Opinions on longevity annuities

[samclem]

We're talking about changes to the current law. People have proposed both:

E2.1 Do not provide benefit credit for earnings above the current-law taxable maximum.
and
E2.2 Provide benefit credit for earnings above the current-law taxable maximum.

The SS actuaries have calculated the financial effect of each: Long Range Solvency Provisions

E2.2 would increase the size of the "trust fund" through 2024, which means (I think) that for the next 9 years the SS program would be self-supporting. After that, the "trust fund" begins a downward trend (i.e. special bonds cashed out over time, money comes from the general fund to pay them), and the "trust fund" reaches zero "balance" at 2060.

Is that a problem? Should the "trust fund" exist forever? IIRC, it was designed as a temporary buffer to hold SS taxes (as special obligation bonds) paid by the bulge of baby boomers and then later pay them out as benefits. I was born on the tail end of the baby boom (early 1960s), and I'll be about 100 years old when the "trust fund" zeros out under proposal E2.2. That (or much sooner) would be a good time to transition SS to a true "pay as you go" system: "Retirees, starting in 2060 your checks will come directly from the contributions of those paying SS taxes. When they make more money, you'll get bigger checks, and when times are rough, we'll all be tightening our belts. But we're done running up bills for people who aren't born yet. People who are born, here, and voting will make the calls and live with the results."

It would be fairly easy to go with option E2.2 but have other aspects of the tax code "claw back" enough of the SS payments to high earners that the actual impact to the budget (not SS in isolation) would look a lot like E2.1. Increase the amount of SS subject to taxation at higher incomes, leave Roth earnings as untaxed ("keeping the promise") but have them increase taxable income dollar-for-dollar for purposes of setting the brackets for income and Cap Gains computations, etc, etc.

Either way, SS will continue to have a terrible "return" for those with high earnings--removing the cap makes it worse than today. It will be highly "progressive" when viewed as ratio of payments to benefits.

 

[Independent]

E2.2 would increase the size of the "trust fund" through 2024, which means (I think) that for the next 9 years the SS program would be self-supporting. After that, the "trust fund" begins a downward trend (i.e. special bonds cashed out over time, money comes from the general fund to pay them), and the "trust fund" reaches zero "balance" at 2060.

Is that a problem? Should the "trust fund" exist forever? IIRC, it was designed as a temporary buffer to hold SS taxes (as special obligation bonds) paid by the bulge of baby boomers and then later pay them out as benefits. I was born on the tail end of the baby boom (early 1960s), and I'll be about 100 years old when the "trust fund" zeros out under proposal E2.2. That (or much sooner) would be a good time to transition SS to a true "pay as you go" system: "Retirees, starting in 2060 your checks will come directly from the contributions of those paying SS taxes. When they make more money, you'll get bigger checks, and when times are rough, we'll all be tightening our belts. But we're done running up bills for people who aren't born yet. People who are born, here, and voting will make the calls and live with the results."

It would be fairly easy to go with option E2.2 but have other aspects of the tax code "claw back" enough of the SS payments to high earners that the actual impact to the budget (not SS in isolation) would look a lot like E2.1. Increase the amount of SS subject to taxation at higher incomes, leave Roth earnings as untaxed ("keeping the promise") but have them increase taxable income dollar-for-dollar for purposes of setting the brackets for income and Cap Gains computations, etc, etc.

Either way, SS will continue to have a terrible "return" for those with high earnings--removing the cap makes it worse than today. It will be highly "progressive" when viewed as ratio of payments to benefits.

I'm not a big fan of the "trust fund" concept, at least not over long periods of time. The fact is that the gov't does/did not invest trust fund assets in private sector bonds, so the concept just moves gov't expenses to a somewhat different sets of taxpayers.

I'd prefer to say that this proposal provides additional tax revenue equal to about 2.3% of currently taxable payroll. That's enough to close the annual cash flow gap through about 2024. And, it closes about half of the eventual annual gap.

 

[gcgang]

Vanguard has a site showing how much you need to put up today to get $1,000/month at various future dates.

https://investor.vanguard.com/annuity/fixed


Sent from my iPad using Early Retirement Forum

 

[nun]

Vanguard has a site showing how much you need to put up today to get $1,000/month at various future dates.

https://investor.vanguard.com/annuity/fixed


Sent from my iPad using Early Retirement Forum

The difficulty with this is the effect of inflation. If you are intending to defer for 20 years you'd better inflate the amount you might want in todays dollars to future dollars and buy that much.

The Vanguard number for a 65 year old male is better than that given on https://www.immediateannuities.com/

With Vanguard the 65 year old male gets $1000/month for $167105 and the same amount would get him $919/month according to Immediate Annuities. Still, lets work out what you are buying from Vanguard.

The payout rate is 7.2%, that's the sort of number that looks good to many people, but the actual interest rate if we assume the guy lives to 84 is 3.6% and if you die before age 79 the interest rate is negative. I'm a fan of annuities, but only at the right price and I would not buy an annuity on the open market today....

 

[Big_Hitter]

The payout rate is 7.2%, that's the sort of number that looks good to many people, but the actual interest rate if we assume the guy lives to 84 is 3.6% and if you die before age 79 the interest rate is negative. I'm a fan of annuities, but only at the right price and I would not buy an annuity on the open market today....

IMO, since we don't know when we are going to die, the way to look at this is what interest rate are they using to calculate the annuity amount? I would assume they are using a very updated (and loaded) mortality table.

I bet it is something north of 3.6% for the 65 yr old male immediate.

 

[gcgang]

The difficulty with this is the effect of inflation. If you are intending to defer for 20 years you'd better inflate the amount you might want in todays dollars to future dollars and buy that much.



The Vanguard number for a 65 year old male is better than that given on https://www.immediateannuities.com/



With Vanguard the 65 year old male gets $1000/month for $167105 and the same amount would get him $919/month according to Immediate Annuities. Still, lets work out what you are buying from Vanguard.



The payout rate is 7.2%, that's the sort of number that looks good to many people, but the actual interest rate if we assume the guy lives to 84 is 3.6% and if you die before age 79 the interest rate is negative. I'm a fan of annuities, but only at the right price and I would not buy an annuity on the open market today....

When you buy a life annuity, the fact that when you "die early" the rate the insurance company pays you is negative doesn't really matter. You're dead.

Inflation is a concern. But converting a portion of other "concernable" investments into an income stream may actually reduce your total concerns.


Sent from my iPhone using Early Retirement Forum

 

[nun]

When you buy a life annuity, the fact that when you "die early" the rate the insurance company pays you is negative doesn't really matter. You're dead.

Inflation is a concern. But converting a portion of other "concernable" investments into an income stream may actually reduce your total concerns.


Sent from my iPhone using Early Retirement Forum

Yes this is a very important point...unless you are concerned about leaving money to heirs. The risk of early death isn't an issue if all you are concerned about is funding your retirement, but it's a difficult thing to ignore for many people as they want to get "value for money". At an implied 3.6% interest rate to age 84 I think it's worth the risk of managing the money yourself.

 

[nun]

IMO, since we don't know when we are going to die, the way to look at this is what interest rate are they using to calculate the annuity amount? I would assume they are using a very updated (and loaded) mortality table.

I bet it is something north of 3.6% for the 65 yr old male immediate.

The 3.6% pops out of an annuity calculator if you assume a 19 year lifespan past 65.

An SPIA is not longevity insurance as it lacks the deferral component.

 

[REWahoo]

I have no interest in deferred annuities/longevity insurance, but if I did I think I'd consider buying a SPIA and saving the monthly income payments in the event I lived a long time.

 

[nun]

Yes longevity insurance is a pure insurance product so I agree that rates of return aren't obviously relevant. But it's interesting to figure out if you could self insure by seeing how much your premium might grow to if you put it in something like a CD ladder. You won't get as much as buying the insurance because of the groups risk aspect, but it's still illuminating to do the calculation. The SPIA is mostly about funding your retirement right now so comparing it to other investment options is a bit more sensible. The SPIA does have an inherent longevity component that you hope to get, but it's not the only thing you are buying. With the SPIA you are mostly spending money to reduce stock market risk.

 

[nun]

I have no interest in deferred annuities/longevity insurance, but if I did I think I'd consider buying a SPIA and saving the monthly income payments in the event I lived a long time.

I just used 20% of my portfolio to buy into my ex employers DB plan which is essentially an annuity. The plan is better that the ones available commercially and I like the fact that I'm buying in at a market peak and that it fits in nicely with Wade Pfaus analysis on using SPIAs instead of bonds in a retirement portfolio. I'm now on a reverse glide path with 80% of my remaining portfolio in equities.

 

[Big_Hitter]

The 3.6% pops out of an annuity calculator if you assume a 19 year lifespan past 65.

An SPIA is not longevity insurance as it lacks the deferral component.

that's now how you calculate the PV of an annuity - people don't die at their life expectancy, a little piece of you dies every day

 

[nun]

that's now how you calculate the PV of an annuity - people don't die at their life expectancy, a little piece of you dies every day

If we knew when we'd die this would all be a lot easier, but we don't, and it would be awful if we did. We make assumptions about life expectancy all the time in retirement planning so I just cranked out the number for the average annuity holder obviously it will only apply for a fraction of the population.

There's no hard and fast rule for buying an SPIA or longevity insurance other than make sure you get value for money. For example I was just offered a lump sum buyout from one ex-employer's pension plan and the chance to buy 10 years of contributions in another ex-employers DB plan. The first was a good lump sum offer given today's low IRS segmented rates. The second was a chance to buy into a COLA'ed pension and get a 7% pay out rate at age 55. I bought one and sold the other.

 

[pb4uski]

I had an idea. Let's say that you wanted a COLAed pension. You could buy a SPIA and then a series of these deferred annuities aka longevity insurance to bump up your benefits for inflation every 5 years.

I took payout rates on a single life of an age 60 male from immediateannuities.com and constructed the following based on a $1,000/month initial benefit.

Inflation	3%
Age	Total monthly benefit	Contract monthly benefit	Payout rate	Premium
60	1,000.00	1,000.00	5.84%	205,479.45
65	1,159.27	159.27	8.09%	23,625.33
70	1,343.92	184.64	11.45%	19,351.16
75	1,557.97	214.05	16.84%	15,253.04
80	1,806.11	248.14	26.68%	11,160.89
85	2,093.78	287.67	47.89%	7,208.19
				282,078.06

I guess that I was surprised that the premium was only 37% higher than the non-COLA benefit premium.

I don't have any particular interest in doing anything like this but was curious what the cost might be.

 

[nun]

I had an idea. Let's say that you wanted a COLAed pension. You could buy a SPIA and then a series of these deferred annuities aka longevity insurance to bump up your benefits for inflation every 5 years.

I took payout rates on a single life of an age 60 male from immediateannuities.com and constructed the following based on a $1,000/month initial benefit.

Inflation	3%
Age	Total monthly benefit	Contract monthly benefit	Payout rate	Premium
60	1,000.00	1,000.00	5.84%	205,479.45
65	1,159.27	159.27	8.09%	23,625.33
70	1,343.92	184.64	11.45%	19,351.16
75	1,557.97	214.05	16.84%	15,253.04
80	1,806.11	248.14	26.68%	11,160.89
85	2,093.78	287.67	47.89%	7,208.19
				282,078.06

I guess that I was surprised that the premium was only 37% higher than the non-COLA benefit premium.

I don't have any particular interest in doing anything like this but was curious what the cost might be.

Interesting. I wonder how it would compare with a true COLA annuity. Your numbers make me feel good about my recent DB purchase, but I knew it was a good deal or I wouldn't have taken it...........$263k at age 54 to buy a COLA single life benefit of $1650/month starting at age 55 vs your constructed COLA annuity that costs $282k at 60 and starts at $1000/month.

 

[Independent]

I had an idea. Let's say that you wanted a COLAed pension. You could buy a SPIA and then a series of these deferred annuities aka longevity insurance to bump up your benefits for inflation every 5 years.

I took payout rates on a single life of an age 60 male from immediateannuities.com and constructed the following based on a $1,000/month initial benefit.

Inflation	3%
Age	Total monthly benefit	Contract monthly benefit	Payout rate	Premium
60	1,000.00	1,000.00	5.84%	205,479.45
65	1,159.27	159.27	8.09%	23,625.33
70	1,343.92	184.64	11.45%	19,351.16
75	1,557.97	214.05	16.84%	15,253.04
80	1,806.11	248.14	26.68%	11,160.89
85	2,093.78	287.67	47.89%	7,208.19
				282,078.06

I guess that I was surprised that the premium was only 37% higher than the non-COLA benefit premium.

I don't have any particular interest in doing anything like this but was curious what the cost might be.

I used incomesolutions that I access through Vanguard.
They had one company (Principal) that has a "COLA" annuity.

Principal's premiums for an initial $1,000 per month:
$208,333 - No changes
$298,525 - Scheduled 3% increases
$300,589 - CPI-U increases

 

[Independent]

that's now how you calculate the PV of an annuity - people don't die at their life expectancy, a little piece of you dies every day

You're thinking about how an insurance company calculates the PV of an annuity. They assume they issue identical policies to thousands of people on the same day, then they assume their total payouts will decrease a little each year as people die.

But, I don't have a reason to do that type of calculation. In terms of my annuity benefits, I'm either 100% alive or 100% dead, there is no middle ground.

I determine the "value of an annuity to me" by asking whether I can meet my financial goals better with a lump sum of invested assets or with a scheduled lifetime income stream.
I've got a bunch of unknowns - future inflation, future investment returns, my date of death, extraordinary end-of-life expenses, ...
I also have various priorities. How willing am I to reduce expenses if things go poorly? How much do I want to leave to my heirs? How much flexibility do I have in that estate? How much cushion do I want? What probabilities seem acceptable?

I need to mash all that together to make a decision.

 

[pb4uski]

I used incomesolutions that I access through Vanguard.
They had one company (Principal) that has a "COLA" annuity.

Principal's premiums for an initial $1,000 per month:
$208,333 - No changes
$298,525 - Scheduled 3% increases
$300,589 - CPI-U increases

The difference between $282k for quinquenial increases equal to 3% compounded and $298k for 3% annual increases seems sensible. I didn't know Principal offered a COLI... good to know.

 

[pb4uski]

....I need to mash all that together to make a decision.

I think a sensible approach would be to calculate expected cash flows (contractual annuity benefit * probability that you will be living) and then calculate an IRR that equates that expected cash flow to the upfront premium.

 

[nun]

I determine the "value of an annuity to me" by asking whether I can meet my financial goals better with a lump sum of invested assets or with a scheduled lifetime income stream.
I've got a bunch of unknowns - future inflation, future investment returns, my date of death, extraordinary end-of-life expenses, ...
I also have various priorities. How willing am I to reduce expenses if things go poorly? How much do I want to leave to my heirs? How much flexibility do I have in that estate? How much cushion do I want? What probabilities seem acceptable?

I need to mash all that together to make a decision.

The bad thing about annuities is they way they are sold. Sales people prey on retirees' desire for guaranteed income because they remember all the times they lost money in the market and they they come along with the "8% variable/hydrid annuity" and the retiree gets scr@wed. Using an SPIA to fund part of retirement might be a good move, but you have to be incredibly sophisticated to do the same with a VA.

 

[nun]

I think a sensible approach would be to calculate expected cash flows (contractual annuity benefit * probability that you will be living) and then calculate an IRR that equates that expected cash flow to the upfront premium.

For the retiree that lives for the moment the only important annuity parameter is the pay out rate. Who cares about whether I'll live long enough to get the money back, heirs schmeirs....and, sure, survivor's benefits are nice...but what do I care, I'll be dead.....give me another gin and tonic. For the 60 year old both of the COLA example pensions have initial payout rates of about 4%, so if you value guaranteed income over leaving money to your heirs, why not use the annuity? It's as good as you'd anticipate with a balanced asset allocation. When I think of it like that a COLA'ed 7% payout rate starting a 55 is a no brainer.

 

[ReadySkiDaddy]

SPIA's and other longevity annuities? You get higher income now, but it comes at a price: Very low return on investment and less income very late in life (if you live that long). Heirs will almost certainly get shafted to one degree or another too.

 

[nun]

SPIA's and other longevity annuities? You get higher income now, but it comes at a price: Very low return on investment and less income very late in life (if you live that long). Heirs will almost certainly get shafted to one degree or another too.

A single premium deferred annuity is more appropriate for pure longevity insurance. You put down a comparatively small sum to ensure an income later in life. A lifetime SPIA will give you some longevity insurance, but without a COLA, inflation will eat into the income. The SPIA is more about current income. But with payout rates in the range of 5.5% for a single male 55 years old without a COLA I find them hard to justify on a financial level. However, if you are risk averse they might be in the mix.

 

[Big_Hitter]

I think a sensible approach would be to calculate expected cash flows (contractual annuity benefit * probability that you will be living) and then calculate an IRR that equates that expected cash flow to the upfront premium.

that's essentially how I do it

 

[Big_Hitter]

You're thinking about how an insurance company calculates the PV of an annuity.

no, I'm thinking about anyone that's ever taken a class in life contingencies and/or interest theory calculates the PV of an annuity