Last Updated: September 6, 2026 Reading Time: 9 min

FedSmith named the three risks of retiring under MRA+30 this week: sequence-of-returns risk in the first five years, the FERS Supplement masking a weakening TSP until it disappears at 62, and the G Fund comfort trap. It did not compute any of them. So we did. One $600,000 TSP, one withdrawal schedule, real C Fund and G Fund returns from tsp.gov, run four ways. The spread between the best and worst outcome at age 66 is $333,270, and the only variable is which five years came first.

The Setup

The retiree separates at 57 with 30 years of service, so the FERS annuity is immediate and unreduced. The TSP holds $600,000, split 50% C Fund and 50% G Fund and rebalanced each year. That is a representative near-retiree mix, not a recommendation.

For years one through five, ages 57 through 61, the retiree withdraws $30,000 a year at the start of each year. The FERS Supplement is running during those years. Using the standard formula, estimated age-62 Social Security benefit × (years of FERS service ÷ 40), a $2,000 monthly Social Security estimate with 30 years of service produces $1,500 a month, or $18,000 a year.

At 62 the supplement stops, the month the retiree turns 62, regardless of when Social Security is claimed. To hold total income flat without filing early, the TSP withdrawal rises to $48,000 a year for years six through ten.

Years one through five use real historical TSP annual returns for two different windows. Years six through ten use a flat 6% for the blended scenarios and the current 4.25% G Fund rate for the all-G column, so the comparison isolates what the first five years did.

The Returns Used

tsp.gov published annual returns, verified September 6, 2026.

Year C Fund G Fund 50/50 blend
2000 -9.14% 6.42% -1.36%
2001 -11.94% 5.39% -3.28%
2002 -22.05% 5.00% -8.53%
2003 28.54% 4.11% 16.33%
2004 10.82% 4.30% 7.56%
2013 32.45% 1.89% 17.17%
2014 13.78% 2.31% 8.05%
2015 1.46% 2.04% 1.75%
2016 12.01% 1.82% 6.92%
2017 21.82% 2.33% 12.08%

The 2000 to 2004 window is the bad-first sequence: three down years to open, then the recovery. The 2013 to 2017 window is the good-first sequence: five straight positive years. Both are real five-year stretches a federal retiree actually lived through.

Four Scenarios, One Starting Balance

FedTools 2026 analysis, $600,000 TSP, 50/50 C and G Fund, $30,000 then $48,000 annual withdrawals.

Scenario Balance at 62 (after 5 draws of $30,000) Balance at 66 (after 5 more draws of $48,000) Change vs. $600,000
Bad-first (2000 to 2004 order) $485,083 $362,335 -$237,665 (-40%)
Flat 5.5% for five years, then 6% $607,534 $526,203 -$73,797 (-12%)
Good-first (2013 to 2017 order) $734,122 $695,605 +$95,605 (+16%)
All-G Fund (2000 to 2004 G returns, then 4.25%) $594,773 $459,982 -$140,018 (-23%)

The bad-first and good-first rows take out identical dollars under identical rules. The gap at 66 is $333,270.

Three cells, worked in full:

Bad-first, end of year one (age 57 to 58). Start $600,000, withdraw $30,000, leaving $570,000. Apply the 2000 blended return of -1.36%: $570,000 × 0.9864 = $562,248.

Good-first, end of year one. Start $600,000, withdraw $30,000, leaving $570,000. Apply the 2013 blended return of 17.17%: $570,000 × 1.1717 = $667,869.

Bad-first, the year-six transition (age 62). The balance entering 62 is $485,083. Withdraw the new post-supplement amount: $485,083 minus $48,000 = $437,083. Apply the flat 6%: $437,083 × 1.06 = $463,308. The good-first retiree takes the same $48,000 jump from a $734,122 balance. Same rule, different landing.

Why the Supplement Cliff Makes It Worse

FedSmith's second risk is the one most retirees underrate. The supplement covers baseline spending from 57 to 62, so a shrinking TSP does not hurt yet. Then the supplement stops and the TSP has to replace it.

In the model that is a 60% increase in the annual draw, from $30,000 to $48,000, at exactly the moment the bad-first retiree's balance is lowest. Five years of $48,000 withdrawals from $485,083, even at a flat 6%, grind the balance down to $362,335 by 66. The same five years from $734,122 leave $695,605.

The retiree who delays Social Security to 67 or 70 for a larger benefit extends this stretch further. Our 62 cliff timing guide runs that break-even; the FERS SRS Calculator gives you your own supplement figure.

The G Fund Comfort Trap, Measured

FedSmith's third risk: the retiree who sees the first bad year and moves everything to the G Fund. The all-G column answers what that costs.

Running the same withdrawals against G Fund returns only, 6.42%, 5.39%, 5.00%, 4.11%, 4.30% for the first five years and 4.25% after, produces $594,773 at 62 and $459,982 at 66. That beats the bad-first diversified path by $97,647. It trails the flat-return diversified path by $66,221 and the good-first path by $235,623.

The G Fund never fell. It also never delivered a 28.54% year or a 32.45% year. The comfort trap is not that the G Fund is bad. It is that a retiree who goes all-G after a crash has locked in the crash and opted out of the recovery, then has to fund $48,000 a year from a balance growing slower than the withdrawals.

What Actually Reduces Sequence Risk

None of this argues against retiring at 57. It argues against planning with an average.

Size a reserve to the pre-62 gap. FedSmith's August 26 point about cash reserves is right in direction. Two to three years of the $30,000 draw, held in the G Fund or outside the TSP, means a 2000-style opening never forces a C Fund sale at the bottom. The reserve is the bucket. Inside the TSP, remember that withdrawals come out of every fund pro rata, so protecting stocks takes an interfund transfer around each draw; our pro-rata withdrawal guide shows the mechanics.

Run your own bad-first case. Take your balance, your planned draw, and your supplement estimate, and apply 2000 to 2004 to the first five years. If the age-62 balance cannot carry the post-supplement withdrawal, the plan needs a different draw, a later date, or a bigger reserve.

Know the earnings test if you plan to work. Only wages count, only before 62, and the 2026 exempt amount is $24,480. TSP withdrawals never reduce the supplement.

Keep the withdrawal method flexible. Installments can be changed; a TSP annuity cannot. The TSP withdrawal guide covers the options.

Model Your Own Sequence

The TSP Calculator projects a balance under your withdrawal rate and return assumption. Run it twice, once at your expected return and once at a bad-first sequence, and read the age-62 balance in each. The FERS Retirement Calculator gives the annuity that sits under all of this, and the FERS SRS Calculator gives the supplement that disappears at 62.

Frequently Asked Questions

If I retire under MRA+30 at 57 with $600,000 in my TSP, is that enough?

There is no single answer. The same $600,000, drawing the same dollar amounts, lands anywhere from $362,335 to $695,605 ten years later in the FedTools table depending only on which five years the early withdrawal period happened to fall on. Enough depends on sequence, not just the starting number.

Does the FERS Supplement really stop at 62 with no way to extend it?

Yes. Under 5 U.S.C. 8421 the supplement ends the last day of the month you turn 62, whether or not you have filed for Social Security. There is no deferral option. The TSP withdrawal has to absorb the gap if you delay Social Security.

Why does losing the supplement at 62 hurt more after a bad market?

Because the higher withdrawal, $30,000 to $48,000 a year in this model, hits whatever balance survived the first five years. In the bad-first sequence that balance is already down to $485,083 when the bigger draw starts. In the good-first sequence it is $734,122. Same policy shock, very different runway.

Is moving everything to the G Fund the safe move at 57?

It avoids the worst sequence outcome, but in this model it also finishes behind both the flat-return path and the good-first path: $459,982 against $526,203 and $695,605. The G Fund removes downside volatility and also removes the recovery years that rebuild a balance after a drop. It protects against the worst case at the price of the recovery.

What is the difference between sequence risk and just bad returns?

Two ten-year periods can average the same return and produce very different outcomes once you are withdrawing. A retiree taking nothing out barely notices order, because the market eventually recovers on the full balance. A retiree pulling $30,000 to $48,000 a year from a shrinking balance sells shares at the bottom and has less money left to ride the recovery.

Does part-time work change this math?

Only through the earnings test on the supplement, and only before 62. In 2026 the exempt amount is $24,480; every $2 of wages above it cuts the supplement by $1. TSP withdrawals themselves do not count as earnings, so the withdrawal schedule in this model is untouched by outside work.

What should I do before retiring under MRA+30 to blunt sequence risk?

Hold a reserve outside the equity funds, two to three years of the pre-62 withdrawal gap, so a year-one or year-two drop never forces a stock sale at a loss. Then model your own number under a bad-first sequence rather than an average return. The $333,270 spread in the table exists because averages hide order.

Sources