Social Security at 62, 66 or 70?

[jdw_fire]

Maybe not, but I also dont come up with hokey ideas based on a bunch fallacious assumptions, then take it personally and whine about it for two frickin years when someone shoots holes in it. Get over it man! Criticizing your ideas is not a criticism of YOU!

First, this just proves my point about who is "being a jerk about it". Second, what are the "fallacious assumptions" you are refering to? There are people who think it is more important to them to spend more money between the ages of 62 & 70 then to leave an estate when they die. Third, you are the one that seem to be taking it personally. Fourth, I have only been providing examples to this idea for a couple of months not years. And fifth, you haven't shot any holes in the idea, you just change the parameters to make your point (as you did in the post I am responding to).

As far as your example...once again loaded with assumptions and not the scenario I've discussed. Just one where you can try to 'win' the debate.

The two examples I presented are not loaded with assumptions, just one, which is that the person making the decision would rather spend more money between the ages of 62 & 70 than leave an estate.

Its as easy as 1-2-3...plug all your data into firecalc, put in your social security statement figures for 62 and 70 where appropriate, and see that your lifetime withdrawal rate is higher and your survival rate is better at 62. Mine says that I could spend an extra 5k a year from age 45 to infinity, if I were willing to count on the social security income...which I dont.

Here is where you are changing the parameters in that you change the age of the person to 45 which is very specific to you. On the other hand my example looks at some one making the decision at 62yo, which will happen to everyone who receives SS. Granted I used specific numbers (yours) in my examples to illustrate the point but the concept stays the same if you change said numbers.

And all this talk of weighing down the portfolio/swr/ss early situation by amplifying the additional delayed social security benefit? Poppycock. Theres absolutely nothing to suggest that the federal government will pay you the full amount 10, 20, 30, 40, 50 years from now...for those insisting on a 120 year run in order to make a delay scenario look better.

The government is unlikely to change the rules for computing your starting SS benefit when you are 62yo.

My example doesn't insist on a 120yr run, in fact it doesn't insist on any particular lifespan. It just uses the 4% SWR for a portfolio that you said you use.

"Your estimated benefits are based on current law. Congress has made changes to the law in the past and can do so at any time. The law governing benefit amounts may change because by 2042 the payroll taxes collected will be enough to pay only about 73 percent of the scheduled benefits"

The old SS is broke argument (see other threads for a hashing of this). First you have to believe that these 25 years in the future estimates are solid enough for you to base a decision you make today. And second, you have to believe that the government would fix SS in the future by cutting benefits to people over the age of 62 (remember you are 62 when you are making the decision to take SS early or postpone it)

Sooo...I think we can quit pretending that the higher payout rate will stick for 40 years.

If you actually read my post you would realize that I said the higher payout was for the years you are 62 to 70, after that the payout is the same whether you start taking SS at 62 or 70.

I'll stick with controlling my own money, investing and withdrawing prudently, taking income streams that may or may not be available in the future as soon as they're available, and as far as what age range i'll set for firecalc when i'm 62? About 25 years and the same 4% thats good enough for me right now. When I'm 72 I'll look at it again.

You must not be reading the threads that are in essence stating that in the future the stock market will likely produce smaller real returns then it has in the past.

 

[indymom]

I am the wife of a CSSR retiree who also has some Social Security coming because of other work. His SS will be reduced because he has a federal pension. Will mine be reduced also if he dies before me, or would I receive the whole amount, since would be the survivor and already receivng a reduced pension? I will be 62 next year and am trying to figure out if taking it then would be in my best interest.

 

[jdw_fire]

I am the wife of a CSSR retiree who also has some Social Security coming because of other work. His SS will be reduced because he has a federal pension. Will mine be reduced also if he dies before me, or would I receive the whole amount, since would be the survivor and already receivng a reduced pension? I will be 62 next year and am trying to figure out if taking it then would be in my best interest.

Are you talking about SS based on your work history or his?

The answer will depend on the rest of your financial picture as well as your motivations about increasing your income stream and leaving an estate when you die.

 

[indymom]

I will get SS on my own, but I haven't made much, so half of his full SS might well be more than my full SS. However, if what I would receive as a surviving spouse will be less than my full SS, then I think I should wait to take SS. As a widow, my share of his pension is about 2/3 of what we receive now. I need to maximize the SS I would receive as a widow. SS is diminished for my husband because he is a federal retiree even tho his SS is based only on what he made under SS, not as a federal employee. Would my SS based on his work history be reduced as well?

 

[jdw_fire]

I think we can quit pretending that the higher payout rate will stick for 40 years.

Giving more thought to the scenario there is a way to get more income for life. Using the numbers from my first example given in this thread and delaying the start of SS till age 70 as in that example but instead of splitting up the $256,500 into 8 equal portions you first split it into two pots, the first equals $189,984 and the second equals $66,516. Then split the first pot into 8 equal portions of $23,748 and put in a MMA for use over the next 8 years to give you an annually inflation adjusted WD of $1979/mo.

You now have three choices as to what to do with the second pot of $66,516 to give you an increased income for life. First you can invest it as you would any portfolio to get you a 4% SWR of $2660/yr adjusted for inflation. Second, you could get $3621.84/yr CPI adjusted from an CPI adjusted immediate annuity. Or third if you want to spend more when you are younger you could buy a fixed payment immediate annuity and get a constant $5093.64/yr which in real terms will decrease as you grow older. Note that all of these three choices produce life time income above the $1979/mo you would get if you took SS at age 62.

 

[jdw_fire]

I will get SS on my own, but I haven't made much, so half of his full SS might well be more than my full SS. However, if what I would receive as a surviving spouse will be less than my full SS, then I think I should wait to take SS. As a widow, my share of his pension is about 2/3 of what we receive now. I need to maximize the SS I would receive as a widow. SS is diminished for my husband because he is a federal retiree even tho his SS is based only on what he made under SS, not as a federal employee. Would my SS based on his work history be reduced as well?

If "half of his full SS" will be "well be more than my full SS" and you want "to maximize the SS I would receive as a widow" you should both delay taking his until he reaches 70yo. This will maximize your benefit both before and after he dies (unless you die first). My statements assume that when you say "half of his full SS" will be "well be more than my full SS" that when you calculated his benefit you took the WEP into account.

 

[Independent]

Its as easy as 1-2-3...plug all your data into firecalc, put in your social security statement figures for 62 and 70 where appropriate, and see that your lifetime withdrawal rate is higher and your survival rate is better at 62. Mine says that I could spend an extra 5k a year from age 45 to infinity, if I were willing to count on the social security income...which I dont.

I did the 1-2-3 and got some interesting results. I think that part of the disagreement here is related to age. I’m thinking of this decision in terms of someone who is turning 62 in 2007, but I think that CFB is quite a bit younger. SS factors vary by year of birth.

So I put JDW's numbers into FireCalc:
Assets of $256,500, “default” investment mix of 75/25, investment expense of 0.18%, a 95% required success rate, and horizon of 30 years. Then I tried four SS strategies:

a) SS of $13,488 starting in 2007
b) SS of $21,348 starting in 2015

c) SS of $17,984 starting in 2011
d) SS of $23,739 starting in 2015

FireCalc returned 95% safe total income amounts of:
a) 23,684
b) 23,332

c) 24,960
d) 24,738

(a) and (b) are the SS amounts that CFB specified. Starting SS immediately gives an extra $352 of total income per year. It also provides a bigger estate – averaging $140,000 extra if you die at 70.

On the other hand, in the 5% of the scenarios where you exhaust your assets, waiting provides an extra $7,860 annually in those later years (if you’re still alive). It seems that most people would start SS immediately.

(c) starts with the $13,488 in (a), but adjusts it for someone turning 62 in 2007 and taking SS at 66. For people born in 1945, the benefit at age 62 is 75% of the benefit at age 66, so the $13,488 at age 62 would convert into $17,984 at age 66. For this person, deferring SS adds $1,256 per year to the total income. Of course, taking SS immediately gives the bigger estate, typically $70,000 extra for death at 70.

In the 5% of the cases where the assets run out, waiting adds $4,496 to the later years. A person who is focused on early income, and believes FireCalc, waits to 66, especially if he’s concerned about the worst 5% possibility.

Finally, (d) is for the person born in 1945 who waits to 70 to start SS. For this person, the SS benefit at 70 is 132% of the benefit at 66. Hence the $24,738. He can get an extra annual total income of $1,054 by waiting (compared to starting at 62). He gives up maybe $150,000 of estate on death at 70.

In the bad 5% of the cases, he has an extra $10,251 in annual income toward the end of his life. Someone who is very focused on the worst case possibilities goes with this option.


Now CFB is going to say that the Spitzer article says you’re almost always better off by taking the money sooner. I think the key is the “almost”. FireCalc’s safe withdrawal rates are driven by the worst 5% of the investment scenarios. It doesn’t care about “most”, just about “worst”. It’s possible (I can’t see the figures) that Spitzer has a few low interest cases where the waiting option wins. If so, that would be consistent with FireCalc.

 

[jdw_fire]

So I put JDW's numbers into FireCalc:
Assets of $256,500, “default” investment mix of 75/25, investment expense of 0.18%, a 95% required success rate, and horizon of 30 years. Then I tried four SS strategies:

a) SS of $13,488 starting in 2007
b) SS of $21,348 starting in 2015

I am going to start this reply with some housekeeping; first the numbers are not mine, I quoted CFB. Second my example used the monthly figures (which when I checked seemed to be in the proper proportion to each other), however one of the annual figures I quoted from CFB is wrong (since I didn't use them I didn't check them til now). The annual figure for SS if started at age 70 would be 1979*12 = 23748 not 21348. Using the correct SS numbers in FIRECalc and all the other parameters the same as Independent stated gives an annual WD of $23,684 when you start taking SS at age 62 and an annual WD of $24,743 when you start taking SS at age 70 (a $1059 difference which is an ~4.5% increase due to waiting).

The next point I'll make is that since in my example the motivation is to spend more between the ages of 62 & 70 (in real terms) then what will be spent after age 70, FIRECalc is not the right tool to use to try and determine how much more can be spent. Also if you will recall the investment I use for the portfolio when SS is delayed till age 70 is a money market account and not 75% stocks. This difference also makes FIRECalc the wrong tool for this side of the analysis. However what I did to make my point was to use simple mathematics with one conservative (meaning the amount I showed could be spent between ages 62 & 70 is lower than what it would really be available to spend) simplification which was that the MMA interest rate would only be equal to the inflation rate instead of something a little higher. Even with this conservative simplification I showed that a person could spend an inflation adjusted $8314.44/yr more (a 35% increase) between ages 62 & 70 by delaying the start of SS till age 70.

Now CFB is going to say that the Spitzer article says you’re almost always better off by taking the money sooner. I think the key is the “almost”. FireCalc’s safe withdrawal rates are driven by the worst 5% of the investment scenarios. It doesn’t care about “most”, just about “worst”. It’s possible (I can’t see the figures) that Spitzer has a few low interest cases where the waiting option wins. If so, that would be consistent with FireCalc.

And I think the key to what is better for a given person is said person's motivation. If said person wants to leave as large an estate as possible when they die then taking SS at 62 is probably a good plan. On the other hand if said person's motivation is to spend as much as s/he safely can between the ages of 62 & 70 then consider my example or some variation of it.

One more thought, this discussion and included examples are only addressing a couple of possible motivations. There are other considerations (eg. a desire to maximize survivor's benefits) that can weigh into the decision of when is the best time to start taking SS.

 

[Independent]

... The annual figure for SS if started at age 70 would be 1979*12 = 23748 not 21348. Using the correct SS numbers in FIRECalc and all the other parameters the same as Independent stated gives an annual WD of $23,684 when you start taking SS at age 62 and an annual WD of $24,743 when you start taking SS at age 70 (a $1059 difference which is an ~4.5% increase due to waiting).

The next point I'll make is that since in my example the motivation is to spend more between the ages of 62 & 70 (in real terms) then what will be spent after age 70, FIRECalc is not the right tool to use to try and determine how much more can be spent. .....

Thanks for the explanation. It looks like I could have saved some time and simplified my post a lot if I would have taken 10 seconds to check the monthly number vs. the annual number. This is especially embarrassing because I noticed that the ratio of the age 70 benefit to the age 62 benefit was odd, but still didn't figure out the problem.

With your correction, my (b) disappears. Note that my (d) is almost identical to the numbers you found. It's tempting to go back and edit the post to get rid of a bunch of extraneous language. But I think the general pros and cons of 62, 66, and 70 are still correct. And it's certainly true that the biggest issue is that different goals will lead to different decisions.

I'll agree that FireCalc (or at least the simple 4% SWR) is not the best tool for retirees who can see that they have more than enough resources to cover their "needs", and would like to use the extra sooner rather than later.

If I'm reading you correctly, your examples seem like "two bucket" strategies. The first bucket gets SS plus enough assets to cover the income level that represents "needs", using conservative assumptions (like MM interest rates, and 4% safe withdrawal rates). The second bucket gets everything else. Both the principle and interest in the second bucket can be spent in whatever pattern you like - like spend it all while you are young and healthy. It can use more aggressive assumptions.

I think that multiple buckets make a lot of sense for putting some structure on the confusing reality that you want a lot of confidence in you basic income, but can live with more risks on the "enjoy life" money. I'm sure I promoted this in some other thread.

However, I can also appreciate the argument that the real "optimal" analysis puts all the money together and uses a withdrawal strategy that varies as the real investment returns play out. I know that FireCalc will allow stepped withdrawals, but I didn't try that to see if I could bet "better" results. FireCalc also has withdrawal strategies that vary with returns, and I haven't tried them. I wouldn't be surprised if people with lots of assets develop a plan where that taking SS at 62 is "optimal" in that kind of analysis.

 

[Cut-Throat]

Good posts! - I have some thoughts on this as well

1.) Since FireCalc is a 'worst case' tool, it might not tell you what you would actually do in the future. And in all probability you will not be in the Worst case situation. So again FireCalc is not the proper tool for this.

2.) The take S.S. now or later is not something you HAVE to decide at age 62. You get to decide every year until you reach age 70. If the market crashes 80-90%, at age 63, you might decide to take S.S. at age 63 and start buying stocks at bargain prices. Which is exactly what FireCalc does! If the market is fairly stable and interest rates are low, you can let your S.S. ride and get about 8% interest for your waiting. And if by chance you live to age 100, you'll be glad you did!

3.) I think the main problem with SS analysis here, is that a 40 year old sits down and runs FireCalc for a decision that does not have to made for over 20 years. He then proceeds to call everyone a moron for not taking it at age 62, and that any discussion about dellaying SS to age 70, involves 'made up math', wrong numbers etc. etc. - There are valid reasons to delay SS or take it early.

I find out more about this topic, everytime someone posts a thread like this and I'm convinced others do also, providing they have an open mind.

 

[d]

There are valid reasons to delay SS or take it early.

ditto

 

[Corporateburnout]

I ran my numbers on Firecalc and It shows that I get 100% success rate at 62 and 66 and 97% at 70.

Since that I plan to leave some money for my heirs I'll take it at 62.

Again the decision will depend on the retiree's own circumstance....

 

[ziggy29]

The right answer for any individual may vary. And the right answer isn't ALWAYS the one that produces the maximum expected income.

 

[haha]

I ran my numbers on Firecalc and It shows that I get 100% success rate at 62 and 66 and 97% at 70.

Since that I plan to leave some money for my heirs I'll take it at 62.

Again the decision will depend on the retiree's own circumstance....

Since FireCalc assumes 70/30 S&P/commercial paper; and since equity returns as well as interest rates have been reasonably good during the 20th century taken as a whole it would be odd indeed if this were not the result returned by Firecalc.

This really cannot be made into a quantitative exercise, because the two paths have fundamentally different goals. One is for greater security and an element of insurance against out of sample outcomes, as well as "longevity insurance". The other is "pedal to the metal it worked before it will work again".

I think for a "do-it-yourselfer" who can afford to wait beyond 62, it is a generally good decision to wait until a) you need the money now b) equity premia have gotten much higher or c) your health has deteriorated meaningfully. I believe this is an even better strategy if you are a woman or the higher earning half of a married couple.

Ha

 

[Mwsinron]

I always figured it would be better to take it early if you are a high income earner. The difference doesnt seem to be as big if you are a lower end or middle income person.

 

[haha]

I always figured it would be better to take it early if you are a high income earner. The difference doesnt seem to be as big if you are a lower end or middle income person.

This is something I hadn't thought of. Could you explain the reasoning underlying this approach?

Ha

 

[Mwsinron]

This is something I hadn't thought of. Could you explain the reasoning underlying this approach?

Ha

What I am talking about is explained here. It made sense to me .

http://www.retireearlyhomepage.com/soc_security.html

 

[haha]

What I am talking about is explained here. It made sense to me .

http://www.retireearlyhomepage.com/soc_security.html

I get it. But it seems that this pertains more to when to retire, rather than when to start receiving benfits.

Ha

 

[d]

I get it.

i don't.

 

[jdw_fire]

i don't.

The article quoted

What I am talking about is explained here. It made sense to me .

http://www.retireearlyhomepage.com/soc_security.html

has nothing to do with

I always figured it would be better to take it early if you are a high income earner. The difference doesnt seem to be as big if you are a lower end or middle income person.

Instead it talks of what effect stopping work before you have 35 years of payment into the SS system has.

 

[New Thinking]

If I may, I will reiterate my findings that have been debated before - but seem to be missed. I will be presenting these next month at The Pension Research Council at Wharton..The working paper will be available for free on their website shortly thereafter.

First and foremost, the calculations for an age 70 amount appear to be wrong when you are putting them in FireCalc..I am just eyeballing them, but there is confusion here. The age 70 amount should not be 76% more than the age 62 amount because the 8 years of COLAs will create a much larger benefit at age 70. So if you are going to compare $13,488 at 62, I would not use $23,738 at 70, but assume a number of inflation such as 3% which puts the age 70 amount up to $30,083..You do not do FireCalc in Today's dollars, so you cannot use the un-inflation-adjusted amount at age 70 either. This is the most common mistake made in calculations and the Spitzer paper may also be making this mistake..I have to review his numbers, but he does state a 76% increase between age 62 and 70, so that is probably what he is doing..As did the University of San Diego professor's article which CFB linked to before (and I am sure he will harrass me for bringing it up again)..But a starting point of $30,083 bersus $23,738 at age 70 is a huge difference..Plug these into FireCalc and rerun the numbers.

• Point number two is that Spitzer ignores the spouse's benefit. You cannot do this and provide an accurate break-even since the spouse will inherit a higher delayed retirement benefit..It must clearly be part of the equation..It also means that the spouse should start her/his benefits sooner rather than later.

• Point three - I think it is not the best way to analyze Social Security break-even points to look at it as a present value..Social security has risk management characteristics and there is a "value" on the fact that it pays someone income for a longer time if they outlive a life expectancy.
  • Point four - Spitzer's paper says that taxes don't matter and he is 100% incorrect on this..And this is probably where our research is unique and we have discussed it on this forum before..You cannot look at the taxation of SS benefits in a vacuum. IRA or pension income causes the taxation of SS income..Marginal federal tax rates of 46.25% and heading higher when Bush's tax cuts sunset. Higher (delayed SS) reverses this "tax torpedo" because IRA income is lesser or eliminated and SS goes into the Combined Income formula at a 50% rate..And even then it is the lesser of three tests, which means that income after age 70 can have dramatically less taxes..This information will be laid out and cites our work in the upcoming book "The Wall Street Journal's Complete Retirement Guide"
  • Point Five - Everyone says that waiting to take SS is for those who don't care as much about leaving a legacy, but I disagree. Sure, it is possible to receive higher returns with the stock market over time, but if you are paying much higher taxes, the returns are often diluted. And we are assuming that we will have the staying power to stay in stocks as we age. Further, Spitzer does cite William Reichenstein's work that SS should be part of the asset allocation equation...and I fully agree. Use a higher amount of SS to lower your bond holdings..That way you don't have to live with lower returns in our current low-yield world and you won't have to live with lower short-term yields in MM if/when short-term rates go lower again..The potential to leave a legacy (or at least the probability) may be increased by waiting to take SS.
    
    Just some additional food for thought for the discussion.
  Everyone's situation is indeed unique but inflation, taxes, longevity and investment risk management, and a low fixed income world should be part of one's thinking.

 

[Running_Man]

• Point Five - Everyone says that waiting to take SS is for those who don't care as much about leaving a legacy, but I disagree. Sure, it is possible to receive higher returns with the stock market over time, but if you are paying much higher taxes, the returns are often diluted. And we are assuming that we will have the staying power to stay in stocks as we age. Further, Spitzer does cite William Reichenstein's work that SS should be part of the asset allocation equation...and I fully agree. Use a higher amount of SS to lower your bond holdings..That way you don't have to live with lower returns in our current low-yield world and you won't have to live with lower short-term yields in MM if/when short-term rates go lower again..The potential to leave a legacy (or at least the probability) may be increased by waiting to take SS.
  

• 
  
  Of course this will be impossible for anyone in the sample that would die I would say between the ages of 62-75. If on the other hand you are assuming that stock market returns less the additional withdrawls needed to cover the SS @ 62 will more than offset the additional portfolio gains made by not needed the additional withdrawl, I would like to see those calculations.

 

[youbet]

• • Point number two is that Spitzer ignores the spouse's benefit. You cannot do this and provide an accurate break-even since the spouse will inherit a higher delayed retirement benefit..It must clearly be part of the equation..It also means that the spouse should start her/his benefits sooner rather than later.

  
  If you have a legal spouse.
  
  If your spouse doesn't die before you.
  
  If your spouse is not disqualified from your SS by the GPO or other factor.

 

[youbet]

The potential to leave a legacy (or at least the probability) may be increased by waiting to take SS.

Or the potential to leave a legacy (or at least the probability) may be decreased by waiting to take SS.

 

[FIRE'd@51]

NT,

I look forward to reading your paper when it is available on the web. It seems as if you are attempting to address a lot of interesting (and often-ignored) issues.

I do want to make a couple of comments, though.

While I'm not sure if FireCalc handles the delayed SS correctly, I agree with the 76% number used in the papers referenced by CFB. Both the age 70 amount and the age 62 amount grow at the inflation rate, so the factor (1+ I)^8 appears in both the numerator and denominator of this ratio, and thus, cancels out, keeping the ratio at 1.76.

With regard to the "tax torpedo", it is not obvious to me that it always favors delaying. You could also have a "tax torpedo" effect with the larger age 70 distribution. Also, the RMD's for those of us with large traditional IRA's (including Rollover IRA's) could push more of that extra SS income into higher brackets (even higher after the Bush tax cuts sunset).

I also would like to point out that someone who takes SS at 62 pays no Medicare Part B premium for the first three years. I believe that premium is currently $93 per month, and I read in another post (I think by OAG) that it is scheduled to increase 17% next year. Another tax issue could be the recently passed means-adjustment to the Medicare Part B premium. I believe this kicks in at AGI's of 80K for singles and 160K for joint returns.

Finally, I just want to point out that, as we age, there is, unfortunately, a higher probability that one spouse will die. The surviving spouse will have to file as a single taxpayer rather than joint. This means all the brackets will "kick in" at half the taxable income, thus greatly raising marginal rates for the suriving spouse. Additionally, the SS provisional income thresholds are lower for a single taxpayer. All of this makes the tax issue very difficult to analyze (at least for me).