Executive Summary
Despite a growing body of research suggesting that most retirees would benefit by delaying the onset of Social Security payments, the majority who are eligible still elect to begin receiving them as early as possible. In no small part, this appears to be attributable to a "take the money and run" mentality from retirees, who simply don't see the value of delaying as being worth the risk of foregoing benefits. And without a doubt, there is a material risk that the retiree will not live to the so-called "breakeven point" where the delay in benefits is worthwhile.
However, what most retirees fail to recognize is that while there is a risk to delaying benefits and never fully recovering them, the upside for living past the breakeven point isn't just that the money is made back; it's that the retiree can make exponentially more. And in fact, these asymmetric results - where the retiree only risks a little by delaying, but stands to gain far more in the long run - are further magnified in situations where the client lives dramatically past life expectancy, experiences high inflation, and/or gets unfavorable portfolio returns - which are, in fact, three of the greatest risks to almost every retiree.
As a result, the reality is that delaying Social Security benefits may actually be one of the best triple-hedges available to any retiree - simultaneously protecting against poor returns, high inflation, and longevity!
The inspiration for today's blog post is the recent Journal of Financial Planning article entitled "How the Social Security Claiming Decision Affects Portfolio Longevity" by William Meyer and William Reichenstein, which was covered in our recent Weekend Reading column. The article shows how delaying Social Security benefits can increase the overall longevity of a retirement strategy that includes Social Security income and a portfolio.
The Impact of Delaying Social Security
The primary reason that delaying Social Security can have such a benefit on increasing the longevity of a portfolio is because of the asymmetric nature of delay decision.
For example, let's look at a scenario where a 66-year-old - currently at Full Retirement Age - with a PIA of $1,000/month, chooses to delay benefits by 1 year, earning a delayed retirement credit of 8%. As a result of the delay, the client will receive a monthly payment of $1,080/month beginning one year from now, at a "cost" of not receiving $1,000/month for the next 12 months. Thus, the client essentially starts out $12,000 in the hole, and makes it up $80/month at a time.
In reality, though, the client recovers the cost of foregone payments slightly more rapidly over time, because the client doesn't merely receive an extra $80/month. The client actually receives $1,080/month increased by annual cost-of-living adjustments, in lieu of receiving $1,000/month payments now, also increasing by annual COLAs. At the margin, this means the client is actually recovering the $12,000 of foregone payments at a pace of $80/month, where the $80/month itself is increasing by COLAs each year thereafter.
How Delaying Social Security Compounds Over Time
Assuming a moderate 8% growth rate on the available funds (e.g., the $1,000/month collected for the first year, compared to the extra $80/month for the client who delays for a year) and a 3% annual COLA, the chart below reveals that it takes just over 20 years for the client to come out ahead due to the delay decision.
However, note what happens in the final years of the chart. While it takes approximately 20 years for the client to initially dig out of the $12,000 hole created by delaying initially, it takes only 6 more years thereafter for the client to go from even to up $12,000. And it takes only another 4 years after that to make another $12,000, and then just over 2 more years to add yet another $12,000. In other words, as the client lives longer, the client doesn't just come out ahead; the client comes out exponentially ahead. Dying 20 years before the breakeven period costs the client $12,000; living just 14 years past the breakeven period brings the client ahead by nearly quadruple that amount, to just shy of $48,000.
Sensitivity to Assumptions
Notably, the value of delaying Social Security benefits is sensitive, both to the level of inflation (which impacts the COLAs) and the growth rate on the investments. Notably, even if the client's plan is to spend the early Social Security benefits, that means other investment funds won't have to be spent down, so the growth rate is relevant regardless of whether the actual Social Security funds will be saved or spent.
As it turns out, the lower the growth rate, the shorter the breakeven period and the greater the value of delaying Social Security benefits (because that initial year's worth of benefits won't have as much time to grow). In addition, the higher the inflation rate, the shorter the breakeven period and the greater the value of delaying, because the higher payments catch up and compound faster.
According, the graph below shows the original benefit delay (at 3% inflation and 8% growth), and an alternative scenario with 4% inflation and only 6% growth. In this case, it takes only 15 years to breakeven, instead of 20; in fact, by year 21, the client is already up over $12,000 in the higher-inflation-lower-growth scenario, and ultimately turns $12,000 at risk into almost $80,000 by the end of the 34-year time horizon. With just a 2% decrease in the growth rate and a 1% increase in the inflation rate, the economic value in the long run nearly doubled.
The True Value of Delaying Social Security
Notably, living 34 years past the starting point to harvest the full value illustrated in these charts is no trivial task. This example assumes an individual who is age 66 at full retirement age (thereby earning an 8% delayed retirement credit for waiting one year), which means reaching age 100 to get the full value shown here.
But on the other hand, that's the whole point to the value of delaying Social Security retirement benefits. The upside for outliving the breakeven isn't just to recover the amount at risk or to make it back again; it's to make exponentially more than the original amount at risk, if the client is fortunate enough to live a long time. Yet for a client who's seeking to hedge against the risk of outliving his/her money and increase the longevity of the portfolio, that's precisely the most desirable outcome; like any other lifetime-annuitized payment, it creates the most value when the client lives the longest, which is exactly when the client needs that upside value the most.
In addition, it's notable that delaying Social Security not only hedges longevity, it also hedges two other adverse scenarios that are otherwise harmful to the retiree: unexpectedly high inflation, and unexpectedly low returns. As noted in the charts above, an adverse outcome with either, or both, can go even further to leverage the value of delaying Social Security, decreasing the breakeven period and increasing the upside for materially outliving life expectancy.
Which means in the end, the true value of delaying Social Security is a triple-benefit of hedging longevity, poor returns, and high inflation, because of the asymmetrical way that delayed higher benefits compound in the later years. It won't necessarily win for every client, but as any good hedge should, it wins the most in the times the client will need it the most.
So what do you think? Have you ever framed the value of delaying Social Security as a longevity, inflation, and bad return hedge for clients? Do you typically recommend that clients delay their benefits? Do you think clients would be more receptive to delaying benefits given all the risks that delaying can hedge?
(Editor's Note: This blog post was featured in the 17th edition of the Carnival of Retirement on Young Cheap Living.)
Good article !!
The benefit of delaying cannot be overstated.
Simple example as follows:
Assumptions: Married person claims early at age 62 in 2008 versus delaying to age 70 in 2016. Also married person collects spousal benefit starting at age 66. Assume COLA for 2013 and later is 1.5%.
Then:
Benefit at age 70 received by delaying is 2.1x larger than benefit collected at age 62. This represents an IRR of 9.72%.
Cumulative breakeven occurs after 15 years (i.e. at age 76 or 2022).
Benefit starting in 2023 or age 77 is 80.6% higher via delaying to age 70 versus collecting early at age 62.
Assuming person lives to age 100 (i.e. 2046) cumulative benefits received via delaying versus collecting early will be 55% larger.
In addition surviving spouse, assuming his/her benefit was lower, would benefit as he/she would receive the higher survivor benefit derived by delaying. This increases the probability of the delaying strategy being successful for a married couple as both spouses must die early to make the claim early strategy a better financial alternative.
Hi, I quickly replied on my FB page. We need to add in one more factor to the benefits of optimizing Social Security: Tax Arbitrage! My words are here http://www.facebook.com/RetireeIncome
Also, we need to be careful about using the right growth rate. In practice this makes sense. But, from an academic perspective the rate should align to the underlying asset being valued. SS is like a TIP so there is an argument that 8% is too high. We have some research coming out on this topic in the Journal of Wealth Management in the next month.
Michael – good post!!!
Bill,
The growth rate should align to how the client will actually invest the money.
If the client would actually take the early benefits and invest them in a balanced portfolio earning 8% – or would spend them in lieu of a portfolio asset earning 8% – the discount rate should be 8%.
The growth rate of Social Security itself is captured in the series of cash flows THAT are being discounted. The discount rate should still reflect the assumed investment rate for the client (which by default I’m assuming is a generic balanced portfolio).
– Michael
Like the article and its premise.
I think one must also consider the fact that Social Security is underfunded and at risk, particularly for the affluent. I’m not sure that I would count on those benefit payouts projected from a delay. If SS were on a supportable path, I might feel more comfortable making the proposed “delay” recommendation. As it stands, that risk must also be factored into the decision. I think many people today prefer a “bird in the hand.”
…8% after tax sure would be nice. I’ll look forward to chatting with you at AICPA or NAPFA.
Note to Bill:
For a lot of people they will not earn 8% per year after tax by delaying. In fact for a lot of people 85% of SS benefits will be taxed. This will occur if AGI + tax-exempt income + 50% of SS security benfits is greater than $34,000 for single people and $44,000 for married people. Assuming these thresholds are exceeded and a 25% marginal tax rate then the after tax rate of return is 6.3% which would represent a 8.4% fully taxable equivalent yield.
Stan,
Give me a call tomorrow, I want to show you something. We have SS taxation (all 3 tests coded) integrated with the tax code and asset location logic.
Have you seen our http://www.ssanalyzer.com application?
913 766 7083.
For the record, we have thousands of people coming through our site for SS advice and most will not be taxed at 85%. You may have an affluent client base, but there are opportunities (withdrawal sequences) to minimize SS taxation even for the mass affluent.
Bill
Would using n+5 for the delay cost be a better comparison? This would assume we planned this decision in advance and bought a bond 5 years before delaying. Having that money invested at 8% for 5 years would be over $17,000 at deferral instead of the $12,000 used.. This adjustment would extend the break even period.
Great article overall and a good start to challenging conventional wisdom (whatever that is). The AICPA Guide to Social Security is another great resource. So much planning based on each spouse’s benefits, delay, etc.
Last thought, at today’s meager interest rates social security income looks pretty attractive.
Follow-Up to Bill
I am aware of these permutations and the planning flexibility they offer. Since I only deal with affluent clients the taxation of 85% of the benefits is an automatic assumption for my clients. I would say this in general though about clients that receive benefits tax free. I would think they would have a more difficult time availing themselves of this delaying strategy in that they must have enough other cash flow to carry them for 8 years from age 62 to age 70 either from other retirement cash flow or from their intent to work till age 70 in these difficult economic times.
Michael, we have used both defined benefit pieces(social security and pensions when applicable) as a fixed income alternative. We discuss whether this commuted value, should color our asset allocation model directly with the client prior to adjusting our allocations. Their response after conversation, dictates whether we count or ignore it as an allocation variable.This then naturally feeds into the conversation of when to draw social security.
I have thoroughly enjoyed reading the article and comments. As a former SS employee of almost 35 years, I travel all over the country encouraging people to take ownership of SS and make an informed decision. What I did not see in any of your responses is the potential entitlement of a spouse’s benefit, divorced spouse’s benefit, divorced widow, or widow’s benefit that can be in the interim from age 66 – 70. This definitely lessens the breakeven point!!!!
Elaine,
Thanks for the comment.
In my more in-depth material on this I spend a lot of time on file-and-suspend and restricted application strategies to get additional dollars and how it shortens the break even periods.
Still so much confusion out there!
-Michael
Another great and relevant post!
I agree with delaying S.S. benefits to hedge against the three risks you discuss, particularly longevity risk.
With that said, I feel that the analysis is somewhat incomplete given that it doesn’t take into consideration the probabilities of future outcomes. In this case the probability of a female living to the BE point at age 86 is only 49% and it’s even less for males at 36%.
While the benefits of delaying SS might be exponentially better than taking SS at FRA, they’re only exponentially better if one actually lives to an age that approximately only 1 in 5 will ever reach (assuming age 90-91 is when benefits begin to grow exponentially). It would seem from a purely statistical point of view that not delaying would in fact be the best option given that the probability of ever reaching BE or beyond is less than a coin flip.
It’d be interesting to see the above analysis with all expected or future outcomes weighted by their respective probabilities of occurring. Given the fairly significant divergence in life expectancy for men and women it’d also be interesting to see how that could effect one’s claiming decision based on gender.
I’m interested to hear your thoughts as I mentioned above I agree with the idea that delaying SS is essentially purchasing longevity insurance, but I think one can make a pretty good argument for getting that “bird in hand” at the first opportunity.
So you guys are in the business of recommending that people should plan on living to 100, and ignoring the possibility that the survival of the Social Security program in the future might require that the rules be changed, like means testing, or reduced COLA’s, or bankruptcy or whatever. Can Uncle Sam afford to keep the hundreds of trillions of dollars of promises he has made for future benefits? Get real!
John,
Actually, I generally recommend against planning for people to live that far beyond life expectancy. See, for instance, http://www.kitces.com/blog/archives/285-Whats-Your-Longevity-Assumption-Are-Planners-Being-Too-Conservative.html
As for potential Social Security changes itself, that’s also been covered on the blog, most recently at http://www.kitces.com/blog/archives/311-Are-We-Overstating-The-Consequences-Of-Social-Securitys-Insolvency.html – which shows the reality that because Social Security remains predominantly a pay-as-you-go system, even “bankruptcy” would only result in a benefits cut of approximately 25%.
In any event, the point of the post is not that you should PLAN to live to age 100 to take advantage of Social Security. The point is that IF you live to 100, it provides a radically superior risk-adjusted return than any alternatives out there today (and although I don’t have the full math here, it actually holds up well even if there IS a 25% cut in 20-30 years).
Simply put, this is a risk management/hedging technique to deal with the potential for unexpected longevity (or inflation, or low returns).
– Michael
The important variable for sensitivity analysis is actually the real rate of return – that is growth minus inflation.
If the real rate of return is 1% (about what TIPS return today) then the breakeven point occurs between 14 and 15 years. This breakeven point doesn’t move for inflation rates of 1 to 8 percent.
If the real return is 5% (as was assumed in the example) the breakeven point is 20 years – and is essentially independent of inflation from 1 to 8%.
Very good article, which appropriately recognizes opportunity costs and the time value of money in evaluating the deferral of SS benefits (as so many do not). I was wondering if you had seen this more recent work by SSA themselves? By considering 1) that that the actual investment alternative may be something other than TIPS (or a single life annuity), and 2) by explicitly accounting for longevity risk (with a statistical Survival Function) they conclude that starting SS benefits early will probably be optimum over ones expected lifetime (and with very little risk of being wrong) — a somewhat surprising conclusion from SSA (with which I tend to agree). Do you have any thoughts on the merits (or demerits) of using your actual investment alternative (e.g. a 60/40 Stock/Bond portfolio) as the alternative investment? Although I know this makes the comparison somewhat “apples to oranges” in terms of market risks, it seems like a more meaningful comparison to me, if this is what one would actually do. Your thoughts?
https://www.ssa.gov/policy/docs/ssb/v76n2/v76n2p1.html
Not understanding the first part of this coment: “Dying 20 years before the breakeven period costs the client $12,000; living just 14 years past the breakeven period brings the client ahead by nearly quadruple that amount, to just shy of $48,000.”
20 years before the breakeven would be 62? He would have gotten nothing?
Correct, if the retiree waited 12 months and did NOT receive $1,000/month, and then died at the end of that year, he/she would have lost out on $12,000 and gained nothing. Thus being $12,000 in the red.
– Michael
I just realized a fourth hedge…the client who delayed SS and used up nest egg money to do so, has a smaller nest egg from which forced RMD’s can be taken if the client subsequently needs LTC (client has no LTC insurance). This is an especially beneficial hedge if the LTC is rehab care after which the client lives quite a while.
Kay C.
I can receive widower SS benefits starting at age 60. If I do this, I can delay taking money out of my taxable retirement account and allow that to grow. My question is, what is better, wait until 67 to increase my monthly SS payment or take the lower value at age 60 and let me portfolio grow for those 7 years?