Mr. FP and I are divorcing and I'm trying to figure out what value should be assigned to his defined benefit pension plan.Much depends on the assumptions used. You could look at the case study spreadsheet (http://forum.mrmoneymustache.com/ask-a-mustachian/how-to-write-a-%27case-study%27-topic/msg274228/#msg274228) to see some of the possibilities.
...
Anyway, one thing I read in the packet they sent me was that if he quit tomorrow and never worked for them again, his monthly benefit in 30 years (when he will be retirement age) would be $303/month.
So I am trying to figure out what amount of money now, left in the market, could be reasonably expected to generate that much money per month, if left in the market long enough. So far I got as far as $90,900 being the sum of money to generate that much monthly income (using 4% safe withdrawal rate), but I don't know how to calculate how much money that is now.
Any suggestions? Did I even phrase the question so it makes sense?
One way to evaluate "pension now" vs. "pension later" |
Compare pension payment promised at the later time to either |
- the "Interest generated by Future Value" (Future Value principal is not touched), or |
- the "Constant withdrawal of FV over time L" (principal goes to zero), or |
- "Trinity-style withdrawal of FV over time L" (annually inflated spending; principal -> zero) |
Lump sum now | PV | $27433 | |
Payment starting now | Pmt_now | 0 | $/payment |
Interest rate | i | 4.0% | /yr |
number of years | n | 30 | yr |
number of payments/year | freq | 12 | /yr |
When payments are made for each n | type | 0 | 0 = at end, 1 = at start |
Future Value | FV | $90900 | |
Interest generated by Future Value | FV(i,n,P) * i | 303 | $/payment |
Longevity of future pension | L | 30 | yr |
Constant withdrawal of FV over time L | Pmt_future | 434 | $/payment |
Spending growth rate (e.g., CPI) | g | 2.0% | /yr |
First year Trinity-style withdrawal | W(FV,L,i,g) | 4037 | $/yr |
336 | $/pmt |
30 years is not the length of the pension; it is the time until he can draw it.
Thanks so much for all the help! I figured out the "solve for any of the five" spreadsheet and got about the same present value that I got by trial and error for generating $309* per month in perpetuity.
I do not understand the "pension now" vs. "pension later" part. I guess the question for that part would be, what amount of money NOW, stored in the market for 29 years, would generate $303 per month over some reasonable number of years if one was OK with drawing down the principle. (20? 25? His family is not very long-lived but mine seems to be.) The $90900 then has no significance, because I wouldn't need as much, right, if I were drawing down the principle? I can't figure out how to extract that number.
It seems like it would yield a lower number than I am getting by assuming $303/month forever. And, umm, frankly I want to show that it's worth MORE.
That $309 per month is a low figure. If he works another 3 years, he'll be eligible for a different kind of calculation that will greatly increase his monthly benefit.
That being the case, is it unreasonable to use the interest-only calc? That would put the current value of his PERA at $12,777 rather than about $9K.
*I remembered it wrong before. It is $309, not $303.
I have been using a 7% interest rate and 4% safe withdrawal rate. Are those reasonable numbers to use?
How long you will live and how long he will live are not relevant. How much you withdraw and swr are not relevant.
The pension calcs are based on acturarial tables that take how long the average person will live into account, as the payouts are spread across a large population. Like figuring out the cost to purchase an annuity. They also use consistent interest rates for everyone.
That age to payout may be 80 or 82 years old... or even may be 90 years old... but is not your number, it is the pension company's number.
The approximately $27K # comes up seems to assume that over the next 29 years, whatever money I get now would earn only 4% interest. Isn't 7% more commonly used?How long you will live and how long he will live are not relevant. How much you withdraw and swr are not relevant.
The pension calcs are based on acturarial tables that take how long the average person will live into account, as the payouts are spread across a large population. Like figuring out the cost to purchase an annuity. They also use consistent interest rates for everyone.
That age to payout may be 80 or 82 years old... or even may be 90 years old... but is not your number, it is the pension company's number.
Pension calcs may be based on actuarial tables... but wouldn't lifespan and SWR be relevant in trying to assign a current lump sum value to a future lifetime benefit?
The approximately $27K # comes up seems to assume that over the next 29 years, whatever money I get now would earn only 4% interest. Isn't 7% more commonly used?How long you will live and how long he will live are not relevant. How much you withdraw and swr are not relevant.
The pension calcs are based on acturarial tables that take how long the average person will live into account, as the payouts are spread across a large population. Like figuring out the cost to purchase an annuity. They also use consistent interest rates for everyone.
That age to payout may be 80 or 82 years old... or even may be 90 years old... but is not your number, it is the pension company's number.
Pension calcs may be based on actuarial tables... but wouldn't lifespan and SWR be relevant in trying to assign a current lump sum value to a future lifetime benefit?
You would think so, but everyone has different numbers, so somewhere we all have to agree, and the default is to use the accountants numbers..
As this is calculation is for a fair division of assets, a legal exercise, you need to use the numbers that are standard practice in pension plans to use.
Hi FP,
Pension actuary here who has been quietly following your journey via your journal and divorce thread. I have a few thoughts.
The problem here, which you already understand (and goldielocks also discussed), is that the portion of Mr. FP's pension that you are truly entitled to is unknowable because it depends on future events (Mr. FP's future pay and future years of contributing to the pension plan). If you agree to receive a low-end value payoff based on Mr. FP's current accrued benefit (the $309), you give up potential value from Mr. FP's future pay increases that you ARE entitled to, which could be a meaningful amount if he works a long career under this pension system. If I were you, it would bug me to leave that potential value on the table.
On the other hand, if I were Mr. FP, I would be less than thrilled to base your payoff on a pension amount that I have not yet earned, and very well may not actually earn if I take a job outside of my current pension system. You said Mr. FP deep down doesn't really want to pay you off at all, so I would guess that he REALLY doesn't want to overpay.
You and goldielocks are both correct in a sense. I didn't make an attempt to follow your estimates, but they aren't "right" not just because the pension amount is unknowable, but because these should be actuarial calculations. On the other hand, yes, of course you can do whatever you want. You can take whatever portion of the 403(b) that you both agree is fair (calculated based on whatever methodology makes sense to you, knowing there is no "right" way to do it at this point in time - can't help you with that) in lieu of the portion of the pension you're entitled to, absolutely.
But I'm having a hard time seeing how you two will come to an agreement on this given all of the unknowns. Which brings me to a question... why the aversion to "signing papers entitling you to a tiny monthly check 30 years from now"? What is so horrible about that? I think that's the best solution for you because it is the solution where you both know that you will end up with exactly what you are entitled to and not a penny more. Easy-peasy.
I do not understand the "pension now" vs. "pension later" part. I guess the question for that part would be, what amount of money NOW, stored in the market for 29 years, would generate $30The answer to that question (see specific numbers used below) is $18,387.39 per month over some reasonable number of years if one was OK with drawing down the principle. (20? 25? His family is not very long-lived but mine seems to be.)
One way to evaluate "pension now" vs. "pension later" |
Compare pension payment promised at the later time to either |
- the "Interest generated by Future Value" (Future Value principal is not touched), or |
- the "Constant withdrawal of FV over time L" (principal goes to zero), or |
- "Trinity-style withdrawal of FV over time L" (annually inflated spending; principal -> zero) |
Lump sum now | PV | $18387 | |
Payment starting now | Pmt_now | 0 | $/payment |
Interest rate | i | 4.0% | /yr |
number of years | n | 29 | yr |
number of payments/year | freq | 12 | /yr |
When payments are made for each n | type | 0 | 0 = at end, 1 = at start |
Future Value | FV | $58541 | |
Interest generated by Future Value | FV(i,n,P) * i | 195 | $/payment |
Longevity of future pension | L | 25 | yr |
Constant withdrawal of FV over time L | Pmt_future | 309 | $/payment |
Spending growth rate (e.g., CPI) | g | 2.0% | /yr |
First year Trinity-style withdrawal | W(FV,L,i,g) | 2981 | $/yr |
248 | $/pmt |
What you want is the present value of that future income stream with an assumed investment rate. Assuming $309/month starting in 30 years and lasting 30 years, with no annual inflation adjustment and 7% discount rate I get present value of $18,890. 7% is a reasonable long term return.
You would be entitled to a portion of this if he was 100% guaranteed to collect it (live to age 95). The life expectancy of a 35 year old, however, is 43 years (age 78). So, that trims the PV back to $11,750. You can argue his family is long-lived but he can argue he could get hit by a bus tomorrow and never receive his pension.
If you are getting divorced, the co-beneficiary options aren't relevant, are they? Aren't they for the situation where he voluntarily reduces his income to ensure that you get more income in the event you outlive him?
At retirement, the portion of Mr. FP's pension that is your share would actually be converted to a monthly benefit payable over your lifetime. (Another of those pesky actuarial calculations...) :) In other words, your payment wouldn't stop upon his death; it would continue for your lifetime.
Another question--when one uses the 7% interest rate,...Depends on what you mean by "the" 7%.
...is that supposed to adjust for inflation? So if I took $4500 and put it in a compound interest calculator to see what it would be worth in 29 years and used 7%, would I be getting today's dollars or (in theory) 2045 dollars?The calculation you describe here tells you the amount you would see in your account in 2045.
At retirement, the portion of Mr. FP's pension that is your share would actually be converted to a monthly benefit payable over your lifetime. (Another of those pesky actuarial calculations...) :) In other words, your payment wouldn't stop upon his death; it would continue for your lifetime.
That's what I would have thought, but this document makes it appear that would be true only if I were the cobeneficiary (see p. 3): https://www.copera.org/sites/default/files/documents/8-145.pdf
And I would not be the cobeneficiary. He only has 2 years in--if he works another 20, say, it would hardly be fair for me to get a cobeneficiary-type share. As Alternate Payee, it appears that benefits would stop with the participant's death.