Argument for taking social security before 70. Comments?

Discuss all general (i.e. non-personal) investing questions and issues, investing news, and theory.
randomguy
Posts: 6504
Joined: Wed Sep 17, 2014 9:00 am

Re: Argument for taking social security before 70. Comments?

Post by randomguy » Wed Sep 12, 2018 1:14 pm

alexander29 wrote:
Wed Sep 12, 2018 12:37 pm
Hope we all live long lives, but the Brookings Institution Hamilton Project calculated that if you've already made it to 65 (which is good, because nearly a fifth of Americans die before that age):

Men have a 90 percent chance of living to 70, 62 percent of living to 80, and just 22 percent of living to 90.

For women it's 93 percent, 71 percent and 34 percent.

Here's the chart: http://www.hamiltonproject.org/ee-ce-im ... 729_80.jpg
Averages are not the right way to look at this as the deviations between populations is too high. Find a calculator that factors in weight, smoking and the like to more closer match your situation. Not going to claim this one (first google hit) is better or worse than others but here are some numbers from https://www.blueprintincome.com/tools/l ... ll-i-live/

65 years old white male, college educated, healthy makes 100k, doesn't smoke. exercises: 75% chance of living to 84. Life expectency of 92
65 years old white male, college educated, healthy makes 100k, smoke 2pakcs. has diabetes, doesn't exercises: 75% chance of living to 71. Life expectency of 77

Second dude should be taking SS asap. First dude needs to debate if the longevity insurance is worth the risk of having less money.

frankbrenowitz
Posts: 56
Joined: Wed Nov 19, 2014 9:53 pm

Re: Argument for taking social security before 70. Comments?

Post by frankbrenowitz » Wed Sep 12, 2018 2:19 pm

orlandoman wrote:
Wed Sep 12, 2018 10:48 am
So, many variables ... let's say you are married & your SS annual benefit would be $25,000 at age 62.

You decide to collect at age 70 (I'm just going to use 25K yr benefit for simplicity):
- age 62-63 would have collected $25,000
- age 63-64 would have collected $25,000
- age 64-65 would have collected $25,000
- age 65-66 would have collected $25,000
- age 66-67 would have collected $25,000
- age 67-68 would have collected $25,000
- you unexpectedly die at age 68

You & your spouse have lost the benefit/use of the $150,000 if you had claimed at age 62 ... not sure how you factor that variable in to a claiming age decision, so many variables.
Many scenarios exist where it as advantageous to claim social security at 62. If I was certain that my wife and I would unexpectedly die at age 68, I would without question claim at 62 and cheerfully withdraw 15% or more from my portfolio every year until that point.

However, I do not know the date of our collective demise so my planning is based on a reasonable assumption that one or the other of us is going to live to a good old age. For this reason, to assure that my SAHM DW is provided with the most secure financial circumstance possible after I die, I am delaying social security to age 70 to receive annual payments that are 76% higher, or $44,000 instead of $25,000. If we both pass before claiming, yes, we will have lost out on the benefit of the foregone payments, but we won't care, we're dead.

User avatar
Phineas J. Whoopee
Posts: 7598
Joined: Sun Dec 18, 2011 6:18 pm

Re: Argument for taking social security before 70. Comments?

Post by Phineas J. Whoopee » Wed Sep 12, 2018 2:58 pm

As others have alluded to, I think before deciding what to do, it's better to evaluate what one wants to accomplish.

OP and many replies in this thread seem to have the unstated objective of collecting as many Social Security retirement dollars as possible before death. If that's you, then fine.

There exist other goals, which cause different evaluations. For one thing, a person could be out of work, nobody's hiring people their age, and they need the money now or soon because of little savings, originally targeted for retirement or otherwise.

Another, and after I thought about what I want and my likely future financial situation (which could come out completely differently), I want to cover a (to me) reasonable lifestyle for as long as I personally live, not for as long as some average table says, or how long somebody else lives.

Accordingly, I saved a lot, and should be able to cover said lifestyle until claiming at 70. Most likely I'll still have savings then, but with the steady payments I might rely less on it. My bequest motive is to send everything that's left to selected charities. I'll live on it, then donate what's left. It could be nothing, but it could be a lot.

I agree married couples have more choices to make, but maybe the desire to get as many dollars paid out as possible could be jointly agreed, rather than simply assumed.

One more thing to point out: I say again I want to cover a (to me) reasonable lifestyle for as long as I personally live, not for as long as some average table says, or how long somebody else lives. If that happens I will have done what I wanted to do.

I know that if I am able to fund the lifestyle, and die at 69, say, after the event I will not feel really, really sad inside about having delayed my claim.

PJW

User avatar
Phineas J. Whoopee
Posts: 7598
Joined: Sun Dec 18, 2011 6:18 pm

Re: Argument for taking social security before 70. Comments?

Post by Phineas J. Whoopee » Wed Sep 12, 2018 3:02 pm

xb7 wrote:
Tue Sep 11, 2018 10:21 pm
Okay, I see a little spreadsheet work in my future --- I had been sort of taking it on faith that waiting to age 70 was the way to go.

One other wild card that I don't know how to 'weight' in the calculation is the possibility of the government changing the rules of the game as we age. I do NOT mean this as a political comment, and apologize if mentioning this violates anything (?) --- but there seems to me to be a very non-zero chance of some of the underlying rules/benefits changing as we go along. Thus perhaps someone starting to collect SS earlier would end up collecting some monies before such changes went into effect.

If anyone is willing to share a spreadsheet that they've already put together to show this break-even point date calculation, I'm sure others besides me would be interested.
If Congress does nothing to change them, the rules as presently ensconced in law result in the projected reduction in benefits sometime in the early- to mid- 2030s. It is changing the rules that could avoid such an outcome, should Congress decide to do so.

PJW

User avatar
willthrill81
Posts: 6173
Joined: Thu Jan 26, 2017 3:17 pm
Location: USA

Re: Argument for taking social security before 70. Comments?

Post by willthrill81 » Wed Sep 12, 2018 3:36 pm

alexander29 wrote:
Wed Sep 12, 2018 12:37 pm
Hope we all live long lives, but the Brookings Institution Hamilton Project calculated that if you've already made it to 65 (which is good, because nearly a fifth of Americans die before that age):

Men have a 90 percent chance of living to 70, 62 percent of living to 80, and just 22 percent of living to 90.

For women it's 93 percent, 71 percent and 34 percent.

Here's the chart: http://www.hamiltonproject.org/ee-ce-im ... 729_80.jpg
Surprisingly, a 65 year old male has a 20% chance of dying before age 74.

I think that many of the people worrying about the '4% rule' surviving for a full 30 years would do well to look at a life expectancy chart. It's very sobering.
“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
munemaker
Posts: 3527
Joined: Sat Jan 18, 2014 6:14 pm

Re: Argument for taking social security before 70. Comments?

Post by munemaker » Wed Sep 12, 2018 3:38 pm

I did not read through this entire thread, but when you did your spreadsheet analysis, did you consider both income streams, i.e. 1) to you and 2) to the surviving spouse?

I only ask because I have a friend who calculated his break even point but did not consider the survivor's benefit.

Dottie57
Posts: 4671
Joined: Thu May 19, 2016 5:43 pm

Re: Argument for taking social security before 70. Comments?

Post by Dottie57 » Wed Sep 12, 2018 3:39 pm

willthrill81 wrote:
Wed Sep 12, 2018 3:36 pm
alexander29 wrote:
Wed Sep 12, 2018 12:37 pm
Hope we all live long lives, but the Brookings Institution Hamilton Project calculated that if you've already made it to 65 (which is good, because nearly a fifth of Americans die before that age):

Men have a 90 percent chance of living to 70, 62 percent of living to 80, and just 22 percent of living to 90.

For women it's 93 percent, 71 percent and 34 percent.

Here's the chart: http://www.hamiltonproject.org/ee-ce-im ... 729_80.jpg
Surprisingly, a 65 year old male has a 20% chance of dying before age 74.

I think that many of the people worrying about the '4% rule' surviving for a full 30 years would do well to look at a life expectancy chart. It's very sobering.
Yes it is sobering. I think I will probably make it to SS at 70.

User avatar
Phineas J. Whoopee
Posts: 7598
Joined: Sun Dec 18, 2011 6:18 pm

Re: Argument for taking social security before 70. Comments?

Post by Phineas J. Whoopee » Wed Sep 12, 2018 3:43 pm

xb7 wrote:
Wed Sep 12, 2018 3:13 pm
It is changing the rules that could avoid such an outcome, should Congress decide to do so.
I was more thinking of rule changes in the other direction --- for example, means-testing perhaps, or at least income-testing, future adjustments for inflation, age of full retirement, etc.
...
Means testing exists, in an indirect way, as legislated by Congress. SS benefits are taxed more and more as income goes up, then plateau. The method we use for inflation adjustments, CPI-W, can only be changed by Congress, along with full retirement age.

What you seem to be thinking about are things only Congress can do.

In terms of deficits, one of the selling points that got the last major SS overhaul passed, in 1983, was it couldn't possibly add to the deficit. It raised taxes, and pushed back the retirement age, and made sure that the SSA can only pay out what it takes in, plus draw from its trust fund and spend the interest it earns there. The projected decrease in payouts is a purposefully designed-in feature, not a bug.

PJW

jalbert
Posts: 3834
Joined: Fri Apr 10, 2015 12:29 am

Re: Argument for taking social security before 70. Comments?

Post by jalbert » Wed Sep 12, 2018 4:04 pm

Enlightening comments for me in this thread. Thanks for posting. I did not realize that spousal benefits froze at FRA if they have less on their own work record
The spouse gets the larger of their own benefit and spousal benefit. Their benefit only increases after FRA if it is based on their own benefit snd they are delaying.

Also, if a spouse has their own benefit, but a spousal benefit is comparable or larger, the spouse may claim at age 62 and draw based on their own benefit, and will convert to a spousal benefit when the primary files. This reduces the cost of the primary delaying for some couples because of the primary's spouse being able to start collecting a benefit at age 62 while waiting for the primary's benefit to start.
Last edited by jalbert on Wed Sep 12, 2018 10:18 pm, edited 1 time in total.
Risk is not a guarantor of return.

User avatar
David Jay
Posts: 5684
Joined: Mon Mar 30, 2015 5:54 am
Location: Michigan

Re: Argument for taking social security before 70. Comments?

Post by David Jay » Wed Sep 12, 2018 4:07 pm

Watty wrote:
Wed Sep 12, 2018 12:24 am
When comparing strategies you have to be careful to not just look at to see what is best, but also how much better.
Agree. My age 70 is $2000 (lifetime!) over age 68.

Spouse is similar age, but my PIA is 4x her PIA. She files early.
Prediction is very difficult, especially about the future - Niels Bohr | To get the "risk premium", you really do have to take the risk - nisiprius

JackoC
Posts: 275
Joined: Sun Aug 12, 2018 11:14 am

Re: Argument for taking social security before 70. Comments?

Post by JackoC » Wed Sep 12, 2018 5:13 pm

willthrill81 wrote:
Wed Sep 12, 2018 3:36 pm
alexander29 wrote:
Wed Sep 12, 2018 12:37 pm
Hope we all live long lives, but the Brookings Institution Hamilton Project calculated that if you've already made it to 65 (which is good, because nearly a fifth of Americans die before that age):

Men have a 90 percent chance of living to 70, 62 percent of living to 80, and just 22 percent of living to 90.

For women it's 93 percent, 71 percent and 34 percent.

Here's the chart: http://www.hamiltonproject.org/ee-ce-im ... 729_80.jpg
Surprisingly, a 65 year old male has a 20% chance of dying before age 74.

I think that many of the people worrying about the '4% rule' surviving for a full 30 years would do well to look at a life expectancy chart. It's very sobering.
To somewhat repeat Randomguy, see post above, general life expectancy tables are not that useful for a given individual, especially not in the socio-demographic trend of the US now where some sub-groups are undergoing a *declining* life expectancy as that of other sub-groups keeps rising.

When I put my particulars into Randomguy's link it says my LE is 93, if I only say my health is 'good'. I've read that people in Japan have a lower average assessment of their own health than Americans do, though of course a notably longer national average life expectancy. I think I might also tend to underestimate my health. If I say 'very good' which is reasonably arguable, LE=96. Male early 60's, good (as in higher LE) answers to all the questions. Play around with plausible other combinations, men who are not unusual in the general population, and you can shave 20-25 yrs off easily.

Believe those particular numbers or not, it's clearly highly variable by personal particulars. A lot of people think it depends a lot on close relatives' longevity (genetically) but it's notable the linked test doesn't even ask. Some others do but AIUI it's a pretty low weighting, family history as distinct from whether family members followed the same healthy/unhealthy choices and/or were in the same shorter/longer lived segments of US society socio-economically.

Anyway I will not take SS any earlier than 70 unless my health goes south before that. As PJW's excellent post above outlined, you must first identify your goal. My goal is not to maximize my take from SS on average across all scenarios. It's to have a relatively big payout in scenario's where I live longer than expected and my investments perform less well than expected, so I actually need Social Security: downside protection.
Last edited by JackoC on Wed Sep 12, 2018 5:19 pm, edited 1 time in total.

grok87
Posts: 8454
Joined: Tue Feb 27, 2007 9:00 pm

Re: Argument for taking social security before 70. Comments?

Post by grok87 » Wed Sep 12, 2018 5:15 pm

MIretired wrote:
Tue Sep 11, 2018 11:07 pm
duffer wrote:
Tue Sep 11, 2018 9:21 pm
The following is copied from a post by an investor at the the Fidelity website. I assume I am not violating any rules by reproducing it here. The author makes an argument for taking social security benefits at age 66. It would be very interesting to hear what some of the sharp minds here have to say about his argument:

I was firmly convinced that delaying Social Security benefits until age 70 was the smart thing to do … until I actually did the math. Now I intend to start benefits next year at age 66.

There are SO many advantages to deferring benefits:
Between ages 66 and 70, your benefit grows roughly 8% a year. I could use those 4 deferral years to maximize my Roth conversions which would reduce the amount of taxes that I paid on my SS benefits later because my RMDs would be less. If I died before my wife, she would receive higher survivor benefits if I claimed at age 70 instead of age 66.

But a couple of hours banging around in an Excel spreadsheet changed my mind. Everyone is different and “your mileage may vary”, but you might find my results interesting. I have come to understand that a key factor in my situation (but not unlike many others) is that my wife has a limited work history and is only eligible for spousal benefits on my work record. I have always read that the “breakeven” age (the age where you catch up on lost benefits when claiming later) is typically the early 80’s. But in my situation this was not the case. My breakeven age was 88. And it didn’t matter what Full Retirement Age benefits I used in the spreadsheet. If I used a 0% return on investment, the breakeven age was always right at 88. If I used a 2% return on investment, the breakeven age was always right at 94. If I used any return great than 2%, neither I nor my wife could possibly live long enough to recoup the loss of waiting to claim at age 70.

Profuse head scratching brought me to two realizations.

When the IRS says that you will receive the same amount of lifetime benefits regardless of the age that you claim (assuming you die when the tables say you should), they mean exactly that. But what they don’t say is that the increased benefit at age 70 does NOT compensate you for the time value of money. It only compensates you for the four years of missed benefits, not the compound rate of return that you could have gotten on those benefits.

Spousal benefits max out at age 66. They do NOT increase at 8% a year between ages 66 and 70 like workers benefits. So by deferring until age 70, our total benefit is not 32% higher but only 21% higher (or slightly over 5% a year).

So then I began doing “what ifs” to see how increased age 70 spousal benefits would affect the breakeven age. To my amazement I could die at any age between 70 and 95 and it would not affect the breakeven age at all. If my assumed return was 0%, breakeven age was 88. If my assumed return was 2%, breakeven age was 94. This defies conventional logic … but I cannot find an error in the calcs.

The next thing I did was use my tax spreadsheet to compare ending (age 95) portfolio value using age 66 benefits vs age 70 benefits. This would tell me how not being able to maximize Roth conversions for the next 4 years would affect my situation. After a bit of spreadsheet modeling I determined that taking benefits at age 66 was less desirable if my assumed investment return was less than 2%. Any return greater than 2% made taking benefits at age 66 more desirable. 2% is a pretty low hurdle even for me.

So there you have it. Even as stubborn as I am, I have changed my mind about when to start taking SS benefits. Probably the most important takeaway is “Don’t assume … do the math”. I would really love it if one of you other spreadsheet geeks (chuckfinn?) could corroborate or refute my findings.
There's another thread that says delaying to 70 And spend more money starting at 62. If you want to see some other math.
viewtopic.php?t=102609 .
Thanks
Keep calm and Boglehead on. KCBO.

User avatar
willthrill81
Posts: 6173
Joined: Thu Jan 26, 2017 3:17 pm
Location: USA

Re: Argument for taking social security before 70. Comments?

Post by willthrill81 » Wed Sep 12, 2018 5:17 pm

JackoC wrote:
Wed Sep 12, 2018 5:13 pm
willthrill81 wrote:
Wed Sep 12, 2018 3:36 pm
alexander29 wrote:
Wed Sep 12, 2018 12:37 pm
Hope we all live long lives, but the Brookings Institution Hamilton Project calculated that if you've already made it to 65 (which is good, because nearly a fifth of Americans die before that age):

Men have a 90 percent chance of living to 70, 62 percent of living to 80, and just 22 percent of living to 90.

For women it's 93 percent, 71 percent and 34 percent.

Here's the chart: http://www.hamiltonproject.org/ee-ce-im ... 729_80.jpg
Surprisingly, a 65 year old male has a 20% chance of dying before age 74.

I think that many of the people worrying about the '4% rule' surviving for a full 30 years would do well to look at a life expectancy chart. It's very sobering.
To somewhat repeat Randomguy, see post above, general life expectancy tables are not that useful for a given individual, especially not in the socio-demographic trend of the US now where some sub-groups are undergoing a *declining* life expectancy as that of other sub-groups keeps rising.

When I put my particulars into Randomguy's link it says my LE is 93, if I only say my health is 'good'. I've read that people in Japan have a lower average assessment of their own health than Americans do, though of course a notably longer national average life expectancy. I think I might also tend to underestimate my health. If I say 'very good' which is reasonably arguable, LE=96. Male early 60's, good answers to all the questions. Play around with plausible other combinations, men who are not unusual in the general population, and you can shave 20-25 yrs off easily.

Believe those particular numbers or not, it's clearly highly variable by personal particulars. A lot of people think it depends a lot on close relatives' longevity (genetically) but it's notable the linked test doesn't even ask. Some others do but AIUI it's a pretty low weighting, family history as distinct from whether family members followed the same healthy/unhealthy choices and/or were in the same shorter/longer lived segments of US society socio-economically.

Anyway I will not take SS any earlier than 70 unless my health goes south before that. As PJW's excellent post above outlined, you must first identify your goal. My goal is not to maximize my take from SS on average across all scenarios. It's to have a relatively big payout in scenario's where I live longer than expected and my investments perform less well than expected, so I actually need Social Security: downside protection.
The numbers I quoted are based on the SSA's actuarial tables and are, strictly speaking, very accurate. Yes, individuals can get a more accurate estimate of their life expectancy by including other factors beyond age and gender, but that doesn't change the accuracy of the general numbers.

I agree that individuals might be well served to estimate their own life expectancy as accurately as they can, keeping in mind that as an individual, you are a sample size of one, and the law of large numbers doesn't apply to you. Case in point, my father-in-law passed away at age 65. One of my grandfathers died at age 43 (car accident), and one of my grandmothers died at age 47 (botched hysterectomy).
“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

JackoC
Posts: 275
Joined: Sun Aug 12, 2018 11:14 am

Re: Argument for taking social security before 70. Comments?

Post by JackoC » Wed Sep 12, 2018 5:37 pm

willthrill81 wrote:
Wed Sep 12, 2018 5:17 pm
JackoC wrote:
Wed Sep 12, 2018 5:13 pm
To somewhat repeat Randomguy, see post above, general life expectancy tables are not that useful for a given individual, especially not in the socio-demographic trend of the US now where some sub-groups are undergoing a *declining* life expectancy as that of other sub-groups keeps rising.

When I put my particulars into Randomguy's link it says my LE is 93, if I only say my health is 'good'. I've read that people in Japan have a lower average assessment of their own health than Americans do, though of course a notably longer national average life expectancy. I think I might also tend to underestimate my health. If I say 'very good' which is reasonably arguable, LE=96. Male early 60's, good answers to all the questions. Play around with plausible other combinations, men who are not unusual in the general population, and you can shave 20-25 yrs off easily.

Believe those particular numbers or not,
1. The numbers I quoted are based on the SSA's actuarial tables and are, strictly speaking, very accurate. Yes, individuals can get a more accurate estimate of their life expectancy by including other factors beyond age and gender, but that doesn't change the accuracy of the general numbers.

2. I agree that individuals might be well served to estimate their own life expectancy as accurately as they can, keeping in mind that as an individual, you are a sample size of one, and the law of large numbers doesn't apply to you. Case in point, my father-in-law passed away at age 65. One of my grandfathers died at age 43 (car accident), and one of my grandmothers died at age 47 (botched hysterectomy).
1. Nobody is questioning the accuracy of SSA's numbers as presented, just their applicability.

2. This jumps too far, from actuarial tables for the whole population on one hand, to the random events which will determine each of our actual dates of death on the other. In between there are systematic differences in life expectancy between identifiable *sub groups* in society, not just between individuals. Those sub-group tables (which is effectively what that links gives you, what it's effectively reading from) might be more subject to questions of accuracy, not necessarily inaccurate but it's a question that could reasonably be raised. Whereas again SSA's table is more undeniably accurate as to what it presents, but seriously compromised as a tool for individual decision by stratified patterns of longevity in US society, and increasingly so. People of sub groups with 'good' (again I emphasize, good with regard to LE, not as any other value judgement) answers to the questions in links like the one above should not be making decisions based on the SSA table. Nor should people with particularly bad answers.

Discussion here AIUI should be focused on actions individuals can take wrt their finances, not issues like the overall health and longevity of the population, important as that is in other settings. The SSA table is not a good tool to decide individual actions wrt finance as it applies to longevity.

User avatar
willthrill81
Posts: 6173
Joined: Thu Jan 26, 2017 3:17 pm
Location: USA

Re: Argument for taking social security before 70. Comments?

Post by willthrill81 » Wed Sep 12, 2018 5:53 pm

JackoC wrote:
Wed Sep 12, 2018 5:37 pm
The SSA table is not a good tool to decide individual actions wrt finance as it applies to longevity.
I largely agree with you, although I wouldn't say that it "is not a good tool" but rather that it is "not the best tool."

My initial assertion was merely that there appears to be a higher than typically expected likelihood of one's dying a relatively early death. Many posters here seem worried about living into their 90s or beyond when the overall likelihood of that happening for most is small. Of course, individualized information improves one's ability to make good decisions, but we cannot discuss all of the relevant factors in play when making generalizable statements. Somewhere along the way, we must aggregate things for the purposes of discussion.
“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
LadyGeek
Site Admin
Posts: 48687
Joined: Sat Dec 20, 2008 5:34 pm
Location: Philadelphia
Contact:

Re: Argument for taking social security before 70. Comments?

Post by LadyGeek » Wed Sep 12, 2018 6:39 pm

I removed an off-topic comment and reply regarding means-testing, deficit spending, and the future of Social Security. As a reminder, see: Politics and Religion
In order to avoid the inevitable frictions that arise from these topics, political or religious posts and comments are prohibited. The only exceptions to this rule are:
  • Common religious expressions such as sending your prayers to an ailing member.
  • Usage of factual and non-derogatory political labels when necessary to the discussion at hand.
  • Discussions about enacted laws or regulations that affect the individual investor. Note that discussions of proposed legislation are prohibited.
  • Proposed regulations that are directly related to investing may be discussed if and when they are published for public comments.
Changing Social Security requires a change in legislation and is therefore a political discussion (proposed legislation).
Wiki To some, the glass is half full. To others, the glass is half empty. To an engineer, it's twice the size it needs to be.

JoeRetire
Posts: 1651
Joined: Tue Jan 16, 2018 2:44 pm

Re: Argument for taking social security before 70. Comments?

Post by JoeRetire » Wed Sep 12, 2018 7:00 pm

duffer wrote:
Tue Sep 11, 2018 9:21 pm
Probably the most important takeaway is “Don’t assume … do the math”.
I agree with this part. The rest, not so much.

My math reached different conclusions than his. And since he doesn't factor survivor's benefits into his math at all and he doesn't include enough details to check his math, I'll just go with mine.

JoeRetire
Posts: 1651
Joined: Tue Jan 16, 2018 2:44 pm

Re: Argument for taking social security before 70. Comments?

Post by JoeRetire » Wed Sep 12, 2018 7:04 pm

David Jay wrote:
Wed Sep 12, 2018 4:07 pm
My age 70 is $2000 (lifetime!) over age 68.
Are you sure? At what age do you project your lifetime to end?

grok87
Posts: 8454
Joined: Tue Feb 27, 2007 9:00 pm

Re: Argument for taking social security before 70. Comments?

Post by grok87 » Wed Sep 12, 2018 7:14 pm

alexander29 wrote:
Wed Sep 12, 2018 12:37 pm
Hope we all live long lives, but the Brookings Institution Hamilton Project calculated that if you've already made it to 65 (which is good, because nearly a fifth of Americans die before that age):

Men have a 90 percent chance of living to 70, 62 percent of living to 80, and just 22 percent of living to 90.

For women it's 93 percent, 71 percent and 34 percent.

Here's the chart: http://www.hamiltonproject.org/ee-ce-im ... 729_80.jpg
i would up those probabilities. it's based on 2005 data. there is a trend toward longer longevity.
Keep calm and Boglehead on. KCBO.

User avatar
David Jay
Posts: 5684
Joined: Mon Mar 30, 2015 5:54 am
Location: Michigan

Re: Argument for taking social security before 70. Comments?

Post by David Jay » Wed Sep 12, 2018 7:38 pm

JoeRetire wrote:
Wed Sep 12, 2018 7:04 pm
David Jay wrote:
Wed Sep 12, 2018 4:07 pm
My age 70 is $2000 (lifetime!) over age 68.
Are you sure? At what age do you project your lifetime to end?
Yes. Because of the spousal issues I laid out in the portion you snipped. This is a combined analysis for two spouses:
Spouse is similar age, but my PIA is 4x her PIA. She files early.
Last edited by David Jay on Wed Sep 12, 2018 7:40 pm, edited 1 time in total.
Prediction is very difficult, especially about the future - Niels Bohr | To get the "risk premium", you really do have to take the risk - nisiprius

radiowave
Posts: 1830
Joined: Thu Apr 30, 2015 5:01 pm

Re: Argument for taking social security before 70. Comments?

Post by radiowave » Wed Sep 12, 2018 7:40 pm

Angst wrote:
Wed Sep 12, 2018 8:00 am
radiowave wrote:
Tue Sep 11, 2018 10:07 pm
My calculated break even at 66 vs 70 came out to 82 yrs and 68 vs 70 breakeven was 95. So I'll probably cut the difference and retire at 67 :)
My suspician is that you need to check your math, unless what you're doing simply escapes me - granted, I have not crunched any of the numbers, but what you say above seems counter-intuitive: Shouldn't the breakeven age fall, not rise as the comparison year rises closer to 70? E.g., if you calculated 69.5 vs 70, the breakeven age is not going to be, say 101, it ought to be close to 70.
I based my calculation on cumulative sum for each starting in the year I turned 66/70 or 68/70. And it does go up the closer i get hence breakeven 68 vs. 70 is 95
Bogleheads Wiki: https://www.bogleheads.org/wiki/Main_Page

User avatar
grabiner
Advisory Board
Posts: 22908
Joined: Tue Feb 20, 2007 11:58 pm
Location: Columbia, MD

Re: Argument for taking social security before 70. Comments?

Post by grabiner » Wed Sep 12, 2018 8:43 pm

Here's how I estimate the break-even. It's easiest to do one year at a time.

Suppose that you have enough money to cover this year's Social Security. You can spend the money this year and delay Social Security to next year, or take Social Security now and buy a TIPS ladder which will cover the future years' payments. The break-even is the point where the TIPS ladder runs out.

If you are 66 and your benefit is $20,000, waiting one year will increase your benefit to $21,600 (all numbers growing with inflation). If you spend $20,000 to buy TIPS, you could get $1600 per year for 12.5 years if the TIPS yields were zero; this would make the break-even 13.5 years. Current yields on a TIPS portfolio with a 7-year duration are 0.85%, which is enough to make the ladder last 13.3 years, for a break-even of 14.3 years. That is to age 80.

If you are 69 and your benefit is $24,800, waiting one year will increase your benefit to $26,400. If you spend $24,800 to buy TIPS, you could get $1600 per year for 15.5 years if the TIPS yields were zero, and 16.5 years at current yields, making the break-even 17.5 years; that is to age 87. 17.5 years is just about the life expectancy for a 69-year-old man, and less than the life expectancy for a 69-year-old woman.

Now, these are average life expectancies. Most people either have higher-than-average life expectancies, or know that they don't (because they have heart disease, cancer, etc.) This makes waiting more attractive.

And even if you have an average life expectancy, there is an advantage for waiting. If you wait, you will have more money if you live longer, and thus have a greater need for the money.

My calculations are only for singles, though. For married couples, the effect on spousal and widow/er's benefits is important. If your spouse is likely to collect widow/er's benefits on your record, that is a strong incentive for you to delay, as the break-even point is based on when you both die. Conversely, if you will switch to a widow/er's benefit if your spouse dies first, the break-even point is when either of you die, which makes claiming at full retirement age likely better.
Wiki David Grabiner

User avatar
#Cruncher
Posts: 2686
Joined: Fri May 14, 2010 2:33 am
Location: New York City
Contact:

Re: Argument for taking social security before 70. Comments?

Post by #Cruncher » Wed Sep 12, 2018 8:50 pm

bradpevans wrote:
Wed Sep 12, 2018 8:45 am
I don't know if the calculations included any investment returns. Ignoring that (i.e. put all your SS money under the pillow), ...
Since SS benefits are inflation-indexed, to not "include any investment returns" is not the same as to "put all your SS money under the pillow". The latter would represent a negative return in real terms. To get a 0% real return requires an investment that matches inflation. E.g., I Bonds with a 0% fixed rate.

Beehave wrote:
Wed Sep 12, 2018 10:03 am
The month one turns 66 and the month one turns 70 are the two most "expensive" (or, put another way, the two most costly) months to delay. That does not mean that these months are wrong or bad, but from a cost perspective (what you pay and what you get) they make the least sense of all.
By "expensive" I assume you mean a small increase versus starting SS the month before. Which particular months have the smallest increase depends on one's Normal Retirement Age (NRA) and the following reductions or additions for claiming before or after it.

Code: Select all

-5/1200 per month: 37 months or more before NRA
-5/900  per month: 1 - 36 months before NRA
+8/1200 per month: each month after NRA
For someone with a NRA of 66, delaying one month to age 70 does indeed have the smallest percentage increase versus claiming a month earlier. But delaying one month to age 63 has the second smallest increase, not delaying one month to age 66.

Code: Select all

  Claim   Mo vs   Pct of
   Age     NRA      PIA     Increase
--------  -----  --------   --------
62+11 mo   -37    79.583%
63         -36    80%         +0.52%
65+11 mo    -1    99.444%
66           0   100%         +0.56%
69+11      +47   131.333%
70         +48   132%         +0.51%
To see all the monthly increases for age 62 to 70 or to see them for NRAs other than 66, Select All, Copy, and Paste [*] the following at cell A1 of a blank Excel sheet. Then format the cells for readability, enter your year of birth in cell B1, and copy the formulas down in row 6.

Code: Select all

Born	1952
NRA	=MIN(67,66+MAX(0,B1-1954)/6)
Start	Pct PIA	Increase
62	=IF($A4<B$2,1-(5/900)*MIN(36,(B$2-$A4)*12)-(5/1200)*MAX(0,(B$2-$A4)*12-36),1+(8/1200)*($A4-B$2)*12)
=A4+1/12	=IF($A5<B$2,1-(5/900)*MIN(36,(B$2-$A5)*12)-(5/1200)*MAX(0,(B$2-$A5)*12-36),1+(8/1200)*($A5-B$2)*12)	=B5/B4-1
=2*A5-A4	=IF($A6<B$2,1-(5/900)*MIN(36,(B$2-$A6)*12)-(5/1200)*MAX(0,(B$2-$A6)*12-36),1+(8/1200)*($A6-B$2)*12)	=B6/B5-1

JW-Retired wrote:
Wed Sep 12, 2018 10:20 am
I'm just now seeing this argument is from the Fidelity website investor and not you duffer. That would seem to partly explain why the guy is leaving out the 2 most important reasons to delay social security benefits to age 70. One being your spouse survivor benefits … The guy mentions that spousal benefits don't increase for the spouse taking the benefit past full retirement age, but doesn't mention that survivor benefits do keep on growing ...
The Fidelity website poster does mention it in the original post:
Fidelity poster quoted in original post wrote:If I died before my wife, she would receive higher survivor benefits if I claimed at age 70 instead of age 66.
And he does assume so in his calculations. See my previous post.

alexander29 wrote:
Wed Sep 12, 2018 12:37 pm
... the Brookings Institution Hamilton Project calculated that if you've already made it to 65 … :
Men have a 90 percent chance of living to 70, 62 percent of living to 80, and just 22 percent of living to 90.
For women it's 93 percent, 71 percent and 34 percent.
For married couples, it's useful to know the probability that either one of the couple will be alive. Here is how it can be computed from the individual probabilities.

Code: Select all

ProbE = 1 - (1 - Prob1) * (1 - Prob2)
  49% = 1 - (1 - 0.22)  * (1 - 0.34)  living to 90
* If you have trouble pasting, try "Paste Special" and "Text".

TravelforFun
Posts: 1521
Joined: Tue Dec 04, 2012 11:05 pm

Re: Argument for taking social security before 70. Comments?

Post by TravelforFun » Wed Sep 12, 2018 9:29 pm

'YMMV' is exactly true. In my case, my wife is 62 and I'm 66 and the significantly higher income earner. All the calculators and my own spreadsheet show my wife should claim her benefits at 62, I should claim my spousal benefits at 66, then I would claim my own benefits at 70 and she would switch to spousal right after that. And that's the path we're on. My wife and I have been received her reduced benefits and my spousal benefits. Then we will make the switch 4 years from now. After I pass, she would get my delayed benefits. Everything is good!

TravelforFun

Yooper16
Posts: 106
Joined: Mon Aug 21, 2017 7:44 am

Re: Argument for taking social security before 70. Comments?

Post by Yooper16 » Wed Sep 12, 2018 9:44 pm

grabiner wrote:
Wed Sep 12, 2018 8:43 pm
Here's how I estimate the break-even. It's easiest to do one year at a time.


My calculations are only for singles, though. For married couples, the effect on spousal and widow/er's benefits is important. If your spouse is likely to collect widow/er's benefits on your record, that is a strong incentive for you to delay, as the break-even point is based on when you both die. Conversely, if you will switch to a widow/er's benefit if your spouse dies first, the break-even point is when either of you die, which makes claiming at full retirement age likely better.
Would the highlighted statement then suggest that if your spouse in not able to collect a widowers benefit, that collecting at or about FRA is probably the better route?

That is the question we have to answer. Both of us are turning 65 next year. FRA for me is 66 with my spouse receiving an old style federal pension.

User avatar
grabiner
Advisory Board
Posts: 22908
Joined: Tue Feb 20, 2007 11:58 pm
Location: Columbia, MD

Re: Argument for taking social security before 70. Comments?

Post by grabiner » Wed Sep 12, 2018 9:51 pm

Yooper16 wrote:
Wed Sep 12, 2018 9:44 pm
grabiner wrote:
Wed Sep 12, 2018 8:43 pm
Here's how I estimate the break-even. It's easiest to do one year at a time.


My calculations are only for singles, though. For married couples, the effect on spousal and widow/er's benefits is important. If your spouse is likely to collect widow/er's benefits on your record, that is a strong incentive for you to delay, as the break-even point is based on when you both die. Conversely, if you will switch to a widow/er's benefit if your spouse dies first, the break-even point is when either of you die, which makes claiming at full retirement age likely better.
Would the highlighted statement then suggest that if your spouse in not able to collect a widowers benefit, that collecting at or about FRA is probably the better route?

That is the question we have to answer. Both of us are turning 65 next year. FRA for me is 66 with my spouse receiving an old style federal pension.
"Old-style federal pension" implies CSRS. If so, your spouse will probably not get any spousal or widow/er's benefit, because the Government Pension Offset will negate the benefit. (The offset is 2/3 of the pension, so there will be no SS benefit at all if the pension is at least 1.5 times the SS benefit.) In that situation, you should optimize Social Security as if you were single.
Wiki David Grabiner

Yooper16
Posts: 106
Joined: Mon Aug 21, 2017 7:44 am

Re: Argument for taking social security before 70. Comments?

Post by Yooper16 » Wed Sep 12, 2018 10:12 pm

grabiner wrote:
Wed Sep 12, 2018 9:51 pm
Yooper16 wrote:
Wed Sep 12, 2018 9:44 pm
grabiner wrote:
Wed Sep 12, 2018 8:43 pm
Here's how I estimate the break-even. It's easiest to do one year at a time.


My calculations are only for singles, though. For married couples, the effect on spousal and widow/er's benefits is important. If your spouse is likely to collect widow/er's benefits on your record, that is a strong incentive for you to delay, as the break-even point is based on when you both die. Conversely, if you will switch to a widow/er's benefit if your spouse dies first, the break-even point is when either of you die, which makes claiming at full retirement age likely better.
Would the highlighted statement then suggest that if your spouse in not able to collect a widowers benefit, that collecting at or about FRA is probably the better route?

That is the question we have to answer. Both of us are turning 65 next year. FRA for me is 66 with my spouse receiving an old style federal pension.
"Old-style federal pension" implies CSRS. If so, your spouse will probably not get any spousal or widow/er's benefit, because the Government Pension Offset will negate the benefit. (The offset is 2/3 of the pension, so there will be no SS benefit at all if the pension is at least 1.5 times the SS benefit.) In that situation, you should optimize Social Security as if you were single.
Thanks It is old time CSRS.

My FRA SS is a bit south of 20K-- his pension is a bit north of 40K-- so no spousal for him. I do get the survivors on his CSRS.
We both have some longevity, my side is a few years longer. I have more earlier deaths than on his side going with aunts, uncles and grandparent other direct blood ancestors, such as great aunts etc.

I don't anticipate that I will file at 65 but will not delay much(if at all) past FRA. I will probably file beginning Jan 2020 anyway that is the present thought. We are drawing down on a small inherited IRA an amount equal to about 75% of my FRA. This IRA will be zeroed by next early fall 2019.

JackoC
Posts: 275
Joined: Sun Aug 12, 2018 11:14 am

Re: Argument for taking social security before 70. Comments?

Post by JackoC » Wed Sep 12, 2018 11:18 pm

willthrill81 wrote:
Wed Sep 12, 2018 5:53 pm
JackoC wrote:
Wed Sep 12, 2018 5:37 pm
The SSA table is not a good tool to decide individual actions wrt finance as it applies to longevity.
I largely agree with you, although I wouldn't say that it "is not a good tool" but rather that it is "not the best tool."

My initial assertion was merely that there appears to be a higher than typically expected likelihood of one's dying a relatively early death. Many posters here seem worried about living into their 90s or beyond when the overall likelihood of that happening for most is small. Of course, individualized information improves one's ability to make good decisions, but we cannot discuss all of the relevant factors in play when making generalizable statements. Somewhere along the way, we must aggregate things for the purposes of discussion.
We need to aggregate, but using stats relevant to the group we're talking about. You say "Many posters here seem worried about living into their 90s or beyond when the overall likelihood of that happening for most is small."
I believe that's significantly inaccurate for 'living into their 90's' (not necessarily inaccurate for 'beyond') when it comes to 'posters here', as opposed to 'Americans'. LE is highly correlated with age (older people beyond relatively earlier causes of death including ones mainly responsible for the *decline* in LE in some US sub groups in recent years), education, socio-econ and other demographics, besides individual lifestyle choices (but which also heavily correlate w/ demographics). It's not just a matter of one person's v another's individual situation but systemic differences in subdivisions of society (I'm not neglecting that some posters here are not from the US :D , but it's not the main statistical issue I see).

We don't have to discuss everyone's individual factors to say that for 'posters here' the chance of living to the 90's probably isn't actually that small. 22% chance of reaching 90 for US'ian male age 65 was quoted. The individualized link said mine is over 50% (age 61 US male LE=93+). I guess the forum (adjusting for male v female) average might be closer to mine than the US average, maybe even similar to mine, I don't know. But I believe we're talking about serious differences in likelihood, otherwise I wouldn't persist with the point. The SSA table is probably not close enough for useful, other than hypothetical, discussion in a group which is not closely representative of the general US population, and the group 'posters here' is not from all I can tell.

bradpevans
Posts: 340
Joined: Sun Apr 08, 2018 1:09 pm

Re: Argument for taking social security before 70. Comments?

Post by bradpevans » Wed Sep 12, 2018 11:46 pm

radiowave wrote:
Wed Sep 12, 2018 7:40 pm
Angst wrote:
Wed Sep 12, 2018 8:00 am
radiowave wrote:
Tue Sep 11, 2018 10:07 pm
My calculated break even at 66 vs 70 came out to 82 yrs and 68 vs 70 breakeven was 95. So I'll probably cut the difference and retire at 67 :)
My suspician is that you need to check your math, unless what you're doing simply escapes me - granted, I have not crunched any of the numbers, but what you say above seems counter-intuitive: Shouldn't the breakeven age fall, not rise as the comparison year rises closer to 70? E.g., if you calculated 69.5 vs 70, the breakeven age is not going to be, say 101, it ought to be close to 70.
I based my calculation on cumulative sum for each starting in the year I turned 66/70 or 68/70. And it does go up the closer i get hence breakeven 68 vs. 70 is 95
I think there is an error here. 68 gets 24 more payments than 70. But 70 get payments of 1.16x rather than just x, and has to catch up 24x, so it’s 150 months, or 12.5 years

rgs92
Posts: 2239
Joined: Mon Mar 02, 2009 8:00 pm

Re: Argument for taking social security before 70. Comments?

Post by rgs92 » Thu Sep 13, 2018 12:06 am

For most people, you have a choice between starting SS payments or drawing down your nest egg each month.

I sure would not like the idea of letting a nest egg get down to a perilously low level (a personal gut feeling amount that may be different for each person). (It's not a great idea to drive around with the gas tank near empty.)

So I think if one's portfolio drops to some uncomfortable level, then it is time to start social security regardless of age.

(And your nest egg could sink for any reason, like a market dip or a need for a lump of money that was not foreseen.)

I believe that most people will do this rather than all the spreadsheet calculations that are basically theoretical with lots of underlying assumptions that are, well, just assumptions (on things like rates of return in a relatively short period).

And also, the calculations of how much you will receive in total at a far-off (or even sooner) age are also just theoretical figures that maybe are not so relevant or of practical use. If you are 80 years old, are you really going to be looking back at some total amount of income you got over the last 15 years and saying to yourself, wow, that amount could have been 20% higher?

Big Dog
Posts: 884
Joined: Mon Sep 07, 2015 4:12 pm

Re: Argument for taking social security before 70. Comments?

Post by Big Dog » Thu Sep 13, 2018 12:22 am

looked at another way, per vanguard:
If the man and woman are married, ... a 45% chance that one will live to age 90.

JoeRetire
Posts: 1651
Joined: Tue Jan 16, 2018 2:44 pm

Re: Argument for taking social security before 70. Comments?

Post by JoeRetire » Thu Sep 13, 2018 6:24 am

David Jay wrote:
Wed Sep 12, 2018 7:38 pm
JoeRetire wrote:
Wed Sep 12, 2018 7:04 pm
David Jay wrote:
Wed Sep 12, 2018 4:07 pm
My age 70 is $2000 (lifetime!) over age 68.
Are you sure? At what age do you project your lifetime to end?
Yes. Because of the spousal issues I laid out in the portion you snipped. This is a combined analysis for two spouses:
Spouse is similar age, but my PIA is 4x her PIA. She files early.
And the number you use for "lifetime" is... ?

User avatar
David Jay
Posts: 5684
Joined: Mon Mar 30, 2015 5:54 am
Location: Michigan

Re: Argument for taking social security before 70. Comments?

Post by David Jay » Thu Sep 13, 2018 9:30 am

JoeRetire wrote:
Thu Sep 13, 2018 6:24 am
David Jay wrote:
Wed Sep 12, 2018 7:38 pm
JoeRetire wrote:
Wed Sep 12, 2018 7:04 pm
David Jay wrote:
Wed Sep 12, 2018 4:07 pm
My age 70 is $2000 (lifetime!) over age 68.
Are you sure? At what age do you project your lifetime to end?
Yes. Because of the spousal issues I laid out in the portion you snipped. This is a combined analysis for two spouses:
Spouse is similar age, but my PIA is 4x her PIA. She files early.
And the number you use for "lifetime" is... ?
opensocialsecurity does not use a single date. As I understand it, Mike’s strategy is to use the spectrum of potential lifespans and weight by likelihood of surviving to various ages.

Without re-running the numbers is seems like both the age 68 and the age 70 filing had present values of just under $780,000 (like 776k and 778K, respectively).
Prediction is very difficult, especially about the future - Niels Bohr | To get the "risk premium", you really do have to take the risk - nisiprius

randomguy
Posts: 6504
Joined: Wed Sep 17, 2014 9:00 am

Re: Argument for taking social security before 70. Comments?

Post by randomguy » Thu Sep 13, 2018 10:01 am

willthrill81 wrote:
Wed Sep 12, 2018 5:53 pm
JackoC wrote:
Wed Sep 12, 2018 5:37 pm
The SSA table is not a good tool to decide individual actions wrt finance as it applies to longevity.
I largely agree with you, although I wouldn't say that it "is not a good tool" but rather that it is "not the best tool."

My initial assertion was merely that there appears to be a higher than typically expected likelihood of one's dying a relatively early death. Many posters here seem worried about living into their 90s or beyond when the overall likelihood of that happening for most is small. Of course, individualized information improves one's ability to make good decisions, but we cannot discuss all of the relevant factors in play when making generalizable statements. Somewhere along the way, we must aggregate things for the purposes of discussion.
When making YOUR choice, you need to think about YOUR situation not the average one. The likely hood of hitting 90 is about 50% for college educated, white male who doesn't smoke, has no know health issues, and isn't obese. That isn't exactly a small probability.

And yes at the other end there are bunches of people who should think about claiming ASAP cause their odds of making to 80 are low. If you have had 2 heart attacks and a stroke, the fact the average life expectancy is like 84 isn't really applicable. You will be lucky to make it to 75.

User avatar
ObliviousInvestor
Posts: 3395
Joined: Tue Mar 17, 2009 9:32 am
Contact:

Re: Argument for taking social security before 70. Comments?

Post by ObliviousInvestor » Thu Sep 13, 2018 10:02 am

David Jay wrote:
Thu Sep 13, 2018 9:30 am
opensocialsecurity does not use a single date. As I understand it, Mike’s strategy is to use the spectrum of potential lifespans and weight by likelihood of surviving to various ages.
That's correct. Open Social Security uses year-by-year probability of being alive (based on mortality table selected).

For a married couple, using fixed death-age assumptions (i.e., using the mean or median life expectancy from the same underlying data) usually has the significant downside of 1) underestimating the value of the high-PIA spouse delaying and 2) overestimating the value of the low-PIA spouse delaying.

Reason for the above is that the spans of time in question are (for the high-PIA spouse) second-to-die joint life expectancy and (for the low-PIA spouse) first-to-die joint life expectancy, and using fixed death-age assumptions for each spouse misstates the expected duration of those two spans of time.

For example, if we have an "average" male/female couple both age 62, their life expectancies are 20 and 22.8 years, respectively. If we used fixed death-age assumptions using those dates, the first-to-die joint life expectancy would be 20 years. In reality, it's only about 15.5 (because there's also a nontrivial possibility that the wife dies early, which is completely ignored when we used fixed dates). So we end up significantly overstating the expected value of the low-PIA spouse delaying benefits.

Of course, it's important to think about which table you want to use.

And then there's also the relevant point that for some people, there is no table that's going to be applicable (e.g., in the case of a health condition with very bad prognosis). So in those cases you would want to use a fixed death-age assumption. But it's still important to recognize that in doing so you're getting somewhat off-the-mark results with respect to joint life expectancies (though the distortion shouldn't be especially large in situations in which one person's life expectancy is quite a lot shorter than the other person's).
Mike Piper, author/blogger

bradpevans
Posts: 340
Joined: Sun Apr 08, 2018 1:09 pm

Re: Argument for taking social security before 70. Comments?

Post by bradpevans » Thu Sep 13, 2018 11:40 am

#Cruncher wrote:
Wed Sep 12, 2018 8:50 pm
bradpevans wrote:
Wed Sep 12, 2018 8:45 am
I don't know if the calculations included any investment returns. Ignoring that (i.e. put all your SS money under the pillow), ...
Since SS benefits are inflation-indexed, to not "include any investment returns" is not the same as to "put all your SS money under the pillow". The latter would represent a negative return in real terms. To get a 0% real return requires an investment that matches inflation. E.g., I Bonds with a 0% fixed rate.
Q. How does SS inflation adjustment work? Suppose at 62 years old I am eligible to collect $1,000/month, then my corresponding age 70 would be $1,750. But if i wait 8 years the $1,750 will have changed. Would the $1,000 change in the same way? In other words, does the ratio of these two payments stay at 1.75? (Not so much a math question but rather how the inflation impacts things)

User avatar
willthrill81
Posts: 6173
Joined: Thu Jan 26, 2017 3:17 pm
Location: USA

Re: Argument for taking social security before 70. Comments?

Post by willthrill81 » Thu Sep 13, 2018 11:42 am

bradpevans wrote:
Thu Sep 13, 2018 11:40 am
#Cruncher wrote:
Wed Sep 12, 2018 8:50 pm
bradpevans wrote:
Wed Sep 12, 2018 8:45 am
I don't know if the calculations included any investment returns. Ignoring that (i.e. put all your SS money under the pillow), ...
Since SS benefits are inflation-indexed, to not "include any investment returns" is not the same as to "put all your SS money under the pillow". The latter would represent a negative return in real terms. To get a 0% real return requires an investment that matches inflation. E.g., I Bonds with a 0% fixed rate.
Q. How does SS inflation adjustment work? Suppose at 62 years old I am eligible to collect $1,000/month, then my corresponding age 70 would be $1,750. But if i wait 8 years the $1,750 will have changed. Would the $1,000 change in the same way? In other words, does the ratio of these two payments stay at 1.75? (Not so much a math question but rather how the inflation impacts things)
Yes, the ratio would stay the same. The $1,750 is in today's dollars, not 2026's dollars.
“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

JoeRetire
Posts: 1651
Joined: Tue Jan 16, 2018 2:44 pm

Re: Argument for taking social security before 70. Comments?

Post by JoeRetire » Thu Sep 13, 2018 11:45 am

David Jay wrote:
Thu Sep 13, 2018 9:30 am
opensocialsecurity does not use a single date.
By default it does not.
But using the Advanced Options checkbox, you can select an age-based longevity.

You could use that for example to compare the lifetime difference between starting at 68 versus 70, if you actually live a long life or a shorter life.
Spousal survivor benefits can be affected significantly in many cases.

User avatar
ObliviousInvestor
Posts: 3395
Joined: Tue Mar 17, 2009 9:32 am
Contact:

Re: Argument for taking social security before 70. Comments?

Post by ObliviousInvestor » Thu Sep 13, 2018 11:45 am

bradpevans wrote:
Thu Sep 13, 2018 11:40 am
#Cruncher wrote:
Wed Sep 12, 2018 8:50 pm
bradpevans wrote:
Wed Sep 12, 2018 8:45 am
I don't know if the calculations included any investment returns. Ignoring that (i.e. put all your SS money under the pillow), ...
Since SS benefits are inflation-indexed, to not "include any investment returns" is not the same as to "put all your SS money under the pillow". The latter would represent a negative return in real terms. To get a 0% real return requires an investment that matches inflation. E.g., I Bonds with a 0% fixed rate.
Q. How does SS inflation adjustment work? Suppose at 62 years old I am eligible to collect $1,000/month, then my corresponding age 70 would be $1,750. But if i wait 8 years the $1,750 will have changed. Would the $1,000 change in the same way? In other words, does the ratio of these two payments stay at 1.75? (Not so much a math question but rather how the inflation impacts things)
Your primary insurance amount ("PIA" -- the amount of your monthly benefit if you file for it at your FRA) is what gets adjusted for inflation each year, beginning at 62.

And, you get a given percentage of your PIA depending on the age at which you claim. For example if you file 4 years prior to your FRA, you get 75% of your PIA. File 4 years after your FRA and you get 132% of your PIA. (In other words, as willthrill81 said, the ratio stays the same.)
Mike Piper, author/blogger

JoeRetire
Posts: 1651
Joined: Tue Jan 16, 2018 2:44 pm

Re: Argument for taking social security before 70. Comments?

Post by JoeRetire » Thu Sep 13, 2018 11:50 am

ObliviousInvestor wrote:
Thu Sep 13, 2018 10:02 am
For example, if we have an "average" male/female couple both age 62, their life expectancies are 20 and 22.8 years, respectively. If we used fixed death-age assumptions using those dates, the first-to-die joint life expectancy would be 20 years. In reality, it's only about 15.5 (because there's also a nontrivial possibility that the wife dies early, which is completely ignored when we used fixed dates). So we end up significantly overstating the expected value of the low-PIA spouse delaying benefits.
I don't understand how you can choose to use fixed death-age assumptions of 20 and 22.8 years, then talk about a nontrivial possibility of the wife dying early. Seems like those are mutually exclusive. If you choose a fixed age, you choose a fixed age, no?

Or am I misunderstanding what you are trying to say here?

User avatar
ObliviousInvestor
Posts: 3395
Joined: Tue Mar 17, 2009 9:32 am
Contact:

Re: Argument for taking social security before 70. Comments?

Post by ObliviousInvestor » Thu Sep 13, 2018 12:04 pm

JoeRetire wrote:
Thu Sep 13, 2018 11:50 am
ObliviousInvestor wrote:
Thu Sep 13, 2018 10:02 am
For example, if we have an "average" male/female couple both age 62, their life expectancies are 20 and 22.8 years, respectively. If we used fixed death-age assumptions using those dates, the first-to-die joint life expectancy would be 20 years. In reality, it's only about 15.5 (because there's also a nontrivial possibility that the wife dies early, which is completely ignored when we used fixed dates). So we end up significantly overstating the expected value of the low-PIA spouse delaying benefits.
I don't understand how you can choose to use fixed death-age assumptions of 20 and 22.8 years, then talk about a nontrivial possibility of the wife dying early. Seems like those are mutually exclusive. If you choose a fixed age, you choose a fixed age, no?

Or am I misunderstanding what you are trying to say here?
What I'm trying to say is that fixed-death assumptions don't reflect joint life expectancies very well.

To back up a step just to explain where I'm coming from:

When the high-PIA spouse delays benefits, it increases the amount the couple receives as long as either person is still alive. So for this person's decision, we want to know the "expected" length of time for which either person will be alive -- the second-to-die joint life expectancy.

When the low-PIA spouse delays benefits, it increases the amount the couple receives as long as both people are still alive. So for this person's decision, we want to know the expected length of time for which both people will be alive -- the first-to-die joint life expectancy.

If we use 20 and 22.8 years, we're assuming 20 years is the first-to-die expectancy and 22.8 years is the second-to-die expectancy.

A given person's life expectancy can be thought of as reflecting the uncertainty in their own lifespan. It's the mean, so it (on balance) reflects die-before-the-mean outcomes and die-after-the-mean outcomes. So the 20-year figure essentially incorporates possibility of the husband dying early. But it doesn't account for the possibility the wife dies early.

In other words, 20 years is the expected length of time for which the husband will be alive. It is not the expected length of time for which they will both be alive, because there's also a possibility that she predeceases him.

The actual expected length of time for which they're both alive is only 15.5 years -- meaning that the cost/benefit analysis for the low-PIA spouse is significantly overstating the benefit of waiting when we use life expectancies as fixed death ages (i.e., it's overstating the expected length of time for which the increased benefit will be received by 4.5 years -- or almost 30%).

(Edited for clarity -- I hope.)
Mike Piper, author/blogger

JoeRetire
Posts: 1651
Joined: Tue Jan 16, 2018 2:44 pm

Re: Argument for taking social security before 70. Comments?

Post by JoeRetire » Thu Sep 13, 2018 12:14 pm

ObliviousInvestor wrote:
Thu Sep 13, 2018 12:04 pm
What I'm trying to say is that fixed-death assumptions don't reflect joint life expectancies very well.
Got it. Thanks. I understand your feeling that the other tables are more accurate.
To back up a step just to explain where I'm coming from:

When the high-PIA spouse delays benefits, it increases the amount the couple receives as long as either person is still alive. So for this person's decision, we want to know the "expected" length of time for which either person will be alive -- the second-to-die joint life expectancy.

When the low-PIA spouse delays benefits, it increases the amount the couple receives as long as both people are still alive. So for this person's decision, we want to know the expected length of time for which both people will be alive -- the first-to-die joint life expectancy.

If we use 20 and 22.8 years, we're assuming 20 years is the first-to-die expectancy and 22.8 years is the second-to-die expectancy.

A given person's life expectancy can be thought of as reflecting the uncertainty in their own lifespan. It's the mean, so it (on balance) reflects die-before-the-mean outcomes and die-after-the-mean outcomes. So the 20-year figure essentially incorporates possibility of the husband dying early. But it doesn't account for the possibility the wife dies early.
I assume you meant to say "high earner" and "low earner" rather than "husband" and "wife" here. But I understand.
In other words, 20 years is the expected length of time for which the husband will be alive. It is not the expected length of time for which they will both be alive, because there's also a possibility that she predeceases him.
I guess here is where we diverge.
If you enter an assumed age of death, then by definition you are assuming there is no possibility of any other ages of death.

I understand your view of what the other longevity tables tell you. But choosing a fixed age but considering other possibilities is "mixing the streams" of the model.
When we go year-by-year, the actual expected length of time for which they're both alive is only 15.5 years -- meaning that the cost/benefit analysis for the low-PIA spouse is significantly overstating the benefit of waiting when we use life expectancies as fixed death ages (i.e., it's overstating the expected length of time for which the increased benefit will be received by 4.5 years -- or almost 30%).
I assume you cannot cure that by choosing 15.5 for the low earner's age of death to account for this and choosing more than 22.8 for the high earner?

User avatar
ObliviousInvestor
Posts: 3395
Joined: Tue Mar 17, 2009 9:32 am
Contact:

Re: Argument for taking social security before 70. Comments?

Post by ObliviousInvestor » Thu Sep 13, 2018 12:22 pm

JoeRetire wrote:
Thu Sep 13, 2018 12:14 pm
ObliviousInvestor wrote:
Thu Sep 13, 2018 12:04 pm
What I'm trying to say is that fixed-death assumptions don't reflect joint life expectancies very well.
Got it. Thanks. I understand your feeling that the other tables are more accurate.
To back up a step just to explain where I'm coming from:

When the high-PIA spouse delays benefits, it increases the amount the couple receives as long as either person is still alive. So for this person's decision, we want to know the "expected" length of time for which either person will be alive -- the second-to-die joint life expectancy.

When the low-PIA spouse delays benefits, it increases the amount the couple receives as long as both people are still alive. So for this person's decision, we want to know the expected length of time for which both people will be alive -- the first-to-die joint life expectancy.

If we use 20 and 22.8 years, we're assuming 20 years is the first-to-die expectancy and 22.8 years is the second-to-die expectancy.

A given person's life expectancy can be thought of as reflecting the uncertainty in their own lifespan. It's the mean, so it (on balance) reflects die-before-the-mean outcomes and die-after-the-mean outcomes. So the 20-year figure essentially incorporates possibility of the husband dying early. But it doesn't account for the possibility the wife dies early.
I assume you meant to say "high earner" and "low earner" rather than "husband" and "wife" here. But I understand.
I don't think I meant to. The 20-year figure is the male life expectancy. So it's accounting for the possibility of him dying early (whether he's higher or lower earner).
JoeRetire wrote:
Thu Sep 13, 2018 12:14 pm
ObliviousInvestor wrote:
Thu Sep 13, 2018 12:04 pm
In other words, 20 years is the expected length of time for which the husband will be alive. It is not the expected length of time for which they will both be alive, because there's also a possibility that she predeceases him.

When we go year-by-year, the actual expected length of time for which they're both alive is only 15.5 years -- meaning that the cost/benefit analysis for the low-PIA spouse is significantly overstating the benefit of waiting when we use life expectancies as fixed death ages (i.e., it's overstating the expected length of time for which the increased benefit will be received by 4.5 years -- or almost 30%).
I assume you cannot cure that by choosing 15.5 for the low earner's age of death to account for this and choosing more than 22.8 for the high earner?
That gets you closer. And yes we have to choose more than 22.8 because we have the same sort of thing going on with the high earner's decision.

22.8 is the expected length of time for which the wife will be alive. But it's not the expected length of time for which either will be alive. That's closer to 26.3.

So we could choose fixed death ages based on 15.5 and 26.3 life expectancies. But that can cause other problems when there are factors that make it matter who is the one still alive (e.g., government pension).
Last edited by ObliviousInvestor on Thu Sep 13, 2018 3:32 pm, edited 1 time in total.
Mike Piper, author/blogger

JoeRetire
Posts: 1651
Joined: Tue Jan 16, 2018 2:44 pm

Re: Argument for taking social security before 70. Comments?

Post by JoeRetire » Thu Sep 13, 2018 12:23 pm

ObliviousInvestor wrote:
Thu Sep 13, 2018 12:22 pm
I don't think I meant to. The 20-year figure is the male life expectancy. So it's accounting for the possibility of him dying early (whether he's higher or lower earner).
D'oh! Of course - my mistake. Thanks!

User avatar
ObliviousInvestor
Posts: 3395
Joined: Tue Mar 17, 2009 9:32 am
Contact:

Re: Argument for taking social security before 70. Comments?

Post by ObliviousInvestor » Thu Sep 13, 2018 12:34 pm

JoeRetire wrote:
Thu Sep 13, 2018 12:14 pm
ObliviousInvestor wrote:
Thu Sep 13, 2018 12:04 pm
In other words, 20 years is the expected length of time for which the husband will be alive. It is not the expected length of time for which they will both be alive, because there's also a possibility that she predeceases him.
I guess here is where we diverge.
If you enter an assumed age of death, then by definition you are assuming there is no possibility of any other ages of death.

I understand your view of what the other longevity tables tell you. But choosing a fixed age but considering other possibilities is "mixing the streams" of the model.
I just noticed this part. I think I agree actually.

Fixed death age assumptions are super useful for doing "what if" analyses.

But trying to use them to reflect uncertainty just doesn't work very well in a couple scenario (which is why you'd have to make strange modifications as discussed above -- it's a square peg into a round hole).

In other words, I think a good approach is to start with the most suitable table to get an idea of "expected" outcomes (accounting for uncertainty and weighting outcomes based on their probability). Then do a bunch of fixed-age what-if scenarios as well, to see what changes.
Mike Piper, author/blogger

JBTX
Posts: 4109
Joined: Wed Jul 26, 2017 12:46 pm

Re: Argument for taking social security before 70. Comments?

Post by JBTX » Thu Sep 13, 2018 1:38 pm

For most Bogleheads, their life expectancy (at age 50) is likely significantly longer than the median.

The top 10% lives about 4-5 years longer than the median wealth, and almost 15 years longer than the poorest.


https://www.nytimes.com/2016/02/13/heal ... owing.html

https://static01.nyt.com/newsgraphics/2 ... oard_3.png

In spreadsheet scenarios that I have run, which include the possiblity of drawing SS benefits for a disabled child, the breakeven for us is mid late 80's - we both work. But it seems like the benefits of waiting start accumulating pretty rapidly after the break-even age, whereas the benefit of retiring early isn't dramatically better a few years before the break-even age. You could have a breakeven age at 87, but if one lives to 92, the benefits of delaying are significant.

I also did a probability / expected value - weighting the additional benefit times the probability you are still living up to age 100, and delaying still won out, by significant but not dramatic differences. There is probably enough play/simplification in my assumptions that the result could materially change.

MikeG62
Posts: 1182
Joined: Tue Nov 15, 2016 3:20 pm
Location: New Jersey

Re: Argument for taking social security before 70. Comments?

Post by MikeG62 » Thu Sep 13, 2018 1:51 pm

JackoC wrote:
Wed Sep 12, 2018 5:13 pm

...Anyway I will not take SS any earlier than 70 unless my health goes south before that. As PJW's excellent post above outlined, you must first identify your goal. My goal is not to maximize my take from SS on average across all scenarios. It's to have a relatively big payout in scenario's where I live longer than expected and my investments perform less well than expected, so I actually need Social Security: downside protection.
This summarizes exactly how I look at it.

The only other thing I would add is that waiting provides me with several additional years of Roth conversions before SS (and RMD's) kick in.
Real Knowledge Comes Only From Experience

Nowizard
Posts: 1596
Joined: Tue Oct 23, 2007 5:33 pm

Re: Argument for taking social security before 70. Comments?

Post by Nowizard » Thu Sep 13, 2018 2:13 pm

Took it at FRA and glad we did. For us there were two factors: 1. We have strong genetic histories that foretell long lives, 2. Most importantly, we did not need the money and could invest it directly or have it in reserve in case of emergencies. For us, the returns have worked out well. You do get a significantly increased percentage for each year delayed, but you also receive compounding on money taken at FRA and invested well.

Tim

BigoteGato
Posts: 20
Joined: Fri Mar 11, 2016 3:05 pm

Re: Argument for taking social security before 70. Comments?

Post by BigoteGato » Thu Sep 13, 2018 2:29 pm

grabiner wrote:
Wed Sep 12, 2018 8:43 pm
Here's how I estimate the break-even. It's easiest to do one year at a time.

Suppose that you have enough money to cover this year's Social Security. You can spend the money this year and delay Social Security to next year, or take Social Security now and buy a TIPS ladder which will cover the future years' payments. The break-even is the point where the TIPS ladder runs out.

If you are 66 and your benefit is $20,000, waiting one year will increase your benefit to $21,600 (all numbers growing with inflation). If you spend $20,000 to buy TIPS, you could get $1600 per year for 12.5 years if the TIPS yields were zero; this would make the break-even 13.5 years. Current yields on a TIPS portfolio with a 7-year duration are 0.85%, which is enough to make the ladder last 13.3 years, for a break-even of 14.3 years. That is to age 80.

If you are 69 and your benefit is $24,800, waiting one year will increase your benefit to $26,400. If you spend $24,800 to buy TIPS, you could get $1600 per year for 15.5 years if the TIPS yields were zero, and 16.5 years at current yields, making the break-even 17.5 years; that is to age 87. 17.5 years is just about the life expectancy for a 69-year-old man, and less than the life expectancy for a 69-year-old woman.

Now, these are average life expectancies. Most people either have higher-than-average life expectancies, or know that they don't (because they have heart disease, cancer, etc.) This makes waiting more attractive.

And even if you have an average life expectancy, there is an advantage for waiting. If you wait, you will have more money if you live longer, and thus have a greater need for the money.

My calculations are only for singles, though. For married couples, the effect on spousal and widow/er's benefits is important. If your spouse is likely to collect widow/er's benefits on your record, that is a strong incentive for you to delay, as the break-even point is based on when you both die. Conversely, if you will switch to a widow/er's benefit if your spouse dies first, the break-even point is when either of you die, which makes claiming at full retirement age likely better.
Grabiner - I find your reasoning clear and compelling. Let's see how it may apply to the OP's situation of a high and a low wage earner. Assume the low wage earner would get $0 SS, the high would get $20K as in your example. If high earner files, then low earner can get $10K (spousal benefit). So your examples are modified to include buying $30K in TIPS, not $20K. Now your 66 yrs old's break-even age becomes 50% higher - 20.25 yrs, or 86.5. The 69 yr old's break-even age is pushed to 95.25.
BigoteGato
(The case where the low earner gets more than $0 but less that the high earne's $20K is left as an exercise to the reader.)

User avatar
galeno
Posts: 1370
Joined: Fri Dec 21, 2007 12:06 pm

Re: Argument for taking social security before 70. Comments?

Post by galeno » Thu Sep 13, 2018 3:13 pm

I played around with this: https://opensocialsecurity.com/

1. Wife and I are 61. Assumed $2000 PIA for me and $1000 PIA for spouse. PIA for both is 66 years 6 months.

Advice: I take SS at 70 years 0 months. Wife takes SS at 62 years 1 month.

NPV of SS for both is $609,934.


2. Wife and I are 61. Assumed $2000 PIA for me and $0 PIA for spouse. PIA for me is 66 years 6 months.

Advice: I take SS at 68 years 6 months. Wife takes SS at 68 years 10 months.

NPV of SS for both is $570,272.
AA = 40/55/5. Expected CAGR = 3.8%. GSD (5y) = 6.2%. USD inflation (10 y) = 1.8%. AWR = 4.0%. TER = 0.4%. Port Yield = 2.82%. Term = 33 yr. FI Duration = 6.0 yr. Portfolio survival probability = 95%.

Post Reply