The "stretch IRA" is a popular estate planning strategy, where the (typically non-spouse) beneficiary of the IRA stretches out required minimum distributions over his/her life expectancy; with a young beneficiary, such as a child or even grandchild, this can result in decades of tax deferral for a large portion of an inherited IRA. However, the planning technique may soon come to an end. As a part of the "Highway Investment, Job Creation and Economic Growth Act of 2012" to reauthorize and replenish the Highway Trust Fund for interstate highway projects, Senate Finance Committee Chairman Max Baucus (D-Mont.) has proposed a provision that would require inherited IRAs to be distributed within 5 years of the original owner's death, eliminating the ability to stretch. If passed, the new rules would take effect for all deaths that occur beginning in 2013. While the legislation - and the amendment to require IRAs to be liquidated within 5 years after death - is still just proposed at this point, and may not ultimately pass in its current form, the fact that an elimination of the stretch IRA rules was on the table at all suggests that the window may soon close on this particular planning technique.
The inspiration for today's blog post was a tweet from IRA expert Bob Keebler regarding the Congressional proposal to eliminate the inherited IRA stretch. The official announcement from the Senate Finance Committee regarding chairman Baucus' proposal to include the stretch IRA provision can be seen here. According to analysis from the Joint Committee on Taxation, the provision is expected to raise $4.6 billion over the next decade, through a combination of accelerated taxation on inherited IRAs, and the annual taxation of investments once pushed out of the IRA.
So how would the new rules work? According to the committee description, the five-year rule (where under IRC Section 401(a)(9) and Treas. Reg. 1.401(a)(9)-5 the inherited IRA must be liquidated by December 31st of the 5th year after death, currently applicable only to situations where there is no designated beneficiary) would become the general rule for all distributions after death. These rules would apply regardless of whether the original IRA owner died before or after the required beginning date.
An exception to the 5-year rule would apply for so-called "eligible" beneficiaries, which includes a surviving spouse, a disabled or chronically ill (as defined under IRC Section 72(m)(7) and 7702B(c)(2), respectively)) individual, someone who is not more than 10 years younger than the original IRA owner, or a child who is still a minor. In those situations, the beneficiary would be eligible to take distributions over life expectancy beginning in the year after death (as currently exists under the law), regardless of whether the original IRA owner died before or after the required beginning date.
If an eligible beneficiary stretches a decedent's IRA but the eligible beneficiary dies, the new rules would require that the IRA becomes subject to the 5-year rule for any successor beneficiary to the IRA; if the eligible beneficiary was a minor, the 5-year rule would begin as of the date the child would have reached the age of majority. This extension of the rule is intended to ensure that the stretch IRA tied to one beneficiary's life expectancy doesn't extend multiple generations beyond that beneficiary.
As proposed, the new rules would only be effective for individuals who died in 2013 or later, grandfathering the current treatment for anyone who already has an inherited IRA, or inherits one due to a death before the end of 2012. However, the provision requiring a 5-year payout to successor beneficiaries after the death of the original beneficiary would apply to beneficiaries who pass away in 2013 or beyond, even if the original decent passed away before 2013. Unfortunately, though, the committee description rules are a bit ambiguous on this point; they also stipulate that a successor beneficiary who passes away in 2013 and beyond may be treated as a eligible beneficiary, potentially allowing the eligible beneficiary stretch provisions to apply (although the committee examples do not support this).
Example 1. Harold dies in 2012 and leaves his IRA to his 40-year-old adult son Chad. Because Harold died in 2012, Chad will be eligible to stretch the IRA over his life expectancy.
Example 2. Harold dies in 2013 and leaves his IRA to his 40-year-old son Chad. Because Harold died in 2013, Chad will not be eligible to stretch the IRA over his life expectancy, unless he otherwise meets one of the requirements for being an "eligible beneficiary" to stretch (spouse of the deceased, disabled, chronically ill, within 10 years of the decedent's age, or a minor child).
Example 3. Harold dies in 2012 and leaves his IRA to his 40-year-old adult son Chad (eligible for stretch), and Chad dies in 2014 just two years into a 40-year distribution period based on his age. Chad names his 28-year-old brother Franklin as his successor beneficiary. Because Chad died in 2013 or later, and Franklin is not an eligible beneficiary, Franklin cannot continue Chad's stretch, and under the new rules must distribute the IRA within 5 years of Chad's death.
Example 4. Harold dies in 2012 and leaves his IRA to his 40-year-old adult son Chad (eligible for stretch), and Chad dies in 2014 just two years into a 40-year distribution period based on his age. Chad names his 8-year-old daughter Samantha as his successor beneficiary. Because Chad died in 2013 or later, but Samantha is an eligible beneficiary, it is possible that Samantha might be able to continue Chad's stretch, but the currently proposed rules are not clear on whether this will be possible, or if a 5-year payout will still be required.
The most direct implication of the new rules is that the opportunity to maximally stretch an IRA for the next generation may be significantly curtailed. And while the planning technique has been popular amongst financial and estate planners, arguably if the public policy purpose of an IRA is to provide for the retirement of an individual, there really isn't a need to allow for the tax consequences to be stretched over multi-decade periods after the IRA owner dies; in point of fact, this ambiguity about whether stretch IRAs serve the public policy purpose of IRAs in general is likely why their elimination is being considered (with an exception for spouses where the IRA would likely have been intended to support both members of the couple).
On the other hand, while the proposed rules nominally eliminate the traditional stretch IRA planning opportunity, the exceptions for an eligible beneficiary are remarkably wide. Spouse beneficiaries will still be eligible to stretch (and ostensibly will still be eligible to roll the IRA over into their own retirement account as exists under current law, although this is technically not addressed in the description of the proposed rules). In addition, beyond hardship rules for disabled and chronically ill beneficiaries, the stretch rules do not apply for any beneficiary old enough to be within 10 years of the decedent's age, nor any beneficiary young enough to be a minor. Thus, the new rules appear to be targeted squarely at preventing the stretch IRA during a generational shift from the original owner to the decedent's adult children (who would be older than the age of minority but more than 10 years younger than the decedent).
Notably, the eligible beneficiary treatment to stretch appears to apply as long as the beneficiary is a minor at the time the IRA is inherited, even though the beneficiary obviously will pass the age of minority at some point thereafter. This creates the odd situation where, in a state that has an age of majority of 18, a client's 19-year-old first child may be ineligible for a stretch and must distribute within 5 years, while the client's 17-year-old second child will be eligible for a multi-decade stretch based on life expectancy, even though neither child would be eligible for a stretch if the decedent passed away one year later (after both children had reached the age of majority). Hopefully, this issue in the proposed rules will be adjusted, as it creates dramatically distorted incentives amongst potential sibling beneficiaries.
Similarly, since the rules as they currently stand allow eligible beneficiary treatment simply for beneficiary who has not reached the age of majority, but not necessarily a beneficiary who is the decedent's own dependent, the incentive will exist to preserve the IRA stretch by finding any eligible minority child in the family. Thus, for example, grandchildren may become the preferred beneficiary of an IRA just to preserve the stretch under the new rules, leaving other assets to the children who are in the "not minors yet more than 10 years younger than the decedent" age window.
The new rules also do not necessarily present a challenge for those who have no children and wish to leave assets to family members of the same generational tier. The proposed rules do still allow a stretch for any beneficiary who is not more than 10 years younger than the decedent; thus, individuals can still leave retirement accounts to siblings who are of similar age or not more than 10 years younger. In fact - and somewhat ironically - an individual could leave an IRA to someone who is older than the decedent to preserve a better stretch than leaving the IRA to one who is younger. For example, a 50-year-old decedent would preserve more of a stretch leaving the IRA to a 70-year-old uncle (life expectancy stretch since not more than 10 years younger than the decedent) than a 30-year-old child (who would have to take the money out within 5 years).
Tax Bracket Issues
The primary risk of the lost stretch IRA rules (beyond just the loss of long-term tax deferral itself) is that beneficiaries will be forced to liquidate the IRA so quickly that it will clump income into a relatively limited number of years, forcing the beneficiary into higher tax brackets. On the other hand, the reality is that the rules do still essentially allow a 5-year stretch under the 5-year rule (which is actually a 6-year stretch including the year of the IRA owner's death and the 5 tax years that follow), which can at least heavily mitigate the tax bracket impact of all but the largest multi-million-dollar IRAs (and in such scenarios, it may simply become even more desirable to find someone who can be an eligible beneficiary by virtue of being young enough, or old enough, to allow for a stretch).
On the other hand, it's also worth noting that in reality, many people don't stretch IRAs at all, but simply take the money and run. For any beneficiary who was not otherwise really going to stretch, the loss of the stretch rule and impact of a 5-year-rule is moot, given that the money was not going to stay in the IRA that long anyway!
What About Roth IRAs?
Thus far, the focus on stretch IRAs and their potential disappearance has focused on traditional IRAs and 401(k) plans, but it appears likely that the proposed legislative change will apply to Roth IRAs as well.
Although the final legislative language has not yet been revealed, the proposed legislation implies it will change the provision of IRC Section 401(a)(9) (dealing with required minimum distributions) generally; yet under Treas. Reg. 1.408A-6, Q&A-14, Roth IRAs after death are subject to the exact same required minimum distribution rules. Accordingly, if the legislation changes the RMD rules in general, it will change the post-death distribution rules for Roth IRAs as well, forcing such accounts to also be distributed within 5 years of death unless an eligible beneficiary exception applies.
If Congress changes IRC Section 401(a)(9) to force a 5-year rule for all but doesn't want to have it impact Roth IRAs, then Congress would actually have to draft a separate legislative adjustment to the tax code just to create such a Roth exception. This seems unlikely since it fails to meet a public policy purpose, not to mention that allowing the new 5-year-rule-for-all rule to apply to Roth IRAs further increases the amount of revenue the Federal government generates over the next decade.
Planning In Anticipation of the Legislation
At this point, major planning changes in advance of any legislation actually passing are probably premature. The Highway Act itself may still change further in the coming weeks, and while the new anti-stretch-IRA provision is in the Senate version of the legislation, it is not in the House version, and could easily be eliminated in committee even if the Senate version passes.
In addition, the reality is that the new rules don't change anything about IRAs during the life of the IRA owner; it is not particularly a disincentive to still contributing to and using IRAs during life. It does, however, mean that if the new rules do pass, more proactive planning about who should inherit the IRAs may be necessary, especially given the significant differences in stretch opportunities between eligible and non-eligible beneficiaries and the somewhat arbitrary way those groups are divided. Loss of the stretch opportunity may also make an IRA an even more appealing asset for charitable bequests for highly affluent clients.
On the other hand, the new rules may be a very slight disincentive for some clients to utilize Roth IRAs, which have been increasingly popular especially as a form of inheritance, to leave the next generation a tax-free retirement account as a legacy (although strictly speaking, if the next generation's tax rate is lower than the original IRA owner, it may still be better to leave them pre-tax accounts). If the reality is that the Roth IRA will be forced out within 5 years as well and there is no stretch for Roth nor traditional IRAs, the focus centers even more squarely on which tax rate is lower, the original owners or the inheritors in the 5 years following death. On the other hand, for scenarios where the tax rate of the current IRA owners is low and the inheritors is high, the new rules may make it even more appealing to do Roth conversions, since the inheritors may no longer be able to stretch the distributions and their tax consequences and would have to recognize the IRA income at the inheritors higher tax rates within a few years of the IRA owner's death.
Ultimately, this may simply turn out to be a bunch of smoke and no fire. But given the need for Congress to raise revenue and the fact that eliminating the stretch does not appear to violate the public policy purpose of IRAs (to provide for the retirement of the original IRA owner), even if the proposed rules don't make it through in the final version of the Highway Act, we may see this proposed again in the coming year. Accordingly, it may be best practice when discussing estate planning matters with clients to at least acknowledge the risk that by the time the client's estate plan is actually implemented, a stretch IRA may no longer be an arrow in the estate planner's quiver.
(Editor's Note: It appears that as of Friday, February 10th, this anti-stretch-IRA provision has been removed from the Senate version of the Highway Act now heading into full chamber reconciliation. However, in response to questions about the potential curtailment of IRAs, and echoing the public policy concerns noted in this blog, Baucus stated "IRAs were meant to be used for retirement. The inherited IRAs are being used not as a retirement tool but more like an estate-planning tool." Which means there's a good chance we'll see this rule crop up again in another piece of legislation sometime in the next year or two.)