SS at FRA vs. at 70 - Unexpected results

[Telly]

FRA for DW & me is 66. DW is one year younger than me.

Using the spreadsheet that I have developed and refined over the years, I ran two plans:

Plan A - I start SS @ FRA; One year later DW at her FRA starts SS on my record.

Plan B - I file-and-suspend at age 67, so DW can file at her FRA on my record. I start benefits at age 70.

Constants: Inflation 3%; SS COLA 2%; Portfolio return 6% nominal; Same portfolio start value; Same tiny pension for me; DW same 1/2 of my FRA amount for her SS, modified same by her GPO.

What was Unexpected? It never crosses over!
Plan A always has the highest amount in the kitty, all the way through age 95 where I stop. The cause seems to be that Plan B requires substantial withdrawals for my age years 66,67,68,69, that Plan A does not. And those substantial withdrawals are a lost opportunity at 6%, that can never be overcome by the larger SS values over time.

I have played making the portfolio start value larger and smaller, the effect is still the same, no crossover.

Now one could say: "SO WHAT!! If you are holding off SS till age 70 to help guaranty DW having a better COLA'd lifetime annuity (and therefore it is assumed more total $$ to live on than not), Plan B will be better if you die before her (which is likely)."

So I then made a Plan A Prime, and a Plan B Prime, the difference from the earlier plans is I die at 73, and DW switches over to my full record in each case. My tiny pension stops either way, DW is GPO'd either way.
And the result was... Taking SS at my FRA rather than at 70 continued to have the highest portfolio amount, even with me dying at 73.
I have looked for errors, and have found none. Interesting, and unexpected!

 

[fh2000]

F Portfolio return 6% nominal;

If you reduce that to say 5%, or 4%, will you get the same conclusion?

 

[Telly]

If you reduce that to say 5%, or 4%, will you get the same conclusion?

That's a good question. So I went and re-ran all four, and found that at 4% portfolio growth, it would crossover. At 5%, it would not.

 

[karluk]

Constants: Inflation 3%; SS COLA 2%;

This is one obvious source of possible errors. Inflation is generally measured using CPI, and SS COLAs are also calculated from CPI. Since both measurements are derived from the same source data, it's hard to see how inflation will average 3% during your retirement, but SS COLAs will only be 2%.

The net result of your assumption that SS COLAs will not keep pace with inflation is that your calculations depreciate the benefit of waiting until age 70 to get a bigger SS payment. That's because you're assuming that the bigger payment will gradually lose purchasing power over the years, so there's not as much value to you in waiting. Without tracing through all of your calculations, I find it entirely believable that this one questionable assumption is the reason that your plan A consistently beats plan B.

 

[athena53]

This is one obvious source of possible errors. Inflation is generally measured using CPI, and SS COLAs are also calculated from CPI. Since both measurements are derived from the same source data, it's hard to see how inflation will average 3% during your retirement, but SS COLAs will only be 2%.

Wanna bet? SS COLA is based on "core inflation" which excludes a few categories that tend to fluctuate more, such as food and energy. Yeah- items that make up a lot of retirees' purchases.

 

[karluk]

Wanna bet? SS COLA is based on "core inflation" which excludes a few categories that tend to fluctuate more, such as food and energy. Yeah- items that make up a lot of retirees' purchases.

I'll take that bet. The following link says that SS COLAs are based on CPI-W. If you have a source that says SS COLAs are based on core CPI, please provide it.

Latest Cost-of-Living Adjustment

 

[braumeister]

Since both measurements are derived from the same source data, it's hard to see how inflation will average 3% during your retirement, but SS COLAs will only be 2%.

I can throw in another data point to support that point.

My military retired pay ("pension" for all practical purposes) is based on the same calculation as Social Security.

With 26 years of experience now, my COLA adjustments cumulatively have kept pace beautifully with CPI-W. I've tracked it every year, and the agreement has been perfect to within a small fraction of a percent.

 

[athena53]

I'll take that bet. The following link says that SS COLAs are based on CPI-W. If you have a source that says SS COLAs are based on core CPI, please provide it.

Latest Cost-of-Living Adjustment

Oops. I remember someone telling me that Core Inflation was used someplace where it didn't really measure the inflation experienced by a particular group and I thought it was SS. Looks like I was wrong.

 

[N02L84ER]

Telly, silly question, but is your age 70 benefit 4 years times 8% (32%) larger than your FRA benefit in your calculations?

 

[karluk]

Telly, silly question, but is your age 70 benefit 4 years times 8% (32%) larger than your FRA benefit in your calculations?

I haven't done SS benefit calculations recently, but I believe that the age 70 benefit is not only increased by 8% per year over the FRA benefit, but is also inflation adjusted. So, using OP's 3% inflation assumption, the age 70 benefit should be around 1.32*1.03^4 = 1.4857 higher than the age 66 FRA benefit. That's nearly a 50% increase in benefits just for waiting four years.

If the post FRA benefit isn't inflation adjusted, please let me know. I have been planning on waiting until age 70 to claim, but if there isn't an inflation adjustment for those four years between 66 & 70, it might affect my plans

 

[bingybear]

apples, oranges and grapes.

plan A you are basing of your individual earnings record, the second is just based on your record. The common crossover is one person taking a one age or another. In general using two different earnings record does not have to cross an any time.... it could.. .just depends on the data.

 

[Katsmeow]

apples, oranges and grapes.

plan A you are basing of your individual earnings record, the second is just based on your record. The common crossover is one person taking a one age or another. In general using two different earnings record does not have to cross an any time.... it could.. .just depends on the data.

I think on both cases it is being assumed that the wife is obtaining benefits on his record. I know in some cases the wife does this until 70 and then starts taking her own benefits. He doesn't mention that, so I guess that means she doesn't have benefits that are more than half of his.

 

[Telly]

Telly, silly question, but is your age 70 benefit 4 years times 8% (32%) larger than your FRA benefit in your calculations?

Yes

This is one obvious source of possible errors. Inflation is generally measured using CPI, and SS COLAs are also calculated from CPI. Since both measurements are derived from the same source data, it's hard to see how inflation will average 3% during your retirement, but SS COLAs will only be 2%.

The net result of your assumption that SS COLAs will not keep pace with inflation is that your calculations depreciate the benefit of waiting until age 70 to get a bigger SS payment. That's because you're assuming that the bigger payment will gradually lose purchasing power over the years, so there's not as much value to you in waiting. Without tracing through all of your calculations, I find it entirely believable that this one questionable assumption is the reason that your plan A consistently beats plan B.

I re-ran it with Inflation = 3%, SS COLA = 3%
Both Plan A & B portfolio amounts were greater over time (expected), but there still was no crossover.

If the post FRA benefit isn't inflation adjusted, please let me know. I have been planning on waiting until age 70 to claim, but if there isn't an inflation adjustment for those four years between 66 & 70, it might affect my plans

COLAs are invoked only once benefits are started and received. They do not affect future benefits not yet started.
For future benefits not started, Wage Indexing can increase the projected amount over the years, even with no more pay-in to SS. However, W.I. stops at age 60. See this for more detail: Indexing Factors for Earnings

Looking into my spreadsheets further, I am rapidly coming to the conclusion that many of us (myself included) have been parroting the "Hold off until age 70, so much more than age 66" mantra, floating out there by itself as a stand-alone entity, not adding the important unstated conditions.

Articles that encourage one to wait till age 70, and show an eventual crossover of total benefits received at some later year, talk about "waiting till 70 to retire". Reading between the lines, that means continuing to work from age 66 to age 70. Therefore, the living expenses for ages 66,67,68,69 are paid from current wages, not pulled from our portfolio!

But that sure ain't my application of it! Those four years of expenses have to come from the portfolio. Just as an example, I left the SS COLA set at 3%, and looked at a few details:

Plan A requires that I pull 18.5k total out of the portfolio over those four years to augment SS benefits received.
Plan B requires that I pull 132k total out of the portfolio over those four years to help pay for living expenses.

Then I looked at the portfolio value, using my Constant Dollars (inflation adjusted) portfolio value column:
At age 70, Plan B was 100k behind Plan A
At age 80, Plan B was 85k behind Plan A
At age 90, Plan B was 66k behind Plan A

This is just a quick example of the effect of pulling a chunk out of the portfolio's 6% assumed growth and burning it for living expenses, to hold off starting SS from age 66 till age 70.
It's the (lost) opportunity cost of money. I'm not saying its a disaster... but rather the delay from age 66 FRA to age 70 needs to be looked at as part of a whole system, not just a mechanism by itself, as so often described. That is what I am learning here. No magic.

 

[Telly]

I think on both cases it is being assumed that the wife is obtaining benefits on his record. I know in some cases the wife does this until 70 and then starts taking her own benefits. He doesn't mention that, so I guess that means she doesn't have benefits that are more than half of his.

Very good Kat!
DW does not have sufficient SSA-able earnings credits to have her own usable record. Medicare credit yes, SS, no.

 

[pb4uski]

Oops. I remember someone telling me that Core Inflation was used someplace where it didn't really measure the inflation experienced by a particular group and I thought it was SS. Looks like I was wrong.

I'm not sure you were. I think the headline CPI number is CPI-U and the SS COLA is based on CPI-W and they are different. Note that the CPI-U number is for a much broader population (89%) and the CPI-W that is the basis for SS COLAs only covers 28% of the population.

The Consumer Price Index (CPI) is a measure of the average change in prices over time of goods and services purchased by households. The Bureau of Labor Statistics publishes CPIs for two population groups: (1) the CPI for Urban Wage Earners and Clerical Workers (CPI-W), which covers households of wage earners and clerical workers that comprise approximately 28 percent of the total population and (2) the CPI for All Urban Consumers (CPI-U) and the Chained CPI for All Urban Consumers (C-CPI-U), which covers approximately 89 percent of the total population and includes, in addition to wage earners and clerical worker households, groups such as professional, managerial, and technical workers, the selfemployed, short-term workers, the unemployed, and retirees and others not in the labor force.

 

[Brett_Cameron]

Yep. One has to look at SS as part of the entire picture. I have found that taking SS at 62 is best for me to avoid withdrawals from other retirement funds.

 

[karluk]

COLAs are invoked only once benefits are started and received. They do not affect future benefits not yet started.

I don't claim to be an expert on SS benefit calculations, but I'm seeing an awful lot of information online that contradicts your assertion. One's starting SS benefit does, indeed appear to be indexed annually for inflation. See, for example, Michael Kitces article on the effects of delaying SS.

At the extreme, that individual would only get a $750/month benefit by starting as early as possible (at age 62), or could get as much as $1,320/month by waiting as long as possible (to age 70). (Notably, these benefits would also be adjusted annually for cost-of-living adjustments, so these represent real inflation-adjusted benefit amounts at various ages).

https://www.kitces.com/blog/how-del...ong-term-investment-or-annuity-money-can-buy/

Looking into my spreadsheets further, I am rapidly coming to the conclusion that many of us (myself included) have been parroting the "Hold off until age 70, so much more than age 66" mantra, floating out there by itself as a stand-alone entity, not adding the important unstated conditions.

Articles that encourage one to wait till age 70, and show an eventual crossover of total benefits received at some later year, talk about "waiting till 70 to retire". Reading between the lines, that means continuing to work from age 66 to age 70. Therefore, the living expenses for ages 66,67,68,69 are paid from current wages, not pulled from our portfolio!

But that sure ain't my application of it! Those four years of expenses have to come from the portfolio.

It is certainly possible that there are articles advocating holding off on claiming SS that carelessly fail to take into consideration the effect of portfolio depletion on the break even point, but this is certainly not true of the Kitces article I quoted above. Kitces explicitly mentions this point and still finds that holding off on SS wins in the long run.

However, the caveat to making the decision to delay is that it also has a non-trivial "cost" in the form of benefits that aren't received (and can't be invested or consumed) today. While this certainly doesn't eliminate the value of delaying Social Security benefits, it does make the trade-off more complex and nuanced than just (incorrectly) equating it to an 8%/year (or 76%-per-8-years) returns.

Those four years of expenses have to come from the portfolio. Just as an example, I left the SS COLA set at 3%, and looked at a few details:
...
At age 70, Plan B was 100k behind Plan A
At age 80, Plan B was 85k behind Plan A
At age 90, Plan B was 66k behind Plan A

Kitces's assumptions are almost identical to yours - 3% inflation and 6% portfolio returns and he gets break even much sooner than you do. Considering that your numbers show Plan B gradually catching up to Plan A, but not passing it until after age 90, I think a reasonable guess is that the only difference between Kitces and you is your difference of opinion on whether the age 70 benefit is inflation adjusted or not. Again, I don't claim to be an expert, but I've read a lot of sources that say delayed benefits are inflation adjusted, so I tend to trust Kitces results more than yours.

 

[karluk]

I think the headline CPI number is CPI-U and the SS COLA is based on CPI-W and they are different.

This is true, but I'm not sure it's relevant. CPI-U and CPI-W are not identical, but they rarely if ever differ by more than a trivial amount and, looking at historical numbers, CPI-W seems to be higher than CPI-U just as often as is lower. That's in marked contrast to core CPI, which frequently differs from CPI-U and CPI-W by wide margins.

 

[fosterscik]

I think that you are missing the guaranteed benefit associated with delayed SS. Although average returns may show that you would benefit by not taking the benefit early they do not address what happens if the market turns down. For me the guarantee provided by a higher permanent (not subject to loss) paycheck far outweighs the benefits you outline based on not waiting until 70.

I plan to delay until 70 unless we have a market crash between 62 and 70. If that occurs I will reevaluate. Higher guaranteed benefits for me and eventually my surviving spouse trump all other considerations.

 

[Ken11]

Seems to me that an increased SS payout is a form of longevity insurance
Insurance against a risk has some value
An analysis of expected portfolio value does not recognize the value of a less risky income stream (the insurance).
This may be a bigger decision factor re: SS strategies than the optimization of expected portfolio value for those carrying uncomfortable levels of longevity or other income risks

 

[Ken11]

I'm 61 my wife depends on my ss record for her benefits
My latest thinking is to buy longevity insurance by delaying SS as long as my withdrawal rate is in my comfort zone. If the market hiccups sending my withftawal rate above my comfort threshold that's when I file
Comments?

 

[karluk]

Telly, I have put together a spreadsheet that resembles your description of yours. I get a break even point between plans A & B around age 86 or 87. That is a similar break even point to what I've seen from online calculators. So, based on what I'm seeing, I think your calculations are well worth going through, but end up just confirming the conventional wisdom that it pays to delay taking SS if you live long enough.

 

[pb4uski]

What I do for now is compare alternatives using the what-if function in Quicken Lifetime Planner. So my base case might be waiting until 70 and then I'll test taking it earlier with all other assumptions held constant and it gives me an idea of the crossover point.

We are a few years away from having to make a decision, but when the time comes we'll probably pony up a little money and have a professional analysis done by socialsecuritysolutions.com or one of the other similar services out there.

 

[RunningBum]

I think that you are missing the guaranteed benefit associated with delayed SS. Although average returns may show that you would benefit by not taking the benefit early they do not address what happens if the market turns down. For me the guarantee provided by a higher permanent (not subject to loss) paycheck far outweighs the benefits you outline based on not waiting until 70.

I plan to delay until 70 unless we have a market crash between 62 and 70. If that occurs I will reevaluate. Higher guaranteed benefits for me and eventually my surviving spouse trump all other considerations.

Yes, this is one decision where I do plan on being a market timer. If I'm unsure, I'll probably wait until 70, but if the market seems low to me at 62 or takes a good fall between 62 and 70 I will probably start early.

I have gone back and forth on this over the years. During the 90s run up, I thought it was a no-brainer to take SS early because I could easily invest and beat the increased later SS payment. After the bust, I got more conservative and liked the guaranteed higher payment for longevity. Now I'm seeing that this could be a good time to try to judge the market to improve on an actuarial draw. My bias is still toward longevity insurance.

 

[REWahoo]

I'm 61 my wife depends on my ss record for her benefits
My latest thinking is to buy longevity insurance by delaying SS as long as my withdrawal rate is in my comfort zone. If the market hiccups sending my withftawal rate above my comfort threshold that's when I file
Comments?

Been there, done that.

I hit the "withdrawal rate above my comfort threshold" in late 2008 and filed for SS at age 62. I'd planned to wait until FRA or later, but the market gods had other plans.