Foreign Pension, US Tax, and Actuarial Tables: Help!

[Garbanzor29]

Situation: Expat American, age 62. Recently retired and living overseas. I'll begin receiving a Japanese national (Social Security-like) pension this November when I turn 63.

According to the IRS, "Unless specified otherwise in an income tax treaty, foreign social security pensions are generally taxed as if they were foreign pensions or foreign annuities." [Source]

Unfortunately, my pension is not exempt from US tax under the US-Japan Tax Treaty. [Source]

I know how much I've paid into the foreign pension system (my total 'premiums', so to speak), and I've already converted that amount into US$ using the US Treasury's annual exchange rates for each year. That'll be my numerator, x.

I plan to divide my total premiums, x, by the actuarially-determined number of years that I have left to live from November 2020.  That'll be my denominator, y, so that x / y = z

Finally, I plan to subtract the result, z, from each annual pension amount to determine how much is taxable in the US.

So far, so good (I think).  Where I get stuck is figuring out what value to use for y.  According to the IRS, I'm supposed to then use "life expectancy (actuarial) tables prescribed by the IRS".  Unfortunately, there seem to be various IRS actuarial tables, each with a different figure for age 63.

QUESTION:  Which IRS actuarial table and value for y  should I use?

Single Life Expectancy   Age 63 =  22.7 years
https://www.irs.gov/publications/p590b#en_US_2019_publink[phone number removed]

Age 63  =  16.2 years (Male)
https://www.irs.gov/publications/p939#en_US_201812_publink[phone number removed]

Age 63 =  18.7731 years
https://www.irs.gov/retirement-plans/actuarial-tables
Or something else????

[jtax]

Thank you for providing citations to your info. That is very helpful.

Unfortunately I do not know the answer for sure. But I will give you some things to consider. I'm not sure you will find the answer in this volunteer forum, though it was certainly worth asking. To get the correct answer you might need to find a CPA, Enrolled Agent, or tax lawyer. And one who is experienced in this area.

The tables you are citing are measuring different things.  E.g.The P590 IRA tables are for computing required minimum distributions. That is not the same as figuring the present value of an income stream.

The "Actuarial Tables" page (https://www.irs.gov/retirement-plans/actuarial-tables) says it is for use in valuing annuities under I.R.C. sec 7520 https://www.law.cornell.edu/uscode/text/26/7520

7520(b) says that it does not apply when specified in the regulations. The publication (ok that is not a regulation but I don't have time to chase down the reg, so keep that in mind) says 

"These actuarial tables do not apply to valuations under Chapter 1, Subchapter D, (relating to qualified retirement arrangements), nor to section 72, (relating to computations for exclusion ratios for annuities), and for certain other limited purposes as provided by regulations at 1.7520-3(a), 20.7520-3(a), and 25.7520-3(a)."

(There are plenty of other uses in the tax code for valuing income streams. E.g. gift and estate taxes.)

https://www.law.cornell.edu/uscode/text/26/72 describes what I think you are trying to do. (a) says annuities are taxable, (b) gives the formula you should use  to proportionately exclude after-tax premium payments and (c)(3) says to use tables provided by the IRS (actually the Treasury Secretary). 

So I think it is clear that the "Actuarial Tables" publication does not apply to sec 72. It seems like sec 72 is what covers your case, but I can't know for certain if there is a special code section or reg that somehow does something different for your kind of pension. Seems unlikely.

So that leaves you at  Pub 939. It feels like it is addressing your question and unless you find something in it that says it doesn't apply, I'd probably suggest you consider going with it. P939 Table 1 is figuring out the present value ("expected return") ... It gives you a "return multiple." That is very different from the Pub 590 IRA RMD divisor. The use of "expected return" in Pub 939 and IRC 72(b)(1),(c)(3) further makes it seem like they are talking about the same thing.

Note of course, that Table 1 only applied for income streams payable for the lifetime of one person. It does not apply to joint annuities or term-certain or other payout periods.

I hope that helps.

**Say "Thanks" by clicking the thumb icon in a post
**Mark the post that answers your question by clicking on "Mark as Best Answer"