Retirement Planning Case Study $900K–$2.8M Households: When “Don’t Retire Yet” Saves Taxes, IRMAA, and Risk

You can be “financially ready” for retirement and still pick the wrong retirement date.

I see it all the time: a household with $900,000 to $2.8 million saved, no debt, and a clear desire to step away… but the first 3–7 years of retirement are set up to trigger avoidable taxes, Medicare premium surcharges, and a fragile withdrawal plan that breaks the first time the market drops.

This is the uncomfortable truth behind “don’t retire yet.” It’s not a moral judgment. It’s math. Sometimes working one more year (or even part-time one more year) isn’t about squeezing out extra savings. It’s about buying a better tax window, a better Social Security strategy, and a safer first chapter of retirement.

In this case study, I’ll walk you through an anonymized couple with about $1.9 million saved. We’ll compare two retirement start dates and two Social Security paths, and we’ll look at what changes in:

- Taxes (ordinary income vs capital gains)

- Medicare IRMAA (income-based Medicare premium surcharges)

- Withdrawal sequencing (which accounts you spend from first)

- Portfolio risk (how much market risk you’re taking when you’re most vulnerable)

I’m Alex Newman at Grape Wealth Management. I’m a fiduciary advisor. That means my job is to put your interests first, and to explain the tradeoffs in plain English so you can make a confident decision.

The household: $1.9M saved, “ready” on paper, but one big timing problem

Meet “Mark” (64) and “Dana” (62). They want to retire soon and travel while they’re healthy.

They have $1.9 million in investable assets:

- $650,000 in a taxable brokerage account (a regular investment account)

- Cost basis: $420,000 (meaning $230,000 of unrealized gains)

- $1,050,000 in traditional IRAs/401(k)s (pre-tax money)

- $150,000 in Roth IRA (tax-free bucket)

- $50,000 in cash

Income and goals:

- Mark earns $185,000 salary. Dana earns $35,000 part-time.

- They spend about $110,000/year after tax today.

- In retirement, they want $95,000/year after tax (they’ll travel early, then slow down later).

- They plan to buy health insurance until Medicare starts (Mark is close; Dana is not).

Portfolio today: 70% stock / 30% bonds.

Their instinct: “We’ve got $1.9M. We’re fine. Let’s retire at the end of this year.”

The timing problem: retiring at the end of this year creates a messy overlap of:

1) A high-income final working year (big tax bracket)

2) A large pre-tax balance (future RMDs)

3) Medicare IRMAA exposure at 65+ (based on income from two years prior)

4) A plan to claim Social Security early “to reduce withdrawals”

5) A portfolio that’s fine for accumulation, but not designed for the first drawdown in retirement

This is exactly the type of household where “don’t retire yet” can be the right advice.

The decision we modeled: retire this year vs work one more year (and why it’s not just about the extra paycheck)

We ran two retirement start dates and two Social Security strategies.

Assumptions (kept conservative and simple):

- Inflation: 3% long-term

- Portfolio return assumption (planning, not prediction):

- Stocks: 6% average

- Bonds: 3% average

- Blended: depends on allocation

- They file taxes as married filing jointly.

- They take the standard deduction.

- They do not have a pension.

- They want to keep a “sleep-at-night” plan, not a maximize-every-last-dollar plan.

Scenario A: Retire at the end of this year

- Mark retires at 65 next year.

- Dana stops part-time work.

- They claim Social Security early: Mark at 66, Dana at 64.

- They spend from taxable first “because it feels safer than touching the IRA.”

Scenario B: Work one more year (then retire with a coordinated tax/Medicare plan)

- Mark works one more year (or negotiates a lighter role) and retires at 66.

- Dana continues part-time for one more year.

- They delay Social Security: Mark to 70, Dana to 67.

- They use the “gap years” (retirement before RMDs) to do planned Roth conversions.

- They build a cash/bond buffer to reduce sequence-of-returns risk.

Important: The “one more year” benefit isn’t primarily the extra savings. It’s the ability to shape income in the years that matter.

Why “retire now” looked fine… until we priced in taxes, IRMAA, and the first bear market

On a simple retirement calculator, Scenario A looked okay.

They have $1.9M. They need about $95,000 after tax. Social Security will cover part of it. The portfolio covers the rest.

But retirement doesn’t fail on averages. It fails on timing.

Here’s what Scenario A quietly created:

1) Higher lifetime taxes due to a bigger future RMD problem

Traditional IRA money is “tax-deferred,” not “tax-free.” Eventually, the IRS forces withdrawals called Required Minimum Distributions (RMDs). Those withdrawals are taxed as ordinary income.

With $1,050,000 pre-tax at retirement, and assuming moderate growth, their IRA could easily be $1.4M–$1.8M by their mid-70s.

That can create a tax spike later, right when:

- Social Security is also coming in

- One spouse may pass away (and the survivor files single, often at higher tax rates)

- Medicare IRMAA surcharges can kick in

If you want a deeper explanation of how this happens, see our supporting post: /post/rmd-tax-spike-planning-1m-ira

2) Medicare IRMAA surprises

IRMAA is an extra charge added to Medicare Part B and Part D premiums if your income is above certain levels.

Plain English: higher income retirees pay more for Medicare.

Two key points retirees miss:

- IRMAA is based on MAGI (Modified Adjusted Gross Income), not just “taxable income.”

- IRMAA uses a two-year lookback. Your Medicare premiums at 65 are based on your income at 63.

So if you retire at 65 after a big final income year, you can get hit with higher Medicare premiums even though you’re “retired now.”

More detail here: /post/medicare-irmaa-magi-planning

3) A fragile withdrawal plan in the first downturn

Sequence-of-returns risk is the risk of getting a bad market early in retirement while you’re withdrawing.

It’s not that markets always go down. It’s that a down market plus withdrawals can permanently shrink the portfolio, making it harder to recover.

If you want a deeper dive, see: /post/sequence-of-returns-risk-first-5-years

In Scenario A, they had no explicit plan for:

- How much cash to hold

- What to sell first in a down market

- When to reduce spending (and by how much)

- How to avoid selling stocks at the worst time

The numbers: side-by-side outcomes with real assumptions (taxes, IRMAA, and income)

Let’s put real numbers on the table. These are planning estimates, not guarantees.

### Social Security assumptions

- Mark’s projected benefit at full retirement age (67): $3,200/month

- Dana’s projected benefit at full retirement age (67): $2,400/month

If they claim early, benefits are reduced.

If they delay to 70, benefits increase roughly 8% per year after full retirement age (not counting inflation adjustments).

More on the tradeoffs: /post/social-security-timing-affluent-retirees

### Scenario A: retire now + claim earlier

Retirement start: next year.

Claiming:

- Mark claims at 66: about $3,000/month ($36,000/year)

- Dana claims at 64: about $1,900/month ($22,800/year)

Total Social Security: ~$58,800/year starting relatively early.

Income gap to fund from portfolio:

- Target spending after tax: $95,000

- Add estimated federal taxes and state taxes (varies): assume $12,000–$18,000 depending on withdrawals

- Rough gross need: ~$110,000–$115,000

- Subtract Social Security: ~$58,800

- Portfolio withdrawals: ~$55,000/year early on

So far, that sounds manageable.

But here’s the catch: the way they planned to pull that $55,000 matters.

Their instinct was “taxable first.” That means selling appreciated investments.

If they realize, say, $45,000 of long-term capital gains in a year, that increases MAGI.

Then add:

- Interest/dividends from the taxable account

- Any IRA withdrawals (which are ordinary income)

- The taxable portion of Social Security (yes, Social Security can become taxable)

Result: they drift into higher Medicare premium territory at exactly the wrong time.

### Scenario B: work one more year + delay Social Security + use the tax window

Retirement start: end of next year.

Claiming:

- Mark delays to 70: about $4,000/month ($48,000/year)

- Dana delays to 67: about $2,400/month ($28,800/year)

Total Social Security later: ~$76,800/year.

That’s about $18,000/year more than Scenario A, for life (again, planning estimate).

Now the key: what happens in the “gap years” between retirement and age 70?

Those years are often your best tax planning window because:

- You’re no longer earning wages

- You may not be taking Social Security yet

- You’re not forced into RMDs yet

That creates room to do Roth conversions.

A Roth conversion is moving money from a traditional IRA to a Roth IRA. You pay tax now, but future growth and withdrawals can be tax-free.

More detail: /post/roth-conversions-gap-years-before-rmds

In this case, we modeled converting $60,000/year for four years (ages 66–69 for Mark).

That’s $240,000 shifted from the future-RMD bucket into the Roth bucket.

Why it mattered:

- It reduced future RMDs

- It created a tax-free “shock absorber” later

- It helped manage IRMAA by targeting specific MAGI levels

The underrated lever: withdrawal sequencing (and why “taxable first” is often misapplied)

Withdrawal sequencing is simply the order you spend from:

- Taxable brokerage

- Traditional IRA/401(k)

- Roth IRA

Most retirees are told a simplistic rule: “Spend taxable first, then IRA, then Roth.”

That rule can be okay.

But for $900K–$2.8M households, it’s often misapplied because it ignores:

- IRMAA thresholds

- Social Security taxation

- Future RMDs

- The value of Roth money later (especially for the surviving spouse)

In Mark and Dana’s case, “taxable first” created two problems:

1) It pushed MAGI up in Medicare years due to capital gains.

2) It preserved the large IRA, which later forced bigger RMDs, which later pushed MAGI up again.

That’s the trap: you avoid IRA taxes now, and you pay more later when you have fewer options.

A more intentional approach looked like this:

- Use taxable for baseline spending, but harvest gains carefully.

- Use IRA withdrawals (or Roth conversions) up to a planned tax bracket ceiling.

- Preserve Roth for later years, big expenses, and survivor planning.

If you want the full framework, see: /post/retirement-withdrawal-sequencing-taxable-ira-roth

Medicare IRMAA: the “stealth tax” that makes some retirement dates objectively worse

Let’s make IRMAA simple.

Medicare has monthly premiums. Most people know Part B has a premium.

IRMAA adds an extra surcharge if your income is above certain thresholds.

Two things make IRMAA feel unfair:

- It’s a cliff system (cross a line, pay more for the whole year).

- It’s based on income from two years ago.

In this case study, Scenario A created a common pattern:

- Big final working year income

- Retirement the next year

- Medicare starts

- Two years later, Medicare premiums reflect that big income year

So Mark could be “retired” and still paying higher Medicare premiums because of income he earned before he retired.

Scenario B didn’t eliminate IRMAA risk, but it allowed us to plan around it:

- We projected MAGI each year.

- We targeted Roth conversion amounts that filled a tax bracket without tripping unnecessary IRMAA tiers.

- We avoided large, unplanned capital gains in Medicare years.

This is where planning becomes premium.

It’s not about avoiding taxes at all costs.

It’s about paying taxes on purpose.

More detail here: /post/medicare-irmaa-magi-planning

Portfolio risk tradeoffs: why “70/30” can be fine… until you start pulling income

A 70/30 portfolio can be perfectly reasonable when you’re earning a paycheck.

In retirement, the question changes.

It’s no longer “What return can we earn?”

It’s “What happens if the market drops 20% in year one and we still need to pay bills?”

In Scenario A, Mark and Dana planned to retire with:

- 70/30 allocation

- Minimal cash reserves

- No guardrails

If the market drops early, they would likely sell investments at depressed prices.

That’s how sequence-of-returns risk becomes real.

In Scenario B, we made two changes:

1) A dedicated cash/bond buffer

We targeted roughly 18–24 months of spending needs in cash and short-term bonds.

Not because cash “earns more.”

Because it buys time.

It lets you avoid selling stocks in a bad year.

2) A written withdrawal policy

We used simple guardrails:

- If the portfolio is down more than 12% from its prior high, pause inflation raises for one year.

- If the portfolio is down more than 20%, reduce discretionary spending (travel, gifting) by a pre-agreed amount.

- Refill the cash buffer only after positive market years.

This is not about fear.

It’s about not improvising under stress.

If you want the deeper explanation, see: /post/sequence-of-returns-risk-first-5-years

When “don’t retire yet” is the right call: 5 red flags I want you to take seriously

I’m going to be direct. If you’re in the $900K–$2.8M range, these are the warning signs that retirement this year may be a bad deal.

1) You have a large pre-tax balance and no Roth conversion plan

If most of your money is in traditional IRAs/401(k)s, you’re sitting on a future tax bill.

Without a plan, RMDs can push you into higher brackets later.

2) You’re planning to claim Social Security early while your IRA is large

Claiming early can reduce your guaranteed lifetime income.

Meanwhile your IRA keeps growing, increasing future RMD pressure.

Sometimes the better move is the opposite: spend from the portfolio in your 60s to buy higher Social Security later.

3) You’re about to start Medicare and you’re ignoring IRMAA

If you don’t know your projected MAGI for the next 3–5 years, you’re flying blind.

IRMAA isn’t the end of the world.

But it’s expensive enough that it should be planned for, not stumbled into.

4) Your spending is too close to portfolio capacity

If your plan only works when markets cooperate, it’s not a plan.

You want margin.

Margin can come from:

- Working one more year

- Reducing fixed spending

- Delaying Social Security

- Building a bigger cash buffer

5) You don’t have a “first downturn” plan

If your retirement plan is basically “we’ll see what happens,” you’re taking more risk than you think.

The first 5 years matter disproportionately.

The questions retirees are asking right now (and the straight answers)

Here are the questions I’m hearing most from retirees and near-retirees with real assets.

1) How can I optimize my retirement withdrawals to minimize taxes?

Start by projecting your income sources year-by-year:

- Wages (if any)

- Social Security

- IRA withdrawals/RMDs

- Dividends/interest

- Capital gains

Then choose a withdrawal sequence that intentionally fills lower tax brackets and avoids unnecessary IRMAA tiers.

For many $900K–$2.8M households, that means some combination of:

- Controlled IRA withdrawals in low-income years

- Roth conversions in gap years

- Careful capital gains realization

Supporting deep dive: /post/retirement-withdrawal-sequencing-taxable-ira-roth

2) What is the best age to start claiming Social Security benefits?

There isn’t one best age.

But there is a best coordination.

If you have enough assets to fund your 60s without claiming, delaying can increase your guaranteed income later. That can reduce pressure on the portfolio in your 70s and 80s.

The decision should consider:

- Health and family longevity

- Survivor needs (the larger benefit often becomes the survivor benefit)

- Tax planning and IRMAA

- Portfolio risk in the first decade

Supporting deep dive: /post/social-security-timing-affluent-retirees

3) How do Medicare IRMAA surcharges affect my retirement planning?

IRMAA can raise your Medicare premiums for a full year if your MAGI crosses certain thresholds.

It’s essentially an extra cost tied to income.

Planning impact:

- Large Roth conversions can trigger IRMAA.

- Large capital gains can trigger IRMAA.

- RMDs can trigger IRMAA.

The goal is not always “avoid IRMAA at all costs.” Sometimes paying one year of IRMAA is worth it to reduce lifetime taxes.

But you should decide that on purpose.

Supporting deep dive: /post/medicare-irmaa-magi-planning

4) What are the risks of withdrawing from my portfolio during market downturns?

The risk is that you sell more shares when prices are down.

That can permanently reduce the portfolio’s ability to recover.

Mitigation tools:

- Cash/bond buffer

- Spending guardrails

- A portfolio allocation designed for withdrawals, not just growth

Supporting deep dive: /post/sequence-of-returns-risk-first-5-years

5) How can I sequence my withdrawals to ensure a sustainable retirement income?

Sustainability is a mix of:

- A realistic spending target

- A tax-aware withdrawal order

- A risk plan for down markets

- A Social Security strategy that reduces long-term pressure

In Mark and Dana’s case, sustainability improved when we:

- Delayed Social Security to increase guaranteed income

- Converted part of the IRA to Roth in gap years

- Reduced future RMD pressure

- Built a buffer for the first downturn

What we recommended for this household (and what changed after one more year)

Here’s the plan we implemented in Scenario B. This is the “Retirement Income & Tax Map” style approach.

1) Retirement date: end of next year

Not because they “needed” another year of savings.

Because it improved the next decade of tax and Medicare outcomes.

2) Social Security: delay strategically

- Mark to 70

- Dana to 67

This increased projected lifetime guaranteed income and strengthened survivor protection.

3) Roth conversions: $60,000/year for four years

We targeted conversion amounts that:

- Filled a planned tax bracket

- Considered IRMAA thresholds

- Reduced future RMD pressure

Supporting deep dive: /post/roth-conversions-gap-years-before-rmds

4) Withdrawal sequencing: intentional, not automatic

In the gap years, we funded spending with a mix of:

- Taxable withdrawals with controlled capital gains

- IRA withdrawals/conversions up to the target bracket

We preserved Roth for later flexibility.

Supporting deep dive: /post/retirement-withdrawal-sequencing-taxable-ira-roth

5) Portfolio risk: shift from “growth default” to “retirement resilient”

We moved from 70/30 toward a more retirement-friendly mix (for this household, closer to 60/40 with a dedicated short-term reserve).

Not because 60/40 is magic.

Because the plan needed to survive the first bad market without panic selling.

6) RMD planning: reduce the future spike

By converting $240,000 over four years, we lowered the projected size of future RMDs.

That doesn’t just reduce taxes.

It reduces the chance of:

- IRMAA later

- Pushing Social Security into higher taxable ranges

- Forcing the surviving spouse into a painful tax bracket

Supporting deep dive: /post/rmd-tax-spike-planning-1m-ira

Your practical action list (do this before you pick your retirement date)

If you’re within a few years of retirement and you have $500K to $5M saved, here are the steps I want you to take.

1) Build a one-page income timeline (ages 60–75)

List, by year:

- When wages stop

- When Social Security starts (each spouse)

- When Medicare starts (each spouse)

- When RMDs start

If you can’t see it on one page, you can’t manage it.

2) Estimate your “gap years” and decide what they’re for

Gap years are the years after you stop working but before Social Security and/or RMDs fully kick in.

For many households, these are your best years for:

- Roth conversions

- Capital gains planning

- Resetting your portfolio risk

3) Set a MAGI target for Medicare years

You don’t need to memorize IRMAA tables.

You do need to know whether your plan is likely to cross IRMAA thresholds.

Start here: /post/medicare-irmaa-magi-planning

4) Write down your withdrawal sequence

Decide in advance:

- Which account funds spending first

- How you’ll realize capital gains

- When you’ll use IRA withdrawals vs Roth

Start here: /post/retirement-withdrawal-sequencing-taxable-ira-roth

5) Stress-test the first 5 years, not just the long-term average

Ask:

- What if the market drops 20% in year one?

- What if inflation runs higher for a few years?

- What if one spouse lives to 95?

Start here: /post/sequence-of-returns-risk-first-5-years

6) Decide what’s overrated and underrated in your plan

In my opinion, for $900K–$2.8M households:

- Overrated: “I’ll just claim Social Security early so I don’t touch investments.”

- Overrated: “Taxable first, always.”

- Underrated: using gap years for Roth conversions.

- Underrated: planning around IRMAA with intention.

- Underrated: a cash/bond buffer and written guardrails.

If you want the same analysis for your household

If you’re sitting on real assets and you’re close to retirement, you don’t need hype. You need a coordinated map.

At Grape Wealth Management, we build a Retirement Income & Tax Map designed for $500K–$5M households. It coordinates:

- Social Security timing

- Roth conversion windows

- IRMAA-aware MAGI targets

- Withdrawal sequencing across taxable/IRA/Roth

- Portfolio risk management for the first decade

If you want to see whether “don’t retire yet” applies to you—or whether you’re clear to retire with confidence—book an appointment here:

Book an appointment: Schedule appointment