoiding penalties imposed under the United States Internal Revenue Code or promoting, marketing or recommending to another person any transaction or tax-related position addressed therein. Further, nothing contained herein is intended to provide legal advice, nor to create an attorney client relationship with any party.  


  

 

 

 EBSA and the IRS issued their long promised Request For Information on annuities-or, should I say, as promised by the EBSA. We had not expected this particular piece to be a joint effort. It shows that Phyllis's and Mark's agenda is getting policy effectively done, not engaging in damning bureaucratic turf warfare. Hmm. With this and the recent DOL/ SEC activities, we may be seeing a trend here somewhere....

The RFI is extensive and well thought out (though they do reference  the GAO report I criticized in an earlier blog). There looks to be a lot of work put into the effort already, as it well identifies the key issues facing the idea of providing lifetime income streams.

Importantly, it does not make the mistake of focusing on "annuities." Instead, it focuses on how an adequate retirement policy addresses three key risks: longevity, Investment and Inflation (OK. So at least on THIS point the GAO report got it right).  I believe the recent attacks by the Investment Company Institute, as well as Jack Brennan, on "annuities" misses the point: there are solutions needed to each one of these risks, solutions which can have a critical role for mutual funds, investment managers and the insurance industry. This is NOT an industry specific effort, as one industry alone cannot address all three of these risks without the others.

As promised, the RFI focuses strongly on transparency (yes, yes, my Annuity Transparency blog is coming), relevancy for the average participant, portability and cost. But I was intrigued by the questions related to 404(c) and IB 96-1 (on participant education).  I am particularly interested in the insurer solvency issue, which to me is the key fiduciary risk (next to transparency), but you need to look closely in the RFI to find that issue.

The substance aside, one must be impressed by the process. We always thought the Borzi/Iwry combination would be an extraordinarily effective one, and this is proving to be true. Seeing these two longtime compatriots openly cooperate with the goal of effective public policy is something for which we have long waited.

______________

 

Any discussion on any tax issue addressed in this blog (including any attachments or links) is not intended to be used, and cannot be used, for purposes of avoiding penalties imposed under the United States Internal Revenue Code or promoting, marketing or recommending to another person any transaction or tax-related position addressed therein. Further, nothing contained herein is intended to provide legal advice, nor to create an attorney client relationship with any party.  

 

 

 

 

Annuitization from DC plans suffers from the lack of clarity on a number of key technical rules, which need to be resolved before such annuities can be widely implemented. The IRS has taken a major step in its issuance of PLR200951039, a complex PLR which- for the first time-defines what an annuity really is for purposes of DC annuitization, and when the annuity election election occurs. This is critical for determining which RMD rule  applies, and when spousal consents will be required. It also, very importantly, recognizes the Plan Distributed Annuity (see my prior blogs)  and the qualification rules which will apply to them.

Even the informed reader is likely to get lost in trying to parse through this particular PLR.  Suffice it to say that there is a highly involved set of facts related to an insurance company's specific group and individual annuity products. The relevant features are:

  • It is an annuity purchased by a DC plan for distribution to participants-either from the group annuity contract held by the plan (and not being a "plan asset", by the way) or as an individual annuity contract purchased by the plan and distributed to a participant-the classic Plan Distributed Annuity.
  • The contracts have account balances within them which are invested in variable separate accounts. The retirement distributions from these contracts are actually treated by the contracts as withdrawals from the account balance. Every dollar taken out reduces the account balance by the same amount.
  • At the time the participant starts taking payments, the participant elects how the amount of the withdrawals will be calculated. The  choice is that the payment will be equal that which would be paid under either a single life annuity or a joint and survivor annuity. This particular product gives the participant the right to actually choose the interest rate at which the annuity will be determined. 
  • The amount of the withdrawal is adjusted every year to reflect investment performance relative to the interest rate selected. It is also adjusted for any "extra" withdrawals taken by the participant during the year. 
  • At a certain age (typically age 85, but the plan can elect the age, within a range), the account balance actually disappears. All payments now come directly from the insurance company, not from the participant's account, and that payment is guaranteed for a lifetime. This particular product has an interesting twist, called "variable annuitization." This feature actually allows the participant to elect to have their annual payment adjusted in accordance with investment performance using a sort of "phantom" set of accounts.

Here's what the IRS has importantly said:

  • Payment as an annuity/not as an annuity. Payments made from the contract after the account balance is "shutdown" IS annuitization. All payments before then are NOT considered annuitization, but systematic or periodic  withdrawals (let's call it the "access period"). Those "access period payments"  are also considered RMDs, but only up to up to the calculated RMD amount. (This, by the way, means that the amounts up to the RMD cannot be rolled over, but the amounts in excess of that can be). 
  • Application /Timing of  spousal consent rules. Spousal consent is required at the time the participants elects distribution from the annuity- even though the payments during the access period are "non-annuity" payments. Electing the form of computing the payment at the time withdrawals begin is necessary under this product to make the systematic withdrawal "match up" with the actual annuity payments, to make it resemble a guaranteed income stream that is set for life. This then makes the election the same thing as currently electing an annuity payout at age 85 (or whatever age is elected), even if the intervening periodic payments are not paid as an annuity. This means that the spousal consent must be received  if the basis for computing the payment (and ultimate annuity payment at a later age) is other than (at least) 50% Joint and Survivor. Though one may quibble whether this is the right decision, we finally have  a rule we can use. As a practical matter, this may cause some problems if there is an intervening divorce and remarriage during the access period.
  • Spousal beneficiary. The account balance during the access period will still be subject to the spousal consent rules on the naming of the beneficiary.
  • RMD.  In determining the RMD, the RMD for the for payments during the access period will be determined using the account balance  under the standard DC rules. AFTER the account balance disappears, the DB method of computing the RMD will apply.

Finally, it is the overall message of the PLR which bears importance: the IRS further affirms the tax treatment of an annuity that was distributed from the plan, for an annuity that meets the requirements of 404(a)(2).

 

 

_________________

 

 Any discussion on any tax issue addressed in this blog (including any attachments or links) is not intended to be used, and cannot be used, for purposes of avoiding penalties imposed under the United States Internal Revenue Code or promoting, marketing or recommending to another person any transaction or tax-related position addressed therein. Further, nothing contained herein is intended to provide legal advice, nor to create an attorney client relationship with any party.