Actionable insights — Vetting an advisor and stress-testing a retirement plan

1. Audit your own plan against the 5 / 15 / 20 buckets

1. Split every retirement dollar into three moments: contribution (5), growth (15), withdrawal (20). The numbers are illustrative — the ratio is the point: what you put in is the smallest number by an order of magnitude.
2. Decide, before looking at your accounts, which of the three you would rather pay tax on. Everyone picks the contribution.
3. Now list every vehicle you actually use — 401k, IRA, non-qualified brokerage, Social Security, pension, CDs, home equity — and assign each to the bucket it is taxed in. Tax-deferred accounts, pensions, Social Security and CDs tax at withdrawal (20); non-qualified growth taxes along the way (15).
4. Compare the two lists. If your stated preference is the 5 and none of your money is taxed there, you have a plan you didn't choose — you inherited a default. Ask why, and be honest that the answer is usually "it's what I'm told to do."
5. Re-run the audit when tax law or your income changes: the deferral trade was correct when current rates were high (the 1990s) and inverts when current rates are historically low.

• Statutory rate changes and scheduled sunsets (which direction do they move your future bracket?); any year your income dips — the cheap window to move dollars from the 20 bucket into the 5.

2. Price the mandatory costs, not just the returns — find the IRMAA-shaped item

1. List the expenses in retirement you cannot opt out of. Medicare Part B is the archetype: skip it and Social Security is withheld, so it is compulsory in practice.
2. Get its inflation rate, not its current price, and compound it over your actual horizon. A 7.2% escalator doubles the cost roughly every decade — which matters far more than the headline premium.
3. Check whether the cost is income-tested. If it is, your withdrawals feed it: more taxable income raises the surcharge, which is deducted from your benefit before you receive it — while you are taxed on the gross.
4. Read the protections narrowly. The Hold Harmless Act shields Part B against COLA only; Part D, supplements and IRMAA itself are unprotected, so the net benefit can fall in real and nominal terms.
5. Judge the plan on net spendable income, never on rate of return: "nobody cares about rates of return if you can't put it in your pocket."

• The annual Part B/Part D premium announcement versus the announced COLA; your own projected MAGI relative to the next IRMAA threshold — a single dollar over moves the whole bracket.

3. The three-question advisor interview — test fluency, not agreement

1. Ask about IRMAA. Not knowing what it is is disqualifying — "that's a red flag. It's a big, big cost coming your way."
2. Ask about annuities. They may dislike them; they must be able to give you the pros, the cons, and a specific reason it doesn't fit your situation.
3. Ask about cash value life insurance. Same standard: explain the mechanics, the trade-offs, and why it does or doesn't apply to you.
4. Grade on fluency and specificity, not enthusiasm. An advisor who can argue coherently against a product understands it; one who dismisses it with a slogan is telling you about their book, not your plan.
5. Calibrate your deference: unlike medicine, the entry bar is "a couple tests." Treat the advice as an opinion to be second-sourced, not a diagnosis to be accepted.

• Answers that are categorical rather than situational ("annuities are always bad," "life insurance is a scam"); an advisor who has never mentioned a Medicare surcharge in a retirement-income conversation.

4. Map the advisor's exit before you take their advice

1. Ask how the person gets paid and how they eventually leave the business — the exit determines the advice more than the fee schedule does.
2. Learn the model: commission products (mutual funds, annuities, life insurance) pay once; assets under management pay ~1% every year the asset stays. Only the recurring stream builds a saleable book.
3. Value the book the way a buyer would — roughly five times annual revenue, or a junior-partner split — and then ask which recommendations shrink it: Roth conversions (lower asset base), annuity purchases, life-insurance premiums leaving the account.
4. Draw the conclusion carefully: this predicts under-emphasis, not fraud. "They're just not as incentivized to position them in the best way all the time."
5. Neutralise it structurally — hire specialists whose incentives differ (an advice-only RIA that cannot earn commission, plus product specialists) and judge each on the person, not the title: "they could both be outstanding. They could also both be very, very bad."

• An advisor nearing retirement or in succession talks (the book is being groomed for sale); recommendations that always keep assets in-house; a "no" on a Roth conversion given without an accompanying report.

5. Build a team that talks — and pre-check the trap doors

1. Split the roles: an asset manager, an annuity specialist, a life-insurance specialist. One generalist cannot be fluent in all three, and a single provider's incentives colour all the advice.
2. Require them to coordinate, and be prepared to convene it yourself. The concrete use case is a down year: draw from the protected (insurance) bucket instead of selling portfolio assets into the drawdown, so a 10% market fall plus a 4% withdrawal doesn't compound to 14%.
3. Before executing any single move, check its timing trap:
   • Roth conversion: the 2-year IRMAA lookback — a conversion at 66 sets the Medicare surcharge at 67–68. Convert early or model the surcharge into the decision.
   • Cash value life insurance: compare several illustrations from different sources ("there's ways to kind of cook those books"), then commit to the premium and loan-repayment discipline — an unrepaid policy loan compounds until the policy "blows up," and that is the buyer's failure, not the product's.
   • Annuities: for non-qualified money, check the exclusion ratio — it determines how much of each payment is taxable and can beat capital-gains treatment.
4. Never accept "it's unavoidable" or "you'll be in a lower bracket" as an answer; both are assertions that a year-by-year report can settle.
5. Reject the deadline excuse. A plan can be materially improved in the last five working years — "if you're going to live to 90, no, it's not [too late]."

• Your age relative to Medicare enrolment minus two years (the conversion window closing); illustrations that differ materially on the same assumptions; a year of portfolio drawdown arriving with no alternative source to draw from.