Social Security Rate of Return

[JohnH]

We had a lengthy discussion on another thread regarding Social Security and the rate of return derived from delaying benefits. As it happens, I received my benefits increase letter yesterday and this issue came back to mind. There is no COLA this year, but anyone who is still working gets an increase if their current earnings replace one of their lower-earning years in the 35 years which are used in the benefits calculation.

Without getting into all the details, last year I had some flexibility in how I could classify about $20,000 of income and so I chose to report it as salary/wages. The income tax would be the same no matter how I reported it, but the combined EE/ER SocSec/Med Tax cost me about $3,000. The resulting increase in monthly SocSec benefit is $26/month. So as I see it, the straight-up break even point is 9.6 years. To me, that implies that it is actuarily neutral.

In looking at the annuity tables, an immediate annuity with a $30,000 payment produces $181 per month in income at current -very low- rates. (I had to use the higher figure because most calculators require at least $ 10,000). Factoring back down to the $3,000, that would equal $18/month for the annuity vs $26/month from SSA. Here's the kicker - in order to get the annuity up to $26/month, the rate on the annuity would have to be --> 7.24%.

I'm not presenting this as an across-the-board analysis with universal application. I didn't take into account that the actual cost is somewhat less than $3,000 because of the tax deductibility of half of the SocSec/Med tax. I also didn't take into account future COLA's if and when inflation fires up again, so the $26/month is actually the equivalent of an indexed annuity, which makes SocSec an even better deal. No one is going to get rich on an additional $26/month with indexing, and they won't starve if it's missing, but the result is close enough to 8% to convince me that there is a general principle at work in SocSec calculations. The math implies a similar result at any point in the process for someone delaying the time they begin to receive benefits (at least until they reach age 70).

[Roland Slugg]

John, the decision you made may have made sense for you, but for someone else, who is exactly the same age you are, the outcome might, and probably would be quite different. That's because your $20,000 of earned income replaced the smallest amount (including the cumulative indexing thereon) from your highest 35 years indexed earnings on your lifetime record. Let's say the amount it replaced in your case was $9,000. That means your base increased by $11,000, resulting in a $26 higher social security benefit.

For someone else, however, that $20,000 would have replaced a different amount ... anywhere between $0 (if the person didn't have 35 years of earnings) and $19,999 ... or no amount at all, if the lowest amount in the 35-year array was already $20,000 or more. For you the $20,000 yielded $26 per month more; for someone else the increase could be more, less or none at all.

Last year I had some flexibility in how I could classify about $20,000 of income and so I chose to report it as salary/wages.

I'm wondering what other classification choice(s) you had. Care to enlighten us?

[Y2KEA]

I'm wondering what other classification choice(s) you had. Care to enlighten us?

Perhaps JohnH is an S-corp. He has the choice to take it as reasonable wages (whatever that is since it is undefined by the code) or as flow through income.

[Rapid Robert]

Or, the activity producing the income could be considered a hobby rather than a trade or business.

[Rapid Robert]

John, the decision you made may have made sense for you, but for someone else, who is exactly the same age you are, the outcome might, and probably would be quite different. That's because your $20,000 of earned income replaced the smallest amoun

Not only what Roland said, but it isn't as if you just wrote a check for $3K, like you would with an annuity. You actually had to do enough work to earn $20K during the year (and pay income tax on it), which to me represents a much, much larger investment than $3K. When you also take that into account, I think your rate of return is significantly smaller.

With that said, I am facing a somewhat similar analysis. While I can't afford to completely retire from work, I am fortunate to not be working purely out of necessity, and have wondered what level of earnings I need in any given year to replace one of my lowest 35 years. I have researched this at SS web site, and while there are some tables available that might help, it appears exceedingly hard to answer this simple question.

[kathyc2]

John, I doubt the one-time 20K is the full reason for the $26 "raise". The 20K divided by 420 months is 48. Depending where your earning fall at the 32% replacement rate an increase of $15 and at the 15% replacement rate $7/mo. SSA takes your last year wages (in this case 2014) and anticipates you will make that same amount every year until FRA. So, the $26 came from the 20K "extra" in all remaining years to FRA.

Robert, SSA used to have software on their site you could download for projections but I found it quite cumbersome to use. It's quite easy to make an Excel sheet to calculate. Enter the years, your actual earnings and the index factor. Then a column for actual x index and isolate to highest 35 years. Then use steps 4 through 6 to calculate your FRA benefits. Once you have it set up you can play around w/ how future earning might affect benefits. Indexes and steps can be found here: https://www.ssa.gov/pubs/EN-05-10070.pdf

[TAXNJ]

Also

John, I doubt the one-time 20K is the full reason for the $26 "raise". The 20K divided by 420 months is 48. Depending where your earning fall at the 32% replacement rate an increase of $15 and at the 15% replacement rate $7/mo. SSA takes your last year wages (in this case 2014) and anticipates you will make that same amount every year until FRA. So, the $26 came from the 20K "extra" in all remaining years to FRA.

Robert, SSA used to have software on their site you could download for projections but I found it quite cumbersome to use. It's quite easy to make an Excel sheet to calculate. Enter the years, your actual earnings and the index factor. Then a column for actual x index and isolate to highest 35 years. Then use steps 4 through 6 to calculate your FRA benefits. Once you have it set up you can play around w/ how future earning might affect benefits. Indexes and steps can be found here: https://www.ssa.gov/pubs/EN-05-10070.pdf

Can also talk to a CFP who specializes in this area. Also, check some articles by Laurence Kotlikoff is a William Fairfield Warren Professor at Boston University.

[kathyc2]

John, as far as the 8% ROI, it really comes down to how long someone lives if it's an actual additional amount. If someone's FRA is 66 and the monthly benefit is 2,000/mo after 198 months (age 82.5) they will have collected 396,000. If instead they start taking benefits at 70 they will receive 2,640/mo (2000 x 1.32). At age 82.5 they will have 150 months of benefits of 396,000.

Of course one should take into account the COLA projections, amounts of SS that would be taxable and different tax rates that may apply between 66 and 70 (if are still working), cash flow needs, etc.

But it someone actually receives more over a lifetime really comes down to if they outlive the actuarial table or not.

[JohnH]

The choice arose because of an income spike in an S corporation. (Y2KEA got it right). I believe I could have justified taking much less in salary and letting it flow through the K-1, but I made the decision to include $20K of the increase in salary/wages instead. Part of the reason for the decision was that I keep a running total of my "best 35" after adjusting for the annual factors, and my base salary for 2014 was close to my lowest year. So the reporting the $20K as salary/wages would have an impact - I just was not sure how much. Since the income was taxed the same either way, the only effect was the additional SocSec/Med tax. When I realized that the actual return apparently worked out so close to the commonly-used 8% figure, I found it interesting.

My point was simply that deferring SocSec is equal to purchasing a Single Premium Income Annuity with inflation protection at a very reasonable price - better than anything the insurance companies offer. And it appears that the 8% which is commonly claimed is probably accurate.

Obviously, one can never know with certainty the optimal financial decision until the day they die. But that is true no matter what one does with their retirement funds. The problem arises because we must make those decisions in real time and then live with the outcome, whatever it may be. Some people are going to die before reaching their life expectancy and leave money on the table. There is no financial risk associated with that outcome. But others are going to exceed their life expectancy and possibly outlive/exhaust their assets - that is where the major risk in retirement lies.

Unless one is in bad health or in need of the money simply to survive, the riskiest action is to begin drawing SocSec benefits early. Every situation is unique and there is no "one size fits all" recommendation. But as a general rule, waiting until full retirement age is a first step in terms of prudent investment planning. Yet roughly 35% of workers fail to do that. Continuing to defer until age 70 is still good economics for many, although only about 2% of retirees actually hold out that long.

[JohnH]

Not only what Roland said, but it isn't as if you just wrote a check for $3K, like you would with an annuity. You actually had to do enough work to earn $20K during the year (and pay income tax on it), which to me represents a much, much larger investment than $3K. When you also take that into account, I think your rate of return is significantly smaller.

With that said, I am facing a somewhat similar analysis. While I can't afford to completely retire from work, I am fortunate to not be working purely out of necessity, and have wondered what level of earnings I need in any given year to replace one of my lowest 35 years. I have researched this at SS web site, and while there are some tables available that might help, it appears exceedingly hard to answer this simple question.

Kathy provided the correct link -> https://www.ssa.gov/pubs/EN-05-10070.pdf.
You can convert that table into a spreadsheet and plug your numbers into it. Once you've done that then you just need to track the "top 35". Once you retire you must do your own tracking, because the benefits estimator on the SSA web site won't work for retirees. There are also some great discussions on this subject at bogleheads.org, including lots of input by some very knowledgeable people who are experts on the subject.

Just as a clarification, I'm 67 and have been receiving Social Security benefits since I reached FRA in 2013 at age 66. So the increase I'm talking about is actual, not theoretical. As I stated, the purpose of the exercise was not focused on the dollar amount, but rather on the "annuity equivalent return" aspect which I'm coming more and more to believe is almost universal at around 8%.

Having said all that, there have been a couple of things pointed out in this thread that are making me re-think the entire exercise. So thanks for all the points and counter points. Seems like I'm talking about these things more and more with clients lately, and I like to know that what I'm saying to them is accurate.

[kathyc2]

If you are already receiving benefits, I'm confused how an additional 20K would increase benefits $26. Like I said the increase to average monthly indexed wages would be $48. An increase of 26 would mean 54% of the 48 going to benefits. Can you mathematically back into the 26 amount?

I've always thought you were talking about the 8% being the increase of delayed benefits. However, it sounds like you are saying that the 8% is the equivalent of the return you would have received it the money had been invested elsewhere other than as a tax????? Is so, that is more than double the rate of returns from what I have seen.

[JohnH]

If you are already receiving benefits, I'm confused how an additional 20K would increase benefits $26. Like I said the increase to average monthly indexed wages would be $48. An increase of 26 would mean 54% of the 48 going to benefits. Can you mathematically back into the 26 amount?

I've always thought you were talking about the 8% being the increase of delayed benefits. However, it sounds like you are saying that the 8% is the equivalent of the return you would have received it the money had been invested elsewhere other than as a tax????? Is so, that is more than double the rate of returns from what I have seen.

I was simply evaluating the increase in benefits vs the amount paid in SocSec tax. Start with the fact that the base salary would not have impacted my benefit significantly because it was near my lowest-of-35. Then add $20K to my salary. The additional $20K causes the 2014 year to pop up and replace my lowest-of-35. The net income tax effect is zero because that $20K would otherwise have been a part of the amount showing on my K-1. So the actual cost I incurred is $1,500 of SocSec/Med withholding plus the matching $1,500 from my S-corp, for a total outlay of $3,000 in SocSec/med taxes. (Technically the $1,500 from my S-corp is tax deductible, so it is actually a net cost of about $1,000, but I ignored that adjustment for simplicity. If work that into the calculation, it becomes even more favorable)

So I made a one-time payment of $3,000 last year and derived a benefit increase of $26/month. This is no different than paying $3,000 to an insurance company to purchase an immediate Single Premium Income Annuity (SPIA) with no term certain and no COLA adjustment. Insurance companies don't deal in numbers that small, so I have to scale up to get started. Multiplying by 10, I find that a $30,000 SPIA purchase would buy me $180/month. Reducing back down to the actual numbers, that scales down to an $18/month benefit for a $3,000 purchase.

SPIA's are not very favorable in this low interest environment - the benefit only equates to about a 3% rate more-or-less. But what I found was that it would require a SPIA with a benefit based on a 7.24% rate to generate the $26/month benefit. Not quite 8%, but close enough. You can't buy a plain vanilla SPIA which comes anywhere near that return right now. And you certainly can't buy one with a COLA.

In summary, my math simply confirms the "8% rule" in this case. And I suspect, because SocSec is actuarily neutral, the investment effect is also uniform across most other choices. In today's low intereest rate environment, it's better than anything the free market has to offer. And this is based on current real-world experience, not some financial planner's assumptions about the future.

[kathyc2]

I understood what you were saying regarding the 3K vs $26. However, if the 26 increase was due partially due to the 20K (3K tax) and partially due to other factors, then attributing the increase due solely to paying 3K more is giving you erroneous numbers.

I still have a few years until FRA but I've been tracking and planning. Earlier this year, after SSA had recorded 2014 earnings, I had ran my projected benefits statement and could tie down the math by my calculations to SSA calculated benefits to $1 rounding difference. Interestingly, when I ran my projected benefits statement today, it's 2.9% higher than it had been
when I tied it out earlier this year. Took me a while to figure it out, but there are new index factors: https://www.ssa.gov/cgi-bin/awiFactors.cgi and new bend points: https://www.ssa.gov/oact/cola/bendpoints.html

Using the new numbers I can again tie my calculations to theirs. My guess is part of your 26 increase is due to 20K part, but also to new indexes and bend points.

[kathyc2]

Taking it a step further: If the 20K was 20K more than the year dropped, then your average monthly wages increased by $48. Assuming you lifetime average monthly wages are greater than 5,157 then you would receive 15% or $7 more based on the 20K, with the rest being due to index and bend point changes.

If that is true, then 27% (7/26) is due to higher wages. Taking the 8% you had calculated times the 27% due to higher wages, it would work out to 2.2% ROI which is more in line to average ROI for SS benefits.

[JohnH]

Taking it a step further: If the 20K was 20K more than the year dropped, then your average monthly wages increased by $48. Assuming you lifetime average monthly wages are greater than 5,157 then you would receive 15% or $7 more based on the 20K, with the rest being due to index and bend point changes.

If that is true, then 27% (7/26) is due to higher wages. Taking the 8% you had calculated times the 27% due to higher wages, it would work out to 2.2% ROI which is more in line to average ROI for SS benefits.

I'm not sure I follow the math. Personally I don't think the $20K affects anything other than the calculation of the SocSec/Med tax. I had the $20K in hand (minus the taxes), no matter how we might slice and dice it. By adding it to salary/wages, I gave up an additional $3,000 in SocSec/Med tax.

Had I reported the $20K as a part of my K-1 income, and had I instead bought a SPIA with the $3,000, the insurance company would have paid me a benefit based on a 3% earnings rate. As it is, I paid SocSec the $3,000 and received the equivalent of a SPIA that the insurance company would not have paid me unless they could earn enough to base the SPIA on a 7.2% rate. Plus it includes a COLA. The SocSec benefit far exceeded anything the market can offer.

Maybe something else is at work here, and maybe it is just a coincidence that the 7.2% return appears in this scenario. Bend points and indexes may have something to do with it, but to me this is a real-world example based on actual numbers and tangible results.

I'll keep pondering it, although nothing so far has changed my overall opinion about the benefits of delaying until full retirement age. (except in cases of bad health). Even when someone must begin drawing SocSec due to job loss or other immediate financial needs, it isn't generally a good long-term financial decision.