RMD

[rkbjr]

I have 2 IRA's in my name and SSN that require RMD's.  I understand that you are allowed to add together the 12/31 Balances from each and then divide the appropriate life expectance factor into the total to determine the total RMD amount.

Once the RMD amount is determined, my further understanding is that you can withdraw the RMD from either IRA account.

My question is:  Does the fact that one of the 2 IRA's is a survivor IRA change what I have described above?

Thanks you.

[Opus 17]

I assume by "survivor" IRA, you are the beneficiary of an IRA.  In that case they can't be combined and the rules are completely different.

You can only combine "like" account.  If you have one IRA in your name, you calculate and take the RMD according to the normal rules.

If you also have a beneficiary IRA, the RMD for that is calculated separately, and using a different formula.  The rules are also different depending on what year the previous owner died, what your relationship was, and what your ages were.  Can you provide those further details?

[rkbjr]

Thanks for your response.

The decedent was my father  DoD 06/24/2018.  My dob  04/15/1950

[Opus 17]

---

@rkbjr wrote:

Thanks for your response.

The decedent was my father  DoD 06/24/2018.  My dob  04/15/1950

---

OK, I may also need to know your father's age when he died. But it seems he was likely over age 70, and therefore he was past his RMD beginning age. 

I am not as familiar with the rules that were in place before 2020, so I am going to ask for another expert to tell you what your options are and how to calculate the RMD.  @dmertz 

However, I believe the rules are as follows:

Because the regular and inherited/beneficiary IRAs can't be combined, for your own regular IRA, you figure the RMD from either Table 2 or Table 3 in publication 590-B, depending on your marital situation and the other factors listed in the tables.

https://www.irs.gov/pub/irs-pdf/p590b.pdf

For the beneficiary IRA, you figure the RMD from Table 1, these are "life expectancy payments" under the old rules.  (This is not how the beneficiary's RMD is calculated after 2019, but I don't think the new rules apply to you.)

[dmertz · Accepted solution]

The RMD for the beneficiary IRA must be satisfied from the beneficiary IRA and the RMD for your own IRA must be satisfied from your own IRA.  The beneficiary IRA is still actually your father's IRA but now maintained for your benefit as beneficiary, so nothing about it can be combined with your own IRA(s).

Since your father died before 2020 (and it's reasonable to assume that he died after his required beginning date for RMDs), your father's age is not relevant to determining your RMD for the beneficiary IRA.  Your life-expectancy factor for the beneficiary RMD is determined from Table 1, the Single Life Expectancy table and your age in 2019 (age 69), reduced by 1 for each subsequent year.  For 2025, the result is 13.6, meaning that the 12/31/24 balance in the beneficiary IRA is to be divided by 13.6 to determine the RMD that must be taken from the beneficiary IRA.

For your own IRA, if you do not have a spouse who is more than 10 years younger, your life-expectancy factor comes from Table 3, the Uniform Lifetime table, based on your age in 2025 (age 75).  That factor is 24.6, which you'll divide into the 12/31/24 balance in your own IRA(s) to determine the RMD that must be taken from your own IRA(s).

[Opus 17]

@dmertz 

I looked at the 2018 pub 590-B and only found the RMD was based on table 1.  The language about using the factor for the first year minus 1 for every subsequent year, was added later.  Does that language still apply to all beneficiary RMDs, or only when the owner died after 2019?

[dmertz]

That applies to all non-spouse beneficiaries who are subject to annual RMDs regardless of the year of the decedent's death.

The reduce-by-one language is present in every version of Pub 590-B and Pub 590 before that all the way back to the original Pub 590 for 1994.  Apparently prior to 2001 you were able to elect to refigure the life expectancy every year, but that language disappeared from Pub 590 in 2001.  I suspect that that change was the result of the Economic Growth and Tax Relief Reconciliation Act of 2001, but I'm not sure.  The language has been essentially the same ever since.