The Survivor Benefit, the HSA, and the Medicare Trap
Frank, 67, was widowed three years ago. His wife had claimed at her full retirement age and was receiving $3,200 a month when she died. Frank has never filed for benefits; his own primary insurance amount at his FRA of 67 is $2,700 a month. He works full time for a 45-employee engineering firm, earns $140,000, is covered by the group health plan, and contributes the family maximum to his HSA. He plans to work to 70 and is in excellent health. His assets are a $900,000 traditional IRA and $300,000 in a brokerage account. He wants to file for the survivor benefit now - "it's free money and I can invest it" - and asks whether he should also sign up for Medicare while he is at it.
1. Frank contributes the family maximum to his HSA. What happens to that when he files for the survivor benefit, and how do you weigh the trade-off?
Read the worked answer
Filing for Social Security makes him entitled to benefits, which triggers automatic enrollment in Medicare Parts A and B because he is past 65. Part A is premium-free, and enrollment in any part of Medicare ends HSA eligibility. Worse, Part A entitlement can be retroactive up to six months before the month of application (not before age 65), so contributions made during those retroactive months become excess contributions.
On this fact pattern: family coverage maximum plus catch-up is roughly $9,500 a year, so about $4,750 would fall in a six-month retroactive window. That is an excess contribution subject to a 6% excise tax each year it stays uncorrected, and the amount must be included in income - unless he removes it (with earnings) by his tax filing deadline. Total cost of the mistake: a few hundred dollars if corrected promptly.
Against $3,200 a month, or $38,400 a year. The comparison is not close.
Judgment: file the survivor claim, stop HSA contributions immediately, and ask the custodian to calculate and remove any excess attributable to the retroactive months. Do not let a ~$300 tax cleanup delay a $38,400-a-year benefit.
Two follow-ons: he can decline Part B while still working and covered by a group plan of 20 or more employees, which avoids premiums he does not need - but Part A alone still kills HSA eligibility, so there is no version of Medicare enrollment that preserves the HSA. And when he does enroll, he gets a special enrollment period running eight months after employment or coverage ends, whichever is first, with no late-enrollment penalty for Part B.
2. Frank asks whether he should convert traditional IRA dollars to Roth between 67 and 70, before he retires. What is your judgment?
Read the worked answer
Be skeptical of a large conversion program here, for a reason specific to his facts: the survivor benefit crowds out the room.
Run the numbers. Salary of $140,000 plus $38,400 of survivor benefits puts him around $178,000 of gross income as a single filer - already inside the 24% bracket, with only a narrow band left before 32%. He is also likely above the first IRMAA tier, so each additional dollar of conversion can trigger a tier jump in Medicare premiums on top of the marginal rate. The conversion "space" that looks available on a bracket chart is largely an illusion once the survivor benefit lands.
What I would actually do:
- Convert only enough to use genuinely unused low-bracket room, and check every increment against the IRMAA tier boundaries rather than the bracket boundary.
- Consider waiting. If he retires at 70, the year of retirement is a life-changing event that permits an IRMAA appeal, and ages 70-74 - after salary stops, before RMDs at 75 - may be a lower-income window than ages 67-70, even with Social Security running. Model both windows rather than assuming earlier is better.
- Keep his asset location honest: his $300,000 taxable account should fund the bridge years, not the IRA, so that conversions are not competing with spending withdrawals for the same bracket room.
- Note the arithmetic asymmetry of his situation: he is single, so there is no surviving-spouse bracket compression later. That removes one of the strongest arguments for aggressive early conversion and makes a moderate, tier-aware program the right answer.