Direct Indexing for Retirees: 0% Gain Harvesting, IRMAA Cliffs, and the §1014 Payoff

Most direct indexing content is written for investors still building wealth — high earners offsetting RSU gains, executives managing K-1 events, tech employees accumulating in taxable accounts. The pitch is straightforward: systematic tax-loss harvesting generates 0.5–1.5%/year of tax alpha, which compounds meaningfully over decades at high income tax rates.

Retirement changes the math. You're no longer primarily accumulating; you're drawing down. Ordinary income from RMDs and Social Security may crowd your tax space. The 0% long-term capital gains bracket — unavailable when you were earning $500K/year — opens up. The §1014 step-up becomes a real end-game payoff you're now close enough to plan around. And IRMAA turns a previously theoretical cliff into a concrete annual premium surcharge that a $5,000 unplanned capital gain can trigger.

None of these shifts make direct indexing less valuable in retirement. They make it differently valuable — in ways that require more active coordination, not less.

Gain Harvesting in the 0% LTCG Bracket

During your working years, the 0% federal long-term capital gains bracket was probably irrelevant — your ordinary income pushed taxable income well above the ceiling. In retirement, that ceiling becomes accessible for the first time.

In 2026, the 0% LTCG rate applies to taxable income up to $49,450 (single) or $98,900 (married filing jointly).¹ A retired couple with $60,000 in Social Security benefits (partially excluded), a small pension, and modest RMDs might find themselves with $70,000–$80,000 of taxable income — inside the 0% bracket or just above it.

In that position, the classic tax-loss harvesting logic flips. Instead of harvesting losses, you intentionally harvest gains — selling appreciated positions in the DI portfolio to reset their cost basis at zero federal tax cost. This is called tax gain harvesting, and a directly indexed portfolio makes it precise:

• You identify specific lots across 200–400 individual stock positions with the smallest embedded gains, sell enough of them to fill the remaining 0% bracket space, and immediately repurchase. The position's weight in the portfolio is unchanged. The cost basis is reset to the current price.
• Future appreciation on those positions now starts from the new, higher basis. If you later sell in a higher-income year, the taxable gain is smaller. And if you never sell — if the position is held until death — the step-up at §1014 applies only to appreciation after the gain-harvest reset, not the full historical gain.
• An ETF investor cannot do this. Selling and repurchasing an ETF triggers wash-sale rules for 30 days. Selling and repurchasing individual stocks with different tickers — across 200+ positions — avoids the wash-sale trap because the securities are not substantially identical.

State taxes still apply. California, New York, New Jersey, and most other states don't have a 0% capital gains rate — they tax all capital gains as ordinary income. Gain harvesting makes the most sense for retirees in zero-income-tax states (Texas, Florida, Nevada, Washington) or those with limited state liability.

IRMAA Cliffs: The Retirement Tax That DI Can Prevent

Medicare's Income-Related Monthly Adjustment Amount (IRMAA) adds a surcharge to Part B and Part D premiums based on MAGI from two years prior. In 2026, the first IRMAA tier begins at $109,000 MAGI for single filers and $218,000 for married filing jointly — adding $81.20/month per person in Part B premium.²

For a couple, crossing the $218,000 threshold by even $1 triggers $1,949/year in additional Part B premiums. Surcharges reach $487/month per person (an additional $11,688/year per couple) at the highest tier.

What's different about IRMAA is that it's calculated on MAGI, which includes:

• Realized long-term capital gains (even those taxed at 0%)
• Capital gain distributions from mutual funds (which you cannot control in an ETF)
• Roth conversion income
• Required minimum distributions

This is where a directly indexed portfolio creates a concrete, dollars-and-cents advantage over ETF-based holdings. In a large ETF position, the fund manager's internal rebalancing generates capital gain distributions passed through to you — gains you didn't choose to realize, that count fully toward IRMAA. In a direct-indexed portfolio holding 200–400 individual stocks, there are no fund-level distributions. You control every taxable event.

At $109,000 MAGI (single) or $218,000 (MFJ), each subsequent IRMAA tier adds further surcharges. A specialist advisor managing a DI portfolio in retirement tracks MAGI throughout the year and can choose which lots to sell, whether to defer a Roth conversion, or whether to use loss harvesting to offset gains — all calibrated to keep MAGI below the next cliff.

Distribution-Phase Lot Selection: Selling Tax-Efficiently

Retirees regularly need to sell portfolio positions to fund living expenses. With a single ETF or mutual fund, selling is simple but blunt: you sell shares in some order (FIFO, average cost) and realize whatever gain or loss that method produces.

A direct-indexed portfolio with 200–400 individual positions gives you a different choice. Instead of selling "the portfolio," you sell specific lots of specific stocks to produce a precise taxable outcome:

• For income in a 0% bracket year: Sell appreciated lots to realize gains at zero federal cost, funding the year's spending while resetting basis.
• For IRMAA cliff protection: Sell high-basis lots first to minimize net gain and keep MAGI below the next threshold.
• For concentrated-sector risk reduction: Use ongoing TLH losses elsewhere in the portfolio to offset gains taken on a sector you're overweight in — reducing concentration without a large tax bill.
• For legacy planning: Deliberately retain the lowest-basis positions — the biggest winners — and sell only from higher-basis lots. The low-basis positions become the step-up candidates at death.

A standard ETF cannot do any of this. An advisor managing your DI portfolio in retirement is essentially running a tax-optimization engine across hundreds of individual positions simultaneously — targeting specific tax outcomes for each year's distribution need.

The §1014 Step-Up: Now You're Close Enough to Plan Around It

If you began direct indexing at 45 or 50 and are now 65 or 70, the §1014 step-up has shifted from theoretical to near-term reality. IRC §1014 eliminates all embedded capital gains on assets held in a taxable account at death — heirs inherit at date-of-death fair market value, with no capital gains tax owed on any appreciation during your lifetime.

In a DI portfolio managed over 15–20 years, this creates a specific pattern: the strongest performers — positions that tripled or quadrupled — were systematically not harvested (because they had gains, not losses). They've been sitting in the portfolio accumulating unrealized gains for years. At death, those gains disappear entirely under §1014.

In retirement, the right strategy is often to never sell the winners. Sell from higher-basis lots to fund spending. Sell overweighted positions that have lost value for TLH. But leave the 20x positions sitting untouched, compounding, waiting for the step-up. An advisor helps you identify which positions fall into each category and ensure your distribution strategy doesn't accidentally realize gains you intended to defer forever.

See our dedicated estate planning guide for the full math on §1014 asymmetry and DAF coordination strategies.

When Direct Indexing Is Less Compelling in Retirement

DI isn't always the right choice for retirees. Three situations where the math doesn't favor it:

• Full distribution phase with large RMDs pushing you into the 20% LTCG bracket. If RMDs from a large IRA put your ordinary income at $400,000+, you're in the same high-rate environment as when you were working. TLH still has value, but if you're actively drawing down the taxable portfolio rapidly, the platform fee may not pay for itself before the account is spent.
• No heirs, no charitable intent, and you'll spend the portfolio in your lifetime. The §1014 step-up benefit only accrues to heirs. If you plan to spend all assets in retirement, the "hold winners until death" strategy doesn't apply. TLH still has value, but the estate-planning multiplier disappears.
• Very low taxable account balance relative to total assets. If most of your wealth is in IRAs (no TLH benefit) and the taxable account is small, the annual platform fee on a $100K–$200K taxable balance may not generate enough tax alpha to justify the complexity.

Why Retirement Requires an Advisor More Than Accumulation Did

During accumulation, the DI strategy is relatively mechanical: harvest losses, avoid wash sales, don't sell winners. A platform's algorithm handles most of it.

In retirement, the decisions require coordination across multiple moving parts:

• Which lots to sell for each year's spending, calibrated to the current year's MAGI and IRMAA position
• Whether to realize gains in the 0% bracket this year or defer them (a decision that depends on your trajectory into future RMD years)
• Whether to do a Roth conversion this year, and whether DI losses can absorb some of the added income
• Which positions are step-up candidates vs. spending candidates, updated as the portfolio changes shape year to year
• Social Security timing (if not yet claimed) and how it interacts with capital gains and provisional income thresholds

A retirement specialist who manages DI portfolios sees all of these simultaneously. The platform algorithm sees none of them.