Review of a DIY Retirement Plan: EDU #2636

Full analysis

Jim Saulnier and Chris Stein walk through a detailed DIY retirement plan submitted by a listener who is retiring in 2027 with $4.6 million across 401(k)s, Roth IRAs, and a brokerage account. The episode covers where her account-by-account structure aligns with their spending-segmented approach, where tax planning could improve it — particularly around charitable giving — and why the mental-decline research should shape how any retirement plan is designed and handed off.

What was covered

• The listener's asset picture: $3 million in two 401(k)s and an HSA (always-taxable), $675,000 in Roth IRA #1, $650,000 in Roth IRA #2 (roughly $1.325 million total in Roth, never-taxable), and $320,000 in a taxable brokerage account — $4.6 million total as of June 30.

• Her account assignments: 401(k)s for "good life" baseline spending; Roth IRA #1 as an emergency/aging/long-term care reserve (what Saulnier calls "SEAL" — Savings for Emergencies, Aging, and Long-term care); the brokerage for charitable giving; and Roth IRA #2 entirely for fun — the "fun number."

• Her delay-period math: Planning to delay Social Security to age 70, she calculated $800,000 in spending needs between retirement and age 70, added a $150,000 buffer, and is holding $950,000 in money-market or stable-value options inside her 401(k)s. The remaining ~$2 million is in low-cost index options and she projects it will grow to roughly $2.75 million by age 70.

• Post-delay math: At 70, she estimates $170,000 in annual spending, minus roughly $60,000 in Social Security, leaving $110,000 needed from the portfolio — which she notes is close to 4% of the projected $2.75 million. She acknowledged this is not guaranteed income and said she will decide then whether to purchase an annuity.

• Jim Saulnier's strong objection to using the brokerage for charitable giving: He argued the brokerage account, which carries a step-up in basis at death, is the best account to leave to heirs tax-free, and that a QCD (qualified charitable distribution — a direct transfer from an IRA to a charity, counted toward required minimum distributions and excluded from taxable income) from a rollover IRA is far more tax-efficient for charitable intent. He recommended moving enough of the 401(k) into an IRA to enable QCDs starting at age 70½.

• The Harvard/Texas Tech research on financial competence and aging: A Harvard study found peak financial decision-making competence averages around age 53, with accelerating decline afterward. A Texas Tech follow-up found that confidence in one's own abilities does not decline in step — the gap between actual and perceived competence is what makes older adults vulnerable to fraud and self-made mistakes.

• Survivorship planning gap: Saulnier flagged that the listener's tax ordering of "2-1-0" (spend for two, then one surviving spouse, then leave nothing specific to heirs) needs stress-testing for the widow/widower tax penalty — losing one Social Security check while RMDs on a large pre-tax account get compressed into single-filer brackets.

Notable claims & predictions

• Jim Saulnier on the 75–80% income rule: "I don't know who came up with that rule… your mortgage, your rent, your utilities pretty much are the same. Maybe your food is a little different… fun will eat up the two main savings [commuting costs and retirement contributions]. You're not going to spend 75 to 80 percent — you're going to spend 100 percent or more, especially early in retirement." (Rule someone could plan around: budget 100% of pre-retirement income, not 75–80%, for the early years.)

• Chris Stein on QCDs as the most tax-efficient charitable vehicle: "There's no more efficient way of moving money to a charity where no one ever paid taxes on the dollars… earnings that were tax deferred, turning those into charitable donations that were also not taxed — you can't get much cleaner than that." (Rule someone could act on: if you are charitably inclined and have a large IRA or 401(k), moving some into an IRA to fund QCDs after age 70½ beats donating appreciated brokerage assets.)

• Chris Stein on peak financial competence: "The Harvard study found that the average [peak] is 53 years old… after 53 there's a decline… the acceleration downward increases. The follow-up study by Texas Tech discovered that people's confidence in their abilities in these areas don't change — and that combination goes a long way toward explaining why there is elder fraud." (Warning: the danger window is not visible to the person inside it.)

• Jim Saulnier on long-term care and tax accounts: "If either her or her husband need long-term care, she may take it from the IRA [instead of the Roth], because most true long-term care expenses will be deductible — even with the 7.5% hurdle — and she can access those pre-tax dollars with little or no tax owed." (Actionable: earmarking Roth for LTC may be less efficient than using pre-tax funds if LTC expenses are large enough to be deductible.)

• Jim Saulnier on the widow/widower tax trap: "At death, those dollars are being jammed through much tighter brackets. You also have a loss of Social Security — one of the two goes away — so you have less secure income and higher taxes. It's a double whammy." (Planning flag for any couple with a large pre-tax account and two Social Security checks.)

Fact check

Harvard study / Texas Tech study on financial competence peak at 53: Saulnier and Stein reference these studies as foundational to their planning approach. A study by David Laibson and colleagues (published in connection with Harvard research) did find that performance on financial literacy and decision-making tasks peaks around the early-to-mid 50s. The Texas Tech confidence-gap finding aligns with published research on the Dunning-Kruger-adjacent phenomenon in financial cognition. Neither study is named precisely or dated in the transcript, so the specific "53" figure and the Texas Tech attribution cannot be fully verified from the transcript alone — but the general finding (peak in mid-50s, confidence does not track decline) is consistent with peer-reviewed literature. Not false, but listeners should know these are averages across populations, individual variation is wide, and the studies are cited from memory without publication dates or links.

QCD eligibility age — "70 and a half": Saulnier correctly states that QCDs are available starting at age 70½. This is accurate under current law. He also correctly notes QCDs cannot be made directly from a 401(k) — funds must first be rolled into a traditional IRA. Both points check out.

Step-up in basis for brokerage accounts: Saulnier states that a brokerage account receives a step-up in basis at death under current law, meaning heirs owe no capital gains tax on appreciation up to the date of death. This is accurate under current law. He noted "at least on the current tax law," which is an appropriate caveat — step-up in basis has periodically been targeted for reform.

IRD (income in respect of a decedent) on 401(k)s and IRAs: Saulnier correctly explains that inherited pre-tax retirement accounts are IRD — the inheriting human owes income tax on withdrawals. This is accurate.

"Etched in Jell-O" tax-ordering flexibility: The broader point — that annual tax planning may justify taking distributions from a different account type than originally planned — is sound retirement-planning advice, not a factual claim requiring verification.

"More money is stolen from seniors through a power of attorney than at the barrel of a gun": Saulnier himself flags this as possibly anecdotal ("I don't know if this was based on statistics or not"). It should be heard as a rhetorical point about family financial exploitation, not a verified statistic. Unverified.

No claims that are clearly false. The main area to note is that specific study citations are from memory and imprecise — listeners interested in the research should seek the primary sources rather than relying on the episode's characterizations.