Friday, September 8, 2017

Social Security Claiming and Pig Races

The retirement planning question I am asked most frequently is when to claim Social Security benefits. I received a letter from a reader and a question from a friend at the local cafe on this topic just in the past three days.

Another friend sitting three rows behind me at a baseball game once stood up in the middle of an inning and shouted, "Hey, Dirk, what is the break-even age for claiming Social Security benefits?"

Since most Americans have not saved nearly enough for retirement, when to claim is often moot. Most people need their benefits right away and can't afford to delay them. But, if you can afford to wait, let me try again to explain why you probably should.

Here's the comment from a reader.
"My wife and I will retire in 4 or five years at about 66. We will each have small pensions and access to a little investment income. Social Security will be the 4th revenue stream. A financial adviser really wants us to wait until 70 to claim Social Security, but I'm concerned that we will have to erode our investments during that 4 or 5 year wait. Trying to forecast what that will cost us in lost investment income vs the higher Social Security benefit when claimed later is like betting on the pig races at the fair. An educated guess, a WAG, a stab in the dark. It doesn't help with the confidence and security we seek in retirement."  – Chris
(Isn't it interesting how often pigs come up in my retirement blog? I need to give that some thought. Think Like a Bayesian Pig.)

Chris, I believe the reason you are not getting a satisfying answer is that you're not asking the right question.

I deduce from your comment that you understand that if you live a long time, you will be better off financially by delaying claiming your Social Security benefits but, if you don't live a long time, you will be better off claiming them right away.

You are having difficulty deciding whether to bet that you will live a long time or a short time after you retire. You might as well bet on that pig race because just as you have no idea which pig is fastest, you have no idea how long you will live.

Social Security retirement benefits are a form of insurance. In fact, Social Security is the commonly used term for the federal Old-Age, Survivors, and Disability Insurance (OASDI) program. Before we delve into old-age insurance, let's consider a more familiar form of insurance, auto insurance.

By car insurance industry estimates, you will file a claim for a collision about once every 17.9 years.[1] That isn't terribly interesting information because you might not be average. You might go 30 years without an accident or have one tomorrow.  Planning your retirement or buying insurance based on averages is a very serious fallacy. Insurance should protect you from catastrophes, not from average losses.

You might ask – though, you probably don't — why you should pay thousands of dollars each year for car insurance when there is a 50% chance that you won't have an accident for nearly 18 years.

The answer is simple. A car accident could be financially catastrophic. Without insurance, it might cost you $40,000 to replace a car, not to mention hospital bills and a million-dollar liability judgment. So, we pay insurance premiums hoping that we never need to file a claim in order to avoid a potentially catastrophic outcome.

That's the definition of insurance. We accept a small, guaranteed loss (the cost of insurance premiums or delayed Social Security benefits) in exchange for protecting us from an unlikely but potentially catastrophic loss.

A reader asks: Is claiming Social Security benefits harder than betting on pig racing?
[Tweet this]

How does that relate to delaying Social Security benefits, a.k.a., Old-Age, Survivors, and Disability Insurance?

Delaying Social Security benefits is like purchasing additional longevity insurance. The equivalent premium payment is forgoing some Social Security benefits that we would otherwise receive if we claimed at age 62.

The equivalent catastrophic loss in retirement is becoming impoverished when we are very old. Delaying claiming Social Security benefits is a way to buy more longevity insurance. Said differently, it's a way to transfer some guaranteed income early in retirement to provide more income later in retirement should we live to be very old.

This is the catch, of course. We don't know if we will live a very long time and, if we don't, we will have given up those early benefits for no reason, like buying auto insurance and never needing to file a claim.

When we buy auto or life insurance, we don't know if we will have a huge accident or die while dependents still need our job income, either. We may be paying those premiums with no return and, in fact, we hope we are. We don't make a bet on when those things might happen, we buy insurance to protect ourselves if they happen because an uninsured loss could be really bad. The alternative is to hope we're lucky.

Short retirements are cheaper than long retirements. Let's assume that you will need $50,000 a year to cover your living expenses after you retire. A retiree who leaves the workforce at age 62 and dies at age 67 will need $250,000 to fund retirement. A worker who retires at age 62 and lives to age 92 will need $1.5M to fund retirement.

(Both numbers are actually smaller than this when we consider the time value of money, but let's keep it simple for now.)

In other words, if you don't live long, your retirement will be relatively cheap and if you become quite old it will be relatively expensive.

When you delay claiming Social Security benefits and only live a short time, you will reduce your income but it won't hurt much because you will also have a cheap retirement. If you live a long time, you will increase your income by delaying claiming and you will need the additional income because you will have a long, expensive retirement.

Now, let's look at the alternative bet of claiming your benefits early.

If you live only a short time after retiring, you will have maximized your Social Security income for a scenario in which your retirement wasn't that long or expensive. If you live a long time, you will have minimized your Social Security income for a scenario in which your retirement is relatively expensive.

The goal of delaying claiming your Social Security benefits is to make sure you have more money in the worst case scenario, living a very long, very expensive retirement. It's like taking some small, certain losses by buying car insurance to avoid an improbable catastrophic loss if you have the big accident.

Now, Chris, I suspect that you're trying to place a bet on whether or not you will live a long time. If you're healthy, you can't know how long you will live. If you're going to place that bet, you're correct, you might as well bet on the pigs.

A better way to frame this decision is as a purchase of additional longevity insurance to protect your household against a very long, expensive retirement instead of framing it as a bet on how long you might live.

Your comment says, "It doesn't help with the confidence and security we seek in retirement."

If it's confidence and security you seek, buy insurance. You get that by delaying claiming Social Security benefits as long as you can afford to do so, thereby taking the catastrophic scenario, inadequate income in old age, off the table.

You also say, "Trying to forecast what that will cost us in lost investment income vs the higher Social Security benefit when claimed later is like betting on the pig races at the fair."

You're trying to compare two very different things. Your investment portfolio is a liquid asset with no longevity guarantee — you can outlive your portfolio. Annuities and Social Security benefits are illiquid but they will provide income if you live to 120. You probably need both.

Like not knowing how long you will live, you can't know how much investment income this decision will cost you. You can only guess. If your investments go south, you will have been much better off with more Social Security income, and vice versa.

Since you don't know how long you will live or how your investments will turn out over the long term, the trick is to make sure you have an adequate floor of Social Security benefits, fixed annuities and the like to protect against a very long life and poor investment results. Invest what's left for upside potential and liquidity.[2] That will tell you whether delaying benefits is a good strategy for your personal situation.

It's a somewhat complicated optimization, but presumably, that's why you asked your advisor for help in the first place.

Delaying claiming Social Security retirement benefits is the most cost-effective longevity insurance you can buy. If you think of it as insurance and not a bet on how long you will live, you may find it makes more sense than pig races.


". . . no one knows the best asset allocation in advance", Does The 4% Rule Work Around The World?, Wade Pfau.

NOTE: My blog seems to be having problems accepting comments, as it does from time to time. Sorry, I hope to have that corrected soon. Meanwhile, you may need to click the "comments:" link in the very last line of the post that begins " 1 comment:" to display the comment entry box. If that doesn't work, you can email me at


[1] How Many Times Will You Crash Your Car? Forbes magazine.

[2] The Retirement Café: Build a Floor, Place a Bet.


  1. This comment has been removed by a blog administrator.

  2. Posted for Jeff Klugman:

    re when to claim social security:
    your own longevity is far from the only factor relevant to this choice.
    choosing when to claim, you are also making an entirely different set of bets about the fiscal environment:

    1. what is the probability that social security receipts decline enough within the next decade so that payments will be cut by 25%?

    2. might your payments be reduced either directly or by increased taxation or means testing?

    3. will the cpi adjustment measure your real cost of living? eg medical expenses are undercounted in the cpi, but rising much faster than cpi. this would imply that dollars received sooner are worth more than the "adjusted dollars" received later. even if you have no need for those dollars, claiming them sooner and investing them in instruments hedging medical and other inflation might be a wiser strategy.

    1. Jeff, these are all somewhat important considerations for your retirement plan, but I disagree that they are important factors in deciding when to claim.

      If “Social Security receipts decline enough within the next decade so that payments will be cut by 25%”, those cuts would presumably apply no matter what age you claimed. Unless the structure of payments encouraging later claiming is also changed, delaying benefits would still provide more income late in retirement, though benefits claimed early and benefits claimed later would both be 25% less than before the cuts.

      Yes, your payments might be reduced someday either directly or by increased taxation or means testing. Again, that would presumably affect you whether you claim early or late. And, tax changes are as hard to predict as lifespans.

      “Will the CPI adjustment measure your real cost of living?” Probably not. It’s an average and your local “personal” inflation will likely be more or less. “Dollars received sooner are worth more than the "adjusted dollars" received later unless your personal cost of living increases slower than the adjustments Congress decides to make. Again, too much uncertainty to bet on.

      When we do that “last great accounting” and measure the ways in which our actual retirement spending varied from what we expected – and it will vary, because retirement expenses are highly unpredictable – the difference between the inflation we experienced and the inflation for which Congress compensated with our Social Security benefits won’t make the top 10 list of greatest variances. When we claimed Social Security benefits will.

      Lastly, all due respect, I think claiming Social Security benefits early and investing them is a terrible idea. You would be taking the most secure retirement asset you have and converting it into a risky portfolio. If you’re married, you would also be exposing your spouse to that additional market risk and, if she is the lower earner, limiting her survivors benefits in the process. Framing this as “hedging” and somehow reducing your risk seems strange to me.

      You are correct that a good retirement plan will consider the risk that Social Security benefits might be reduced and that inflation adjustments might not keep up with our personal inflation rates but I don’t see either as having much impact on deciding your claiming age.

      Thanks for the discussion and sorry about the “comments posting” problem.

    2. While I default to the "claim later" view as well, is it really so unreasonable to claim early and invest? Consider:

      1) you are not "giving up" the entire annuity; a substantial secure stream remains to "insure" against longevity

      2) If you are able to invest at more than the 2.9% (I believe) real rate used to discount the actuarially "equivalent" benefits you would presumably come out ahead. Not an insignificant hurdle, granted.(You could be helped by a lower tax bill on investment income if your benefits would be otherwise taxed.)

      3) Since you don't know you longevity, you are hedging your bets by splitting the benefit into a portion that wins if you live long and a portion that wins if you die early.

      It's tough to make categorical statements on this topic.

    3. I don't find it tough.

      You are suggesting converting the most secure retirement asset available and the most cost-effective longevity insurance – an asset that most people should be delighted to purchase more of, if they could – into a bet that one needs to "hedge."

      Hedging your bets by splitting them into a portion that pays off if you live long and one that pays off if you don't is an excellent idea. It's the basis of floor-and-upside strategies. But, converting some of the absolute best floor asset (Social Security benefits), and one that is relatively scarce, into part of the upside bet will seldom be the best allocation of assets.

      I don't believe people should bet that their house will not burn down, that they won't have an automobile accident, or that they won't outlive their standard of living when cost-effective insurance is available.

      But I do agree that all retirees should choose the bets that make them comfortable and if this is one that you're happy with, I hope you win it.

      Thanks for the comment!

  3. Dirk,

    The insurance focus is on point - I like the parallels with auto insurance. SS claiming is a direct response to risks with aging, and as you have pointed out, longevity acts as a sort of multiplier for a host of other risks - more exposure and typically fewer options and resources to respond to them.

    One part of this puzzle is people don't seem to have much grasp on their own longevity. In discussion they mention their own parents, and maybe a couple of other family members who come to mind, but often don't seem to think even those apply to them.

    This comme ci comme ça attitude does not help decision making.

    Any thoughts about the usefulness of longevity estimators in this context? e.g.,

    1. Hi, Tom, great to hear from you! I assume the changes in the HECM world are keeping you busy, though I do recall that you predicted them.

      So, you raise an interesting question regarding longevity estimators.

      I believe that some outcomes are unacceptable and that becoming impoverished in old age is one of them. For risks with unacceptable outcomes, I like to rely on insurance and not probabilities. At some point, the magnitude of the risk becomes so great that simply asserting that "it probably won't happen to me" is inadequate mitigation.

      I have no idea why some people are convinced that they can predict whether they will or will not live a long life, despite all evidence to the contrary, but I run into it all the time. And, they say it with such certainty!

      I knew a man who had a serious childhood illness, never met a green vegetable he liked, got little exercise during retirement, and had long-term exposure to the sun with no protection (his father died from melanoma). His wife was 10 years younger and lived a far healthier lifestyle throughout her entire life.

      He passed away at age 98 and she succumbed to an automobile accident at age 80. Not what the statistics would infer.

      On the other hand, the drug Gleevec was introduced in 2001 turning a fatal form of cancer (CML) into a condition manageable with a pill and instantly increasing the life expectancy for thousands of people.

      Longevity estimators demonstrate a common fallacy in retirement planning: basing an individual's plan on population averages. Life insurance companies can predict fairly accurately how many of their customers will die each year but they have no idea which individuals that will be. It doesn't much matter to the insurance company which individual policy holders will die, but it matters a great deal to the policy holders.

      Basing life expectancy on a sample of 400,000 people, as this longevity estimator site claims, probably provides a pretty good estimate of how long lots of people like you will live. But, it can't tell you how long you will live and that's the only important life expectancy prediction for your retirement plan.

      I think longevity estimators can have two impacts. They might convince some people who are not planning for a long retirement to rethink their longevity expectations. That would be safe and good. They might also convince others that they are unlikely to live a long life, influencing them to accept a riskier retirement plan with a shorter life expectancy.

      I think everyone should assume, for planning purposes, a long life expectancy to at least age 95 regardless of their personal beliefs or the conclusions of life expectancy estimators.

      I also think that this is an excellent scenario in which to use a dynamic planning strategy as championed by people like David Blanchett, Larry Frank, Sr., and others. With a dynamic strategy, retirees will modify their plan annually to compensate for a constantly-decreasing life expectancy instead of guessing when they'll shuffle off this mortal coil.

      Thanks for the question!

      (Here's a clickable version of Tom's link.)

    2. Tom, just got my results back from the life expectancy estimator. Half of people like me will live to age 90. It would be more helpful to know which half I'm in! :-)

      By the way, it recommends that, based on my life expectancy, I postpone claiming Social Security benefits until age 70 (which I plan to do).

      Thanks again!

  4. Dirk,

    I agree with your penultimate recommendation that everybody SHOULD plan to 95, but that's hardly the coming in assumption for most. Before I retired as practicing financial advisor my take was our relatively well off clients tended to be those who have longer than average life expectancies, and of those who tried a life expectancy estimator, tended to get encouragement to delay. I suspect that applies to a much broader set than my clients. I concur with the recommendation for everyone to delay as the "safest" approach. As we all have seen, "spend now" is compelling.

    And I personally advocate, and practice, your ultimate recommendation - a dynamic strategy.

    (ps. It helps to be older when you do the life expectancy estimators. The older you are, the longer your life expectancy estimate. I'm not going to do an update until next year when I hit 70. A good day to run it would be the day I start my SS. My wife and I both are delaying ours, and I am enjoying getting monthly checks on her records. (for the SS savvy, that subtlely conveys she's older than I am.))

  5. I thought this was an excellent post, and only wish more would read it. I grow weary of the "breakeven" argument offered for claiming SS at 62; it indicates a willingness to gamble one's late-life financial well-being against being able to say - post mortem - that one didn't 'leave any money on the table'.

    I am not a financial planner, but as a reverse mortgage originator find myself often in conversations with clients about SS deferral in conjunction with discussions about HECM strategies. I have always used the insurance analogy - small payment to mitigate an unlikely but possible catastrophic event - in the context of setting up a HECM line of credit as a financial safety net and in the absence of any current need. Most - financial professionals and consumers alike - don't "get it", which continues to baffle me as they go on writing their checks for auto, home, health, disability, LTC and other insurances, hoping none of they never need to make a claim on any of those policies.

  6. Dirk, couldn't claiming at 70/70 actually create more longevity risk for a couple than claiming at 62/70? One spouse might pass relatively early and the other live a very long life. The surviving spouse would be left with less assets because the deceased spouse delayed benefits.

    1. I answered this below, but want to add that a lower-earning spouse claiming benefits at 62 or any other age below full retirement age (FRA) has complications you should consider. Please see my response below with a link to an article containing an example.

      The surviving spouse in your example would receive less income only if the higher-earning spouse died at age 67, 68 or 69 before claiming because an earlier death would pay out as if the deceased had claimed at FRA (66). Worst case (death at 69 or age 70 before claiming) would pay out only about 16% less and, of course, the couple's retirement expenses would decline.

      I don't think it would create a lot more longevity risk to postpone benefits, at least nowhere near as much as longevity risk as not delaying creates.

      Thanks for asking!

  7. I'm not sure both spouses claiming at 70 increases longevity risk by much, but it certainly leaves money on the table. For married couples, it will generally be advantageous for the higher earner to delay claiming as long as possible and for the lower earner to claim at full retirement age. (Claiming earlier than FRA, about age 66, can cause problems.)

    A lower-earning survivor will generally receive a larger survivor benefit if the higher earner delays those retirement benefits. If the higher earner hasn't claimed before he dies, survivors benefits are based on the deceased' FRA, so it would depend on the specifics.

    I recommend running an optimizer like to be sure. But your point is a good one. It typically isn't optimal for both spouses to claim late.

    Thanks for the comment!

  8. Hi Dirk. Thank you for your insightful blog. Your stories are always on spot for me.

    Yesterday I read your blog about freezing our credit reports. I tried to follow your advice and was successful freezing Experian and TransUnion, but the Equifax site refuses to process my request. I get a computer error every time I try. I've been trying for two days now.

    It figures that the guilty agency is the one I can't freeze. Plus I've confirmed that my identity was stolen on their site.

    1. According to Experian's website, "You can request a security freeze be added to your credit report by going online to Experian’s Freeze Center, by phone at 1 888 EXPERIAN (1 888 397 3742), or by mail to Experian Security Freeze, P.O. Box 9554, Allen, TX 75013."

      I would assume their hair is on fire and the website is suffering as a result. Snail mail might be the best option.

  9. Hi, Dirk:
    Would you please let us know what problem it causes by claiming earlier than FRA?
    (Claiming earlier than FRA, about age 66, can cause problems.)


    1. I don't claim to be a Social Security expert, so I am going to suggest you get advice elsewhere.

      I can tell you that there is an earnings test if the person on whose earnings the benefits are paid is younger than full retirement age (FRA). See Social Security Earnings Test Ensnares a Spouse.

      Depending on your age, claiming before FRA can affect a restricted-application strategy. See Life After 'File and Suspend'.

      Widows benefits are also reduced if claimed before FRA.

      I ran several of my own scenarios through the retirement optimizer at Financial Engines and the optimal strategy never included claiming spousal benefits at age 62.

      That doesn't mean yours won't. The rules are quite complex and I gave up on mastering and remembering them long ago. As I see it, there are basically three ways to analyze your Social Security-claiming strategies: become an expert, hire an expert, or use software.

      Here is a list of free optimizers, though I use MaximizeMySocialSecurity, a paid service.

      If you want to become an expert yourself, I recommend Mike Piper's book, Social Security Made Simple, and his blog, Oblivious Investor.

      I recommend that you use an optimizer with your own specific earnings estimates, etc., rather than trying to understand all the rules. It's easy to miss one.

      Thanks for writing.

    2. Perhaps this article will help explain.

      Unless you were grandfathered into the file-and-suspend strategy (too late, now), a lower-earning spouse (typically the wife, sad to say) cannot claim spousal benefits until the higher-earning spouse claims retirement benefits. She would have to claim her own benefit, which might be lower than half the higher-earner's benefit (the spousal benefit).

      She can switch to spousal benefits once her spouse claims if it would larger. It will equal her already-claimed benefit plus the difference between her spouse's full retirement age (FRA) benefit and her own FRA benefit and that will be less than the spousal benefit she would have received by waiting to claim FRA (66 for people about to retire).

      The article provides an example.

      Thanks for the question!

  10. Dirk,
    I agree that the question is comparing an insurance plan (waiting to 70) with an investment plan (Money spent from 66-70) and like the extended floor benefits of waiting.
    Is it possible to create a comparison that might help? I thought to consider the additional funds at age 70 above FRA funds as an annuity called 70A. There is some value of an annuity we could approximate. There is also retirement funds spent from 66-70 to cover expenses that now have to be spent called FRA66. This cash stream of expenses could be considered as a separate amount.
    Is it possible to ask the question of what would be the % return of the cash stream FRA66 over the 4 years that would be required to be able to purchase an annuity 70A at the age of 70. In this way you are asking if your investment money lost by spending it from 66-70 could grow to buy the additional annuity income available by waiting and what that return would have to be.
    Thanks for all your posts.

    1. Frances, I plan to do that analysis in an upcoming blog. Spoiler alert: you will find that the deal you get from Social Security is way better than anything you can buy on the market.

      I don't think you need to look at expenses; they will be the same in either scenario. It's income that changes.

      The last part of your suggestion, the return you need to "break even" gets to the core of the problem of comparing insurance to stock returns – just because you need it doesn't mean you will get it. And, predicting stock returns over a short period not exceeding 8 years is impossible. You might do better and you might do worse with investments, which is fine if that's a bet you want to make.

      That's identical to exploring how long you need to live for delaying benefits to break even. Needing to live a certain number of years doesn't seem to be optional.

      I believe you're correct about comparing the cost of an annuity, though. Breakeven analysis pretends that the retiree who is delaying gets nothing in return, but she is actually buying insurance and that makes "break even" shorter.

      Thanks for the question!


  11. Dick
    I agree with you about postponing Social Security benefits as long as it does not put a financial burden on you. The idea that folks would claim Social Security Benefits at FRA or earlier and would invest those funds and come out ahead is not realistic for most people.
    Most people talk about investing but do not execute on their ideas. I administered a Pension Plan and found that most employees went with the default option rather than make choices. The Plan ended up changing the default option to a target funds from a money market fund as large percentage of employees were going with the default account rather than selecting from the other investment options.
    Of course if a person has health issues or needs the money for bills, the person may decide to take the SS benefits before 70 years of age.


  12. This comment has been removed by a blog administrator.