[Nominal] Annuities are a bet on future inflation rates and should not be purchased by retirees

Discuss all general (i.e. non-personal) investing questions and issues, investing news, and theory.
randomguy
Posts: 7974
Joined: Wed Sep 17, 2014 9:00 am

Re: Annuities are a bet on future inflation rates and should not be purchased by retirees

Post by randomguy » Thu May 30, 2019 6:31 pm

dbr wrote:
Thu May 30, 2019 6:24 pm
Ben Mathew wrote:
Thu May 30, 2019 6:22 pm

Even if it's empirically true that retirees spend less over time, I wonder if it could be driven by their finances getting tighter over time. i.e. Older retirees might choose not to go out and eat as often, not necessarily because that's what they want, but because they can't (or feel they can't) afford to take an Uber to their favorite restaurant.

I have read somewhere that this result in the data on spending by the elderly is a result of the elderly being on average too impoverished by lack of retirement resources to maintain the spending they would prefer.
I have read the exact opposite. Retiree spending dropped despite the retirees account balances going up. It could very well be different subsets if the retire populations.

User avatar
Ben Mathew
Posts: 751
Joined: Tue Mar 13, 2018 11:41 am
Location: Seattle
Contact:

Re: Annuities are a bet on future inflation rates and should not be purchased by retirees

Post by Ben Mathew » Thu May 30, 2019 6:34 pm

vineviz wrote:
Thu May 30, 2019 4:37 pm

It should be immediately clear that the trajectory of this portion of spending is entirely dependent on the personal circumstances of the retiree. It could be growing in real terms, but that's unlikely for most people: it's almost always going to either be declining in real terms or declining in nominal terms.
Even if the trajectory of expenses is declining in real terms, the ideal annuity would be one whose payouts follow that trajectory in real terms. A nominal payout that you now expect to follow that real trajectory, will not follow the real trajectory if expected inflation does not coincide with actual inflation. A real annuity that features a declining payout does not exist. But you can try to approximate it with a combination of a real annuity and TIPS ladders. Say your expected real consumption starting age 65 is $100K per year, declining at $2K per year, so it's is down to $60K by age 85. At age 65, you could buy an annuity that pays $60K per year, and combine that with a TIPS ladder to finance the extra consumption between ages 65 and 85.
Last edited by Ben Mathew on Thu May 30, 2019 6:39 pm, edited 1 time in total.

User avatar
willthrill81
Posts: 11875
Joined: Thu Jan 26, 2017 3:17 pm
Location: USA

Re: Annuities are a bet on future inflation rates and should not be purchased by retirees

Post by willthrill81 » Thu May 30, 2019 6:35 pm

randomguy wrote:
Thu May 30, 2019 6:31 pm
dbr wrote:
Thu May 30, 2019 6:24 pm
Ben Mathew wrote:
Thu May 30, 2019 6:22 pm

Even if it's empirically true that retirees spend less over time, I wonder if it could be driven by their finances getting tighter over time. i.e. Older retirees might choose not to go out and eat as often, not necessarily because that's what they want, but because they can't (or feel they can't) afford to take an Uber to their favorite restaurant.

I have read somewhere that this result in the data on spending by the elderly is a result of the elderly being on average too impoverished by lack of retirement resources to maintain the spending they would prefer.
I have read the exact opposite. Retiree spending dropped despite the retirees account balances going up. It could very well be different subsets if the retire populations.
That was my understanding as well. Retirees tend to reduce their spending when their portfolio suffers, but they don't increase their spending much, if any, when their portfolio does well.
“It's a dangerous business, Frodo, going out your door. You step onto the road, and if you don't keep your feet, there's no knowing where you might be swept off to.” J.R.R. Tolkien,The Lord of the Rings

User avatar
Ben Mathew
Posts: 751
Joined: Tue Mar 13, 2018 11:41 am
Location: Seattle
Contact:

Re: Annuities are a bet on future inflation rates and should not be purchased by retirees

Post by Ben Mathew » Thu May 30, 2019 6:47 pm

dbr wrote:
Thu May 30, 2019 6:24 pm
Ben Mathew wrote:
Thu May 30, 2019 6:22 pm

Even if it's empirically true that retirees spend less over time, I wonder if it could be driven by their finances getting tighter over time. i.e. Older retirees might choose not to go out and eat as often, not necessarily because that's what they want, but because they can't (or feel they can't) afford to take an Uber to their favorite restaurant.

I have read somewhere that this result in the data on spending by the elderly is a result of the elderly being on average too impoverished by lack of retirement resources to maintain the spending they would prefer.
Such a result would not be surprising, given what we know about human behavior. Most people are terrible at financial planning, and will not give up consumption today to get more tomorrow. It would be surprising if people who did not save enough during their working years will suddenly become fiscally responsible upon retirement and will refrain from overconsuming in the early retirement years in order to leave enough for late retirement.

User avatar
Ben Mathew
Posts: 751
Joined: Tue Mar 13, 2018 11:41 am
Location: Seattle
Contact:

Re: Annuities are a bet on future inflation rates and should not be purchased by retirees

Post by Ben Mathew » Thu May 30, 2019 6:54 pm

ResearchMed wrote:
Thu May 30, 2019 5:26 pm
Why are "graded" (or "fixed % COLA") being so disparaged here? If one doesn't have a true COLA option, isn't having some regular increase over time still much better than no increase at all?
I'm not disparaging fixed COLAs--just agreeing with JackoC that it's a misleading term. No harm if people fully understand the difference between CPI COLA and a fixed 2% or 3% COLA. One provides protection from unexpected inflation, the other does not.

If a CPI COLA is not available, or is too expensive, then a 2% or 3% COLA would be better than a flat nominal annuity. And a flat nominal annuity would be better than none (if longevity is a concern and there is a risk of running out of money at old age).

User avatar
vineviz
Posts: 4659
Joined: Tue May 15, 2018 1:55 pm

Re: Annuities are a bet on future inflation rates and should not be purchased by retirees

Post by vineviz » Thu May 30, 2019 8:00 pm

Ben Mathew wrote:
Thu May 30, 2019 6:22 pm
nisiprius wrote:
Thu May 30, 2019 4:16 pm
I seriously distrust the sudden "discovery" a few years ago that suggests that spending and needed income decreases through retirement. It's "very conVEEEENient" as the Church Lady used to say. At least one of the studies that was cited explicitly said it did not include out-of-pocket healthcare spending. And even if it is true, I think it is likely that it is just a broad statistical average over many retirees, not something that one can count on as a planning guide for me, an individual retiree. I mean, us old folks feel as if we "need" to fly first class.
Even if it's empirically true that retirees spend less over time, I wonder if it could be driven by their finances getting tighter over time. i.e. Older retirees might choose not to go out and eat as often, not necessarily because that's what they want, but because they can't (or feel they can't) afford to take an Uber to their favorite restaurant.
I'm sure that many retirees are feeling the pressure of not having enough resources, but the research I've read specifically identified a decreasing need and/or desire to spend even among those who had the ability to spend more.
"Far more money has been lost by investors preparing for corrections than has been lost in corrections themselves." ~~ Peter Lynch

User avatar
Ben Mathew
Posts: 751
Joined: Tue Mar 13, 2018 11:41 am
Location: Seattle
Contact:

Re: Annuities are a bet on future inflation rates and should not be purchased by retirees

Post by Ben Mathew » Thu May 30, 2019 9:09 pm

vineviz wrote:
Thu May 30, 2019 8:00 pm
Ben Mathew wrote:
Thu May 30, 2019 6:22 pm
nisiprius wrote:
Thu May 30, 2019 4:16 pm
I seriously distrust the sudden "discovery" a few years ago that suggests that spending and needed income decreases through retirement. It's "very conVEEEENient" as the Church Lady used to say. At least one of the studies that was cited explicitly said it did not include out-of-pocket healthcare spending. And even if it is true, I think it is likely that it is just a broad statistical average over many retirees, not something that one can count on as a planning guide for me, an individual retiree. I mean, us old folks feel as if we "need" to fly first class.
Even if it's empirically true that retirees spend less over time, I wonder if it could be driven by their finances getting tighter over time. i.e. Older retirees might choose not to go out and eat as often, not necessarily because that's what they want, but because they can't (or feel they can't) afford to take an Uber to their favorite restaurant.
I'm sure that many retirees are feeling the pressure of not having enough resources, but the research I've read specifically identified a decreasing need and/or desire to spend even among those who had the ability to spend more.
Interesting. I think it will be useful to get the perspectives of some of the older retirees on this board as to what their experience has been. I have started a separate thread requesting their input:

Older retirees: Do you have lower expenses now than you did in early retirement?

AlohaJoe
Posts: 4671
Joined: Mon Nov 26, 2007 2:00 pm
Location: Saigon, Vietnam

Re: Annuities are a bet on future inflation rates and should not be purchased by retirees

Post by AlohaJoe » Thu May 30, 2019 10:39 pm

ResearchMed wrote:
Thu May 30, 2019 5:26 pm
Why are "graded" (or "fixed % COLA") being so disparaged here? If one doesn't have a true COLA option, isn't having some regular increase over time still much better than no increase at all?

Worst case, there was no inflation at all, and you didn't need the increase.
I don't think that's the worst case. The worst case is that inflation is more than the graduated payments and you have to cut your standard of living. Here's an example showing how graduated payments actually provide less inflation protection than flat payments over the first many years:

You and your wife are both 70, you go to immediateannuities and price out a $250,000 annuity both with & without 2% graduated payments. With flat payments you get $15,768 a month for life. With graduated payments you get $12,324 the first year and then it increases by 2% a year after that. Then let's assume inflation turns out to be 2.5% a year (so only a very small difference from the graduated payments).

Image

See how in year 7, with graduated payments, you have to cut your standard of living? But with flat payments you are still "protected" from the inflation simply due to the fact that your payments were bigger than you needed to start with for the same amount of money.

It isn't until the 14th year (when you are 84) that the graduated payments finally provide more inflation protection than the flat payments. (That cross-over point is the same regardless of how high inflation gets.)

So flat payments provide more inflation protection over the first 14 years at the expense of less inflation protection over the remaining years.

That seems to me like a non-trivial decision for a retiree to make, seeing as they need to balance conflicting goals/fears about inflation risk aversion and longevity risk aversion. Two people with different risk aversions could make different decisions and both be right, neither be wrong, I think.

User avatar
ResearchMed
Posts: 9052
Joined: Fri Dec 26, 2008 11:25 pm

Re: Annuities are a bet on future inflation rates and should not be purchased by retirees

Post by ResearchMed » Thu May 30, 2019 10:45 pm

AlohaJoe wrote:
Thu May 30, 2019 10:39 pm
ResearchMed wrote:
Thu May 30, 2019 5:26 pm
Why are "graded" (or "fixed % COLA") being so disparaged here? If one doesn't have a true COLA option, isn't having some regular increase over time still much better than no increase at all?

Worst case, there was no inflation at all, and you didn't need the increase.
I don't think that's the worst case. The worst case is that inflation is more than the graduated payments and you have to cut your standard of living. Here's an example showing how graduated payments actually provide less inflation protection than flat payments over the first many years:

You and your wife are both 70, you go to immediateannuities and price out a $250,000 annuity both with & without 2% graduated payments. With flat payments you get $15,768 a month for life. With graduated payments you get $12,324 the first year and then it increases by 2% a year after that. Then let's assume inflation turns out to be 2.5% a year (so only a very small difference from the graduated payments).

Image

See how in year 7, with graduated payments, you have to cut your standard of living? But with flat payments you are still "protected" from the inflation simply due to the fact that your payments were bigger than you needed to start with for the same amount of money.

It isn't until the 14th year (when you are 84) that the graduated payments finally provide more inflation protection than the flat payments. (That cross-over point is the same regardless of how high inflation gets.)

So flat payments provide more inflation protection over the first 14 years at the expense of less inflation protection over the remaining years.

That seems to me like a non-trivial decision for a retiree to make, seeing as they need to balance conflicting goals/fears about inflation risk aversion and longevity risk aversion. Two people with different risk aversions could make different decisions and both be right, neither be wrong, I think.
We'd be looking at it considering the *later* years, which is why we'd be getting a life annuity in the first place.
It's not the early years that are the concern wrt life annuities. At least, not for us. It's the "what if we live very, very long" (as relatives have and are doing, approaching 100).

But yes, "different strokes for different folks", etc., here!

RM
This signature is a placebo. You are in the control group.

User avatar
willthrill81
Posts: 11875
Joined: Thu Jan 26, 2017 3:17 pm
Location: USA

Re: Annuities are a bet on future inflation rates and should not be purchased by retirees

Post by willthrill81 » Thu May 30, 2019 10:48 pm

AlohaJoe wrote:
Thu May 30, 2019 10:39 pm
So flat payments provide more inflation protection over the first 14 years at the expense of less inflation protection over the remaining years.
Good analysis.
ResearchMed wrote:
Thu May 30, 2019 10:45 pm
We'd be looking at it considering the *later* years, which is why we'd be getting a life annuity in the first place.
It's not the early years that are the concern wrt life annuities. At least, not for us. It's the "what if we live very, very long" (as relatives have and are doing, approaching 100).
My thoughts precisely. As I've said before, most people who are attracted to life annuities are trying to insure against longevity. As such, it would seem prudent to me to buy either a CPI-linked or at least a graduated-payment annuity. Otherwise, you pretty well know that if the risk you're attempting to insure against manifests itself (i.e. you live a long time), you're very likely to get a nasty kick in the teeth from inflation.
“It's a dangerous business, Frodo, going out your door. You step onto the road, and if you don't keep your feet, there's no knowing where you might be swept off to.” J.R.R. Tolkien,The Lord of the Rings

3-20Characters
Posts: 571
Joined: Tue Jun 19, 2018 2:20 pm

Re: Annuities are a bet on future inflation rates and should not be purchased by retirees

Post by 3-20Characters » Fri May 31, 2019 1:17 am

Ben Mathew wrote:
Thu May 30, 2019 6:22 pm
nisiprius wrote:
Thu May 30, 2019 4:16 pm
I seriously distrust the sudden "discovery" a few years ago that suggests that spending and needed income decreases through retirement. It's "very conVEEEENient" as the Church Lady used to say. At least one of the studies that was cited explicitly said it did not include out-of-pocket healthcare spending. And even if it is true, I think it is likely that it is just a broad statistical average over many retirees, not something that one can count on as a planning guide for me, an individual retiree. I mean, us old folks feel as if we "need" to fly first class.
Even if it's empirically true that retirees spend less over time, I wonder if it could be driven by their finances getting tighter over time. i.e. Older retirees might choose not to go out and eat as often, not necessarily because that's what they want, but because they can't (or feel they can't) afford to take an Uber to their favorite restaurant.

Healthcare expenses get more expensive over time. Property taxes seem to go up in real terms.

Also an income that stays flat in dollars when all prices are rising would seem to be psychologically stressful. That alone is probably worth getting the COLA.
My anecdotal observation is that most of the elderly (over 80) choose to spend significantly less across the board in almost all discretionary categories. They choose to eat out less, travel less, don’t buy large ticket items, and even put off any “optional” house repairs when they can afford to make them (often at the consternation of their children). What I see is that past a certain age, the desire to do and consume seems to abate drastically and the comfort of simplicity and routine in day-to-day life is paramount.

Medical expenses are a worrying category but Medicare plus medigap seem to cover them very well in most cases. In fact, at 61 and with coverage in the individual market (aca), I’m worse off in this department.

The biggest expense that seems to come up again and again is extended end of life care. A year or more in a long term care facility (or needing full time aides at home) often depletes their savings and leaves the remaining spouse with little or no money. (I’m talking about retirees that have savings in the few hundreds of thousands—not some of the lofty multi-million dollar nest eggs reported here.)

User avatar
ResearchMed
Posts: 9052
Joined: Fri Dec 26, 2008 11:25 pm

Re: Annuities are a bet on future inflation rates and should not be purchased by retirees

Post by ResearchMed » Fri May 31, 2019 6:01 am

willthrill81 wrote:
Thu May 30, 2019 10:48 pm
AlohaJoe wrote:
Thu May 30, 2019 10:39 pm
So flat payments provide more inflation protection over the first 14 years at the expense of less inflation protection over the remaining years.
Good analysis.
ResearchMed wrote:
Thu May 30, 2019 10:45 pm
We'd be looking at it considering the *later* years, which is why we'd be getting a life annuity in the first place.
It's not the early years that are the concern wrt life annuities. At least, not for us. It's the "what if we live very, very long" (as relatives have and are doing, approaching 100).
My thoughts precisely. As I've said before, most people who are attracted to life annuities are trying to insure against longevity. As such, it would seem prudent to me to buy either a CPI-linked or at least a graduated-payment annuity. Otherwise, you pretty well know that if the risk you're attempting to insure against manifests itself (i.e. you live a long time), you're very likely to get a nasty kick in the teeth from inflation.
By the way, we prefer the "less insurancy" wording of "insuring FOR longevity", rather than "insuring AGAINST longevity". :happy
It's counter to traditional "insurance speak", because it's different from most other insurance types. Generally, what one is insuring is something one hopes will not happen (auto accident, house fire, serious illness or death, etc.)
But with life annuities, one would typically hope that "long life" DOES happen!

RM
This signature is a placebo. You are in the control group.

Leesbro63
Posts: 5907
Joined: Mon Nov 08, 2010 4:36 pm

Re: Annuities are a bet on future inflation rates and should not be purchased by retirees

Post by Leesbro63 » Fri May 31, 2019 7:02 am

ResearchMed wrote:
Fri May 31, 2019 6:01 am
willthrill81 wrote:
Thu May 30, 2019 10:48 pm
AlohaJoe wrote:
Thu May 30, 2019 10:39 pm
So flat payments provide more inflation protection over the first 14 years at the expense of less inflation protection over the remaining years.
Good analysis.
ResearchMed wrote:
Thu May 30, 2019 10:45 pm
We'd be looking at it considering the *later* years, which is why we'd be getting a life annuity in the first place.
It's not the early years that are the concern wrt life annuities. At least, not for us. It's the "what if we live very, very long" (as relatives have and are doing, approaching 100).
My thoughts precisely. As I've said before, most people who are attracted to life annuities are trying to insure against longevity. As such, it would seem prudent to me to buy either a CPI-linked or at least a graduated-payment annuity. Otherwise, you pretty well know that if the risk you're attempting to insure against manifests itself (i.e. you live a long time), you're very likely to get a nasty kick in the teeth from inflation.
By the way, we prefer the "less insurancy" wording of "insuring FOR longevity", rather than "insuring AGAINST longevity". :happy
It's counter to traditional "insurance speak", because it's different from most other insurance types. Generally, what one is insuring is something one hopes will not happen (auto accident, house fire, serious illness or death, etc.)
But with life annuities, one would typically hope that "long life" DOES happen!

RM
Kinda like how “life insurance” is actually death insurance.

longinvest
Posts: 3754
Joined: Sat Aug 11, 2012 8:44 am

Re: Annuities are a bet on future inflation rates and should not be purchased by retirees

Post by longinvest » Fri May 31, 2019 7:11 am

AlohaJoe wrote:
Tue May 28, 2019 9:25 pm
$100,000 for a male/female couple that are 65/65 and live in Colorado pays $470/month for life (nominal).

With 2% adjustments it pays $350.

With CPI adjustments it pays $305.
Given the Federal Reserve's 2% inflation target, is a (($350 / $305) - 1) = 15% premium reasonable for CPI indexing over 2% indexing? In all likelihood, inflation will be close to 2% during the retiree's remaining life. The additional 15% cost of a CPI indexed SPIA will only benefit the retiree if inflation significantly increases beyond the Federal Reserve's inflation target.

I'm of the opinion that, in absence of a competitively-priced CPI-indexed annuity, a 2%-indexed annuity is a good enough alternative to insure a base of stable lifelong income (in addition to Social Security) when the Federal Reserve communicates its intention to keep inflation on a 2% target to allow the public "to make accurate longer-term economic and financial decisions".
Bogleheads investment philosophy | single-ETF balanced portfolio | VBAL

BigJohn
Posts: 1735
Joined: Wed Apr 02, 2014 11:27 pm

Re: Annuities are a bet on future inflation rates and should not be purchased by retirees

Post by BigJohn » Fri May 31, 2019 7:45 am

longinvest wrote:
Fri May 31, 2019 7:11 am
I'm of the opinion that, in absence of a competitively-priced CPI-indexed annuity, a 2%-indexed annuity is a good enough alternative to insure a base of stable lifelong income (in addition to Social Security) when the Federal Reserve communicates its intention to keep inflation on a 2% target to allow the public "to make accurate longer-term economic and financial decisions".
I agree with this perspective. While inflation is a significant risk in retirement, it's not the only risk you have to worry about and sometimes managing for one risk doesn't help (or can even hurt) the others. I decided years ago that I was going to take a balanced approach, doing something to manage all the major risks. However, this means that none of the risks are managed to the max extent possible. While annuities are currently not in my plans, if I ever decide to add one, a 2% indexed version would be my choice.

longinvest
Posts: 3754
Joined: Sat Aug 11, 2012 8:44 am

Re: Annuities are a bet on future inflation rates and should not be purchased by retirees

Post by longinvest » Fri May 31, 2019 7:46 am

longinvest wrote:
Fri May 31, 2019 7:11 am
AlohaJoe wrote:
Tue May 28, 2019 9:25 pm
$100,000 for a male/female couple that are 65/65 and live in Colorado pays $470/month for life (nominal).

With 2% adjustments it pays $350.

With CPI adjustments it pays $305.
Given the Federal Reserve's 2% inflation target, is a (($350 / $305) - 1) = 15% premium reasonable for CPI indexing over 2% indexing? In all likelihood, inflation will be close to 2% during the retiree's remaining life. The additional 15% cost of a CPI indexed SPIA will only benefit the retiree if inflation significantly increases beyond the Federal Reserve's inflation target.

I'm of the opinion that, in absence of a competitively-priced CPI-indexed annuity, a 2%-indexed annuity is a good enough alternative to insure a base of stable lifelong income (in addition to Social Security) when the Federal Reserve communicates its intention to keep inflation on a 2% target to allow the public "to make accurate longer-term economic and financial decisions".
Here's a numerical analysis. I assumed that despite the Federal Reserve's 2% inflation target, inflation grows 3%/year during the remaining lifetime of the joint annuitants. Here's an analysis of two income streams in real dollars:

Code: Select all

 Age  2%-indexed   Cumulative   CPI-indexed   Cumulative  
   65      $350.00      $350.00       $305.00      $305.00
   66      $346.53      $696.53       $305.00      $610.00
   67      $343.10    $1,039.64       $305.00      $915.00
   68      $339.71    $1,379.34       $305.00    $1,220.00
   69      $336.34    $1,715.69       $305.00    $1,525.00
   70      $333.01    $2,048.70       $305.00    $1,830.00
   71      $329.72    $2,378.42       $305.00    $2,135.00
   72      $326.45    $2,704.87       $305.00    $2,440.00
   73      $323.22    $3,028.09       $305.00    $2,745.00
   74      $320.02    $3,348.11       $305.00    $3,050.00
   75      $316.85    $3,664.96       $305.00    $3,355.00
   76      $313.71    $3,978.67       $305.00    $3,660.00
   77      $310.61    $4,289.28       $305.00    $3,965.00
   78      $307.53    $4,596.81       $305.00    $4,270.00
   79      $304.49    $4,901.30       $305.00    $4,575.00
   80      $301.47    $5,202.77       $305.00    $4,880.00
   81      $298.49    $5,501.26       $305.00    $5,185.00
   82      $295.53    $5,796.79       $305.00    $5,490.00
   83      $292.61    $6,089.39       $305.00    $5,795.00
   84      $289.71    $6,379.10       $305.00    $6,100.00
   85      $286.84    $6,665.94       $305.00    $6,405.00
   86      $284.00    $6,949.94       $305.00    $6,710.00
   87      $281.19    $7,231.13       $305.00    $7,015.00
   88      $278.40    $7,509.54       $305.00    $7,320.00
   89      $275.65    $7,785.19       $305.00    $7,625.00
   90      $272.92    $8,058.10       $305.00    $7,930.00
   91      $270.22    $8,328.32       $305.00    $8,235.00
   92      $267.54    $8,595.86       $305.00    $8,540.00
   93      $264.89    $8,860.76       $305.00    $8,845.00
   94      $262.27    $9,123.02       $305.00    $9,150.00
It takes until age 94 for the CPI-indexed annuity to cumulatively pay as much as the 2%-indexed annuity in real dollars. In other words, assuming a 0% real discount rate, it takes 29 years of survival for the CPI-indexed annuity to be "financially fair" in a constant 3% inflation world (ignoring the higher duration of the CPI-indexed stream which should logically command a higher discount rate than the 2%-indexed stream).

At age 94, the 2%-indexed income stream is still paying ($262.27/$305) = 86% of the value of the CPI-indexed SPIA, while it has lost (($262.27/$350) - 1) = 25% of its initial purchase power.

I'll leave it to readers to decide if the fear of runaway inflation justifies an additional 15% cost for a CPI-indexed joint SPIA over a 2%-indexed joint SPIA.
Bogleheads investment philosophy | single-ETF balanced portfolio | VBAL

User avatar
vineviz
Posts: 4659
Joined: Tue May 15, 2018 1:55 pm

Re: Annuities are a bet on future inflation rates and should not be purchased by retirees

Post by vineviz » Fri May 31, 2019 8:22 am

BigJohn wrote:
Fri May 31, 2019 7:45 am
longinvest wrote:
Fri May 31, 2019 7:11 am
I'm of the opinion that, in absence of a competitively-priced CPI-indexed annuity, a 2%-indexed annuity is a good enough alternative to insure a base of stable lifelong income (in addition to Social Security) when the Federal Reserve communicates its intention to keep inflation on a 2% target to allow the public "to make accurate longer-term economic and financial decisions".
I agree with this perspective. While inflation is a significant risk in retirement, it's not the only risk you have to worry about and sometimes managing for one risk doesn't help (or can even hurt) the others.
So true.

Additionally, inflation risk is in reality a symmetrical risk (i.e. it is not the risk that inflation is unexpectedly high but is actually the risk that the realized rate of inflation differs from the expected rate of inflation).

The risk of deflation, for example, is often overlooked by individual investors even though it is probably a more likely scenario than runaway inflation. There is a fascinating working paper at NBER by Fleckenstein, Longstaff, and Lustig on deflation risk.
We actually find direct evidence from inflation options that market participants are pricing in large deflation in disaster states, because the risk-neutral probability of a deflation is much larger than the actual probability. This actually makes nominal bonds less risky because they provide a hedge against large consumption disasters.
"Far more money has been lost by investors preparing for corrections than has been lost in corrections themselves." ~~ Peter Lynch

JackoC
Posts: 872
Joined: Sun Aug 12, 2018 11:14 am

Re: Annuities are a bet on future inflation rates and should not be purchased by retirees

Post by JackoC » Fri May 31, 2019 8:35 am

vineviz wrote:
Thu May 30, 2019 4:37 pm
JackoC wrote:
Thu May 30, 2019 3:46 pm
That more clearly states you idea, which I think is clearly wrong in general.
I didn't state it clearly enough, apparently.

Most retirees (or retirees on average, take your pick) have annual spending that grows in nominal terms but declines in real terms. In other words, spending grows at less than the rate of inflation.

Furthermore, for the purposes of this discussion let's assume that their annual spending will be funded by some combination of the following: Social Security payments, portfolio withdrawals, and/or cash flow from annuities & pensions.

Social Security payments will grow with inflation and (using very common assumptions) portfolio withdrawals are expected to grow with inflation. That leaves some portion of spending that must be covered by cash flow from annuities and/or pensions.

It should be immediately clear that the trajectory of this portion of spending is entirely dependent on the personal circumstances of the retiree. It could be growing in real terms, but that's unlikely for most people: it's almost always going to either be declining in real terms or declining in nominal terms.

Only by assessing the probability and impact of each possible outcome can the investor decided whether a flat nominal annuity, a stepped annuity, a true CPI-linked annuity, or no annuitization at all is more likely to help them achieve their goals. In other words, it's an empirical - not theoretical - question.
It's beside the point to bring in Social Security. It's already been mentioned a bunch of times that Social Security is a real annuity and hedges against inflation for whatever portion of the retiree's needs it meets. We're always talking about which needs are *not* met by SS, or actually CPI adjusted pensions. If those sources meet all or almost all needs, it's likely the retiree has no need to consider annuities, also already mentioned.

Likewise it's circular to refer to "portfolio withdrawals expected to grow with inflation" without specifying what this portfolio is. Part of this 'portfolio' is what the retiree would annuitize. To the extent that's a nominal (flat or step up) annuity, it will *not* increase in response to an increase in inflation, which is the whole point. For some portion of the portfolio a fixed SPIA might still be OK. Again, broadly speaking that's all that's available, fixed SPIA or no SPIA (with perhaps one company offering real annuities, plus a rare product is less likely to be highly competitively priced). But the fact the fixed SPIA doesn't adjust for unexpected inflation is significant in determining how much of a solution it can be.

And again your point about retiree spending tending to decline in real terms is clear, it's just wrong as a reason to say the retiree is not exposed to inflation. Because the real decline, if so, of 'natural' spending has to do with aging (being less active), not inflation. Again, assume the real spending decrease is 2%pa*. Assume inflation is 2% pa. OK in that case a flat fixed annuity is most convenient *assuming there's no risk of much higher inflation*. But if inflation goes up to 4% pa, the retiree is then *forced* to reduce real spending at 4% pa. The fact that real spending might naturally decline at 2% pa due to aging is absolutely zero hedge against inflation *risk*.

*as in discussion above, we don't necessarily know what retiree real spending decline is in general, or how much might be due to involuntary acceptance of a lower living standard due to poor financial condition, and even so no general average applies to everyone. But just assuming 2% pa real spending decline for illustration.

User avatar
vineviz
Posts: 4659
Joined: Tue May 15, 2018 1:55 pm

Re: Annuities are a bet on future inflation rates and should not be purchased by retirees

Post by vineviz » Fri May 31, 2019 8:49 am

JackoC wrote:
Fri May 31, 2019 8:35 am
It's beside the point to bring in Social Security. It's already been mentioned a bunch of times that Social Security is a real annuity and hedges against inflation for whatever portion of the retiree's needs it meets. We're always talking about which needs are *not* met by SS, or actually CPI adjusted pensions.
This is contradictory. For one thing, you can't discuss the trajectory of the portion of spending *not* met by other sources of income without considering those other other sources of income. And even more importantly, you can NOT generalize about the trajectory of the portion of spending *not* met by other sources of income: that trajectory will be unique to each investor.
JackoC wrote:
Fri May 31, 2019 8:35 am
For some portion of the portfolio a fixed SPIA might still be OK.
Thats not really what I was discussing. I distinguish between income streams (i.e. pensions, Social Security, annuities) and an investment portfolio (i.e. marketable assets like stocks, bonds, mutual funds). An annuity isn't a marketable asset so (in my world) isn't part of the portfolio to which I referred.
JackoC wrote:
Fri May 31, 2019 8:35 am
our idea that a natural real spending decline insulates retirees from inflation risk is just completely wrong, sorry to be blunt.
Being blunt isn't a problem. Being blunt AND mistaken is.

Let's get back to basics. The only really relevant question is this: which type of annuity (nominal or real) provides an income stream that best matches the spending it is designed to cover?

A lot of people seem to be assuming that a real annuity will always be the best match, and I'm merely trying to interject a dose of reality to the discussion.
"Far more money has been lost by investors preparing for corrections than has been lost in corrections themselves." ~~ Peter Lynch

randomguy
Posts: 7974
Joined: Wed Sep 17, 2014 9:00 am

Re: Annuities are a bet on future inflation rates and should not be purchased by retirees

Post by randomguy » Fri May 31, 2019 9:09 am

longinvest wrote:
Fri May 31, 2019 7:11 am
AlohaJoe wrote:
Tue May 28, 2019 9:25 pm
$100,000 for a male/female couple that are 65/65 and live in Colorado pays $470/month for life (nominal).

With 2% adjustments it pays $350.

With CPI adjustments it pays $305.
Given the Federal Reserve's 2% inflation target, is a (($350 / $305) - 1) = 15% premium reasonable for CPI indexing over 2% indexing? In all likelihood, inflation will be close to 2% during the retiree's remaining life. The additional 15% cost of a CPI indexed SPIA will only benefit the retiree if inflation significantly increases beyond the Federal Reserve's inflation target.

I'm of the opinion that, in absence of a competitively-priced CPI-indexed annuity, a 2%-indexed annuity is a good enough alternative to insure a base of stable lifelong income (in addition to Social Security) when the Federal Reserve communicates its intention to keep inflation on a 2% target to allow the public "to make accurate longer-term economic and financial decisions".
In all likely hood a 50/50 portfolio will pay out 350/month stable income for life with CPI adjustments. Is that also good enough? You are getting into pretty personal choices. I am sure 90% of the time 2% will be plenty. How much do you worry about living the 10% when your income drops 50% over the first 10 years?

longinvest
Posts: 3754
Joined: Sat Aug 11, 2012 8:44 am

Re: Annuities are a bet on future inflation rates and should not be purchased by retirees

Post by longinvest » Fri May 31, 2019 9:13 am

randomguy wrote:
Fri May 31, 2019 9:09 am
longinvest wrote:
Fri May 31, 2019 7:11 am
AlohaJoe wrote:
Tue May 28, 2019 9:25 pm
$100,000 for a male/female couple that are 65/65 and live in Colorado pays $470/month for life (nominal).

With 2% adjustments it pays $350.

With CPI adjustments it pays $305.
Given the Federal Reserve's 2% inflation target, is a (($350 / $305) - 1) = 15% premium reasonable for CPI indexing over 2% indexing? In all likelihood, inflation will be close to 2% during the retiree's remaining life. The additional 15% cost of a CPI indexed SPIA will only benefit the retiree if inflation significantly increases beyond the Federal Reserve's inflation target.

I'm of the opinion that, in absence of a competitively-priced CPI-indexed annuity, a 2%-indexed annuity is a good enough alternative to insure a base of stable lifelong income (in addition to Social Security) when the Federal Reserve communicates its intention to keep inflation on a 2% target to allow the public "to make accurate longer-term economic and financial decisions".
In all likely hood a 50/50 portfolio will pay out 350/month stable income for life with CPI adjustments. Is that also good enough? You are getting into pretty personal choices. I am sure 90% of the time 2% will be plenty. How much do you worry about living the 10% when your income drops 50% over the first 10 years?
Has the Federal Reserve made any communication about a target return for a 50/50 stocks/bonds portfolio, and does spending a constant inflation-indexed withdrawal (SWR) from this portfolio provide guaranteed lifelong income? If not, I think that you're comparing apples and oranges.

I've discussed, earlier in this thread, Canada's 28 year history of inflation targeting. Unlike future returns, which are determined by markets, inflation can be controlled when a central bank is willing to do what this requires. Since 2012, the US Federal Reserve has adopted a 2% inflation target.

ADDED: I've just realized that my earlier post was a reply to you. Also, despite our numerous previous discussions in other threads, you come back with a suggestion to use SWR in retirement. Obviously, you want to use SWR in your retirement. Good luck!
Last edited by longinvest on Fri May 31, 2019 10:32 am, edited 8 times in total.
Bogleheads investment philosophy | single-ETF balanced portfolio | VBAL

randomguy
Posts: 7974
Joined: Wed Sep 17, 2014 9:00 am

Re: Annuities are a bet on future inflation rates and should not be purchased by retirees

Post by randomguy » Fri May 31, 2019 9:17 am

vineviz wrote:
Fri May 31, 2019 8:49 am

Being blunt isn't a problem. Being blunt AND mistaken is.

Let's get back to basics. The only really relevant question is this: which type of annuity (nominal or real) provides an income stream that best matches the spending it is designed to cover?

A lot of people seem to be assuming that a real annuity will always be the best match, and I'm merely trying to interject a dose of reality to the discussion.
It comes down to what inflation fears you worry about. If 1971-1981 happens, a 2% COLA isn't going to match your income requirements. Nobody wants to cut spending by 2/3rds over the first 20 years of retirement (call it 60-80). Get more normal inflation (i.e. the normal 2-3%) and then maybe getting 50% of what you started with when your 90 might be an acceptable tradeoff.

randomguy
Posts: 7974
Joined: Wed Sep 17, 2014 9:00 am

Re: Annuities are a bet on future inflation rates and should not be purchased by retirees

Post by randomguy » Fri May 31, 2019 9:50 am

longinvest wrote:
Fri May 31, 2019 9:13 am


Has the Federal Reserve made any communication about a target return for a 50/50 stocks/bonds portfolio, and does spending a constant inflation-indexed withdrawal (SWR) from this portfolio provide guaranteed lifelong income? If not, I think that you're comparing apples and oranges.

I've discussed, earlier in this thread, Canada's 28 year history of inflation targeting. Unlike future returns, which are determined by markets, inflation can be controlled when a central bank is willing to do what this requires. Since 2012, the US federal reserve has adopted a 2% inflation target.

ADDED: I've just realized that my earlier post was a reply to you. Also, despite our numerous previous discussions in other threads, you come back with a suggestion to use SWR in retirement. Obviously, you want to use SWR in your retirement. Good luck!
The point is the word likely hood. If you are willing to accept 90% success, a lot of things work. You have to pay a lot to get to 95%. And even more for 99%. I don't consider the odds of inflation targeting working to be 99%. You might.

It all comes down to your faith in the ability of the government. You look at the last 28 years and say, things are working. I look at the last 28 years and compare the countries that have stated inflation targets and the ones that don't and go, inflation just hasn't been around for the last 28 year in the developed world. When the next crisis hits and tough choices need to be made, I am not sure if the government will favor keeping inflation low versus say increasing the gdp growth or keeping unemployment low. Or someone deciding that we need to print money to pay the debt off cause they don't want to raise taxes/cut services and a bit of inflation lets them kick the can down the road 8 years.

I sure don't think the odds of high inflation are high. But I am not willing to put them at 0% either. So you are left at deciding how much you want to pay to protect against a rare event but one that could drastically affect you.

JackoC
Posts: 872
Joined: Sun Aug 12, 2018 11:14 am

Re: Annuities are a bet on future inflation rates and should not be purchased by retirees

Post by JackoC » Fri May 31, 2019 9:51 am

vineviz wrote:
Fri May 31, 2019 8:49 am

Being blunt isn't a problem. Being blunt AND mistaken is.

Let's get back to basics. The only really relevant question is this: which type of annuity (nominal or real) provides an income stream that best matches the spending it is designed to cover?

A lot of people seem to be assuming that a real annuity will always be the best match, and I'm merely trying to interject a dose of reality to the discussion.
Again a nominal annuity or no annuity is the typical real choice, real annuities barely exist in the USD market.

The more general point where you'e very mistaken, though you do seem to have backed off on this point, is where you said retirees have 'little net exposure to inflation'. That's seriously wrong if any retiree were to accept it as an idea in general, though I guess it's obviously wrong enough few people would.

Starting at point zero, people who no longer work are much *more* exposed to inflation risk than people who are working. Wages tend as a first order approximation to rise with inflation, though inflation also acts as a 'lubricant' to let the job market more easily reduce the real wages of jobs which aren't worth as much as they used to be, so depends on the job. But starting with no job and a pile of cash, you are much more exposed to inflation. The fact that your real spending might naturally decline is wholly irrelevant to that risk. As everyone already knows, most US retirees receive Social Security which is a real annuity. But basically all 'retirement finance' discussions take this as given and focus on financial needs *beyond* SS. Some people have pensions but not a large % anymore.

That heightened inflation risk in retirement can be countered directly with TIPS, though those don't do anything to address longevity risk. Stocks are to some degree hedges against moderate increases in inflation but as the performance of the US stock market in the late 1960's-early 1980's periods shows, not a great hedge against serious unexpected inflation. Real estate might be another. But anyway real annuities would in fact generally be much better suited to most people's annuity needs than fixed (*if they have any annuity need, which some people don't*). They are the only product that can hedge inflation and longevity risk at the same time. And both are serious risks for retirees. However, though an equally competitively priced real SPIA would almost always be superior to a fixed one...again real ones barely exist. So the practical point is mainly that retirees are in fact vulnerable to unexpected inflation, natural decline in real spending does not hedge against inflation risk, whatsoever, so some solution has to be found in a combination of assets which separately hedge inflation risk (TIPS, stocks to a limited degree) *or* hedge longevity risk (fixed SPIA).

Phil DeMuth
Posts: 141
Joined: Thu Mar 06, 2008 7:02 pm

Re: Annuities are a bet on future inflation rates and should not be purchased by retirees

Post by Phil DeMuth » Fri May 31, 2019 10:08 am

Just to stir the pot:

Wouldn't one really want an annuity indexed to CPI-E, which is the rate of inflation experienced by the elderly?

Also, what about standard of living? You could have a real annuity that completely fails to keep up with the rising standard of living. In thirty years people could be driving flying cars while you are sitting there with your fax machine and 8-track tape.

Robert Merton has designed government bonds for South Korea that will let their citizens insure against both inflation and standard of living decline in their old age. They also help answer the insurance company credit risk. We need those in the USA.

User avatar
vineviz
Posts: 4659
Joined: Tue May 15, 2018 1:55 pm

Re: Annuities are a bet on future inflation rates and should not be purchased by retirees

Post by vineviz » Fri May 31, 2019 10:14 am

JackoC wrote:
Fri May 31, 2019 9:51 am
The more general point where you'e very mistaken, though you do seem to have backed off on this point, is where you said retirees have 'little net exposure to inflation'.
I've run out of both patience and new ways of explaining this.

Honestly, it's a pretty basic concept so I'm going to assume that anyone who wants to understand it probably already does.
"Far more money has been lost by investors preparing for corrections than has been lost in corrections themselves." ~~ Peter Lynch

User avatar
nisiprius
Advisory Board
Posts: 38740
Joined: Thu Jul 26, 2007 9:33 am
Location: The terrestrial, globular, planetary hunk of matter, flattened at the poles, is my abode.--O. Henry

Re: Annuities are a bet on future inflation rates and should not be purchased by retirees

Post by nisiprius » Fri May 31, 2019 10:39 am

And yet, these results have been pounced upon by retirement planning writers, who have already opined that retirement has three phases, the "go-go years," "slow-go years," and "no-go years" and used to justify higher safe withdrawal rates initially, on the assumption that spending will ramp down over time. Some retirement planner even talked about "hedonic

Even if it is a) true on average, and b) driven by health and energy, not by impoverishment, it is a case where the averages surely are not a good representation of the spread. The CEO of a firm that I worked for actually set a Guinness record for stairclimbing (in skyscrapers) in his late seventies, and climbed Mount Whitney with his son in his eighties. Another couple I know really enjoyed going on those alumni travel tour deals, well into their late eighties--you know, the ones that last three weeks and involve comfortable travel in fairly remote and exotic locations.They're not cheap. The point is, they had the energy to do it. And they would probably would not have had the energy to do it cheap.

And even if one becomes frail, I am not yet in the world of chairvans and stairlifts and ramps, and while I think they are covered by Medicare, I'll just bet they aren't fully covered by Medicare, and that out-of-pocket expenses for mobility and comfort probably often increase.
Annual income twenty pounds, annual expenditure nineteen nineteen and six, result happiness; Annual income twenty pounds, annual expenditure twenty pounds ought and six, result misery.

User avatar
ResearchMed
Posts: 9052
Joined: Fri Dec 26, 2008 11:25 pm

Re: Annuities are a bet on future inflation rates and should not be purchased by retirees

Post by ResearchMed » Fri May 31, 2019 10:49 am

nisiprius wrote:
Fri May 31, 2019 10:39 am
And yet, these results have been pounced upon by retirement planning writers, who have already opined that retirement has three phases, the "go-go years," "slow-go years," and "no-go years" and used to justify higher safe withdrawal rates initially, on the assumption that spending will ramp down over time. Some retirement planner even talked about "hedonic

Even if it is a) true on average, and b) driven by health and energy, not by impoverishment, it is a case where the averages surely are not a good representation of the spread. The CEO of a firm that I worked for actually set a Guinness record for stairclimbing (in skyscrapers) in his late seventies, and climbed Mount Whitney with his son in his eighties. Another couple I know really enjoyed going on those alumni travel tour deals, well into their late eighties--you know, the ones that last three weeks and involve comfortable travel in fairly remote and exotic locations.They're not cheap. The point is, they had the energy to do it. And they would probably would not have had the energy to do it cheap.

And even if one becomes frail, I am not yet in the world of chairvans and stairlifts and ramps, and while I think they are covered by Medicare, I'll just bet they aren't fully covered by Medicare, and that out-of-pocket expenses for mobility and comfort probably often increase.
I've seen a few describe the cost graph as more of a "U" shaped. The "go go" years are just that, as are the "go slow" years. But they recognize that the "no go" years can be quite costly.

But, as with the examples you give, these averages are just that: AVERAGES. They may not describe most people; indeed, "the average" may not even describe a single actual case. And just because "the average model" is "close enough" for "most people", that's small comfort for those who are well outside those numbers...

OTOH, most models and discussions do not seem to take into account that at least for many of these years, we are intelligent and observant. That is, one can cut back - and most of us here are more than comfortable enough to have some extra to cut if needed - if the returns aren't quite what we had hoped. Being able to take one vacation every other year, or perhaps a shorter vacation on the planned annual schedule, or to buy a new car, and a less expensive new car, less frequently, etc...
Most of us probably aren't going to spend our way into oblivion without making any adjustments until falling off the cliff.

RM
This signature is a placebo. You are in the control group.

User avatar
nisiprius
Advisory Board
Posts: 38740
Joined: Thu Jul 26, 2007 9:33 am
Location: The terrestrial, globular, planetary hunk of matter, flattened at the poles, is my abode.--O. Henry

Re: Annuities are a bet on future inflation rates and should not be purchased by retirees

Post by nisiprius » Fri May 31, 2019 10:55 am

Phil DeMuth wrote:
Fri May 31, 2019 10:08 am
...Wouldn't one really want an annuity indexed to CPI-E, which is the rate of inflation experienced by the elderly?...
Sure, but that's getting close to a technicality. The amount of error between my personal future inflation and zero (flat payments) is sure to be huge. The error in an annuity that increases by exactly 3% compounded annually is likely to be much less. The error in one indexed to CPI-U is likely to be less yet. And CPI-E, less yet. But the difference between all those flavors of CPI--CPI-U, used by TIPS and I bonds; CPI-W, used by Social Security; CPI-E, advocated by AARP for Social Security; and "chained" versions of the CPI's, advocated by politicians hoping to save money on CPI-linked benefits--are small.
...Also, what about standard of living? You could have a real annuity that completely fails to keep up with the rising standard of living. In thirty years people could be driving flying cars while you are sitting there with your fax machine and 8-track tape....
Yes, this. To make it a little more concrete, thirty years ago you "needed" electricity and telephone. Today, it is not an exaggeration to say you "need" high-speed Internet and WiFi; certainly a cell phone; and probably it "needs" to be a smartphone. This is not a huge additional expense, but it is there. Of course other "needs" drop out with time. You no longer need a physically delivered paper newspaper, probably.
Annual income twenty pounds, annual expenditure nineteen nineteen and six, result happiness; Annual income twenty pounds, annual expenditure twenty pounds ought and six, result misery.

User avatar
Topic Author
CULater
Posts: 2109
Joined: Sun Nov 13, 2016 10:59 am

Re: Annuities are a bet on future inflation rates and should not be purchased by retirees

Post by CULater » Fri May 31, 2019 11:17 am

Well now, if my mother's experience is any guide you can expect your cost of living to go up a whole bunch as you approach the end of your mortal coil. She moved from living in her own home, to living in a 55+ apartment, to a 55+ senior facility with meals, to assisted living, to memory care, and now to a nursing home. Each move came with a doubling of her monthly living expenses. The problem with annuities as longevity insurance is that you might actually end up needing that insurance and it will be paying out pocket change in real money just when you need lots and lots of real money. Best hope when you buy that longevity insurance that you'll never need to collect it.
On the internet, nobody knows you're a dog.

User avatar
nisiprius
Advisory Board
Posts: 38740
Joined: Thu Jul 26, 2007 9:33 am
Location: The terrestrial, globular, planetary hunk of matter, flattened at the poles, is my abode.--O. Henry

Re: Annuities are a bet on future inflation rates and should not be purchased by retirees

Post by nisiprius » Fri May 31, 2019 2:30 pm

CULater wrote:
Fri May 31, 2019 11:17 am
Well now, if my mother's experience is any guide you can expect your cost of living to go up a whole bunch as you approach the end of your mortal coil. She moved from living in her own home, to living in a 55+ apartment, to a 55+ senior facility with meals, to assisted living, to memory care, and now to a nursing home. Each move came with a doubling of her monthly living expenses. The problem with annuities as longevity insurance is that you might actually end up needing that insurance and it will be paying out pocket change in real money just when you need lots and lots of real money. Best hope when you buy that longevity insurance that you'll never need to collect it.
And what that tells us is that there is a huge amount of variability in the retirement experience, longevity being one, and financial needs with age being another, and the big take-home is that you can't eliminate this uncertainty and the only way to deal with it is to accept the prudence of keeping a big cushion which you probably won't use. If your expenses in retirement were stable, then an annuity allows you to efficiently use all your money for yourself, "die broke," and live so that "the check to the undertaker bounces." But since it isn't, even an annuity does not eliminate the need for a big cushion and the likelihood of "leaving money on the table."
Annual income twenty pounds, annual expenditure nineteen nineteen and six, result happiness; Annual income twenty pounds, annual expenditure twenty pounds ought and six, result misery.

User avatar
Topic Author
CULater
Posts: 2109
Joined: Sun Nov 13, 2016 10:59 am

Re: Annuities are a bet on future inflation rates and should not be purchased by retirees

Post by CULater » Fri May 31, 2019 2:49 pm

nisiprius wrote:
Fri May 31, 2019 2:30 pm
CULater wrote:
Fri May 31, 2019 11:17 am
Well now, if my mother's experience is any guide you can expect your cost of living to go up a whole bunch as you approach the end of your mortal coil. She moved from living in her own home, to living in a 55+ apartment, to a 55+ senior facility with meals, to assisted living, to memory care, and now to a nursing home. Each move came with a doubling of her monthly living expenses. The problem with annuities as longevity insurance is that you might actually end up needing that insurance and it will be paying out pocket change in real money just when you need lots and lots of real money. Best hope when you buy that longevity insurance that you'll never need to collect it.
And what that tells us is that there is a huge amount of variability in the retirement experience, longevity being one, and financial needs with age being another, and the big take-home is that you can't eliminate this uncertainty and the only way to deal with it is to accept the prudence of keeping a big cushion which you probably won't use. If your expenses in retirement were stable, then an annuity allows you to efficiently use all your money for yourself, "die broke," and live so that "the check to the undertaker bounces." But since it isn't, even an annuity does not eliminate the need for a big cushion and the likelihood of "leaving money on the table."
I guess life insurance works the same way. When you take out a $1M policy and die 30 years later, that $1M doesn't "insure" as much as you were thinking. I always get a smile out of Alex Trebec when he reassures folks that their insurance premium will "never increase". But he fails to mention that your benefit will "never increase" as well, and will be worth a lot less down the road. Money Illusion.
On the internet, nobody knows you're a dog.

User avatar
#Cruncher
Posts: 2810
Joined: Fri May 14, 2010 2:33 am
Location: New York City
Contact:

Re: Annuities are a bet on future inflation rates and should not be purchased by retirees

Post by #Cruncher » Fri May 31, 2019 4:50 pm

Ben Mathew wrote:
Wed May 29, 2019 7:21 am
… Given that the expected yield of TIPS isn't significantly less than that of nominal bonds ..., it's not clear to me why most people choose to hold nominal bonds instead of real. TIPS offer free, or almost free, insurance against unexpected inflation. Why take a risk we don't have to? Same with real vs nominal annuities. If there isn't a significant cost, real is better than nominal. (underline added)
Ah, but that's the rub! There is a "significant cost" of a CPI-linked annuity versus a nominal one. Using the following figures from this post by Aloha Joe [1], here are the returns [2] of a nominal and real annuity for various life expectancies.

Code: Select all

Cost         100,000
             Nominal    Real
Payment          470     305
Years              Return       Diff
-----          --------------   ----
  15           (2.2%)  (7.3%)   5.1% 
  20            1.2%   (3.0%)   4.2% 
  25            2.9%   (0.7%)   3.6%
  30            3.9%    0.6%    3.2% 
  35            4.4%    1.5%    3.0% [2]
  40            4.8%    2.0%    2.8% 
  45            5.1%    2.4%    2.6% 
  50            5.2%    2.7%    2.5%
Even if the annuities ran 35 years (well beyond the expected time that either of the couple would still be alive [3]), the nominal return is 3.0% points higher than the CPI-linked real annuity. By comparison, the nominal yield of a 30-year nominal Treasury is only about 1.84% points higher than that of a 30-year TIPS. (2.58% vs 0.74% according to the 5/31/2019 Treasury yield curves, nominal and real.)
  1. AlohaJoe wrote:
    Tue May 28, 2019 9:25 pm
    $100,000 for a male/female couple that are 65/65 and live in Colorado pays $470/month for life (nominal). … With CPI adjustments it pays $305.
  2. Example return calculation if annuities ran for 35 years, using Excel RATE function:
    4.4% = RATE(35 * 12, 470, -100000, 0, 0) * 12
    1.5% = RATE(35 * 12, 305, -100000, 0, 0) * 12
  3. According to the 2016 SSA Period Life Table there is only a 4.5% chance that at least one of a male / female couple both age 65 will survive to age 100.

    Code: Select all

                         Men    Women
                       ------  ------
    Alive at age  65   79,893  87,574   
    Alive at age 100      994   2,892
    Percent alive       1.24%   3.30%
    Odds both dead    = 95.5% = (1 - 0.0124) * (1 - 0.0330)
    Odds either alive =  4.5%

columbia
Posts: 1738
Joined: Tue Aug 27, 2013 5:30 am

Re: Annuities are a bet on future inflation rates and should not be purchased by retirees

Post by columbia » Fri May 31, 2019 4:59 pm

It seems like it would be cheaper to achieve increasing return above the baseline of a SPIA (than buying inflation-linked): buy less of the nominal SPIA and put the rest of that bucket in corporate bonds.

I think there’s a reason why there’s a low demand for inflation-related SPIA: it’s hard to make the numbers work.

randomguy
Posts: 7974
Joined: Wed Sep 17, 2014 9:00 am

Re: Annuities are a bet on future inflation rates and should not be purchased by retirees

Post by randomguy » Fri May 31, 2019 5:26 pm

columbia wrote:
Fri May 31, 2019 4:59 pm
It seems like it would be cheaper to achieve increasing return above the baseline of a SPIA (than buying inflation-linked): buy less of the nominal SPIA and put the rest of that bucket in corporate bonds.

I think there’s a reason why there’s a low demand for inflation-related SPIA: it’s hard to make the numbers work.
It is both cheaper and riskier. If in 1971 you would have bought a nominal SPIA and invested the rest in Corporate bonds, would you have maintained your purchasing power to say 1991? Maybe something along the lines of more mortality credits and higher payouts would get you close but I sort of doubt it.

Inflation adjusted annuities (and the 2% ones for that matter) are a tough sell because all the gains are back loaded. A 65 year old buying an annuity probably has a life expectancy of around 88-90 (i.e. obese, smokers who have had 2 heart attacks aren't buying them:)). You shouldn't expect a break even to past that point. And with the inflation adjusted ones, you should only expect them to be a good deal when inflation shows up. .

In the end, we all have personal risk tolerances and fear different things. And we all try and come up with ways of dealing with those fears. Some we rationalize away(this time is different:)). Some we buy insurance. And some we decide are too expensive to protect against and have to accept.

Longtermgrowth
Posts: 619
Joined: Thu Nov 26, 2015 1:59 pm

Re: Annuities are a bet on future inflation rates and should not be purchased by retirees

Post by Longtermgrowth » Fri May 31, 2019 5:29 pm

#Cruncher wrote:
Fri May 31, 2019 4:50 pm
here are the returns [2] of a nominal and real annuity for various life expectancies.


That chart is interesting, but what if inflation spikes in the first few decades of payouts?
Todays low interest rates and current low inflation may look quite different even just a decade down the road :?

afan
Posts: 4269
Joined: Sun Jul 25, 2010 4:01 pm

Re: Annuities are a bet on future inflation rates and should not be purchased by retirees

Post by afan » Fri May 31, 2019 9:10 pm

Regarding present values:

Although people often calculate them assuming the discount rate is constant across years, this is not necessary. One can have different discount rates for different time periods. In that case, one may well have a higher discount rate for more remote times. This would be particularly applicable to an income stream that is only paid while an individual is alive. The farther in the future, the lower the probability of survival and hence the higher the discount rate to apply.

In the "B" example, at least as it seemed to be presented, a 75 year old person would have to wait 50 years for a payout. That means the probability of getting anything is zero. If there will be no future cash flows then the present value is zero.

If option "A" also had a zero payout, and hence a present value of zero, then the investor would be indifferent between one investment with a zero value and another also with zero value.

In this example either

1. both have zero present value, which repeats the tautology vivenix noted,

or

2. "A" has future positive cash flows while "B" does not. In that case the present values are not equal.

Not only can discount rates vary for different time periods, they can vary across individuals. A 20 year old might have a very different set of time dependent discount rates than a 75 year old. That 50 year wait for a payout does not equate to a probability of zero for the younger person. So an expected payment 50 years later would be worth something more than zero.
We don't know how to beat the market on a risk-adjusted basis, and we don't know anyone that does know either | --Swedroe | We assume that markets are efficient, that prices are right | --Fama

longinvest
Posts: 3754
Joined: Sat Aug 11, 2012 8:44 am

Re: Annuities are a bet on future inflation rates and should not be purchased by retirees

Post by longinvest » Fri May 31, 2019 10:54 pm

#Cruncher wrote:
Fri May 31, 2019 4:50 pm
Ah, but that's the rub! There is a "significant cost" of a CPI-linked annuity versus a nominal one. [...]
Even if the annuities ran 35 years (well beyond the expected time that either of the couple would still be alive [3]), the nominal return is 3.0% points higher than the CPI-linked real annuity. By comparison, the nominal yield of a 30-year nominal Treasury is only about 1.84% points higher than that of a 30-year TIPS.[...]
Should a similar analysis be done with fire insurance premiums and payouts? What if the conclusion was that one shouldn't insure the house against fire because one is likely to lose the premiums in most cases? :wink:

A SPIA* is an insurance product. I would only buy as little inflation-indexed SPIA as strictly necessary within my retirement plan to complement Social Security and provide a lifelong a stable income base to combine with variable portfolio withdrawals (VPW) during retirement. In situations where an inflation-indexed SPIA doesn't exist or isn't competitively priced, I would consider buying a 2%-indexed SPIA, instead, because the Federal Reserve has adopted a 2% inflation target. I'll also consider buying an inflation-indexed (or 2%-indexed) SPIA with part (not all!) of my residual portfolio, at age 80, if necessary to dampen the low-probability but severe financial risk of long life (life beyond age 100).

* Single-Premium Immediate Annuity.
Bogleheads investment philosophy | single-ETF balanced portfolio | VBAL

User avatar
Topic Author
CULater
Posts: 2109
Joined: Sun Nov 13, 2016 10:59 am

Re: Annuities are a bet on future inflation rates and should not be purchased by retirees

Post by CULater » Sat Jun 01, 2019 9:08 am

Even if inflation only runs the present 2% annually, in 30 years the purchasing power of a nominal annuity would be 45% less than now and you can also reckon that the living expenses you'll be incurring by then will be a gob higher then in real dollars too because of health care needs. That's why I say that you better hope you won't actually be around long enough to benefit from "mortality credits", which is the whole idea of buying an annuity in the first place isn't it? If I handed you a dollar and received 65 cents or less back later, when I really needed that dollar (or more) should you be arrested?
On the internet, nobody knows you're a dog.

User avatar
Horton
Posts: 254
Joined: Mon Jan 21, 2008 3:53 pm

Re: Annuities are a bet on future inflation rates and should not be purchased by retirees

Post by Horton » Sat Jun 01, 2019 9:12 am

A real annuity hedges against not only inflation risk but also sequence of inflation risk - that is, the risk that inflation over the entire retirement period may be tame (2-3%) but could be characterized by extremely high periods of inflation for the first 5-10 years followed by subsequent long periods of very low inflation (or even deflation). The fact that the Principal annuity does not decrease in the event of deflation, would be a favorable outcome in this scenario.

Social Security is the best real annuity one can "buy" (i.e., by delaying Social Security). By doing this, an individual may be able to cover most/all their real non-discretionary expenses and, therefore, may have no use for a real annuity.

With nominal and fixed COLA annuities, issuers are likely pricing the products using corporate bond yields (as David Blanchett did in this article) with an additional cost load added to cover profit and reserves. This is possible because the value of the cash flows are known in advance and longevity is pooled among the group. In effect, the issuer needs to immunize against interest rate risk but not inflation risk, which can be done with either Treasuries or corporates (although using corporates will likely increase reserve requirements). With real (inflation adjusted) annuities, the value of the cash flows is not known - i.e., it is dependent upon future inflation. Therefore, the issuer's only option is to price the product using TIPS. The issuer in this case needs to immunize against both inflation and interest rate risk. Indeed, the cost of a real annuity is very similar to the cost of a TIPS ladder (or duration matched TIPS fund portfolio).
"You must know that there is nothing higher and stronger and more wholesome and good for life in the future than some good memory, especially a memory of childhood, of home." - Dostoyevsky

User avatar
#Cruncher
Posts: 2810
Joined: Fri May 14, 2010 2:33 am
Location: New York City
Contact:

Re: Annuities are a bet on future inflation rates and should not be purchased by retirees

Post by #Cruncher » Sat Jun 01, 2019 9:18 am

Longtermgrowth wrote:
Fri May 31, 2019 5:29 pm
#Cruncher wrote:
Fri May 31, 2019 4:50 pm
here are the returns ... of a nominal and real annuity for various life expectancies.
That chart is interesting, but what if inflation spikes in the first few decades of payouts?
In that case the CPI-indexed annuity may very well have a better outcome in spite of the fact that it costs significantly more than the nominal annuity. Again using the figures from this post by AlohaJoe [1], for a $100,000 purchase in 1964 the real annual return after 35 years [2] from the $470 / month nominal annuity would have been -0.8%. This would have been 2.2% points worse than the +1.4% real annual return over the same period of the $305 / month CPI-indexed annuity.

Code: Select all

    Col C    Col D  Col E ColF     Col G   Col H   Col I   Col J
     Ann                         - Nom Annuity -   RAnn
Row  Year  Apr CPI   Chg  Year   Real CF   RIRR    RIRR     Diff

Code: Select all

  7  1964   30.900           0  (100,000)
  8  1965   31.400   1.6%    1     5,550 
  9  1966   32.300   2.9%    2     5,396 
 10  1967   33.100   2.5%    3     5,265 
 11  1968   34.400   3.9%    4     5,066 
 12  1969   36.300   5.5%    5     4,801 
 13  1970   38.500   6.1%    6     4,527 
 14  1971   40.100   4.2%    7     4,346 
 15  1972   41.500   3.5%    8     4,199 
 16  1973   43.600   5.1%    9     3,997 
 17  1974   48.000  10.1%   10     3,631  (12.5%) (15.1%)   2.5% 
 18  1975   52.900  10.2%   11     3,294  (10.8%) (12.8%)   2.0% 
 19  1976   56.100   6.0%   12     3,107   (9.3%) (10.8%)   1.5% 
 20  1977   60.000   7.0%   13     2,905   (8.1%)  (9.2%)   1.1% 
 21  1978   63.900   6.5%   14     2,727   (7.1%)  (7.9%)   0.8% 
 22  1979   70.600  10.5%   15     2,468   (6.3%)  (6.7%)   0.4% 
 23  1980   81.000  14.7%   16     2,152   (5.6%)  (5.7%)   0.1% 
 24  1981   89.100  10.0%   17     1,956   (5.1%)  (4.8%)  (0.3%)
 25  1982   94.900   6.5%   18     1,836   (4.6%)  (4.1%)  (0.5%)
 26  1983   98.600   3.9%   19     1,768   (4.2%)  (3.4%)  (0.8%)
 27  1984  103.100   4.6%   20     1,690   (3.8%)  (2.8%)  (1.0%)
 28  1985  106.900   3.7%   21     1,630   (3.4%)  (2.3%)  (1.2%)
 29  1986  108.600   1.6%   22     1,605   (3.1%)  (1.8%)  (1.3%)
 30  1987  112.700   3.8%   23     1,546   (2.8%)  (1.4%)  (1.4%)
 31  1988  117.100   3.9%   24     1,488   (2.6%)  (1.0%)  (1.5%)
 32  1989  123.100   5.1%   25     1,416   (2.3%)  (0.7%)  (1.6%)
 33  1990  128.900   4.7%   26     1,352   (2.1%)  (0.4%)  (1.7%)
 34  1991  135.200   4.9%   27     1,289   (1.9%)  (0.1%)  (1.8%)
 35  1992  139.500   3.2%   28     1,249   (1.7%)   0.2%   (1.9%)
 36  1993  144.000   3.2%   29     1,210   (1.6%)   0.4%   (2.0%)
 37  1994  147.400   2.4%   30     1,182   (1.4%)   0.6%   (2.0%)
 38  1995  151.900   3.1%   31     1,147   (1.3%)   0.8%   (2.1%)
 39  1996  156.300   2.9%   32     1,115   (1.1%)   1.0%   (2.1%)
 40  1997  160.200   2.5%   33     1,088   (1.0%)   1.2%   (2.2%)
 41  1998  162.500   1.4%   34     1,072   (0.9%)   1.3%   (2.2%)
 42  1999  166.200   2.3%   35     1,049   (0.8%)   1.4%   (2.2%) [3]
 43  2000  171.300   3.1%   36     1,017   (0.7%)   1.6%   (2.3%)
 44  2001  176.900   3.3%   37       985   (0.6%)   1.7%   (2.3%)
 45  2002  179.800   1.6%   38       969   (0.5%)   1.8%   (2.3%)
 46  2003  183.800   2.2%   39       948   (0.4%)   1.9%   (2.4%)
 47  2004  188.000   2.3%   40       927   (0.4%)   2.0%   (2.4%)
 48  2005  194.600   3.5%   41       896   (0.3%)   2.1%   (2.4%)
 49  2006  201.500   3.5%   42       865   (0.2%)   2.2%   (2.4%)
 50  2007  206.686   2.6%   43       843   (0.2%)   2.3%   (2.4%)
 51  2008  214.823   3.9%   44       811   (0.1%)   2.3%   (2.4%)
 52  2009  213.240  (0.7%)  45       817   (0.1%)   2.4%   (2.5%)
 53  2010  218.009   2.2%   46       799   (0.0%)   2.5%   (2.5%)
 54  2011  224.906   3.2%   47       775    0.0%    2.5%   (2.5%)
 55  2012  230.085   2.3%   48       757    0.1%    2.6%   (2.5%)
 56  2013  232.531   1.1%   49       749    0.1%    2.6%   (2.5%)
 57  2014  237.072   2.0%   50       735    0.2%    2.7%   (2.5%)
longinvest wrote:
Fri May 31, 2019 10:54 pm
#Cruncher wrote:
Fri May 31, 2019 4:50 pm
Ah, but that's the rub! There is a "significant cost" of a CPI-linked annuity versus a nominal one. ...
Should a similar analysis be done with fire insurance premiums and payouts? What if the conclusion was that one shouldn't insure the house against fire because one is likely to lose the premiums in most cases? :wink:
I wasn't arguing against buying a CPI-indexed annuity. I was arguing that its cost relative to a nominal annuity is much greater than the "cost" of TIPS relative to nominal treasuries. Just as with TIPS vs nominal Treasuries, if the CPI rises more than the difference in returns (i.e., the breakeven inflation rate), then the CPI-indexed annuity will have the better outcome. One needs it to protect against that possibility.
longinvest in same post wrote:In situations where an inflation-indexed SPIA doesn't exist or isn't competitively priced, I would consider buying a 2%-indexed SPIA, instead ...
Depending on the specifics the 2% compounded annuity may not protect against severe inflation any better than the straight nominal annuity. Again using the figures from this post by AlohaJoe, after 35 years the annual return of the $350 / month +2% compounded annuity purchased in 1964 was only -0.9% [4], a touch worse than the -0.8% annual return of the $470 / month straight nominal annuity.

Code: Select all

    Col C    Col D  Col E ColF     Col G   Col H   Col I   Col J
     Ann                         - Nom Annuity -   RAnn
Row  Year  Apr CPI   Chg  Year   Real CF   RIRR    RIRR     Diff

Code: Select all

  7  1964   30.900           0  (100,000)
  8  1965   31.400   1.6%    1     4,216 
  9  1966   32.300   2.9%    2     4,180 
 10  1967   33.100   2.5%    3     4,161 
 11  1968   34.400   3.9%    4     4,084 
 12  1969   36.300   5.5%    5     3,947 
 13  1970   38.500   6.1%    6     3,796 
 14  1971   40.100   4.2%    7     3,718 
 15  1972   41.500   3.5%    8     3,664 
 16  1973   43.600   5.1%    9     3,557 
 17  1974   48.000  10.1%   10     3,296  (14.8%) (15.1%)   0.2% 
 18  1975   52.900  10.2%   11     3,050  (12.8%) (12.8%)  (0.1%)
 19  1976   56.100   6.0%   12     2,934  (11.2%) (10.8%)  (0.3%)
 20  1977   60.000   7.0%   13     2,798   (9.8%)  (9.2%)  (0.6%)
 21  1978   63.900   6.5%   14     2,680   (8.7%)  (7.9%)  (0.8%)
 22  1979   70.600  10.5%   15     2,474   (7.7%)  (6.7%)  (1.0%)
 23  1980   81.000  14.7%   16     2,200   (6.9%)  (5.7%)  (1.2%)
 24  1981   89.100  10.0%   17     2,040   (6.3%)  (4.8%)  (1.5%)
 25  1982   94.900   6.5%   18     1,953   (5.7%)  (4.1%)  (1.6%)
 26  1983   98.600   3.9%   19     1,917   (5.2%)  (3.4%)  (1.8%)
 27  1984  103.100   4.6%   20     1,870   (4.7%)  (2.8%)  (1.9%)
 28  1985  106.900   3.7%   21     1,840   (4.3%)  (2.3%)  (2.0%)
 29  1986  108.600   1.6%   22     1,847   (3.9%)  (1.8%)  (2.1%)
 30  1987  112.700   3.8%   23     1,816   (3.5%)  (1.4%)  (2.1%)
 31  1988  117.100   3.9%   24     1,783   (3.2%)  (1.0%)  (2.2%)
 32  1989  123.100   5.1%   25     1,730   (2.9%)  (0.7%)  (2.2%)
 33  1990  128.900   4.7%   26     1,685   (2.6%)  (0.4%)  (2.2%)
 34  1991  135.200   4.9%   27     1,638   (2.4%)  (0.1%)  (2.3%)
 35  1992  139.500   3.2%   28     1,620   (2.1%)   0.2%   (2.3%)
 36  1993  144.000   3.2%   29     1,600   (1.9%)   0.4%   (2.3%)
 37  1994  147.400   2.4%   30     1,595   (1.7%)   0.6%   (2.3%)
 38  1995  151.900   3.1%   31     1,579   (1.5%)   0.8%   (2.3%)
 39  1996  156.300   2.9%   32     1,565   (1.4%)   1.0%   (2.3%)
 40  1997  160.200   2.5%   33     1,557   (1.2%)   1.2%   (2.3%)
 41  1998  162.500   1.4%   34     1,566   (1.0%)   1.3%   (2.3%)
 42  1999  166.200   2.3%   35     1,562   (0.9%)   1.4%   (2.3%) [4]
 43  2000  171.300   3.1%   36     1,545   (0.8%)   1.6%   (2.3%)
 44  2001  176.900   3.3%   37     1,526   (0.6%)   1.7%   (2.3%)
 45  2002  179.800   1.6%   38     1,532   (0.5%)   1.8%   (2.3%)
 46  2003  183.800   2.2%   39     1,529   (0.4%)   1.9%   (2.3%)
 47  2004  188.000   2.3%   40     1,524   (0.3%)   2.0%   (2.3%)
 48  2005  194.600   3.5%   41     1,502   (0.2%)   2.1%   (2.3%)
 49  2006  201.500   3.5%   42     1,480   (0.1%)   2.2%   (2.3%)
 50  2007  206.686   2.6%   43     1,471   (0.0%)   2.3%   (2.3%)
 51  2008  214.823   3.9%   44     1,444    0.1%    2.3%   (2.3%)
 52  2009  213.240  (0.7%)  45     1,484    0.1%    2.4%   (2.3%)
 53  2010  218.009   2.2%   46     1,480    0.2%    2.5%   (2.3%)
 54  2011  224.906   3.2%   47     1,464    0.3%    2.5%   (2.2%)
 55  2012  230.085   2.3%   48     1,459    0.4%    2.6%   (2.2%)
 56  2013  232.531   1.1%   49     1,473    0.4%    2.6%   (2.2%)
 57  2014  237.072   2.0%   50     1,473    0.5%    2.7%   (2.2%)
To repeat this calculation with other annuity figures or with a different starting year, follow these steps:
  • Select All, Copy, and Paste [5] the following at cell A1 of a blank Excel sheet:

    Code: Select all

    Cost	100000
    Nom mo pmt	350
    Nom COLA	0.02
    Real mo pmt	305
    		Ann				--- Nom Annuity ---		RAnn	
    Year	Apr CPI	Year	Apr CPI	Chg	Year	Real CF	RIRR	RIRR	Diff
    1913	9.8	1964	=VLOOKUP(C7,A$7:B$113,2,FALSE)		0	=-B1
    1914	9.8	=C7+1	=VLOOKUP(C8,A$7:B$113,2,FALSE)	=D8/D7-1	=F7+1	=12*B$2*(1+B$3)^F8*(D$7/D8)	=IRR(G$7:G8)	=RATE(F8,12*B$4,-B$1,0,0)	=H8-I8
    1915	10
    1916	10.6
    1917	12.6
    1918	14.2
    1919	16.7
    1920	20.3
    1921	18.1
    1922	16.7
    1923	16.9
    1924	17
    1925	17.2
    1926	17.9
    1927	17.3
    1928	17.1
    1929	16.9
    1930	17
    1931	15.5
    1932	13.9
    1933	12.6
    1934	13.3
    1935	13.8
    1936	13.7
    1937	14.3
    1938	14.2
    1939	13.8
    1940	14
    1941	14.3
    1942	16.1
    1943	17.4
    1944	17.5
    1945	17.8
    1946	18.4
    1947	21.9
    1948	23.8
    1949	23.9
    1950	23.6
    1951	25.8
    1952	26.4
    1953	26.6
    1954	26.8
    1955	26.7
    1956	26.9
    1957	27.9
    1958	28.9
    1959	29
    1960	29.5
    1961	29.8
    1962	30.2
    1963	30.5
    1964	30.9
    1965	31.4
    1966	32.3
    1967	33.1
    1968	34.4
    1969	36.3
    1970	38.5
    1971	40.1
    1972	41.5
    1973	43.6
    1974	48
    1975	52.9
    1976	56.1
    1977	60
    1978	63.9
    1979	70.6
    1980	81
    1981	89.1
    1982	94.9
    1983	98.6
    1984	103.1
    1985	106.9
    1986	108.6
    1987	112.7
    1988	117.1
    1989	123.1
    1990	128.9
    1991	135.2
    1992	139.5
    1993	144
    1994	147.4
    1995	151.9
    1996	156.3
    1997	160.2
    1998	162.5
    1999	166.2
    2000	171.3
    2001	176.9
    2002	179.8
    2003	183.8
    2004	188
    2005	194.6
    2006	201.5
    2007	206.686
    2008	214.823
    2009	213.24
    2010	218.009
    2011	224.906
    2012	230.085
    2013	232.531
    2014	237.072
    2015	236.599
    2016	239.261
    2017	244.524
    2018	250.546
    2019	255.548
  • Format for readability
  • Copy cells C8:J8 down to row 57
  • Revise annuity figures in cells B1:B4 if desired
  • Revise starting year for the annuity in cell C7 if desired
  1. AlohaJoe wrote:
    Tue May 28, 2019 9:25 pm
    $100,000 for a male/female couple that are 65/65 and live in Colorado pays $470/month for life (nominal). With 2% adjustments it pays $350. With CPI adjustments it pays $305.
  2. Over the 35 years from April 1964 to April 1999 the CPI increased at a compounded annual rate of 4.9%. (See series CUUR0000SA0 from BLS Top Picks.)
    4.9% = (166.2 / 30.9) ^ (1 / 35) - 1
  3. Example calculations for 35 years using Excel IRR and RATE functions:

    Code: Select all

    1,049 = 470 * 12 * (30.9 / 166.2)
    -0.8% = IRR(G$7:G42)
    +1.4% = RATE(35, 305 * 12, -100000, 0, 0)
  4. Example calculation for 35 years:

    Code: Select all

    1,562 = 350 * 12 * 1.02 ^ 35 * (30.9 / 166.2)
    -0.9% = IRR(G$7:G42)
  5. If you have trouble pasting, try "Paste Special" and "Text".

User avatar
ResearchMed
Posts: 9052
Joined: Fri Dec 26, 2008 11:25 pm

Re: Annuities are a bet on future inflation rates and should not be purchased by retirees

Post by ResearchMed » Sat Jun 01, 2019 9:19 am

What were the SS COLA's back when inflation was sky high, say, 1980 - 1982 (or a year or two later, given any lag)?

And what was the "official inflation" rate?
All I remember are the high mortgage rates and the high money market rates, both even more astonishing in retrospect (at least from this vantage point now).

RM
This signature is a placebo. You are in the control group.

longinvest
Posts: 3754
Joined: Sat Aug 11, 2012 8:44 am

Re: Annuities are a bet on future inflation rates and should not be purchased by retirees

Post by longinvest » Sat Jun 01, 2019 9:36 am

#Cruncher,
#Cruncher wrote:
Sat Jun 01, 2019 9:18 am
longinvest wrote:
Fri May 31, 2019 10:54 pm
#Cruncher wrote:
Fri May 31, 2019 4:50 pm
Ah, but that's the rub! There is a "significant cost" of a CPI-linked annuity versus a nominal one. ...
Should a similar analysis be done with fire insurance premiums and payouts? What if the conclusion was that one shouldn't insure the house against fire because one is likely to lose the premiums in most cases? :wink:
I wasn't arguing against buying a CPI-indexed annuity. I was arguing that its cost relative to a nominal annuity is much greater than the "cost" of TIPS relative to nominal treasuries. Just as with TIPS vs nominal Treasuries, if the CPI rises more than the difference in returns (i.e., the breakeven inflation rate), then the CPI-indexed annuity will have the better outcome. One needs it to protect against that possibility.
Insurance in inherently expensive. The insurance company has to collect more, in premiums, than it pays back to insurees because it has to cover its costs and make a profit, otherwise it will go out of business.

A sensible approach to insurance is to never buy more insurance than strictly needed.
#Cruncher wrote:
Sat Jun 01, 2019 9:18 am
longinvest in same post wrote:In situations where an inflation-indexed SPIA doesn't exist or isn't competitively priced, I would consider buying a 2%-indexed SPIA, instead ...
Depending on the specifics the 2% compounded annuity may not protect against severe inflation any better than the straight nominal annuity. Again using the figures from this post by AlohaJoe, after 35 years the annual return of the $350 / month +2% compounded annuity purchased in 1964 was only -0.9% [4], a touch worse than the -0.8% annual return of the $470 / month straight nominal annuity.
#Cruncher, you've cut an important part of my statement. I wrote:
longinvest wrote:
Fri May 31, 2019 10:54 pm
In situations where an inflation-indexed SPIA doesn't exist or isn't competitively priced, I would consider buying a 2%-indexed SPIA, instead, because the Federal Reserve has adopted a 2% inflation target.
Backtesting US inflation over a period before 2012, like your 1964 backtest, is inappropriate. I've explained this in details in a previous post.
Bogleheads investment philosophy | single-ETF balanced portfolio | VBAL

randomguy
Posts: 7974
Joined: Wed Sep 17, 2014 9:00 am

Re: Annuities are a bet on future inflation rates and should not be purchased by retirees

Post by randomguy » Sat Jun 01, 2019 9:47 am

ResearchMed wrote:
Sat Jun 01, 2019 9:19 am
What were the SS COLA's back when inflation was sky high, say, 1980 - 1982 (or a year or two later, given any lag)?

And what was the "official inflation" rate?
All I remember are the high mortgage rates and the high money market rates, both even more astonishing in retrospect (at least from this vantage point now).

RM
https://www.ssa.gov/cola/

People talk about rising costs in retirement but some of them aren't inflation related. Buying an annuity isn't the way to deal with expensive nursing homes. On average medical costs aren't that bad (something like 150k for the average 65 years. Sounds like a lot til you realize it is spread out over 20-30 years). There are some outliers who get hit with big bills but that isn't really an inflation issue. And there is definitely a lot of uncertainty (i.e. what will medicare look like in 20-30 years).

User avatar
ResearchMed
Posts: 9052
Joined: Fri Dec 26, 2008 11:25 pm

Re: Annuities are a bet on future inflation rates and should not be purchased by retirees

Post by ResearchMed » Sat Jun 01, 2019 10:02 am

randomguy wrote:
Sat Jun 01, 2019 9:47 am
ResearchMed wrote:
Sat Jun 01, 2019 9:19 am
What were the SS COLA's back when inflation was sky high, say, 1980 - 1982 (or a year or two later, given any lag)?

And what was the "official inflation" rate?
All I remember are the high mortgage rates and the high money market rates, both even more astonishing in retrospect (at least from this vantage point now).

RM
https://www.ssa.gov/cola/

People talk about rising costs in retirement but some of them aren't inflation related. Buying an annuity isn't the way to deal with expensive nursing homes. On average medical costs aren't that bad (something like 150k for the average 65 years. Sounds like a lot til you realize it is spread out over 20-30 years). There are some outliers who get hit with big bills but that isn't really an inflation issue. And there is definitely a lot of uncertainty (i.e. what will medicare look like in 20-30 years).
Thanks.

I would not have guessed that it got as high as 14.3% at the peak of that time period.

RM
This signature is a placebo. You are in the control group.

User avatar
Ben Mathew
Posts: 751
Joined: Tue Mar 13, 2018 11:41 am
Location: Seattle
Contact:

Re: Annuities are a bet on future inflation rates and should not be purchased by retirees

Post by Ben Mathew » Sat Jun 01, 2019 12:00 pm

#Cruncher wrote:
Fri May 31, 2019 4:50 pm
Ben Mathew wrote:
Wed May 29, 2019 7:21 am
… Given that the expected yield of TIPS isn't significantly less than that of nominal bonds ..., it's not clear to me why most people choose to hold nominal bonds instead of real. TIPS offer free, or almost free, insurance against unexpected inflation. Why take a risk we don't have to? Same with real vs nominal annuities. If there isn't a significant cost, real is better than nominal. (underline added)
Ah, but that's the rub! There is a "significant cost" of a CPI-linked annuity versus a nominal one.
I agree. I concluded that in this post upthread. Given that TIPS are not much more expensive than nominal treasuries, I think real annuity premiums are higher because there is only one provider (Principal). That in turn is probably because there is little demand from consumers, likely because of money illusion.

User avatar
Topic Author
CULater
Posts: 2109
Joined: Sun Nov 13, 2016 10:59 am

Re: Annuities are a bet on future inflation rates and should not be purchased by retirees

Post by CULater » Fri Jun 07, 2019 8:45 am

This from Oblivious Investor, argues that annuities with a fixed COLA actually are actually worse than nominal annuities for most people when there is high inflation:
The key point here is that, the higher inflation is over this person’s lifetime, the worse the annuity with the COLA does as compared to the annuity without the COLA.

Why is this? It’s because the annuity with the COLA has a greater portion of its payout occurring later in the annuitant’s life (due to the fact that its payout starts lower, but climbs over time). And in a scenario in which dollars are declining in value over time due to inflation, the annuity that front-loads the payout (i.e., the annuity without the COLA) does better.

But annuities with fixed cost of living adjustments do not protect against inflation. Not only do they not keep up with high rates of inflation, they actually perform worse in the face of inflation than annuities without COLAs.
https://obliviousinvestor.com/annuities ... inflation/
On the internet, nobody knows you're a dog.

longinvest
Posts: 3754
Joined: Sat Aug 11, 2012 8:44 am

Re: Annuities are a bet on future inflation rates and should not be purchased by retirees

Post by longinvest » Fri Jun 07, 2019 8:59 am

The Oblivious Investor article doesn't mention the Federal Reserve's 2% inflation target.

As I wrote earlier, I'm of the opinion that, in absence of a competitively-priced CPI-indexed annuity, a 2%-indexed annuity is a good enough alternative to insure a base of stable lifelong income (in addition to Social Security) when the Federal Reserve communicates its intention to keep inflation on a 2% target to allow the public "to make accurate longer-term economic and financial decisions".
Bogleheads investment philosophy | single-ETF balanced portfolio | VBAL

Leesbro63
Posts: 5907
Joined: Mon Nov 08, 2010 4:36 pm

Re: Annuities are a bet on future inflation rates and should not be purchased by retirees

Post by Leesbro63 » Fri Jun 07, 2019 9:08 am

I've helped an aging relative with this decision. She's now age 83. 11 years ago, we looked at annuity costs and determined that the most prudent approach, in our opinion, was to buy a little then, a little more later and even more later. Of fixed payment annuities. Because the inflation adjusted ones, even with a fixed inflation kicker, was so much more money.

It's worked out well, although the original annuity was with AIG which almost went away shortly after we bought the annuity (perhaps that's why the price was so much better). But AIG survived and we purchased more SPIAs about 5 years later and again 2 years ago. Might convert more nest egg to SPIAs in the future.

My point is that a good alternative to highly priced inflation-adjusted SPIAs is to ladder purchases of non-inflation adjusted SPIAs you age, starting just after age 70.
Last edited by Leesbro63 on Fri Jun 07, 2019 9:35 am, edited 1 time in total.

azanon
Posts: 2534
Joined: Mon Nov 07, 2011 10:34 am
Contact:

Re: Annuities are a bet on future inflation rates and should not be purchased by retirees

Post by azanon » Fri Jun 07, 2019 9:17 am

I apologize upfront if you guys already knew this (and I only had time to scan the thread), but was this article by chance a response to David Blanchett's article at the same site entitled: "Inflation-Linked SPIAs Are a Bad Deal" published 5/20/19(link: https://www.advisorperspectives.com/art ... a-bad-deal )

So David (PhD, CFA, CFP® is the head of retirement research for Morningstar Investment Management LLC.), says SPIAs linked to inflation don't make sense for the vast majority of retirees.

Post Reply