When To Take Social Security? Study takes a statistical approach.

[DrRoy]

I read an analysis once that instead of trying to determine the maximum payout from SS filing, one should look at it as longevity insurance and file as late as you can (up to 70).

 

[TheWizard]

I read an analysis once that instead of trying to determine the maximum payout from SS filing, one should look at it as longevity insurance and file as late as you can (up to 70).

Correct, assuming you have more than sufficient funds to live on while delaying.

So the optimization algorithm is simple:
Rich people in good health: delay until 70.
Poor people OR poor health: claim at 62...

 

[TheWizard]

Correct, assuming you have more than sufficient funds to live on while delaying.

So the optimization algorithm is simple:
Rich people in good health: delay until 70.
Poor people OR poor health: claim at 62...

And what does "more than sufficient" mean?

Assuming near maximum age 70 SS benefit of $50k per year (presently), then 8-year delay from 62 to 70 takes $400k.
But if someone has only $400k in investible assets, that's a 12.5% WR to start with, meaning high risk of having very little left at age 70, even if invested at 60/40 AA.

$1M invested 60/40 means 5% WR to start with, and a good chance of a decent stash remaining at age 70.

But who wants to live in SS alone, even $50k per year for a single person? Let's say desired total income is $100k per year.
If our retiree has a $50k pension, then he's All Set with $1M and delaying SS to 70.

Otherwise, a $2M stash allows a somewhat risky $100k withdrawal for 8 years from 62 to 70 and a $50k withdrawal from age 70 on.

So those are good minimums for defining who can afford to delay SS to 70...

 

[ShokWaveRider]

There is also another tact. If, and that is a big IF, there will be a ~20% haircut in 203x or whenever, then taking it as early as possible holds a bit of water.

 

[Rianne]

I decided to see what firecalc says:
Spend: $120,000
Portfolio: $2,000,000
30 years
SS at FRA - all other being equal - AA 45/55, inflation CPI
Results:
FRA
End of Retirement: $961,982 - $9,956361
Average at end: $3,190,942

SS at 70, all equal to above except the SS age:
End of Retirement: $274,116 - $8,403,887
Average at end: $2,533,100

If you apply SS at FRA all others being equal you come out ahead during and after retirement. I also changed the inflation rate to 3%, still ahead at FRA (100%), but only 98.4% success at 70.

Edit: Constant spending at both SS: FRA and 70

 

[USGrant1962]

I decided to see what firecalc says:
Spend: $120,000
Portfolio: $2,000,000
30 years
SS at FRA - all other being equal - AA 45/55, inflation CPI
Results:
FRA
End of Retirement: $961,982 - $9,956361
Average at end: $3,190,942

SS at 70, all equal to above except the SS age:
End of Retirement: $274,116 - $8,403,887
Average at end: $2,533,100

If you apply SS at FRA all others being equal you come out ahead during and after retirement. I also changed the inflation rate to 3%, still ahead at FRA (100%), but only 98.4% success at 70.

Edit: Constant spending at both SS: FRA and 70

Interesting. What amount(s) did you use for SS? And what year(s)?

 

[Midpack]

If one has no other choice but to take SS at a certain age like many people do, then one isn't really choosing.

The study defines optimal as lifetime income and presumes that the subject has a choice, so I don't think of it as a flaw.

+1. Several people here didn't read the article or had an answer in mind outside what the linked study was aimed at. The article is about optimizing lifetime Soc Sec income, it does not suggest that's the only consideration re: when to start Soc Sec.

Clearly, lifetime income is an important factor, something very worth knowing - the article and underlying study never said it was the only one.

The purpose of this analysis was to determine if these roughly 20,000 retirees had made an optimal claiming decision. For United Income, an "optimal" decision was one that resulted in the highest possible lifetime benefits for a claimant. Note, highest lifetime benefit may not be synonymous with highest possible monthly benefit.

 

[Rianne]

Interesting. What amount(s) did you use for SS? And what year(s)?

Recalculate numbers to be sure, changed:
FRA $60K - 2024

End of Retirement
$278,021 - $10,943,437
Average: 3,342344
100%

Age 70 $68K - 2028

Edit: Our numbers, based on that I took SS at age 65, then used totals with spouse starting at FRA and 70. Portfolio and spending I made up as an average. Our portfolio obviously fluctuates with the market.

End of Retirement
-$119,855 - $9,724,411
99.2%

 

[pb4uski]

There is also another tact. If, and that is a big IF, there will be a ~20% haircut in 203x or whenever, then taking it as early as possible holds a bit of water.

I dunno about that. I would you rather receive 80% of 124% than 80% of 70%.

 

[Out of Steam]

My wife filed a few months before turning 63 because of her gut feeling on longevity--though retirement has agreed with her health-wise. I think she's being unduly pessimistic.

I'm also in the 62-65 window, and am committed to delaying only until a few months after 65. Though, I don't think I will delay more than a few months after FRA, which is when Open Social Security estimates that we'll collect 95% of our maximum.

It's a balance between money being more enjoyable now than later, and the fact that, if one is running short of money, it's a lot easier to get a j*b at 65 than 85.

 

[11522914]

In my situation there are so many different variables. How will it affect ACA subsidies, FASFA loans and grants for my daughter, federal taxes and Roth conversions.

But the bottom line is I simply don't need the money right now. Just watching it month to month. Nice to know it's there if needed.

 

[Rianne]

Using Firecalc and another calculator, financial mentor, both results based on an overall portfolio over 30 years come out better taking SS at FRA, for us anyway.

 

[street]

I dunno about that. I would you rather receive 80% of 124% than 80% of 70%.

Depends on how long I live.

 

[Midpack]

Using Firecalc and another calculator, financial mentor, both results based on an overall portfolio over 30 years come out better taking SS at FRA, for us anyway.

I'd run it through https://opensocialsecurity.com including the select situations (mortality comparable to whatever you entered in FIRECALC?) if applicable if you haven't already. More focused on optimizing lifetime income from Soc Sec than FIRECALC. FWIW

 

[qwerty3656]

Recalculate numbers to be sure, changed:
FRA $60K - 2024

End of Retirement
$278,021 - $10,943,437
Average: 3,342344
100%

Age 70 $68K - 2028

Edit: Our numbers, based on that I took SS at age 65, then used totals with spouse starting at FRA and 70. Portfolio and spending I made up as an average. Our portfolio obviously fluctuates with the market.

End of Retirement
-$119,855 - $9,724,411
99.2%

Doesn't SS grow by 8% per year? So if FRA is 60k, age 70 would be more than 68k. Also, you should apply the 3 years of inflation to the year 70 SS.

 

[Rianne]

Doesn't SS grow by 8% per year? So if FRA is 60k, age 70 would be more than 68k. Also, you should apply the 3 years of inflation to the year 70 SS.

Not necessarily, I took mine at 65. And this is where I have an issue with opensocialsecurity.com. Based on my PIA ($1012/month), my annual benefit in the OSS calculation is $10,930. I am earning $13,644 per year. This is what goes into my checking account minus Medicare. My spousal benefit when taken at 69 and 10 months is $7062 and is the same if DH takes at FRA 66.5 yrs. No change in my spousal benefit if DH takes at 69 and 10 months or FRA.

When you calculate the 8% increase from FRA to 70, it does not make sense because of my numbers. It is not cut and dried as everyone thinks.

For one, I am currently making $2714 more per year today than OSS says I'm making.

So together at 70 filings according to their strategy, our SS income is:
$65, 879. But it's really $68,593.

Filing at FRA according to OSS is $56,404. But really it's $56,404 + $2714. $59,118.

I'm saying, plug your numbers into Firecalc and see the difference in the outcome after 30 years. My calculations show overall we are better off taking SS at DH FRA.

 

[RunningBum]

I don't see any recognition given to the "time value of money" in the study.

I didn't read the study closely, but I assume it allows for the same amount of spending regardless of when you take SS. This is what those of us who advocate holding off on taking SS until age 70, if you can.

So if I can spend the same amount every year, and have more money at some target age if I wait until age 70, where does the "time value of money" come into play?

 

[MarieIG]

This article takes an interesting statistical, back-testing approach to the much debated question of when/what age to claim Social Security.

An excerpt:

“Statistically speaking, there is a best age to claim Social Security benefits. In 2019, online investment management and financial planning company United Income released a study that analyzed the retired-worker claiming decisions of approximately 20,000 respondents using the University of Michigan's Health and Retirement Study.

The purpose of this analysis was to determine if these roughly 20,000 retirees had made an optimal claiming decision. For United Income, an "optimal" decision was one that resulted in the highest possible lifetime benefits for a claimant. Note, highest lifetime benefit may not be synonymous with highest possible monthly benefit.

What United Income's study showed was that claiming age and optimal claiming decisions were almost a perfect inverse of each other. In other words, there were very few optimal claims, with seniors regularly leaving large sums of Social Security income on the proverbial table.”

This approach is definitely not perfect. I think I’d prefer a more diverse, larger sample.

What do y’all think?

I’m trying to find the original study.

I would be curious to see if this holds up based upon updated (current) life expectancy tables.

(For DH and I, we are claiming based upon longevity insurance as a couple vs. optimization.)

 

[ERD50]

There is also another tact. If, and that is a big IF, there will be a ~20% haircut in 203x or whenever, then taking it as early as possible holds a bit of water.

I dunno about that. I would you rather receive 80% of 124% than 80% of 70%.

But if you take it early (and I'm not suggesting you do, I'm just going through the numbers), you'd be getting 100% of 70% up until 203x. So depends on when you start, and how close to 203x (if it happens), but it is a factor, right?


-ERD50

 

[SevenUp]

And this is where I have an issue with opensocialsecurity.com. Based on my PIA ($1012/month), my annual benefit in the OSS calculation is $10,930. I am earning $13,644 per year. This is what goes into my checking account minus Medicare. My spousal benefit when taken at 69 and 10 months is $7062 and is the same if DH takes at FRA 66.5 yrs. No change in my spousal benefit if DH takes at 69 and 10 months or FRA.

When you calculate the 8% increase from FRA to 70, it does not make sense because of my numbers. It is not cut and dried as everyone thinks.

For one, I am currently making $2714 more per year today than OSS says I'm making.

So together at 70 filings according to their strategy, our SS income is:
$65, 879. But it's really $68,593.

Filing at FRA according to OSS is $56,404. But really it's $56,404 + $2714. $59,118.

Might you post details in "Open Social Security" calculator: feature requests, bug reports, etc.?

If you do, either you or Mike Piper will learn something.

 

[pb4uski]

Might you post details in "Open Social Security" calculator: feature requests, bug reports, etc.?



If you do, either you or Mike Piper will learn something.

It looks like the differences are nominal vs real number differences. OSS is all real so if a couple years later one is comparing what you are actually collecting to what OSS predicted then the difference is likely due to COLA adjustments which OSS ignores because they are common to all alternatives.

 

[Midpack]

Not necessarily, I took mine at 65. And this is where I have an issue with opensocialsecurity.com. Based on my PIA ($1012/month), my annual benefit in the OSS calculation is $10,930. I am earning $13,644 per year. This is what goes into my checking account minus Medicare. My spousal benefit when taken at 69 and 10 months is $7062 and is the same if DH takes at FRA 66.5 yrs. No change in my spousal benefit if DH takes at 69 and 10 months or FRA.

When you calculate the 8% increase from FRA to 70, it does not make sense because of my numbers. It is not cut and dried as everyone thinks.

For one, I am currently making $2714 more per year today than OSS says I'm making.

So together at 70 filings according to their strategy, our SS income is:
$65, 879. But it's really $68,593.

Filing at FRA according to OSS is $56,404. But really it's $56,404 + $2714. $59,118.

I'm saying, plug your numbers into Firecalc and see the difference in the outcome after 30 years. My calculations show overall we are better off taking SS at DH FRA.

That’s odd. OSS gets my benefit calculation on the button with the Soc Sec estimate. I can’t understand how you could have a discrepancy that large (incorrect/old PIA?) but since we don’t know your particulars who knows.

 

[FREE866]

That’s odd. OSS gets my benefit calculation on the button with the Soc Sec estimate. I can’t understand how you could have a discrepancy that large (incorrect/old PIA?) but since we don’t know your particulars who knows.

Mine is spot on too.


Like someone else suggested, contact Mike Piper--he actually responds to emails.
mike@simplesubjects.com



Be curious his response.

 

[G-Man]

Using Firecalc and another calculator, financial mentor, both results based on an overall portfolio over 30 years come out better taking SS at FRA, for us anyway.

I'm not seeing a huge difference in the retirement planning software I'm using either (taking SS at FRA vs 70).

 

[camfused]

Ok, so I thought I would try too, with FireCalc. I got my numbers for DW and I from the ssa.gov site, for ages 67-70.

Plugging 77% (haircut) of each them into FC, I got:

Age Average Amount
67 $9.11M
68 $9.17M
69 $9.25M
70 $9.42M

So, for the average of 117 scenarios, age 70 comes out ahead (for us). I *still* standby our plan to pull at 70.

I had always assumed that the values at various ages take the extra 8% per year into account. Is that true? But, I would not think it takes COLAs into account, since that is unknown yet.