Value of a pension

[TomTX]

So, I'm notionally vested in a pension, which I will be able to draw at age 65, or sooner* if I keep working for the State of Texas.

As I assembled a family Net Worth spreadsheet, I was puzzling over how to determine the value of the pension.

The simplest way I came up with is to go use the pension website, and have it calculate what purchasing** a year of service credit would cost me. The amount paid is supposed to end up being revenue-neutral to the pension, and so should be a "true" value of one year of service. I then multiply by my current years of service.

So, if I have 10 years of service and another year will cost me $11,200 - I figure the current value at $112,000.

*"Sooner" is complicated. Age + years of service = 80. I can buy 3 years of service and therefore bring in the date by 18 months. Accumulated/unspent leave time becomes service time - if and only if you retire directly into the pension. I currently have about 2 months of unspent leave time.

** Yes, I can purchase service time.

[arebelspy]

Previously discussed here: https://forum.mrmoneymustache.com/investor-alley/accounting-question-pension-as-an-asset/

As I said there, there's two strategies I think are valid:
1) Take the NPV of the income stream, or
2) Price out a comparable annuity (in this case a deferred annuity) from an insurance company and value it at that (market) price.

Hope that helps!  :)

[beltim]

I like your idea but I'd strongly suspect that value is far from linear.  To test this, try seeing what the price of 2 and 3 years of service would be.  If the price of 2 years is twice that of 1, and the price of buying 3 years is thrice that of 1, then your calculation makes sense.  But I'd guess that the cost is not a linear function.

[arebelspy]

I like your idea but I'd strongly suspect that value is far from linear.  To test this, try seeing what the price of 2 and 3 years of service would be.  If the price of 2 years is twice that of 1, and the price of buying 3 years is thrice that of 1, then your calculation makes sense.  But I'd guess that the cost is not a linear function.

Purchasing years of service is linear for my pension.  Regardless, like I said in the other thread, I don't think contribution amount gives you a very valid or meaningful number.

[TomTX]

I like your idea but I'd strongly suspect that value is far from linear.  To test this, try seeing what the price of 2 and 3 years of service would be.  If the price of 2 years is twice that of 1, and the price of buying 3 years is thrice that of 1, then your calculation makes sense.  But I'd guess that the cost is not a linear function.

It's linear.  Notionally, it is supposed to be an accurate estimate of the cost to the pension.

The problem of using the annuity method is that I'm too far from being able to claim the pension. Small changes in assumption of future inflation will make big differences in the value. And it's more effort ;)

I'm not all THAT attached to a particular value on the pension. I'm just looking for a "reasonable" number to plug in.

Edit: Hm, last time you said "pick whatever makes you feel warm and fuzzy." ;)

[brewer12345]

I like your idea but I'd strongly suspect that value is far from linear.  To test this, try seeing what the price of 2 and 3 years of service would be.  If the price of 2 years is twice that of 1, and the price of buying 3 years is thrice that of 1, then your calculation makes sense.  But I'd guess that the cost is not a linear function.

It's linear.  Notionally, it is supposed to be an accurate estimate of the cost to the pension.

The problem of using the annuity method is that I'm too far from being able to claim the pension. Small changes in assumption of future inflation will make big differences in the value. And it's more effort ;)

I'm not all THAT attached to a particular value on the pension. I'm just looking for a "reasonable" number to plug in.

Edit: Hm, last time you said "pick whatever makes you feel warm and fuzzy." ;)

It is possible to get quotes for a deferred payout annuity, but assuming you are under 50 you will get weird numbers.  I would just cuff it at the # years of service times the cost of buying another year.

[beltim]

I like your idea but I'd strongly suspect that value is far from linear.  To test this, try seeing what the price of 2 and 3 years of service would be.  If the price of 2 years is twice that of 1, and the price of buying 3 years is thrice that of 1, then your calculation makes sense.  But I'd guess that the cost is not a linear function.

Purchasing years of service is linear for my pension.  Regardless, like I said in the other thread, I don't think contribution amount gives you a very valid or meaningful number.

Now that I think about it more, its linearity makes sense.  Normally I think about it nonlinearly because pensions are linked to final (X years) salary, and most positions with pensions also have nonlinear increases in salary based on years of service or promotions.  But if you're buying years of service at the end, there aren't any more chances for salary increases or promotions.

I agree with you that contribution amounts don't give you a meaningful number.  But that's not what Tom is calculating.  Tom is calculating what the NPV of one year of service is, as calculated by the provider of the pension, and then multiplying by the number of years of his service.  It seems like that's even better than the two strategies you suggested, and certainly much easier to figure out.

[TomTX]

Exactly. I am not using contribution amounts.

I am going to the pension website, punching in current salary* and years of service. They calculate the NPV of one, two and three additional years of service. That will be the price to buy the years of service.


*Okay, I am actually doing the average of the highest 36 months, which is what would actually be used to calculate my actual pension. Buying in shortly after a raise is a net gain for the pension, as you pay based on current salary, they pay out based on 36 month average.

[arebelspy]

I follow, I was thinking earlier you were actually purchasing years and counting those.  It clicked later what you were doing.

I still think calculating the NPV yourself will be more accurate based on your situation.

Let me give you an example: If I work 5 years in my pension I'm vested and can purchase up to 5 years, and can collect at age 65.  If I work 10 years I can collect at age 60.  Yet the cost to purchase is based solely on my age and salary.

Thus if I've worked 5 years and purchase a year, they say it is worth $X.  If I've worked 9 years and purchase a year, they value it at that same $X, although that extra year in the 9th year adds WAY more to the NPV of the pension, as I can collect it 5 years earlier.

Plus it can be quite instructive to calculate it yourself based on various discount rates and longevity estimations both for future planning and just for fun to compare to theirs and see what assumptions they are using.

Further, I'd rather do it myself because I can then set the discount rate.  Pensions are often wildly optimistic (woefully so, from the viewpoint of those who depend on the payouts, and/or the taxpayers who become responsible for bailing them out) in their projections.  That's the second reason I'd rather figure out my own NPV than use their number.

But ultimately it doesn't matter.  They'll probably come close enough to each other's values anyways, and it's a meaningless number within a meaningless number.

Just fun to discuss.  :)

[beltim]

I follow, I was thinking earlier you were actually purchasing years and counting those.  It clicked later what you were doing.

I still think calculating the NPV yourself will be more accurate based on your situation.

Let me give you an example: If I work 5 years in my pension I'm vested and can purchase up to 5 years, and can collect at age 65.  If I work 10 years I can collect at age 60.  Yet the cost to purchase is based solely on my age and salary.

Thus if I've worked 5 years and purchase a year, they say it is worth $X.  If I've worked 9 years and purchase a year, they value it at that same $X, although that extra year in the 9th year adds WAY more to the NPV of the pension, as I can collect it 5 years earlier.

In which case it's worth going to Vegas to bet the value of one year's service!

[arebelspy]

I follow, I was thinking earlier you were actually purchasing years and counting those.  It clicked later what you were doing.

I still think calculating the NPV yourself will be more accurate based on your situation.

Let me give you an example: If I work 5 years in my pension I'm vested and can purchase up to 5 years, and can collect at age 65.  If I work 10 years I can collect at age 60.  Yet the cost to purchase is based solely on my age and salary.

Thus if I've worked 5 years and purchase a year, they say it is worth $X.  If I've worked 9 years and purchase a year, they value it at that same $X, although that extra year in the 9th year adds WAY more to the NPV of the pension, as I can collect it 5 years earlier.

In which case it's worth going to Vegas to bet the value of one year's service!

lol.

[TomTX]

I follow, I was thinking earlier you were actually purchasing years and counting those.  It clicked later what you were doing.

I still think calculating the NPV yourself will be more accurate based on your situation.

Let me give you an example: If I work 5 years in my pension I'm vested and can purchase up to 5 years, and can collect at age 65.  If I work 10 years I can collect at age 60.  Yet the cost to purchase is based solely on my age and salary.

Thus if I've worked 5 years and purchase a year, they say it is worth $X.  If I've worked 9 years and purchase a year, they value it at that same $X, although that extra year in the 9th year adds WAY more to the NPV of the pension, as I can collect it 5 years earlier.

Ah, there is no significant step function like that in my pension. Any early collection date is changed month-by month using the Rule of 80. When Age + Years of Service = 80, I can collect the pension. Otherwise, collect at 65.  Based on some quick BOE, If I keep working and do nothing else, I can start collecting at age ~55.5. If I buy max service and keep working, i can start collecting around my 54th birthday.

My job is actually useful, I have a good boss, the hours are flexible, and I get comp time when I work extra. Heck, the likely successor to my boss seems pretty good - though I doubt he's going anywhere for at least 5 years, more likely at least 10.

The currently calculated value of the pension adds about 18% to my net worth.

...but yeah, mostly just fun to kick around. It doesn't really matter for a long time.

[brewer12345]

For me this is a much easier matter.  The small pension I am vested in has a cash lump sum option.  Since I am not taking this at time of separation, it accrues interest every year until I either take the lump sum or start the pension (like a stable value fund).  I can't start the pension for at least 10 years and I probably will hold off for a long time.  In the meantime I just look at whatever the accumulated cash balance is and call it a day.

[ender]

When I get vested into my pension, I will have as part of our pension estimation tool a present value.

This is how I plan on valuing it.

[TomTX]

For me this is a much easier matter.  The small pension I am vested in has a cash lump sum option.  Since I am not taking this at time of separation, it accrues interest every year until I either take the lump sum or start the pension (like a stable value fund).  I can't start the pension for at least 10 years and I probably will hold off for a long time.  In the meantime I just look at whatever the accumulated cash balance is and call it a day.

Sounds like a good method if the cashout is fairly valued and is easy info to find. Mine is neither, until I actually reach retirement age.

[arebelspy]

So I was thinking about adding in the value of the pension into my net worth, because I don't currently do so.

One issue it presents though is projections.

It may be ~100k now, but that will decline in real terms.  As the rest of my portfolio grows over the next few months, years, and decades, that will stay the same amount as whatever it ends up at when I FIRE.  So it will be a smaller and smaller portion of my net worth, and when I finally collect the tiny pension, that 100k won't be worth much.  That all makes sense.

When making projections about Net Worth, however, you generally assume it will grow at X% (or whatever your assumptions are).  Only now you have 100k in there that not only won't grow at that rate, but will actually decline in real terms by whatever the inflation rate is.

So if you're doing any net worth projections, you need to back out that 100k and not count it (so why add it in the first place?) or you need to count it at its negative real return.  Either is a bit odd.

In any case, it was something I thought of while noodling over the issue, figured I'd post it in case it was relevant to anyone else.

[arebelspy]

People don't typically count social security in their net worth, right?  How is this that much different?

[fiveoh]

Is this the TRS system?  My wife is in this plan and I've looked over the documents but am still fuzzy on how it works.  Do you mind if I pm you some questions?(if this is TRS)

[RoseRelish]

I don't count Social Security because I highly doubt that it will exist in its current form by the time I'm 67.

And the way I'd value a pension is to divide the expected annual cash flow by the rate on current 30-year US Treasury Bonds. As rates move, so too would the value of your pension (assuming a fixed cash flow).

[DoubleDown]

People don't typically count social security in their net worth, right?

Ha, I hope not. Although it would be pretty great to be penniless but declare yourself a millionaire because you expect to collect 33k in SS for 30 years!

I'm guessing people want to calculate the "value" of a pension in order to reverse-engineer it into some sort of 4% SWR/25x-your-annual-expenses figure, and see if they've hit the magic number??

I don't personally count my future pension in any kind of net worth assessment, I just count it as a source of income later on that will reduce the amount I have to withdraw from other assets. Since my pension will also be calculated based on my average highest salary now (and not my inflation-adjusted salary in the future), I just apply a negative rate of return for inflation to the annual amount, compounded over the years until I withdraw it. Since I'm already 47, it's at least not a killer that makes the pension completely trivial.

[Psychstache]

Is this the TRS system?  My wife is in this plan and I've looked over the documents but am still fuzzy on how it works.  Do you mind if I pm you some questions?(if this is TRS)

The TRS manual reads like stereo instructions. I've basically figured out I am just gonna cash out due to having to wait forever to start (if retire at 40 like plan) and no COLA adjustments.