Discrepancy in Social Security COLA calculation

[Retireatee1]

I'm doing some testing with my Retireator Social Security calculator and have encountered a mystery.  I compare the results of my calculations with those on the SSA.gov website.  One useful test is Maximum-Taxable Earnings using this page:

https://www.ssa.gov/OACT/COLA/examplemax.html

So assume a birthdate of 1/2/55, which would put the FRA at 66 and 2 months (January 1st triggers a special case).  This individual's base year would be 2015 (year turning 60) and would use the following bend points:

Bend Point 1: $826
Bend Point 2: $4,980

https://www.ssa.gov/OACT/COLA/bendpoints.html

I can recreate the maximum AIME of $9,979 for 35 years of maximum earnings.  You can then calculate the base year PIA as follows:

90% of $826 = 743.4
32% of $4,980-$826 = 1329.28
15% of $9,979-$4,980 = 749.85

Base year PIA = $2822

https://www.ssa.gov/OACT/COLA/piaformula.html

The trouble is in getting from the base year PIA to the PIA at FRA (no reduction factors or delayed credits in this case study).  I believe that there is no inflation adjustment from age 60 to age 62.  Starting at age 62, there would then be four years of COLA to get to 66:

2017->2018 2.00%   $2878
2018->2019 2.80%   $2959
2019->2020 1.60%   $3006
2020->2021 1.30%   $3045

https://www.ssa.gov/OACT/COLA/colaseries.html

I get to $3045, which is short of the SSA prediction of $3113.  This is too large to be a rounding error.

Any ideas?

[Retireatee1]

Well it was incorrect to use bend points from the base year, it should be base year + 2.  I had that right originally, then broke it.

So now:

Base year (+2) PIA = $2917.27

2017->2018 2.00%      $2975
2018->2019 2.80%      $3058
2019->2020 1.60%      $3107
2020->2021 1.30%      $3148

So now I am 1.1% too big.

[Retireatee1]

OK I've got it.

The "Retirement at age 66" column on the SSA page really does mean retirement at age 66.  So the year 2021 retirees were born in 1955, and reduction factors now apply at age 66 which didn't occur in 2020/1954.  Once I adjust for that, I'm within a few bucks.

Close enough for government work.

[MDM]

You could compare with the 'SocialSecurity' tab in the case study spreadsheet.

That matches the https://www.ssa.gov/OACT/COLA/examplemax.html numbers exactly.

[iluvzbeach]

@MDM, would you consider adding a “person 2” to the SS page of the workbook? Or, is it as easy as just copying the tab itself and duplicating it onto another tab? I didn’t want to screw up anything in the workbook by making the change without asking first.

If you prefer that I ask the question in the other thread, let me know and I can do that.

[MDM]

@MDM, would you consider adding a “person 2” to the SS page of the workbook? Or, is it as easy as just copying the tab itself and duplicating it onto another tab? I didn’t want to screw up anything in the workbook by making the change without asking first.

If you prefer that I ask the question in the other thread, let me know and I can do that.

Copying to another tab would work well.

After you get both PIAs (I'm assuming this is for a couple), head over to Open Social Security: Free, Open-Source Social Security Calculator for an excellent tool on when each of the couple might want to start SS benefits.

[iluvzbeach]

TYVM!

[Retireatee1]

That's all sorted out in my 2020.1228 release.

[Retireatee1]

OK I've got a new minor release out.  I've worked out a strategy to appraise the present value of Social Security benefits.  The formula is pretty simple actually.  I add up the annuity factors from FRA or current age (whatever is larger) to age 120 based on the SSA Period Life Table.  There is no adjustment for interest rates, as SS benefits are inflation-adjusted.  This sum is then multiplied by the PIA * 12.

So for a male retiring at 66 in 2021, the maximum his benefits can be worth is $626,730.  For a female, it is $719,860.  The data is attached.

This sum can then be plugged back in as a balance on a variable annuity with a spread fee of 0% and no guaranteed term, and this will generate a comparable income stream.

Note that I do not take into account the value of spousal/survivor benefits.  That would be a bit convoluted to do.  So that equivalent annuity that you buy might no be so equivalent if those are a factor.

In reality though, annuity companies always charge a spread fee.  So the appraisal above is actually a bit low compared to the street price of an equivalent annuity.  But it is interesting nonetheless.