Order of RMD Distributions

Submitted by Jake Brennan on Tue, 2020-01-21 18:18

Per this: https://irahelp.com/slottreport/2020-life-tables-and-rmds-todays-slott-report-mailbag, an RMD is the first distribution out of an IRA account but funds after that are not (necessarily) RMD for the other account. That does not seem correct to me. E.g., in this article: https://irahelp.com/slottreport/rmds-must-be-taken-doing-rollover it says that the full $1,000 is an excess contribution. However if the mailbag post is accurate then only the portion of the $1,000 attributable to CD #2 would be excess because, per the mailbag entry: ‘The RMD that must be satisfied is only from the specific IRA being converted. You do not have to take the RMD from the other IRA that is not being converted and can wait to take that “second” RMD later in the year.’ Can someone confirm which is correct or if I’m missing some finer point?



Submitted by Alan - IRA critic on Tue, 2020-01-21 18:54

  • The recent article is correct per IRS Reg 1.408-8 QA 4. The amount of a distribution that is not eligible for rollover is limited to the RMD for only the particular IRA issuing the distribution. This is true even though that IRA might have distributed an amount large enough to otherwise satisfy the total RMDs of several IRA accounts. In fact, by partitioning your IRA into accounts of optimal size, you can arrange to do a sizeable conversion early in an RMD year and still complete the RMDs for other IRA accounts late in that same year. The IRA account being converted must only distribute the RMD for THAT IRA account before converting.
  • Looking back at the 2012 article, it is not clear whether it is incorrect or just poorly stated. If that 1000 was the RMD for CD 2 only, then 29,000 can be rolled over and the 1000 could not, and becomes an excess IRA contribution. Since 1,000 is actually less than the lowest possible RMD for CD 2 it cannot represent the RMD for other accounts. Still, the article after stating the person was aggregating RMDs over several accounts, should have clearly stated that 1000 was only the RMD for CD 2.
  • Put another way, if you have 6 accounts and the total RMD for all of them is 20,000, and you distribute 20,000 from one account, under the aggregation rules your total RMD is satisfied. But you could then actually roll over  (“unsatisfy some of the total RMD”) the RMDs amounts from all the other accounts other than the distributing one if you wanted to. A conversion is just a rollover to a Roth IRA of a TIRA distribution.
  • So perhaps the IRS does not really want to know the dates of distributions given the enforcement implications that would trigger……………

Submitted by David Mertz on Tue, 2020-01-21 19:08

My response was going to be the same, except that I realized that, even though one of the CDs had a distributable value in 2012 of $30,000, an aggregate RMD of only $1,000 is possible if the individual’s spouse is more than 10 years younger, so I was going to leave that out (although I doubt that the hypothetical was actually created with that consideration in mind).

Add new comment

Log in or register to post comments