Dr. Pfau table of implied returns for delaying SS

[Animorph]

Just a chart of the real annualized returns of delaying from age 62 to age 70 for each year you live past 71. I think this makes the financial tradeoff of this simplest case very clear. The traditional crossover age of around 80, without investment returns considered, shows up in this chart as a roughly 0% return by age 80. Before that age the returns are negative, after 80 the returns are positive. Even at age 95 the implied return is only 6%. And the risk of delay is shown very clearly by a -22.4% return if you die at 72.

http://wpfau.blogspot.com/2014/04/delaying-social-security-what-investment.html

Doesn't change anything about the insurance value of delaying SS of course. I'm still planning to delay.

 

[samclem]

Doesn't change anything about the insurance value of delaying SS of course.

Yes, and that's the biggest point about the whole idea of delaying SS: It's not the absolute value amount of the anticipated returns that is most important (to me), it is the type of returns: COLA adjusted and a dependable monthly amount as long as I live. If a bigger % of my living expenses in later years are coming in like that, I can spend down that nest egg a bit.
Of course, that depends on the whole "deal" remaining as it is now.

 

[ronin]

Odd that the title of the table states that it is a comparison of starting at 66 with 62 or 70 but the numbers in the table don't seem to reflect that. Maybe I am missing something.

 

[Animorph]

Odd that the title of the table states that it is a comparison of starting at 66 with 62 or 70 but the numbers in the table don't seem to reflect that. Maybe I am missing something.

It gives the FRA benefit as of age 66, but the next line says 62 vs. 70.

 

[gerntz]

I take Real Return to be annual return, not total return, starting at age 62. Correct?

 

[gerntz]

And I don't think this accounts for spouse or spousal benefits before age 70 for marrieds.

 

[rayvt]

Some simple math:
8 yrs * $7500 = $60,000
$60,000 / $5700 = 10.5 years to break even.

But if you can afford to not collect SS for those 8 years, then by definition you didn't need the $7,500 to live on. Yet you are so financially clueless that you just stuck the money into an account earning 0%.

If instead you invested it in, say, bonds earning 5% it would have built up to $76,200.
$76,200 / $5700 = 13.3 years.

But of course you'd really be drawing down at the rate of $5700/yr and the rest would still be earning 5%.
That requires a calculator rather than a simple division. The retirement calculator I used shows that this will take 21.6 years to exhaust the account.

So now the actual break-even age is 92 years old.

Pfau surely knows how to compute compounded growth and compounded drawdown, so why didn't he?

Perhaps 5% is too optimistic. Refigure at 3%.
Grows to $68,700.
Lasts 15 years.
So the breakeven age is 85.

 

[mathjak107]

the problem with figuring bonds is inflation and other investments is lack of guarantee and risk factor .

sure we can throw money into equities for a bigger return but basically you are taking a no risk, almost guaranteed investment and trading it for a high risk investment.

about as close as you can get to being protected against inflation as pfau said in forbes would be TIPS and they pay about 1/2 of what delayed ss will if you live to the average age for a healthy person.

.

 

[ronin]

See Michael Kitces blog on this topic including an analysis on opportunity cost of investment returns:

"While much has been written about the inherent benefits of delaying Social Security benefits to age 70, a fundamental challenge in the real world is that the decision cannot be viewed in the abstract. The decision to delay Social Security isn't just about the value of delaying, but also about the money that must be spent from the portfolio to sustain spending in the meantime, and/or the decision to allocate money towards delaying Social Security and not towards other fixed income investments or a commercially available lifetime immediate annuity."

how-delaying-social-security-can-be-the-best-long-term-investment-or-annuity-money-can-buy

 

[Brett_Cameron]

See Michael Kitces blog on this topic including an analysis on opportunity cost of investment returns:

"While much has been written about the inherent benefits of delaying Social Security benefits to age 70, a fundamental challenge in the real world is that the decision cannot be viewed in the abstract. The decision to delay Social Security isn't just about the value of delaying, but also about the money that must be spent from the portfolio to sustain spending in the meantime, and/or the decision to allocate money towards delaying Social Security and not towards other fixed income investments or a commercially available lifetime immediate annuity."

how-delaying-social-security-can-be-the-best-long-term-investment-or-annuity-money-can-buy

This aspect makes delaying SS to age 70 prohibitively expensive for us.

 

[ronin]

This is something that I am working through now. I turn 62 this month. But I get the full WEP haircut and the annual % reductions pre-66 and and increases post-66 are different than the simple examples for non-WEPers. If my calculations are correct the benefit of delaying is more in the 10-11% annual range after FRA due to the effects of the WEP. Fortunately, we don't need the SS benefit so it one less thing in the "get it early" column.

 

[Animorph]

Some simple math:
8 yrs * $7500 = $60,000
$60,000 / $5700 = 10.5 years to break even.

But if you can afford to not collect SS for those 8 years, then by definition you didn't need the $7,500 to live on. Yet you are so financially clueless that you just stuck the money into an account earning 0%.

If instead you invested it in, say, bonds earning 5% it would have built up to $76,200.
$76,200 / $5700 = 13.3 years.

But of course you'd really be drawing down at the rate of $5700/yr and the rest would still be earning 5%.
That requires a calculator rather than a simple division. The retirement calculator I used shows that this will take 21.6 years to exhaust the account.

So now the actual break-even age is 92 years old.

Pfau surely knows how to compute compounded growth and compounded drawdown, so why didn't he?

Perhaps 5% is too optimistic. Refigure at 3%.
Grows to $68,700.
Lasts 15 years.
So the breakeven age is 85.

Not sure I follow you.

That's the beauty of this chart. It shows the annualized return (fully compounded) you earn by "investing" the $7.5k into SS for eight years and getting a "return" of $5.7k for life starting at age 70. If you expect to earn a better return for your portfolio, take SS early. Otherwise you can delay.

You can see roughly your numbers in the chart. Die at age 80 and your SS "investment" made 0.5%. Die at 85 and your SS investment made 3.6%. Die at 92 and your SS investment makes 5.6%. Very easy to see what the breakeven age is for any portfolio real annualized gain you would like to check.

 

[rayvt]

Kitces's article said the same thing I said. He did it a different way but came up with the same thing I did. If you invest the early money, it takes 22 years to get back to breakeven. 70 + 22 = 92.

He does go on to say "Delaying Social Security As The Best Long-Term Return Money Can Buy" -- if you live longer than the breakeven period.

"in the later years the real returns become very large, crossing about 1.3% after 20 years, 4% after 25 years, 5% after 30 years, and 6% after 34 years."

Problem is, even if the return is a big percentage the dollar amount that the percentage is applied to isn't very large. This is the same problem as the now-defunct "return and reset" tactic. You don't get to pick your dollar amount, it is determined by the SSA and isn't really all that large.

Pfau's example was a delta of $5700/yr or $475/mo. So, hmmm after 30 years -- which is age 100? -- you see an effective 5% P.A. return. Which is $285/yr or $24/mo. This is peanuts.
If I am going to make an investment that takes 15-20 years to break even, I want to get a heck of a lot more than $24/mo in 30 years.
And how many of us are going to see 100 anyway?

 

[Chuckanut]

Well.... $24 is $24. Pizza and beer for this week's Monday night game.

I see your point and certainly won't argue with your choice for what is good for you.

But, some of us managed to achieve this anointed state of life by doing just that - saving $24 at a time until it added up to real money. $30 a month cut from the cable bill, $15 in gasoline saving from buying a higher milage car, [-]$20 a month from less expensive wine,[/-] $20 a month making coffee at home rather than buying a latte, $20 a month from eating out less, and an extra $15 a month from handling investments a bit better than the other guy. Hey, now it's $100 a month.

I have a number of acquaintances who will be working long past 65 (forget about early retirement) because they can't save small amounts.

 

[mathjak107]

Kitces's article said the same thing I said. He did it a different way but came up with the same thing I did. If you invest the early money, it takes 22 years to get back to breakeven. 70 + 22 = 92.

He does go on to say "Delaying Social Security As The Best Long-Term Return Money Can Buy" -- if you live longer than the breakeven period.

"in the later years the real returns become very large, crossing about 1.3% after 20 years, 4% after 25 years, 5% after 30 years, and 6% after 34 years."

Problem is, even if the return is a big percentage the dollar amount that the percentage is applied to isn't very large. This is the same problem as the now-defunct "return and reset" tactic. You don't get to pick your dollar amount, it is determined by the SSA and isn't really all that large.

Pfau's example was a delta of $5700/yr or $475/mo. So, hmmm after 30 years -- which is age 100? -- you see an effective 5% P.A. return. Which is $285/yr or $24/mo. This is peanuts.
If I am going to make an investment that takes 15-20 years to break even, I want to get a heck of a lot more than $24/mo in 30 years.
And how many of us are going to see 100 anyway?

but at 70 with all that extra dough coming in demand falls off for selling investments as well as you can take on more risk with your portfolio with the reliance on it cut so much.

there is a whole other investing side that you can do taking it at 70 which if you take at 62 may not exist because you may be to afraid being so reliant on your own portfolio for a life time.

i know i intend to use the rising glide path method . as my risk and allocation increases by 70 i will be less reliant on my portfolio and not so worried about taking on to much risk.

the two mesh very nicely..

 

[haha]

Well.... $24 is $24. Pizza and beer for this week's Monday night game.

I see your point and certainly won't argue with your choice for what is good for you.

But, some of us managed to achieve this anointed state of life by doing just that - saving $24 at a time until it added up to real money. $30 a month cut from the cable bill, $15 in gasoline saving from buying a higher milage car, $20 a month from less expensive wine, $20 a month from eating out less, and an extra $15 a month from handling investments a bit better than the other guy. Hey, now it's $100 a month.

I have a number of acquaintances who will be working long past 65 (forget about early retirement) because they can't save small amounts.

I'm with you. But please leave wine quality off the table.

Ha

 

[rayvt]

Right. But that's people who haven't made it.

When you are starting out and broke, $25 is a big deal, as you said. A few of those and you're saving $100/mo and doubling your net worth. When you are F.I.R.E.'d $25 disappears in the round-off of your net worth.

If you have retired at 62 and and have enough money that you can forgo $7500/yr from Social Security, then you are among those who have made it. And if you've got it made, then getting an extra $24 at 92 is peanuts. So, you splurge and order the $250 bottle of wine instead of the $225 one?

 

[racy]

Maybe I don't get it, but it seems to me that the discussion about "delaying to get the most from SS" doesn't account for the impact on one's portfolio. I've created a year by year cash flow spreadsheet comparing taking SS at 62, 66 or 70 and calculating ending portfolio value. Taking it at 62 is the best option up until I'm 74 yrs old. Taking SS at 66 wins up until I reach age 90. Taking it at 70 is best if I live beyond 90, but not greatly more the 66 decision. I think I'll go with the mortality tables and take it at 66.

 

[Animorph]

Kitces's article said the same thing I said. He did it a different way but came up with the same thing I did. If you invest the early money, it takes 22 years to get back to breakeven. 70 + 22 = 92.

He does go on to say "Delaying Social Security As The Best Long-Term Return Money Can Buy" -- if you live longer than the breakeven period.

"in the later years the real returns become very large, crossing about 1.3% after 20 years, 4% after 25 years, 5% after 30 years, and 6% after 34 years."

Problem is, even if the return is a big percentage the dollar amount that the percentage is applied to isn't very large. This is the same problem as the now-defunct "return and reset" tactic. You don't get to pick your dollar amount, it is determined by the SSA and isn't really all that large.

Pfau's example was a delta of $5700/yr or $475/mo. So, hmmm after 30 years -- which is age 100? -- you see an effective 5% P.A. return. Which is $285/yr or $24/mo. This is peanuts.
If I am going to make an investment that takes 15-20 years to break even, I want to get a heck of a lot more than $24/mo in 30 years.
And how many of us are going to see 100 anyway?

I don't think you understand this chart. And I'm surprised no one has called it out yet.

At age 62 you take your $7500 and stick it in a bank, at say 5% interest (good luck, but it's only an example). You do the same thing for the next 7 years. That's 8 deposits over 8 years of $7500 each. Total of deposits is 8*$7500 = $60000. You're making 5% annual interest on that, which adds up to another $11618. So you have a total of $71618 in the bank. That's what you're making 5% on, not $5700.

Then you start taking $5700 out of your account each year. With the bank still paying you 5% interest per year, you can withdraw $5700/year for a little over 20 years, through age 89. That's 10 more years of $5700/year than if you received 0% from the bank. Where the heck does $285/year come from? Ten extra years of $5700 (COLA'd since we're talking real interest rates here) is a significant amount. The $5700, in real dollars, is fixed the entire time.

Rather than specifying an interest rate and telling you how long your payments will last, Dr Pfau's chart specifies the age you survive to and tells you the interest rate required to make those payments last your lifetime.

Or even more relevant, the constant investment gains you would have to make in order to match the return of delaying SS. If you live to age 90 and your portfolio provided annualized gains of less than about 5% (real, subtracting out inflation) you would have done better to take SS at age 70. If your portfolio gains were better than 5% you would have done better to take early SS and leave more in your portfolio. Each row in the chart gives you the portfolio annualized real return you should expect to exceed if you are going to take SS early. The longer you live, the higher the return, and the better delaying SS looks.

The calculation is very simple in Excel. Eight years of $7500. Then the IIR of getting one future payment of $5700 if you live just one extra year (up to your 71st birthday). The IIR with two future payments of $5700 if you live two extra years (up to your 72nd birthday), and so on.

So what's your expected real annualized investment return? Do you think you'll live long enough for delaying SS to pay off better than that?

 

[springnr]

I just whipped up a excel spreadsheet: Single person IRA + SS withdrawals 62 vs 66 vs 70
Ran different scenarios - varying IRA return, withdrawal amount
Unless you are living on the edge of your assets. (withdrawing too much,long term return is below 2%)
It looks like a wash. You may leave a couple 100K more or less to heirs or charity.

This does not take into account other concerns, different schemes for maximizing benefits, if you loose all your money could you get by on just the SS amount, or due to say dementia would a higher SS get you in a better facility, those kinds of considerations.

This is a thrown together spreadsheet so it may act up easily depending on inputs. and the last time I really used spreadsheets was with Lotus 123.

 

[Independent]

Some simple math:
8 yrs * $7500 = $60,000
$60,000 / $5700 = 10.5 years to break even.

But if you can afford to not collect SS for those 8 years, then by definition you didn't need the $7,500 to live on. Yet you are so financially clueless that you just stuck the money into an account earning 0%.

If instead you invested it in, say, bonds earning 5% it would have built up to $76,200.
$76,200 / $5700 = 13.3 years.

But of course you'd really be drawing down at the rate of $5700/yr and the rest would still be earning 5%.
That requires a calculator rather than a simple division. The retirement calculator I used shows that this will take 21.6 years to exhaust the account.

So now the actual break-even age is 92 years old.

Pfau surely knows how to compute compounded growth and compounded drawdown, so why didn't he?

Perhaps 5% is too optimistic. Refigure at 3%.
Grows to $68,700.
Lasts 15 years.
So the breakeven age is 85.

He did. It looks to me like your numbers agree with his.

Looking at his table, if I earn a constant 5%, my breakeven is between 89 and 90.
If I earn a constant 3%, breakeven is between 83 and 84.
If I earn 0%, breakeven is between 79 and 80.

He breaks even a little sooner than you. That may be monthly vs. annual calculations or beginning vs. ending cash flows.

Note that all these percents are "real" not "nominal", so the 5% means "CPI + 5%". I think any US retiree would be hard pressed to get CPI + 3% today. Ten year TIPS are yielding about CPI + 0.6% and 30 year are around CPI + 1.3%

 

[rayvt]

Rather than specifying an interest rate and telling you how long your payments will last, Dr Pfau's chart specifies the age you survive to and tells you the interest rate required to make those payments last your lifetime

.

Ahhh! Now I understand his table. Either he didn't explain it very well, or I just miscomprehended what he was saying.

Indeed my calculations just about equal his.
Re-reading his table with an eye to how you explained it, we agree.
At 0% earnings rate, the breakeven age is 80.
At 3%, age 85.
At 5%, age 91.
At 6.5%, age 105.
At 5.3%, age 92.
All net of inflation.

Vanguard says that the average return (1926-2013) of a 50/50 balanced portfolio was 8.3%. If you subtract 3% inflation and use 5.3% earnings and add 3% SS COLA, it just about cancels out and the breakeven age is 92 1/2.

So there's the thing, and everybody who does the math right comes up with pretty much the same answer.
No earnngs, breakeven is 10 years
3% (after inflation) the breakeven is 15 years.

What is different is which aspect of the situation is given the emphasis.

One camp says, "If you live to 90, you'll get more SS if you delay."

The other camp says, "If I earn nothing, I'll be 80 before I break even. If I save or invest and earn 4% (after inflation) I'll be 84 when I break even."
It is worth noting that the SSA reports life expectancy is 84 years old. So on average the break-even date is just about the same as the date of death.

The reason the debate sometimes gets so heated is that there is really no difference financially. The only difference is in which aspect is more important to you.

I am in the second camp, but acknowledge that many reasonable people are in the first camp. My thinking is that I would continue to use the same investment skills that got me to early retirement in the first place, and apply those skills to the early SS money. So even a low-risk 50/50 investment would easily push the breakeven date to long after my life expectancy. And if I went with a riskier 80/20 investment there is NEVER a breakeven.

 

[teejayevans]

I'll wait because:
it's a form of longevity insurance
allows me to spend down IRA, lower RMDs, possibly lower taxes
A guaranteed increase of ~8%/yrincrease payouts, with a COLA...

BUT, if things change like how they calculate the COLA, or possible cutting of SS benefits, I take it at 62

 

[Independent]

.
The reason the debate sometimes gets so heated is that there is really no difference financially. The only difference is in which aspect is more important to you.

I am in the second camp, but acknowledge that many reasonable people are in the first camp. My thinking is that I would continue to use the same investment skills that got me to early retirement in the first place, and apply those skills to the early SS money. So even a low-risk 50/50 investment would easily push the breakeven date to long after my life expectancy. And if I went with a riskier 80/20 investment there is NEVER a breakeven.

I agree with the first paragraph above. The mathematical difference between the two is so slight that it's easy for different people with slightly different priorities to be on opposite sides.
Hence, there is no "right" answer.

I'll complicate this anyway by looking at the other part of my retirement plan, my regular withdrawals from my investment portfolio.

I note that a low-risk 50/50 investment allocation has historically earned inflation + 5.3%. I also note that if I'm starting at 62, 30 gets me to age 92, which is long after my life expectancy.

So, if I have a $1 million portfolio, I should be able to make withdrawals that start at $63,000 and increase with inflation. I can do the math and show that such a plan leaves me with $66,000 after 30 years.

But, people here don't start their retirement withdrawals at 6.3%. Somehow, that calculation isn't convincing when they are looking for a "safe" withdrawal plan.

I'm not saying that people should use a 6.3% rate, or they should approach SS differently than your calculation. I do feel like I should have some consistency between my approaches for the two decisions, or at least I should be able to explain why I approach them differently.

(Of course, many people here get that consistency by doing integrated calculations in FireCalc or some similar system.)

 

[mathjak107]

i was listening to an interview with jason zweig on this topic. he brought up a very important point.

social security is an income stream that really can be converted to an asset and what he means is this:

if you take your ss and muliply by 30x it is a very rough idea of the equivalent in bonds it would take to generate that income.

if you are getting 1k a month in ss that is 12k a year. as a very rough rule of thumb ,you have an income stream coming in that would be like having 350k in bonds to duplicate.

so if you get 22k at 62 and 40k at 70 like i would that represents having 660k in bonds vs having the equivalent of 1.2 million in bonds.

having the equivalent of 1.2 million in bonds lets you actually cut back on your fixed income holdings and put much more in to equities than only 350k in bonds would allow.

the end result is the higher level of aggressiveness allowed can refill savings much faster than just looking at how long the additional ss stream alone would bring you to break even.