VVested
RSU Management··18 min read·Reviewed June 2026

The RSU tax-saving playbook for US employees: harvesting, donor-advised funds, NUA and the 83(b) myth

The legitimate tax-saving plays for US employees with RSUs: tax-loss harvesting around the wash-sale rule, donating appreciated shares via a donor-advised fund, NUA on 401(k) company stock, tax-gain harvesting, and why 83(b) doesn't apply to RSUs.

Share:XLinkedInWhatsApp

There is a wide gap between aggressive tax schemes that get people audited and the legitimate, boring tax-saving plays that high-earning employees routinely leave on the table. RSU holders are especially prone to the second problem: they pay six figures in tax on vests every year, then never harvest a loss, never donate a share, and pay full freight on a charitable giving budget they fund with cash. None of the moves below are clever or risky. They are the standard toolkit — and most people use none of them.

The 30-second answer: The legitimate RSU tax-saving plays for US residents are: harvest losses in your diversified portfolio to offset gains (mind the 30-day wash-sale rule by swapping into a similar-but-not-identical fund); donate appreciated long-term shares directly — ideally through a donor-advised fund — to deduct full fair market value and skip capital gains tax; use NUA if you hold employer stock inside your 401(k); and tax-gain harvest in the 0% long-term bracket during low-income years. One myth to kill: you cannot make an 83(b) election on standard RSUs — that election is for restricted stock awards and early-exercised options, not RSUs.

This is the companion to the diversification playbook and the retirement-corpus guide. Those cover where the money goes; this covers how to keep more of it from the IRS along the way. It is written for US residents filing a US return.

Kill the 83(b) myth first

Search "RSU tax tips" and you will eventually be told to "file an 83(b) election" to lock in a low tax. For standard RSUs, this is wrong, and acting on it is impossible.

An 83(b) election lets you choose to be taxed at grant on restricted property, betting that the value at grant is low and future appreciation gets pushed into capital-gains treatment. It works because actual property — shares — is transferred to you at grant, subject to vesting. RSUs are different: they are an unfunded promise to deliver shares in the future. Nothing is transferred at grant, so there is nothing to make an election on. The IRS taxes RSUs at vest, full stop.

The 83(b) election is real and valuable — but for restricted stock awards (RSAs) and early-exercised stock options, which startup founders and very early employees receive. If you have those, an 83(b) filed within 30 days of grant can matter enormously. If you have RSUs, it simply does not apply. Knowing the difference stops you from chasing a phantom and missing the plays that do work.

The distinction comes down to one word: property. An 83(b) election is an election under Section 83 to be taxed when restricted property is transferred, even though it has not yet vested. RSAs and early-exercised options put real shares in your name on day one — those shares can be forfeited if you leave, but they exist, and you own them. RSUs put nothing in your name. You hold a contractual right to receive shares later. There is no property to be taxed early, so there is no election to make, and a brokerage cannot file one for you even if you ask.

It helps to see the three side by side:

Equity typeProperty transferred at grant?83(b) available?When taxed (default)
Standard RSUNo (unfunded promise)NoAt vest, as ordinary income
Restricted stock award (RSA)Yes (shares, subject to vesting)Yes (file within 30 days)At vest, unless 83(b) elected
Early-exercised stock optionYes (shares, subject to vesting)Yes (file within 30 days)At vest of the exercised shares, unless 83(b) elected

The 30-day window is unforgiving: it runs from the grant or exercise date, there are no extensions, and a late filing is simply void. So the only useful action for an RSU holder is to confirm what you actually hold. Read the grant agreement. If it says "Restricted Stock Unit," there is no 83(b) to make, and the next four plays are where your attention belongs.

Bottom line: You cannot 83(b) a standard RSU because nothing is transferred to you at grant — that election is only for RSAs and early-exercised options, filed within 30 days, and confusing the two costs you nothing but wasted effort chasing a phantom.

Play 1 — tax-loss harvesting, around the wash-sale rule

When you diversify your RSU proceeds into a broad portfolio, that portfolio becomes your harvesting engine. In any year, some holdings are underwater. Selling them realises a loss that offsets your realised gains — including the gains from trimming your concentrated employer position — and then up to $3,000 of ordinary income per year, with the remainder carried forward indefinitely.

The constraint that trips people up is the wash-sale rule: if you buy a "substantially identical" security within 30 days before or after the loss sale, the IRS disallows the loss. The clean workaround is to harvest and immediately replace with something highly similar but not identical.

Worked example. Harvesting to offset an RSU sale gain

In 2026 you trim your concentrated employer position and realise a $40,000 long-term gain. Separately, a total-US-market index fund you bought last year is down $12,000.

You sell the index fund, booking a $12,000 loss, and on the same day buy a different provider's S&P 500 index fund — similar exposure, not substantially identical, so no wash sale. Your taxable gain for the year drops from $40,000 to $28,000. At a 15% LTCG rate plus 3.8% NIIT (18.8%), that is tax of $5,264 instead of $7,520 — a $2,256 saving for a few minutes of work, with your market exposure unchanged.

The ordering rules, and why losses are worth more than they look

Capital losses are spent in a fixed order. First they offset capital gains of the same character (long-term loss against long-term gain, short-term against short-term), then the categories net against each other, then up to $3,000 of leftover loss offsets ordinary income per year, and anything still left carries forward indefinitely. That carryforward is the part people undervalue. A loss harvested today does not expire; it sits on your return as ammunition for every future gain until it is used up.

Worked example. A big-loss year carried forward over four years

Suppose a sharp drawdown lets you harvest a $60,000 long-term loss in 2026, a year in which you have no gains to offset. You cannot deduct it all at once. Here is how it drains over the following years as you realise gains and take the annual ordinary-income allowance.

YearGains realisedLoss applied to gainsOrdinary-income offsetLoss used this yearCarryforward remaining
2026$0$0$3,000$3,000$57,000
2027$20,000$20,000$3,000$23,000$34,000
2028$10,000$10,000$3,000$13,000$21,000
2029$18,000$18,000$3,000$21,000$0

Over four years the $60,000 loss wipes out $48,000 of capital gains and $12,000 of ordinary income. At an 18.8% effective LTCG rate, sheltering $48,000 of gains is worth about $9,024; the $12,000 of ordinary income offset, at a 35% marginal rate, is worth about $4,200 — roughly $13,200 of tax deferred or saved from a single harvest, spread across years you would otherwise have paid in full.

Keeping the loss real: the wash-sale traps

The wash-sale rule is also why you keep harvesting separate from your employer stock: if you are selling company shares at a loss and your next vest lands within 30 days, that vest can trigger a wash sale on the loss. Watch the calendar around vest dates.

Three edge cases catch careful people:

  • The vest is a purchase. An RSU vesting is treated as acquiring shares. Harvesting a loss on company stock within 30 days before or after a vest of the same stock disallows part or all of the loss.
  • Dividend reinvestment counts. If a fund you harvested has automatic dividend reinvestment switched on, a reinvest inside the 30-day window buys back a substantially identical security and triggers a partial wash sale. Turn off reinvestment before harvesting.
  • The IRA leak. Buying the substantially identical security inside your IRA or 401(k) within the window also triggers the rule — and worse, the disallowed loss vanishes rather than adjusting basis. Coordinate across every account, not just the taxable one.

Bottom line: Harvested losses offset gains first, then $3,000 of ordinary income a year, then carry forward forever — so a single big-loss year can quietly shelter five figures of future tax, provided you avoid rebuying anything substantially identical within 30 days in any account, including your retirement accounts and your next vest.

Play 2 — donate appreciated shares through a donor-advised fund

If you give to charity at all, this is the highest-leverage move in the playbook, and almost nobody does it correctly. The mistake is giving cash while holding appreciated shares. The fix: give the appreciated shares directly.

When you donate long-term appreciated stock to a public charity, two things happen. You deduct the full fair market value, and you never pay capital gains tax on the built-in appreciation, because the charity is tax-exempt and sells with no tax. Donating cash instead means you first pay capital gains to free up that cash, then donate less.

Worked example. Same $50,000 gift, stock versus cash

You want to give $50,000 to charity. You hold appreciated shares worth $50,000 with a cost basis of $10,000 — so $40,000 of long-term gain is built in. Compare giving those shares directly against selling them and giving cash, assuming an 18.8% LTCG rate and a 35% marginal income-tax rate.

Donate shares directlySell shares, donate cash
Gift value to charity$50,000$50,000
Capital gains tax you pay$0$7,520 (18.8% of $40,000)
Charitable deduction$50,000$50,000
Income-tax value of deduction (35%)$17,500$17,500
Net cost of the gift$32,500$40,020

Donating the shares directly costs you $32,500 out of pocket against $40,020 for the sell-then-give route — a $7,520 difference, which is exactly the capital gains tax you avoided. The charity receives the same $50,000 either way; you simply keep more by never triggering the gain. (Note: when you give the shares, you should buy them back later only with cash you would have donated anyway — there is no wash-sale concern on a gift, but rebuying resets your basis to the new, higher price.)

A donor-advised fund (DAF) makes this practical. You contribute appreciated shares to the DAF in one transaction, take the full deduction in that year, and then recommend grants to actual charities over months or years. This unlocks a second technique — bunching: concentrate several years of intended giving into a single high-income year (a big vest year, an IPO, a bonus), clear the standard deduction by a wide margin, and then grant it out gradually.

Worked example. Bunching two years of giving into one

Suppose you give about $20,000 a year and your only other itemizable deduction is $10,000 of state and local tax (capped). The 2025 standard deduction is roughly $30,000 for a married couple. Compare giving $20,000 each year against bunching two years' worth ($40,000) into a single year through a DAF, using a 37% marginal rate for illustration.

Without bunching (each year): itemized deductions are $10,000 SALT + $20,000 gift = $30,000, which only ties the standard deduction. You get no extra benefit from the gift in either year — you would have taken the $30,000 standard deduction regardless. Two-year deduction total: $60,000.

With bunching (year one): you contribute $40,000 to the DAF in year one. Year-one itemized deductions are $10,000 SALT + $40,000 gift = $50,000, beating the standard deduction by $20,000. Year two you give nothing fresh (you grant out of the DAF) and take the $30,000 standard deduction. Two-year deduction total: $50,000 + $30,000 = $80,000.

Bunching produced $20,000 more in deductions across the two years. At a 37% marginal rate that is about $7,400 of additional tax saved — for giving the same total amount, just front-loaded. The DAF lets the charities still receive their grants on the old yearly schedule.

The limits to know: the deduction for appreciated property to a public charity is capped at 30% of your adjusted gross income in a year, with a five-year carryforward for anything above that. Cash gifts get a higher 60%-of-AGI limit, but lose the capital-gains advantage. For shareholders sitting on large gains, the 30% appreciated-stock route almost always wins. One practical sequencing point: pick the lots with the lowest basis and longest holding period to give away first — that maximizes the capital gains you sidestep per dollar donated.

Bottom line: Give appreciated long-term shares, not cash, so you skip the capital gains tax entirely and still deduct full market value — and route them through a DAF to bunch several years of giving into one high-income year, which can add five figures of deductions for the same total generosity.

Play 3 — NUA on employer stock inside your 401(k)

This one is narrow but valuable if it applies. If you hold actual employer shares inside your 401(k) — common where the plan offers company stock as an investment option or where ESPP/match came in as stock — Net Unrealized Appreciation (NUA) can cut the tax sharply.

On a qualifying lump-sum distribution, you pay ordinary income tax only on the cost basis of those shares, and the appreciation (the NUA) is taxed at long-term capital gains rates when you sell, rather than all of it being taxed as ordinary income on withdrawal like the rest of your 401(k). For someone with heavily appreciated company stock in the plan, the difference between LTCG rates and ordinary rates on the appreciation is large.

Worked example. NUA distribution versus rolling everything to an IRA

Say you hold company stock inside your 401(k) worth $300,000, with a cost basis of $50,000 — so $250,000 is net unrealized appreciation. Compare two paths at the same 35% ordinary rate and an 18.8% LTCG rate. (For simplicity, assume you sell the shares soon after distribution; the NUA is taxed at long-term rates regardless of how long you held inside the plan.)

NUA distributionRoll to IRA, then withdraw
Cost basis$50,000$50,000
Net unrealized appreciation$250,000$250,000
Tax on the basis$17,500 (35% ordinary)
Tax on the appreciation$47,000 (18.8% LTCG)
Tax on full $300,000 withdrawal$105,000 (35% ordinary)
Total tax$64,500$105,000

The NUA route pays $64,500 against $105,000 for treating the whole balance as ordinary income on withdrawal — a $40,500 saving, driven entirely by moving the $250,000 of appreciation from a 35% rate to an 18.8% rate. The bigger the gap between basis and value, the larger the win; NUA does little when the stock has barely appreciated.

When NUA does not pay

NUA does not apply to RSUs sitting in a brokerage account — only to employer stock inside the qualified plan — and the lump-sum-distribution rules are strict (it must follow a triggering event such as separation from service, reaching 59 and a half, death, or disability, and the entire account must be distributed in one tax year). Several things can sink the benefit:

  • High basis. If the basis is a large fraction of the value, you pay ordinary tax on most of it up front and the LTCG advantage is thin.
  • A broken lump sum. Take any partial distribution in a prior year, or fail to empty the account in one year, and the lump-sum requirement is blown.
  • Rolling the stock into an IRA by mistake. Once company shares land in an IRA, the NUA election is gone permanently — the appreciation becomes ordinary income on withdrawal.

Model it with a CPA before pulling the trigger, because a botched distribution forfeits the benefit and there is no do-over.

Bottom line: If low-basis employer stock is sitting in your 401(k), a qualifying NUA lump-sum distribution taxes only the basis at ordinary rates and the appreciation at long-term rates — often a five-figure saving — but the rules are strict and one wrong step, especially an accidental IRA rollover, erases the benefit for good.

Play 4 — tax-gain harvesting in low-income years

Most RSU holders are in high brackets and this does not apply — but the years it does apply are exactly the years people forget to act. In a low-income year — a sabbatical, a gap between jobs, unpaid leave, the first year of a startup, early retirement — your taxable income can drop into the 0% long-term capital gains bracket, roughly up to $48,350 of taxable income for single filers and $96,700 for married filing jointly in 2026 (indexed annually).

In that window, you can deliberately realise long-term gains that fit inside the 0% band and immediately rebuy the same security. There is no wash-sale rule on gains, so you reset your cost basis higher at zero federal tax cost. Do this in a planned low-income year and you quietly erase future capital gains tax on a chunk of your portfolio.

The mechanic that matters is that the 0% bracket is measured against taxable income, and your long-term gains stack on top of your ordinary income. So the headroom for free gains is the top of the 0% band minus your ordinary taxable income — not the full bracket.

Worked example. A sabbatical year inside the 0% bracket

You take a single-filer sabbatical in 2026. Your only ordinary income is $20,000 of consulting work. After the roughly $15,000 standard deduction, your ordinary taxable income is about $5,000. The top of the 0% LTCG band is about $48,350 of taxable income.

That leaves roughly $43,350 of room for long-term gains taxed at 0% ($48,350 minus the $5,000 of ordinary taxable income). You hold an index fund with $43,000 of long-term gain, so you sell it, pay $0 federal tax on the gain, and immediately rebuy at the new price. Your cost basis resets upward by $43,000.

The payoff comes later. In a future high-income year, selling that same fund would have triggered tax at 18.8% on the $43,000 — about $8,084. By harvesting the gain for free during the low-income year, you erased that future bill entirely while keeping the exact same position. Watch two limits: gains spilling above the $48,350 ceiling are taxed at 15%, and a large realised gain can raise your AGI enough to affect ACA premium subsidies, so size the harvest to stay inside the band.

Bottom line: In a sabbatical or gap-year when your taxable income is low, you can realise long-term gains that fit under the 0% bracket (about $48,350 single, $96,700 married, minus your ordinary income), pay zero federal tax, and rebuy immediately to reset basis higher — turning a future five-figure capital gains bill into nothing.

A note on startup and pre-IPO equity (QSBS)

If your "RSUs" are actually early-stage startup shares — RSAs or early-exercised options in a qualified small business — you may be in Qualified Small Business Stock (Section 1202) territory, where a large share of the gain can be excluded from federal tax if holding-period and company-size tests are met. The 2025 tax law revised QSBS with tiered exclusions, a higher dollar cap, and shorter minimum holding tiers. This is a specialist area with real money at stake; if it might apply to you, get a tax adviser who handles QSBS specifically rather than relying on general guidance.

Bottom line: Standard RSUs are not QSBS, but if you hold genuine early-stage startup stock, the 2025 changes to Section 1202 may let you exclude a large share of the gain — this is worth a specialist's review rather than a rule of thumb.

The closing read

None of this is exotic. Harvest your losers to pay for trimming your winners. Give shares, not cash, and route them through a donor-advised fund so the deduction and the giving can happen on different schedules. Use NUA if company stock is trapped in your 401(k). Realise gains for free in the years your income dips. And stop trying to 83(b) an RSU — it cannot be done. The tax code rewards the employee who does the boring things consistently far more than the one chasing a clever structure once. The plays are sitting there every year. Use them.

Cross-references

Critical disclaimer: this article reflects US federal tax rules as of June 2026 and is general information, not personalised advice. Capital gains brackets, NIIT thresholds, AGI deduction limits, NUA distribution rules, the wash-sale rule, and QSBS (Section 1202) treatment depend on your specific facts and can change. State tax treatment varies. Consult a licensed CPA or CFP before acting. Nothing here is a recommendation to buy, sell, or donate any specific security.

Found this useful? Share it.

Help another Indian working with US RSUs or LRS not get blindsided by this stuff.

Share:XLinkedInWhatsApp

About the author

Shivang Badaya
Shivang Badaya

Co-Founder & Chief Executive Officer, Rovia

CFA charterholder with 10+ years across hedge funds and NRI fintech. Covers RSU taxation, equity comp, and cross-border investing for Indian residents. Ex-JP Morgan, Makrana Capital, Zolve.

More about Shivang

Get more like this in your inbox

One practical post a week on US investing & RSU strategy.