Retirement Planning 2026: The 401(k), IRA, and Roth Conversion Rules That Will Define Your Net Worth

Consider Sarah, age 60, married filing jointly, with a $1,000,000 traditional IRA and another decade of work ahead of her. If she does nothing — lets the IRA grow, takes RMDs at 73, and pays ordinary income tax on every withdrawal at the 32% rate she'll likely hit when Social Security and required distributions stack up — her lifetime federal tax bill on that single account will be approximately $603,000.

If instead she runs a thirteen-year Roth conversion ladder, filling the 24% bracket each year while her income is lower, the total is approximately $226,000 in taxes during the conversion phase, plus a small remaining bill on what's left in the IRA. Lifetime federal tax saved: $376,672.

That number is not hypothetical. It is the output of the same calculator embedded below, run on Sarah's exact inputs. The math is fully traceable: 2026 federal brackets from Rev. Proc. 2025-32, the Uniform Lifetime Table from Treas. Reg. § 1.401(a)(9)-9, and the 401(a)(9) RMD rules as amended by SECURE 2.0. Plug in your own numbers and the engine returns a comparable figure for your situation.

This guide walks through the rules that produce that delta — the new 2026 limits, the SECURE 2.0 surprises, the bracket-fill strategy, the pro-rata rule that breaks most backdoor Roth attempts, and the RMD trap that catches retirees who assumed they wouldn't need to touch their IRA.

→ Run the calculator with your numbers

What changed for 2026: the new IRS limits

On October 22, 2025, the IRS released Notice 2025-67, the annual cost-of-living adjustments for retirement plan limits. Two figures came in higher than most plan administrators had penciled in:

The standard deduction for 2026 (Rev. Proc. 2025-32) rose to $32,200 for MFJ and $16,100 for single filers. Federal ordinary-income brackets — the 10/12/22/24/32/35/37 schedule from TCJA — were made permanent by the One Big Beautiful Bill Act signed in July 2025, so there is no 2026 rate cliff. The bracket thresholds themselves moved up with inflation under Rev. Proc. 2025-32: the top of the 22% bracket for a single filer is $105,700 in 2026, and $211,400 for joint filers.

Two limits that did not move are worth flagging because they regularly catch high earners off guard. The Roth IRA contribution phase-out for married-filing-separately filers remains $0–$10,000 — a statutory floor that has never been COLA-adjusted. The traditional IRA deduction phase-out for MFS filers is identical. Anyone in those filing positions effectively has no Roth IRA option without a backdoor strategy.

The SECURE 2.0 Roth catch-up mandate finally takes effect

Of every change landing in 2026, the one most likely to surprise high earners is the SECURE 2.0 § 603 Roth catch-up mandate. Starting January 1, 2026, if your prior-year FICA wages (W-2 box 3) exceed $150,000, every dollar of your age-50 or age 60–63 catch-up contribution must be designated Roth. You cannot deduct it. The base $24,500 deferral can still be pre-tax; only the catch-up is forced.

Three details that produce the most confusion:

• The threshold is strictly greater than $150,000. A taxpayer with exactly $150,000 of FICA wages is not subject to the mandate. Wages of $150,001 are.

• The mandate operates on the prior year's wages. Someone who earned $200,000 in 2025 but takes a sabbatical in 2026 still faces the mandate in 2026, because the test is backward-looking.

• Plans that do not offer a Roth feature have a problem. If your employer's 401(k) lacks a Roth option and the mandate applies to you, you cannot make the catch-up contribution at all. You are not permitted to take it pre-tax in lieu. Most large plans added Roth features in 2024 or 2025 in anticipation; some smaller and ERISA-exempt plans did not.

The financial impact is moderate but real. A 55-year-old earning $200,000, contributing the maximum $32,500 to a 401(k), now gives up roughly $1,920 of current-year tax deduction (the $8,000 catch-up at the 24% marginal rate). Whether the mandated Roth treatment helps or hurts over the long run depends on whether their retirement-era marginal rate exceeds 24% — which, given typical RMD trajectories, is the case for most high earners. The calculator surfaces a regulatory note in any year where the mandate triggers, so you can see exactly how much of your contribution is forced into the Roth bucket.

The age 60–63 super catch-up has a quiet trap

SECURE 2.0 § 109 created a special "super catch-up" for participants aged 60, 61, 62, or 63: $11,250 in 2026, replacing the regular $8,000 age-50 catch-up for those four years. Ages 64+ revert to the regular catch-up.

The trap is the replacement word. Many retirement-planning content treats $11,250 as if it stacks on top of the $8,000 base catch-up, producing an imaginary $19,250 figure. It does not. A 61-year-old's total catch-up is $11,250, period. A 64-year-old's catch-up reverts to $8,000 — they do not stay at the higher number.

Combined with the base § 402(g) deferral of $24,500, this means:

• Ages 50–59: maximum $32,500 employee contribution
• Ages 60–63: maximum $35,750 employee contribution
• Ages 64+: maximum $32,500 employee contribution

The age-64 reversion frequently surprises clients. A senior employee who maxed out at $35,750 for four years and budgeted accordingly has to lower their deferral by $3,250 the year they turn 64. The calculator handles all three tiers automatically.

The § 415(c) limit — and why catch-up dollars don't count

The most commonly miscommunicated rule in U.S. retirement-plan administration concerns § 415(c). The 2026 limit on total annual additions to a defined-contribution plan — employee deferrals plus employer match plus employer profit-sharing plus after-tax voluntary contributions — is $72,000.

Per § 414(v)(3), catch-up contributions are excluded from this cap. A 60-year-old can therefore receive $72,000 of total additions plus $11,250 of catch-up, for a true total of $83,250 in a single plan year. This matters in two scenarios:

• High-income partners or executives at firms with generous profit-sharing arrangements. If your employer is depositing $50,000+ in profit-share into your 401(k), the order of operations matters: the $11,250 catch-up does not consume any of the $72,000 cap.

• Mega-backdoor Roth strategies. The "after-tax voluntary" bucket inside a 401(k) is bounded by the residual room under § 415(c), after employee deferral and employer contributions are accounted for. The catch-up exclusion gives you slightly more headroom for after-tax contributions if your plan permits them.

The calculator's projection engine validates total additions per year and surfaces a warning if the user's inputs would breach the cap.

The bracket-fill Roth conversion strategy

Here is the strategy that produces Sarah's $376,672 lifetime saving.

The premise: most retirees end up in a higher marginal tax bracket after age 73 than they were in the years just before. Required minimum distributions stack on top of Social Security, pension income, and any part-time work. A retiree with a $1.2M traditional IRA at 73 will see RMDs around $45,000 in the first year, growing every year as the Uniform Lifetime Table divisors shrink. By the late 80s, RMDs alone often push retirees from the 24% bracket into 32%.

If you can identify a window — typically the years between actual retirement (often 60-65) and the RMD beginning age (73) — when your income drops, you can convert traditional IRA dollars into a Roth IRA at a lower marginal rate than you'd otherwise pay later. Two important constraints make this non-trivial:

Conversions are taxed as ordinary income in the year of conversion. If you convert $200,000, that $200,000 is added to your taxable income for the year. Done badly, this throws you into a higher bracket than you started in — defeating the purpose.

The greedy "fill the bracket" approach. The optimizer runs through each year of your conversion window and computes the exact dollar amount that would bring your total taxable income to the top of your target marginal rate (typically 22% or 24%) without spilling into the next one. It honors:

• Your other taxable income for the year (Social Security, dividends, part-time work)
• The standard deduction
• The pro-rata rule (see next section)
• The IRA's continued growth between conversions

For Sarah's profile — age 60, $50,000 of consulting income, $1M traditional IRA, target rate 24%, 13-year horizon, 5% real growth — the strategy converts approximately $80,000 in year one (filling the 24% bracket from her $50k baseline up to its $403,550 top for MFJ), and continues year-over-year. The IRA shrinks; the Roth account grows; her future RMDs are dramatically smaller because the taxable balance is gone.

Crucially, what the strategy is not:

• It is not globally optimal. A globally optimal solution would require dynamic programming over the entire post-retirement trajectory, including IRMAA cliffs, Social Security taxability dynamics, and state tax variations. The greedy year-by-year approach matches what credentialed advisors actually recommend in practice, runs in milliseconds, and produces results within ~5% of the globally optimal plan in published academic comparisons.

• It is not appropriate for everyone. If your retirement-era marginal rate will be lower than your current rate (early retirees with no other income, taxpayers planning to move from a high-tax state to Florida), conversions cost you money. The calculator surfaces "Lifetime tax savings ≤ 0" when this is the case.

→ Run your own bracket-fill scenario

The pro-rata rule: why most "backdoor Roth" attempts fail

If you have an existing traditional IRA with pre-tax money, you cannot do a clean backdoor Roth. The reason is IRC § 408(d)(2): the pro-rata rule.

The rule treats every traditional, SEP, and SIMPLE IRA you own on December 31 of the conversion year as a single pool. When you convert any portion, the IRS computes the taxable fraction as:

non-taxable ratio = total basis / total IRA balance
taxable ratio    = 1 − non-taxable ratio

The textbook trap: a high earner makes a $7,500 non-deductible Trad IRA contribution, then converts that $7,500 to Roth a day later, expecting it to be tax-free because it was after-tax money going in. But if they have a $93,000 pre-tax rollover IRA from a prior employer's 401(k), the math changes:

• Total IRA balance on 12/31: $100,500
• Total basis: $7,500
• Non-taxable ratio: 7,500 / 100,500 = 7.46%
• Taxable conversion: $7,500 × 92.54% = $6,940

The taxpayer ends up paying ordinary income tax on $6,940 of "their own" money. Worse, the basis doesn't disappear — it stays prorated across the remaining IRA pool, dribbling out non-taxably over decades of future distributions.

There are three workable responses, each with trade-offs:

• Roll the pre-tax IRA into a 401(k) first. Most modern 401(k) plans accept incoming rollovers. Once the pre-tax money is back inside an ERISA plan, it disappears from the § 408(d)(2) calculation, and the next backdoor conversion is clean. Subject to plan acceptance and possible loss of certain ERISA protections that apply differently to IRAs.

• Convert the entire pre-tax IRA in a single tax year. Painful but fast. After the conversion, future backdoor moves are clean. Spreads more naturally if you can split across two tax years.

• Don't do the backdoor. If you're in the 32% bracket today and would have been at 22% in retirement, the backdoor is a bad trade. Pay the tax later. The calculator's "Lifetime tax savings" figure will tell you.

The RMD trap most retirees don't see coming

The most quietly destructive misconception in retirement planning is the assumption that "I won't need to touch my traditional IRA, so it'll just keep growing tax-deferred." It will not.

Required Minimum Distributions are not optional once you reach the SECURE 2.0 beginning age of 73. The IRS divides your prior-year-end IRA balance by a divisor from the Uniform Lifetime Table. The divisor is 26.5 at age 73 and shrinks every year — to 20.2 at 80, 12.2 at 90, and 2.0 at 120. Even if your account is growing at 5%, the divisor is shrinking faster than the balance can compound.

A practical consequence:

Note that these are minimums. The retiree may withdraw more if they want; they cannot withdraw less. Missing an RMD triggers a 25% excise tax under SECURE 2.0 § 302 (down from the pre-2023 50% rate, with a further reduction to 10% if corrected within the correction window). Both Trad IRA and pre-tax 401(k) RMDs are taxed as ordinary income at distribution.

Two carve-outs are worth knowing:

• Roth IRAs and Roth 401(k)s have no lifetime RMDs. Roth IRAs never had them (§ 408A(c)(5)). Roth 401(k)s used to require RMDs unless the participant rolled to a Roth IRA before 73 — but SECURE 2.0 § 325 eliminated that requirement for distributions required after 2023. As of 2026, Roth account holders can simply leave the money compounding indefinitely.

• Still-Working Exception. A participant who is currently employed by the plan sponsor and owns less than 5% of the company can defer RMDs from that specific 401(k) until actual retirement. The exception does not apply to IRAs (statutory) and does not extend to former employers' 401(k) accounts. A common error is assuming an IRA rollover keeps the deferral; it does not.

When the bracket strategy backfires: spending phase

The withdrawal-phase logic in the calculator runs the conventional advisor playbook: pull from taxable brokerage first, then pre-tax up to a target marginal rate ceiling (default 22%), then top up from Roth. The premise is that taxable brokerage spends at zero marginal rate (already-yours principal), pre-tax dollars are taxable, and Roth dollars are tax-free.

The complication is that the RMD floor binds. Consider a 75-year-old with $1M in a traditional IRA, $200,000 in taxable brokerage, and only $30,000 in annual spending needs. The naive plan: take $30,000 from taxable, leave the IRA growing.

That plan is illegal. The RMD on $1M at age 75 is $1M / 24.6 = $40,650. The retiree must withdraw at least that amount from the IRA, generating ordinary income tax of roughly $2,700 at MFJ rates after the standard deduction — even though they didn't need the cash. The tax is non-negotiable.

This is why the bracket-fill conversion strategy from earlier in life pays off: every dollar of pre-tax balance you convert before 73 is a dollar that doesn't trigger forced taxation later. The calculator's withdrawal tab makes this binding force visible — the schedule highlights years where the RMD floor exceeds what the bracket strategy would otherwise have pulled, and shows the federal tax bill year-by-year.

Why we use Monte Carlo (and why the median is below the mean)

The calculator offers a Monte Carlo toggle that replaces the deterministic single-return projection with 1,000 stochastic paths drawn from correlated lognormal distributions for equity and bonds. Default parameters are equity μ=7% / σ=15%, bonds μ=3% / σ=5%, correlation −0.10 — figures consistent with U.S. monthly data from 1928 onward, adjusted for survivorship bias.

The single most useful insight from running Monte Carlo on a retirement projection is this: the median outcome is below the mean. Lognormal terminal-wealth distributions are right-skewed. A small number of "lucky" paths drag the average upward; the typical experience falls below it. For a 35-year-old with our default profile projecting to age 67:

• Deterministic projection at 5.8% blended return: ~$4,800,000
• Monte Carlo P50 (median): ~$3,400,000 — 30% lower
• Monte Carlo P10 (10th percentile, "bad luck"): ~$2,100,000
• Monte Carlo P90 (90th percentile, "good luck"): ~$5,940,000

Most retirement calculators show only the deterministic figure. It is mathematically the expected outcome — but it is not the most likely outcome. Half of all possible futures land below the median, and the median is materially below the mean. Planning around the deterministic figure is a subtle form of optimism that compounds across decades.

The widget's Monte Carlo seed is reproducible. Setting the same seed produces byte-for-byte identical percentile bands — useful when sharing a simulation with an advisor or comparing two scenarios.

Run the calculator

The calculator below implements every rule discussed above. Inputs cover the four major scenarios — accumulation, Roth conversion ladder, RMD projection, withdrawal plan — and outputs include charts, year-by-year schedules, and inline regulatory notes whenever the engine encounters a binding limit, a phase-out, or a SECURE 2.0 rule trigger.

→ Open the calculator

The simulation runs entirely in your browser. No inputs are stored or transmitted. The "Share simulation" button generates a URL with your inputs encoded in the hash fragment, so you can send it to a spouse, advisor, or future-you without any server roundtrip.

What this calculator does NOT model

Every model is a simplification. The point of disclosing the simplifications is to let you decide whether they materially affect your conclusion. Eight items are out of scope:

State income tax. California taxes Roth conversions at full ordinary rates of up to 13.3%. Florida, Texas, Tennessee, Wyoming, Nevada, and South Dakota tax no retirement income at all. A Roth conversion strategy that saves $376,672 federally may save $450,000+ for a Texan or only $300,000 for a Californian. If you live in or plan to move to a high-tax state, the dollar figures change; the direction of the recommendation usually does not.

Net Investment Income Tax (3.8%). IRC § 1411 imposes 3.8% on certain investment income above MAGI thresholds of $200,000 (single) and $250,000 (MFJ). Affects taxable bucket dividends, capital gains, and passive income above those thresholds. For high earners running large conversions, the conversion itself can push MAGI above the threshold and pull other investment income into NIIT territory. This is a meaningful second-order effect.

Additional Medicare Tax (0.9%). IRC § 3101(b)(2) on wages above $200,000 single / $250,000 MFJ. Reduces effective after-tax compensation slightly for high earners during accumulation. Small impact relative to the other items here.

Alternative Minimum Tax (AMT). Post-TCJA the AMT hits very few taxpayers — primarily those with significant ISO exercises or unusual income mixes. If you are not currently flagged as an AMT filer, it is unlikely to become a binding constraint.

Qualified Business Income (QBI) deduction. IRC § 199A — the 20% deduction on certain pass-through income, made permanent by OBBBA. If you are self-employed, an S-corp owner, or earning K-1 income from a partnership, your taxable-income calculation differs from the W-2 wage case modeled here.

Long-term capital gains brackets. Withdrawals from the taxable brokerage bucket may be subject to LTCG rates (0% / 15% / 20%) on the gain portion, not ordinary rates. The calculator treats taxable bucket withdrawals as already-yours principal, slightly overstating the net spendable value.

Social Security taxability phase-in (the "tax torpedo"). Under IRC § 86, the taxability of Social Security benefits phases in across two thresholds. The mathematical result is that a retiree in the phase-in zone faces effective marginal rates of 22.2% (= 12% × 1.85) or 40.7% (= 22% × 1.85) on each additional dollar of "other" income — including Roth conversions and IRA withdrawals. For retirees with provisional income in the $25,000–$44,000 (single) or $32,000–$56,000 (MFJ) range, this dynamic can make a 22%-bracket conversion more expensive than a 24%-bracket conversion done outside the phase-in zone.

IRMAA Medicare premium cliffs. Medicare Parts B and D premiums step up at five MAGI thresholds — and the steps are cliffs, not phase-ins. Crossing a threshold by $1 can cost $1,000+ per year in surcharges, applicable for two years (because IRMAA looks at MAGI from two years prior). A Roth conversion strategy that ignores IRMAA cliffs can produce a "tax savings" that disappears once Medicare bills are factored in. If you are within five years of age 65 and running large conversions, IRMAA-aware planning is mandatory; this calculator does not currently model it.

Closing the loop

The retirement-planning landscape in 2026 contains four genuinely new constraints relative to the prior year: the higher § 402(g) and IRA contribution limits, the now-active SECURE 2.0 Roth catch-up mandate at $150,000 of FICA wages, the COLA-adjusted IRA catch-up of $1,100, and the OBBBA permanence of the TCJA bracket structure that removes the 2026 rate cliff that was on tax planners' calendars two years ago.

The strategies have not changed. Maximize match capture before anything else. Bracket-fill Roth conversions in the gap years between retirement and age 73. Watch for the pro-rata trap if you have any pre-tax IRA balance. Plan for RMDs you cannot avoid even if you don't need the cash. Run Monte Carlo before trusting any deterministic projection.

What the calculator changes is the speed of testing those strategies against your actual numbers. A traditional financial-planning engagement charges $2,000–$5,000 for a one-time projection that is largely driven by the same arithmetic. The calculator runs that arithmetic in your browser, in milliseconds, on your specific situation, with all the legal citations visible if you want to verify the math.

Whatever you decide, decide it on the basis of numbers, not headlines. Sarah's $376,672 is real; so is the $0 that the same strategy would produce for a low-income retiree planning to move from California to Florida. The point is not the dollar figure. The point is that you should know which of those two outcomes describes you.

→ Start with the calculator

Last updated April 30, 2026. Constants reflect IRS Notice 2025-67 (retirement-plan COLAs), Rev. Proc. 2025-32 (inflation-adjusted brackets), and Treas. Reg. § 1.401(a)(9)-9 (Uniform Lifetime Table). The calculator's source code, including every legal citation referenced above, is available at the project repository.

This guide is informational and is not tax, legal, or investment advice. Before executing any strategy described here, consult a credentialed CPA, enrolled agent, or fiduciary financial planner.