Direct Indexing and Roth Conversions

The most common misconception: "I'll harvest $50,000 of capital losses with my DI portfolio to offset my Roth conversion income." That's not how the tax code works — capital losses and ordinary income live in different character silos. But direct indexing does supercharge your Roth conversion strategy through three mechanisms that most advisors miss. Here's the real math.

The character mismatch: why capital losses don't offset Roth conversions

IRC § 408A treats a Roth conversion as a taxable distribution followed by a contribution — the amount converted is ordinary income in the year of conversion, added directly to your AGI. It is taxed at your marginal ordinary income rate.

Capital losses from tax-loss harvesting are capital in character. Under IRC § 1211(b), an individual can deduct net capital losses against ordinary income only up to $3,000 per year. Losses beyond that carry forward to future years as capital losses — they do not convert into ordinary income deductions.¹

Understanding this character distinction matters because it shapes how you coordinate the two strategies. You are not harvesting losses to pay for your conversions. You are doing two parallel things that interact at the margin — and the interactions are worth money, just not in the way most people expect.

Three ways direct indexing actually helps Roth conversions

1. DI eliminates mutual fund capital gains distributions

This is the biggest, most underappreciated benefit — and it has nothing to do with tax-loss harvesting directly.

When you hold mutual funds in a taxable account, the fund manager realizes gains internally (when rebalancing or meeting redemptions from other investors) and distributes them to all shareholders at year-end. You owe capital gains tax on those distributions even if you never sold a share. For actively managed funds, these annual LTCG distributions can run 5–15% of the fund's value in a strong market year. Even broad index funds occasionally distribute meaningful gains.

A direct-indexed portfolio does not work this way. You own the individual stocks. Gains are realized only when you direct a sale. In years when you're doing Roth conversions, you typically don't want to sell appreciated stocks — so you simply don't. No distribution. No surprise taxable event.

The practical effect: mutual fund investors planning Roth conversions have less control over their taxable income than DI investors do. If your $1.5M equity portfolio is in mutual funds and the funds distribute $30,000 in LTCG in December, that's $30,000 added to your MAGI whether you wanted it or not. It consumes LTCG bracket space, potentially raises your NIIT exposure, and may push you into a higher IRMAA tier the following year. A DI investor avoids this entire problem.

2. TLH-generated losses suppress net investment income

In years when the DI portfolio does generate capital gains (from rebalancing, closing positions at a gain, or raising cash), the loss bank built through ongoing tax-loss harvesting offsets those gains dollar-for-dollar. The result is lower net capital gains — and lower net investment income subject to the 3.8% Net Investment Income Tax (NIIT) under IRC § 1411.

This matters specifically during Roth conversions because a large conversion spikes your ordinary income, which in turn makes it more likely that any capital gains you realize will be taxed at 15–18.8% rather than 0%.² DI's systematic loss harvesting keeps capital gains low, so the investment income side of your tax picture stays lean while the conversion income fills your bracket.

3. The $3,000 annual ordinary income offset — small but consistent

When your DI portfolio generates more capital losses than capital gains (a net loss year), IRC § 1211(b) allows a $3,000 deduction against ordinary income. That's $720 of saved taxes per year at the 24% bracket — not transformative, but it's real and it compounds over a 10-year conversion window. $7,200 in cumulative tax savings from this mechanism alone, before accounting for compounding.

Additionally, the capital loss carryforward builds over time, so even in years when the DI portfolio has net gains, prior-year carryforwards are available to offset them. You enter each year of the conversion window with a growing reservoir of loss carryforwards that keeps investment income low.

The Roth conversion window and why DI is uniquely valuable in it

For HNW investors, the Roth conversion window typically falls between retirement (when earned income drops) and when Social Security and Required Minimum Distributions begin pushing income back up.

Under SECURE 2.0, RMDs now begin at age 73 for those born between 1951 and 1959, and at age 75 for those born in 1960 or later.³ If you retire at 60 or 62, you may have a 10–15 year window where your taxable income is dramatically lower than your peak earning years — and dramatically lower than it will be once RMDs force $80,000–$200,000+ per year out of a large IRA.

The math on converting in the window vs. taking RMDs later can be significant. At a $2M IRA growing at 6% annually, the RMD at age 75 (Uniform Lifetime Table divisor of ~22) is roughly $150,000+ in ordinary income, stacked on top of Social Security — likely pushing a significant portion into the 32%+ bracket. Converting $150,000/year in the window at the 22–24% bracket saves 8–10 percentage points on that income permanently.

DI amplifies the window in two ways:

• Income control. You can keep your taxable investment income low each year by not realizing gains — letting DI losses offset any gains that do occur. This preserves bracket capacity for the Roth conversion itself.
• § 1014 step-up preservation. The DI portfolio benefits from a full step-up in cost basis at death — gains never need to be realized at all if you hold to estate. This means DI-taxable and Roth-IRA are the two most tax-efficient wealth containers at death. Building both during the conversion window is the optimal long-run structure.

Worked example: Michael and Jennifer, ages 63

Setup: Retired couple in Texas (no state income tax). Both age 63, both retired in 2025. Michael has a $2M traditional IRA; Jennifer has a $500K traditional IRA. Combined, they plan to convert $200,000/year until RMDs begin at age 75 for both. Their goal: reduce projected RMDs and let Roth assets compound tax-free for heirs.

Income during the window:

• Part-time consulting (Jennifer): $40,000
• Rental income (net): $30,000
• Roth conversion: $200,000
• Taxable account income: variable (depends on whether they hold mutual funds or use DI)

Scenario A — $1.8M held in mutual funds:

The fund portfolio distributes $28,000 in LTCG and $12,000 in qualified dividends at year-end.

MAGI = $310,000 → above $250K NIIT threshold → NIIT applies to $40K of investment income.
LTCG + NIIT on $40,000: $40,000 × 18.8% = $7,520

Scenario B — $1.8M held in direct indexing portfolio:

The DI portfolio generates $38K in gross capital gains (from rebalancing) and $55K in capital losses (from TLH). Net: -$17,000 capital loss.

• $17K net loss: $3,000 deducted against ordinary income; $14,000 carried forward.
• DI qualified dividends (from individual holdings): ~$12,000 (similar to mutual fund dividends)
• Zero LTCG distributions (DI doesn't distribute gains)

MAGI = $279,000 → above $250K NIIT threshold → NIIT on $12K investment income.
LTCG + NIIT on $12,000: $12,000 × 18.8% = $2,256
Ordinary income difference (lower by $3,000 from §1211(b) offset): $3,000 × 24% = $720 savings

Annual tax savings with DI vs. mutual funds:

Compounded over 12 years of conversion activity (ages 63–75) at a 6% discount rate: ~$100,000 in after-tax wealth difference from this mechanism alone — before accounting for the additional compounding of Roth assets, the $14K/year capital loss carryforward building in the DI portfolio, and the § 1014 step-up on the DI portfolio at death.

2026 bracket reference for Roth conversion planning

Conversion math depends on where your ordinary income stacks in the bracket. The 24%/32% boundary is the critical decision point for most HNW investors doing conversions — converting at 24% and stopping there is typically the right call unless the IRA is so large that future RMDs will be taxed at 37%.

2026 standard deduction: $32,200 MFJ / $16,100 single. Brackets above are taxable income (after deductions). Source: IRS Rev. Proc. 2025-32 / Tax Foundation.⁴

LTCG thresholds (2026, married filing jointly):

NIIT threshold ($250K MFJ / $200K single) is not indexed for inflation. Source: IRC §1411(b); Kiplinger / IRS Rev. Proc. 2025-32.²

IRMAA: the Roth conversion trap DI can't fully solve

Medicare's Income-Related Monthly Adjustment Amount (IRMAA) assesses Medicare Part B and Part D surcharges based on your MAGI from two years prior. For 2026 Medicare premiums, IRMAA looks at 2024 tax returns. A large Roth conversion in 2026 will affect your 2028 Medicare costs.

IRMAA surcharges are tier-based and can add $1,000–$5,000+ per person per year in Medicare costs once you cross a threshold. For a couple, that's $2,000–$10,000+ per year — costs that last as long as MAGI remains elevated.

DI helps by keeping your investment income lower (eliminating fund distributions, keeping capital gains low). But the Roth conversion itself adds directly to MAGI and is the primary IRMAA driver. For any conversion above ~$150,000, you need to model whether the conversion is occurring near an IRMAA tier boundary. Crossing a tier line by $5,000 can trigger $2,000+ in additional Medicare costs — a marginal effective rate on that $5,000 that is much higher than the 24% you thought you were paying.

Your advisor should run both a Roth conversion bracket projection and an IRMAA tier projection together, not in isolation.

Questions to ask an advisor about DI + Roth conversion coordination

1. How do you build a 10-year Roth conversion model? You want an advisor who projects IRA balance growth, future RMDs, Social Security timing, and conversion tax cost in the same spreadsheet — not separately.
2. How do you coordinate the DI portfolio's tax activity with the Roth conversion calendar? Loss harvesting is more valuable in high-conversion years. Gain realization should be minimized or deferred. These decisions should be made together.
3. Do you do a Q4 income projection before I execute the final conversion tranche? Year-end fund distributions are often unknown until late November. An advisor who finalizes your conversion amount before knowing your DI portfolio's year-end activity leaves money on the table.
4. How do you handle IRMAA tier management in the conversion plan? A good answer includes modeling the two-year lag and avoiding tier crossings when the incremental conversion income is small.
5. Do you have access to institutional DI platforms? Consumer platforms (Wealthfront, Schwab, Frec) have lower minimums but limit cross-account wash-sale coordination. An advisor with access to Parametric or Aperio can coordinate wash-sale avoidance across your entire household — including the Roth IRA.

Do capital losses from direct indexing offset Roth conversion income?

Not directly. Roth conversions create ordinary income (IRC § 408A). Capital losses can offset capital gains in full, but can only deduct up to $3,000/year against ordinary income under IRC § 1211(b). If you harvest $60,000 of capital losses in the same year as a $200,000 Roth conversion, only $3,000 of that reduces ordinary income. The remaining $57,000 carries forward as capital loss carryforward to offset future capital gains.

How does DI help with Roth conversions if capital losses can't offset the income?

Three mechanisms: (1) Eliminating uncontrolled capital gains distributions from mutual funds keeps your MAGI lower and more predictable. (2) TLH losses offset capital gains you do realize, suppressing net investment income and the 3.8% NIIT. (3) The $3,000 annual ordinary income offset from net capital losses — modest but consistent over a 10-year conversion window. Together these effects can be worth $5,000–$10,000/year depending on portfolio size.

What is the Roth conversion window for direct indexing investors?

The conversion window runs from retirement until RMDs begin — age 73 for born 1951–1959, age 75 for born 1960+ under SECURE 2.0. During this window, ordinary income from work typically drops while the IRA continues to grow untaxed. Converting while marginal rates are lower than they'll be when RMDs force distributions is the core logic. DI investors can keep investment income artificially low during this window, maximizing the bracket space available for conversions.

Should I prioritize Roth conversions or building a DI loss bank?

Both. They serve different parts of your balance sheet. The DI portfolio lives in taxable and builds toward § 1014 step-up. The Roth conversion shifts IRA assets from future-taxable to future-tax-free. Starting both early — funding the DI account with available taxable cash and converting from the IRA at the same time — is the optimal strategy. The DI loss bank takes 1–3 years to build meaningfully; starting it before large gain events (including the year you start running high conversion income) is ideal.

Does DI make IRMAA worse with Roth conversions?

No — DI helps by keeping capital gains distributions out of MAGI. But DI cannot offset the Roth conversion's direct MAGI impact. The conversion income itself is the primary IRMAA trigger. Anyone converting $100,000+ per year should model IRMAA tier boundaries as part of the conversion plan. Avoiding a tier crossing by keeping the conversion $5,000 smaller is sometimes worth more than the marginal tax benefit of converting that additional $5,000.

When in the year should I execute a Roth conversion if I have a DI portfolio?

Wait until Q4 when your DI portfolio's year-to-date net capital activity is clearer. Fund distributions (if you still hold any funds) are usually announced in November. Do a full income projection with your advisor in late November or early December — then execute the final conversion tranche in the final weeks of December if the numbers support it. Converting a lump sum in January is simpler but gives up income optimization for the full year.

Can DI capital losses fund a larger Roth conversion at the same tax cost?

Indirectly, yes. If your DI portfolio eliminates $30,000 of capital gains distributions that would otherwise cost $5,250 in taxes (15% LTCG), that $5,250 in saved taxes can fund a modestly larger conversion at no net increase in total tax cost. The mechanism is different from "losses offset conversions," but the economic outcome is similar: lower total tax bill, more assets shifted to tax-free Roth accounts.