Investing in Opportunity Zones

In November 2019 ProPublica reported that the Rybovich superyacht marina on the West Palm Beach waterfront, with luxury apartment towers planned beside it, sat inside a census tract that Governor Rick Scott had nominated as an Opportunity Zone. The nomination came a little over a week after a direct appeal from one of the project’s backers, Wayne Huizenga Jr., whose family had given at least $1 million to Scott and the Florida Republican party over the previous decade. Poorer tracts the city itself had asked for were passed over. The New York Times reported separately that Treasury designated a tract in Storey County, Nevada containing land owned by Michael Milken, on instructions from Treasury Secretary Steven Mnuchin and over an internal IRS memo warning that the decision was “demonstrably unfair to every other state”. Mnuchin said he had no knowledge of Milken’s investments, and Treasury disputed the account, calling the Times report inaccurate. Neither story involved anything illegal. Both show the programme working as written: a governor draws a map, Treasury certifies it, and whoever already owns land inside the lines receives a federal subsidy on the gain.

That is one half of the record. The other half is that a great deal got built. Treasury’s own analysts, David Coyne and Craig Johnson, counted $112 billion of qualified Opportunity Zone property deployed in designated communities through 2024, reaching 77% of the tracts in the fifty states and the District of Columbia, and the accounting firm Novogradac has tracked more than 203,000 homes financed by the funds it follows. The Economic Innovation Group, which invented the policy and lobbied it into law, credits the incentive with 416,000 new housing units and a 70% increase in residential construction inside the zones. Whether those units would have been built anyway is the question the economists have spent seven years failing to settle.

This guide is about the instrument rather than the argument. Opportunity Zones are the only mechanism in the American tax code that lets a seller of any asset — a company, a portfolio of stock, a painting, a bitcoin position — defer the tax on the gain and then, if the money stays put for a decade, take the second gain out entirely untaxed. What follows is the statute with its dates, what expired and what the 2025 Act replaced, the fund market and its fee stack, the evidence on where the money went, and a $500,000 gain worked in both directions. Investing in Luxury Real Estate is the hub’s flagship, and Investing in 1031 Exchanges, Investing in Real Estate Syndications and Private REITs, Investing in Net-Lease Commercial Property, Investing in Short-Term Rentals and Buying a Second Home Abroad cover the neighbouring machinery; this one covers the zone.

An Opportunity Zone is a line on a map, not a product. Congress created two new sections of the Internal Revenue Code in the Tax Cuts and Jobs Act of December 2017 — §1400Z-1, which defines the geography, and §1400Z-2, which defines the tax break — and everything sold to you as an “OZ investment” is a private fund that has volunteered to operate inside those lines under those rules. Three objects have to line up, and a failure at any one of them costs the investor the whole benefit.

The tract

The geography is drawn once and then frozen. In the first round, governors nominated tracts in early 2018 from a pool of 42,176 eligible low-income communities, capped at 25% of each state’s eligible tracts, and Treasury certified them in IRS Notice 2018-48. Treasury’s official count is 8,764 designated tracts. The tally published through Mission Investors Exchange gives 8,762, and is the more useful of the two for seeing how the map was drawn: it splits the total into 8,532 tracts that qualified as low-income communities in their own right and 230 that qualified only because they touched one. The two-tract discrepancy runs through the academic literature and changes nothing. What matters is that the designation was a political act performed once, by fifty governors, on data from the 2011–2015 American Community Survey, and that it then governed a decade of federal subsidy.

The contiguous-tract allowance is how a waterfront marina came to sit in a poverty programme. The GAO, reviewing the round in October 2021, found that the tracts governors chose had on average higher poverty rates and larger non-White populations than the eligible tracts they passed over — a finding about which places were picked, not about where the money later went.

The fund

The wrapper is a Qualified Opportunity Fund, a corporation or partnership that self-certifies on IRS Form 8996 and must hold at least 90% of its assets in qualified Opportunity Zone property, tested twice a year on the average of two testing dates. There is no application and no approval; a fund exists because it filed a form saying it does. Miss the 90% test and the fund owes a monthly penalty equal to the shortfall multiplied by the federal underpayment rate and divided by twelve, unless it can show reasonable cause; a sustained failure can decertify the fund and unwind every investor’s benefit at once.

The business

Below the fund sits either property held directly or, far more commonly, an interest in a Qualified Opportunity Zone Business, which exists because the business-level tests are looser than the fund-level ones. A QOZB must keep at least 70% of its tangible property inside the zone, derive at least 50% of its gross income from the active conduct of the business there, hold no more than 5% of the unadjusted basis of its property in non-qualified financial property, and stay out of the statutory list of excluded trades that §1400Z-2 borrows from §144(c)(6)(B) — private or commercial golf courses, country clubs, massage parlours, hot-tub and suntan facilities, racetracks and other gambling facilities, and stores whose principal business is selling alcohol for consumption off the premises. It also gets the 31-month working-capital safe harbour: with a written plan and a schedule of projected uses, a QOZB may hold cash for 31 months while it builds without failing the financial-property test, which is what makes ground-up development possible at all.

The final regulations allow sequential or overlapping safe harbours out to 62 months where later cash infusions form part of the original plan.

The property itself must satisfy one of two conditions. Either its original use in the zone begins with the fund — new construction, or a building vacant long enough to reset — or the fund substantially improves it, which under the original rules means adding to the building’s basis more than the basis it started with, excluding the land, within 30 months. Doubling your basis in a building inside 30 months is a demanding test, and it is the reason the OZ market became a development market rather than a market in existing rental property.

The incentive is three separate benefits bolted to one investment, and they behave very differently. Understanding which one is doing the work is the difference between a rational decision and a marketing brochure, because two of the three are small and the third is enormous.

The mechanism starts with a realised capital gain. You sell something — anything — and within 180 days of the sale you contribute an amount up to that gain into a Qualified Opportunity Fund and elect deferral under §1400Z-2(a). Only the gain need go in; unlike a 1031 exchange, the return of your original basis is yours to keep and spend, which is the single most useful structural difference between the two regimes. Gains from the sale of a business, of publicly traded stock, of crypto, of art, and of real estate all qualify.

Business property is the one case worth spelling out, because the rule that governs it was rewritten in the taxpayer’s favour and the old version is still quoted. Gain on depreciable property and land used in a trade or business — a warehouse, a fleet, a farm’s equipment — is taxed under §1231. The proposed regulations of April 2019 would have made such a seller wait until the last day of the tax year before the 180-day clock started, and would have let a taxpayer invest only the net §1231 gain, meaning the year’s §1231 gains less its §1231 losses, which cannot be known until the year is over. The final regulations reversed both points. A taxpayer may invest the gross gain from each individual sale, ignoring losses elsewhere in the year, and the 180 days run from the date of that sale exactly as they do for any other gain. A business seller in March no longer waits nine months to act.

Deferral

The first benefit postpones the tax. Under the original 2017 statute the deferred gain was included in income on the earlier of the date you disposed of the fund interest or a hard calendar date, December 31, 2026, no matter when you invested. Deferral was therefore a wasting asset: an investor entering in 2018 got eight and a half years of postponement, one entering in 2025 got barely eighteen months. The benefit is the time value of the money, nothing more. On a $500,000 gain taxed at the top federal long-term rate of 20% plus the 3.8% net investment income tax, the tax is $119,000, and postponing it five years at an 8% opportunity cost is worth about $38,000 in present value.

One protection is built into the deferral and is easy to miss, because the marketing never mentions it and the criticism rarely does either. When the deferred gain comes back into income, §1400Z-2(b)(2)(A) sets the amount included as the excess, over your basis in the investment, of the lesser of the gain you originally deferred or the fair market value of your fund interest on the inclusion date. That lesser-of test is the only downside protection in the statute. If the fund is worth less than the gain you rolled into it when the bill falls due, the bill falls with it. You are postponing tax on a gain you have already made; you are not guaranteeing that gain to the Treasury out of a position that has since lost value. The rule applied to the first round’s fixed December 31, 2026 date and it applies to the rolling five-year inclusion that replaced it.

The step-ups

The second benefit forgives part of the old gain by increasing your basis in the deferred gain itself. The 2017 schedule granted 10% of the deferred gain after a five-year hold and a further 5% after seven years, for 15% in total. Because both step-ups had to be earned before the December 31, 2026 inclusion date, they came with silent deadlines: the 15% required an investment by the end of 2019 and the 10% by the end of 2021. Neither was extended. On a $500,000 gain the full 15% was worth $17,850 of tax; the surviving 10% is worth $11,900. This is the smallest of the three benefits and it is the one the brochures lead with.

The ten-year exclusion

The third benefit is the reason the market exists. Hold the fund interest for at least ten years and you may elect under §1400Z-2(c) to treat your basis in it as equal to its fair market value on the day you sell, which means the entire appreciation between the day you invested and the day you exit is excluded from federal tax. Not deferred — excluded. Because the basis steps up to fair market value, the election also washes out the depreciation the fund took along the way: the §1245 and §1250 recapture that would normally follow a decade of depreciating a building is folded into the stepped-up basis and disappears with it, including the unrecaptured §1250 gain that would otherwise be taxed at a maximum rate of 25% under §1(h). On $500,000 compounding at 8% net for ten years, the excluded gain is $579,462 and the federal tax avoided is $137,912 before any recapture. Everything else in the statute is rounding.

The order of that chart is the whole argument. The parts of the incentive that reward you for having a gain are worth tens of thousands; the part that rewards you for the investment working is worth a multiple of them, and it pays nothing at all if the investment does not work. An Opportunity Zone break converts an ordinary real-estate development into a tax-free one. It does not convert a bad development into a good one, and the record of the first round is largely a record of investors who confused the two.

The original statute was written with an expiry built into every clause, and by 2026 most of it had run out. The dates matter because an investor holding a first-round position and an investor entering the second round face two different laws with the same name, and because the handover between them is governed by a single piece of administrative guidance that decides what a reader holding a 2026 gain can still do.

The step-ups that closed

The 15% step-up required a seven-year hold before December 31, 2026 and therefore closed to new money at the end of 2019. The 10% step-up required five years and closed at the end of 2021. From January 1, 2022 an OZ 1.0 investment carried deferral and the ten-year exclusion and nothing else, and the deferral shrank by a day for every day that passed. That is the whole reason the first round’s last years were poor ones to enter: the buyer was paying a development sponsor’s fee stack for a benefit that had two of its three legs cut off.

The inclusion date, and the ceiling on it

The consequence for anyone still holding a first-round position is a tax bill without a sale. The deferred gain is included in income on the 2026 return, filed in 2027, whether or not the fund has distributed a dollar. The lesser-of ceiling of §1400Z-2(b)(2)(A) applies here in full: the amount included is the excess, over your basis in the investment, of the lesser of the gain you originally deferred or the fair market value of the fund interest on the inclusion date. If the buildings are worth less than they cost, the tax falls with them — which is why a defensible year-end valuation of an illiquid, minority fund interest became a live 2026 planning question rather than a formality, and why a first-round holder who has not had one prepared is leaving money on the table.

None of that solves the cash problem. The fund interest is illiquid by construction, most sponsors do not distribute during a development, secondary markets in QOF interests are thin and a buyer does not inherit your Opportunity Zone benefits, so the investor must find the money elsewhere. That is the characteristic failure of the first round, and it is entirely foreseeable from the statute: an inclusion event with a fixed date and no matching liquidity.

The bridge into 2027

The ability to elect deferral under the old rules ran until the end of 2026, which leaves an obvious question: what becomes of a gain realised late in 2026 whose 180 days run past December 31? Notice 2026-40, issued June 18, 2026, answers it. Treasury and the IRS confirmed that a pre-2027 gain invested in a fund on or after January 1, 2027, within its own 180-day window, is deferred under the new regime instead. The gain is not stranded between the two laws. The practical effect is that a gain realised from roughly July 2026 onward can fall under either law, and on the tax alone the new one is better: a rolling five years of deferral measured from the investment rather than a stub running to December 31, 2026, and a 10% basis step-up that an OZ 1.0 entry at that date could no longer earn at all.

Two conditions sit under that, and both are the reader’s to check rather than the sponsor’s to assert. The first is arithmetic: the 180 days have to reach January 1, 2027, which for a gain realised in the first days of July 2026 is a matter of days, so the sale date decides whether there is a choice at all. The second is that a fund has to be open on the other side of the date, in a tract that is still designated, and willing to take the money then. The tax treatment is not the deal, and a January fund that does not exist is worse than a December one that does. What is worth resisting is the pressure rather than the December close itself: a sponsor who tells a reader with a live 180-day window that the closing date is neutral is wrong, and the difference is five years of deferral and a tenth of the deferred gain.

What outlives the map

The ten-year exclusion outlives all of this. Under the original rules the election to step basis up to fair market value remains available to first-round investors on a sale through December 31, 2047. A 2019 investor may therefore hold until 2047 and still exit tax-free, having paid the deferred tax in 2027. The tracts are the part with a shorter life: first-round designations run to December 31, 2028 (December 31, 2027 for Puerto Rico), not to the end of 2026 as the original ten-year designation period implied. Notice 2026-40 also says Treasury intends to let qualifying funds and businesses go on treating those expired tracts as Opportunity Zones for certain tangible-property, active-business, gross-income and intangibles tests through 2047, so that a fund does not fail a compliance test in 2029 because a line on a map lapsed. That is a stated intention rather than a published regulation, and it belongs on the watch list until it is one.

The One Big Beautiful Bill Act, Public Law 119-21, signed July 4, 2025, did something Congress rarely does with a tax expenditure of this size: it made it permanent, and then it tightened it. The new rules govern investments made on or after January 1, 2027, and they change the shape of the deal in five ways.

A rolling deferral

The fixed December 31, 2026 inclusion date is gone. From 2027 a deferred gain is recognised in the taxable year containing the fifth anniversary of the investment. Every investor gets the same five years regardless of entry date, which removes the wasting-asset problem and, more importantly, removes the year-end stampede that let sponsors raise money from people working to a deadline rather than to a spreadsheet. The lesser-of ceiling travels with the new timetable: at the five-year mark the amount brought into income is still capped by the fair market value of the fund interest on that date, so an investor whose fund has fallen in value pays on the lower figure.

One step-up, and a bigger rural one

The seven-year 5% step-up is repealed. A standard fund carries a 10% basis step-up on the deferred gain at five years. A new category, the Qualified Rural Opportunity Fund, carries 30%. A QROF must hold at least 90% of its assets in rural zone tracts, with rural defined in the statute as an area that is not in, and not immediately adjacent to, a city or town of 50,000 or more inhabitants. Rural projects also get a relaxed rehabilitation test: the substantial-improvement threshold falls from more than 100% of basis to more than 50%, which makes the renovation of an existing building in a small town economic in a way it never was under the first round.

A thirty-year ceiling

The ten-year exclusion is no longer open-ended. Under the amended §1400Z-2(c), for investments made after December 31, 2026 the step-up happens on the earlier of the date of sale or the thirtieth anniversary of the investment; appreciation accruing after that thirty-year mark falls outside the election and is taxable. This replaces the December 31, 2047 outer date that governs first-round investments with a rolling thirty-year window measured from each investor’s own entry. Theoretical on a ten-to-fifteen-year hold; a hard stop for a family planning to hold through a generation.

A narrower map

Eligibility tightened twice. The low-income community threshold falls from 80% of area or statewide median family income to 70% — a tract may still qualify on the alternative test of a poverty rate of at least 20% with median family income no higher than 125% of the state or metropolitan median — and the ability to designate contiguous tracts, the provision that produced the marina, is repealed outright. Both changes cut against the sort of designation that made the first round notorious, and both reduce the number of tracts a governor may choose from; commentators expect the designated universe to shrink by something close to a fifth.

Reporting, at last

The first round was designed with essentially no reporting, which is why the academic literature had to reconstruct what happened from tax returns years after the fact. The 2025 Act adds Code §§6039K and 6039L, requiring funds and investors to report, and §6726, which prices the failure: $500 a day up to $10,000 per return, rising to $50,000 for funds with more than $10 million of gross assets, and $2,500 a day up to $50,000 or $250,000 for intentional disregard. For an investor this is a diligence item — a fund that cannot describe its reporting process has told you something about its back office.

A designation is worth money to whoever owns land inside it, which is why the map is the most consequential and least examined part of the programme. The second round is being drawn on a decennial cycle written into the 2025 Act, and the mechanics are public.

Treasury and the IRS opened the cycle by publishing the eligible universe in Rev. Proc. 2026-14: 25,332 census tracts qualify for nomination as low-income communities under the tightened test, of which 8,334 are comprised entirely of a rural area and so also qualify for the rural benefits. Governors may nominate up to 25% of their state’s eligible tracts, with floors for the smallest states. The statutory nomination window runs 90 days from July 1, 2026, closing September 28, 2026, with a single 30-day extension available on request that would carry a state to October 28; Treasury then certifies, with official designations expected by around November 28, 2026, to take effect January 1, 2027. As of this guide’s September 10, 2026 as-of date the window had not moved and only a handful of jurisdictions had filed. Thereafter the exercise repeats every ten years, so the 2027 map governs through December 31, 2036.

The old map does not vanish when the new one arrives. Second-round designations take effect January 1, 2027, while the first-round tracts, under Notice 2026-40, remain designated until December 31, 2028, so for two years both are live. A 2027 or 2028 investment can be made into either map, and Treasury’s intended safe harbour is meant to keep a fund that bought into a first-round tract compliant after that tract lapses. What it cannot do is find a buyer in 2035 for a building in a tract nobody chose to designate twice.

Two things follow for an investor. The first is that eligibility is not a permanent attribute of a place: a tract designated in 2018 may not be designated in 2027, and a fund raising money in 2026 for a 2027 deployment is betting on a map that does not yet exist. The second is that nomination is a lobbying process. Governors’ offices take submissions from mayors, developers and economic-development agencies, and the tracts nominated tend to be the ones with someone advocating for them.

Because the whole benefit hangs on the tract, verify the designation yourself against Treasury’s published list before you wire money, and verify it against the specific parcel rather than the neighbourhood. Census tract boundaries do not follow the way people describe places, and “in the Opportunity Zone” is a phrase that appears in offering materials for buildings across the street from one.

The market that grew around the incentive is larger than most alternative-asset niches and considerably less transparent, because self-certification means nobody was required to tell anyone anything until 2025. Two sources measure it and they measure different things.

The complete count comes from tax returns. Treasury’s Office of Tax Analysis, in Working Paper 128 published in June 2026, reported that Qualified Opportunity Funds held $116 billion of total assets including $112 billion of qualified Opportunity Zone property through tax year 2024, that roughly 12,800 funds and about 41,000 taxpayers had used the incentive, that around 85% of those taxpayers were individuals, and that roughly $75 billion of deferred capital gains sat inside the structure. Qualified property grew from about $4 billion across some 1,300 funds in 2018 to $112 billion across 12,800 funds in 2024. This is the authoritative number and it is two years stale by construction, because it comes from filed returns.

The visible count comes from Novogradac, the accounting firm that has maintained a public survey since 2019. Its tracked universe — 2,203 funds, of which 1,758 report a figure — had raised $43.88 billion in equity by the middle of 2026. The gap between $43.88 billion of tracked equity and $112 billion of tax-return property is the answer to a question worth asking: most of this market is not funds sold to outside investors at all. It is captive vehicles, single-project funds and family structures set up by the people who already owned the deal, which Novogradac does not survey and which no broker will ever show you.

Concentration

Both datasets say the money is concentrated. Novogradac’s “super QOFs” are those raising $100 million or more, and at the end of 2022 there were 68 of them among the 1,274 funds that reported an equity figure — 5.3% of the reporting funds, holding 61% of the tracked equity. Three years of new entrants barely moved it. At the end of 2025 the 82 funds at or above $100 million still accounted for 59.3% of all reported equity, and the four above $1 billion for 13.2% of it on their own. The largest raisers Novogradac identifies are Bridge Investment Group at roughly $3.7 billion, CIM Group at roughly $2.3 billion and Griffin Capital at roughly $1.63 billion — sponsor totals for which no as-of date is published, so treat them as orders of magnitude. Below them sit several dozen sponsors at $50 million to $500 million, and below those a long tail of single-asset funds whose economics are the sponsor’s economics.

What it builds

The asset mix is not diversified in any meaningful sense. In Novogradac’s tracked investment, purely residential projects accounted for $14.27 billion, about 45.9%, and projects including a residential component accounted for $24.02 billion, about 77.3%; within residential, 91.2% was multifamily. As of July 30, 2025 the firm could identify 203,560 homes in 238 cities financed by the funds it tracks. The corollary is the criticism, and it is worth getting the provenance right, because the figure is usually quoted as though Treasury had published it. It comes from Novogradac’s tracking, reported by the Center for American Progress, which puts the share of Opportunity Fund equity that went into operating businesses at less than 2%; the Tax Policy Center, citing an Opportunity Zone accounting firm, puts it under 3%. Either measure says the same thing. The statute was drafted to fund businesses in poor places. The market it produced funds apartment buildings.

One series shows how completely this asset class is a creature of its deadlines. Novogradac-tracked funds raised $850.8 million in the first quarter of 2026 and $264.0 million in the second, the second-lowest quarterly figure since the survey began, for a first-half total of $1.11 billion. The cause was not real-estate fundamentals: the December 31, 2026 deferral deadline had made a 2026 entry nearly worthless while the 2027 rules were not yet usable, so capital waited. An asset class whose fundraising can halve on a calendar technicality is one in which the sponsor’s need to close is frequently more urgent than yours.

The tax break is gross; the return is net. Between the two sits a fee stack that is, in the third-party fund market, heavier than in almost any other real-estate wrapper, and a set of frictions that no brochure quantifies. The reason is structural: a sponsor who can tell an investor “the exit is tax-free” faces less price resistance than one who cannot.

The fee stack

Published terms across the third-party QOF market cluster in these ranges, which we state as ranges rather than a single schedule because the funds are private placements with no standard form. An acquisition fee of 1% to 2% of purchase price is charged when the fund buys. An annual asset-management fee of 1% to 2% of committed or invested capital runs for the life of the fund, and independent surveys put all-in annual fees for third-party OZ funds at 1.5% to 3%. A promote, or carried interest, most commonly 20% of profits, takes effect above a preferred return generally set at 8% to 10%; one industry estimate puts the sponsor’s share of total appreciation in third-party funds at 15% to 35%. Development fees, disposition fees, financing fees and property-management fees appear on top, each defensible on its own and cumulatively material.

Run those through a ten-year hold and the drag is legible. Take a fund charging a 1% acquisition fee, a 1.5% annual asset-management fee and a 20% promote over an 8% preferred return, and ask what gross asset-level return produces what net return to the investor.

The shape of that drag is worth reading carefully. It is roughly 1.6 percentage points a year at every outcome below the preferred return and widens above it as the promote engages, so the sponsor’s share rises with success — which is the point of a promote — but the investor pays a flat 1.6 points for failure as well. Over ten years, 1.6 points a year on $500,000 is about $150,000 of terminal value.

The frictions the fee schedule omits

Three further costs sit outside the fee table. The first is illiquidity of an unusual kind: the ten-year hold is a statutory condition rather than a soft target, and selling in year eight forfeits the exclusion that was the entire reason to invest. There is no meaningful secondary market in QOF interests, and where transfers are permitted the buyer does not inherit your holding period. The second is the year-five tax with no matching cash, a permanent feature of the rolling deferral from 2027 onward: the deferred gain is taxed in the fifth year, and a development fund in its fifth year is generally not distributing. The lesser-of ceiling caps that bill at the fund interest’s fair market value, so the worst case is smaller than the headline gain, but it is a cap and not a payment plan. Reserve the cash separately at the outset.

The third friction is state tax, which does not follow federal treatment everywhere. California is the most-cited non-conforming state, and New York, New Jersey and Massachusetts are commonly listed alongside it; a resident of a non-conforming state pays state tax on the gain on the ordinary schedule and again on the appreciation at exit. Conformity is stated here from practitioner summaries rather than a state-by-state review; check your own.

The academic literature that has accumulated since 2018 gives a reasonably clear answer to the question of whether Opportunity Zones moved capital, and a much murkier one to the question of whether the capital did any good. The first answer is yes, at a scale that surprised the sceptics. The second is that the effects on the people living in the zones are, so far, difficult to detect.

Where the capital went

Kennedy and Wheeler, working with the first tranche of tax data and published through Berkeley’s Opportunity Lab, found the flow was real and extraordinarily concentrated: 84% of the 8,764 tracts in their sample appear to have received no investment at all, the top 1% of tracts took 42% of the capital, and the top 5% took 78%. They also found that investment went disproportionately to tracts with higher incomes, higher home values, higher educational attainment and pre-existing income and population growth. Treasury’s later and more complete data revises the reach upward considerably — Coyne and Johnson report investment reaching 77% of designated tracts through 2024 — while confirming the direction of the distributional finding: investment was “more heavily concentrated in less-distressed areas and areas already trending toward economic growth.” Both findings can be true. Kennedy and Wheeler measured the early years with partial data; Treasury measured six years with complete returns; the money spread out over time but it started, and largely stayed, in the better tracts.

What changed for residents

Chen, Glaeser and Wessel tested the most direct market signal, house prices, and found essentially nothing: their preferred estimates rule out price effects greater than 0.5 percentage points at 95% confidence, which is to say that buyers did not believe the designation would transform the neighbourhood. Freedman, Khanna and Neumark, in the Journal of Urban Economics in 2023, examined outcomes for the residents themselves using restricted-access American Community Survey microdata for 2013 to 2019, and found little or no evidence of positive effects on the employment, earnings or poverty of zone residents — with the caveat that their data end in 2019, before the regulations were even finalised that December. That caveat is doing real work: the null results describe a programme that had barely started.

The strongest affirmative evidence is about buildings rather than people. The Economic Innovation Group, in a February 2025 working paper by Benjamin Glasner, Adam Ozimek and John Lettieri, applied difference-in-differences methods to HUD’s aggregated USPS data on residential addresses and attributed more than 416,000 new residential addresses between 2019 and the first quarter of 2025 to the incentive, along with a 70% increase in new residential construction inside designated tracts; OZ-linked units account for 48% of new housing in those tracts, 16% across all low-income communities and 4% of total US supply. Lettieri is EIG’s chief executive, and EIG designed the policy and campaigned for its renewal, so this is advocacy research; it is also the only large-scale attempt to measure the housing counterfactual, and its data sources are public.

What it cost

On the other side of the ledger, the Joint Committee on Taxation estimated that extending the expiring Opportunity Zone provision through 2034 would cost about $70 billion, a figure reported by the Tax Foundation in its analysis of the 2025 tax debate. Set that against 416,000 units on the most generous available estimate and the implied subsidy is on the order of $170,000 a unit — a rough order of magnitude rather than a finding, since the two estimates cover different periods and neither was built for this comparison.

The number nobody publishes

There is no audited, survivorship-corrected return series for Qualified Opportunity Funds, and after seven years there should be. Funds are private placements that self-certify; they are not required to report performance, most have not reached a ten-year exit, and the ones that have are not obliged to say what happened. Sponsor marketing quotes target IRRs; where a realised number does appear on a sponsor’s page it is typically a gross, self-computed figure across a self-selected set of deals, not an audited net-of-fee series. Any figure you are shown for “OZ fund returns” should be labelled that way in your own notes. This is the single largest gap in the evidence on this asset class and it will not close before the early 2030s, when the first-round funds complete their ten-year holds — 2028 is the first year in which the earliest investments can be sold into the ten-year exclusion at all.

Almost every dollar of Opportunity Zone equity went into American housing, so the honest benchmark for the asset class is not a venture return or an index of distressed neighbourhoods but the national residential market over the same years. Invest Alternative maintains a housing sub-index inside its composite, and two of its underlying series bracket the period well.

Our longest housing series is the Zillow top-tier US home value measure, stored monthly back to January 2000, which stood at 715,426 on July 31, 2026. Read against December 2017, the month the Tax Cuts and Jobs Act was signed, when it stood at 465,947, the series is up 53.5%, or about 5.1% a year over eight and a half years. That is the passive alternative an Opportunity Zone investor was implicitly betting against: build nothing, take no development risk, own American housing, compound in the low single digits with no tax break at all. A fund charging 1.6 points of annual drag has to beat that by its fee load and its risk premium before the exclusion adds a cent. The hub’s flagship guide, Investing in Luxury Real Estate, reads the same top-tier series from the owner-occupier’s side.

The tape since the end of 2024 is flatter, and more instructive about the exit first-round funds face. Our stored Zillow top-tier level was 714,222 at the end of December 2024 and 715,426 in July 2026 — a rise of 0.17% across nineteen months. Our shorter, higher-frequency series, price per square foot from Parcl Labs, ran from 463.57 on June 30, 2026 to 453.12 on September 7, 2026, a fall of 2.25% across the 68 observations we hold; the Invest Alternative Housing sub-index, built on it and weighted at 12% of our composite, stood at 97.746 on September 8, 2026. Its trailing-year field also reads −2.25%, because the underlying series only begins on June 30, 2026 — the sub-index has no twelve-month history yet, and that number is the same ten-week move, not a year of it.

These are our own series, generated by our collection engine on September 8, 2026, describing the national housing tape and not the Opportunity Zone tracts specifically; no public index measures OZ tracts as an asset. What they say is that funds which raised money in 2019 and 2020 on rising-rent pro formas will be exiting into a market that has gone sideways for two years, and that the tax exclusion applies to whatever appreciation there turns out to be.

A gain looking for a home has three plausible destinations, and the choice between them turns on what you own now, what you want to own next, and how long you are willing to be locked up. The differences are structural rather than a matter of degree.

Against paying the tax

Paying is the baseline and it is not a bad one. You surrender 23.8% at the federal top rate plus state, keep complete liquidity, and can put the remainder into anything, including a listed REIT or an index fund, with daily pricing and no ten-year condition. The Opportunity Zone route wins only if the fund’s net return, after its fee stack and after the year-five tax on the deferred gain, exceeds what the after-tax money would have compounded at. The next section computes exactly where that line sits.

Against a 1031 exchange

Section 1031 defers real-estate gain indefinitely and, held to death, historically eliminated it through the basis step-up; the Opportunity Zone route defers for five years, then taxes, then exempts the second gain. Four differences decide it. A 1031 exchange requires you to reinvest the entire sale proceeds to defer the whole gain, while an OZ investment requires only the gain, leaving your basis free. A 1031 exchange works only for real property traded for real property; an OZ investment accepts a gain from any asset. A 1031 defers forever and an OZ deferral ends at five years. And a 1031 exchange preserves nothing of the appreciation from tax if you eventually sell — the whole chain unwinds — while the OZ exclusion ends the chain permanently at ten years. The hub’s guide to 1031 exchanges covers that regime’s own machinery, its qualified intermediaries and the Delaware Statutory Trust industry that sells access at loads which can exceed 15%.

Against a syndication or a non-traded REIT

The economics of an OZ fund are the economics of a real-estate syndication with a tax wrapper attached, and the hub’s guide to syndications and private REITs describes those fee and governance patterns. The wrapper adds one constraint a syndication does not have: an early exit destroys the reason you invested. A syndication investor unhappy in year four can push for a sale; an OZ investor unhappy in year four is a captive.

The estate question

A transfer at death is not an inclusion event, so the fund interest passes to heirs, but the deferred gain travels with it as income in respect of a decedent under §691 and gets no step-up under §1014 — the reverse of the 1031 chain’s traditional appeal. The heir takes the decedent’s basis and holding period, owes the deferred tax when the inclusion date arrives, and may still make the ten-year fair-market-value election on a later sale. Anyone planning around this needs an estate lawyer, not a guide.

The arithmetic below is the whole decision in one page. It uses a $500,000 long-term capital gain realised in 2027 by a top-bracket taxpayer, a 23.8% federal rate, an eight-percent alternative portfolio, a ten-year horizon and the representative fee schedule from the costs section. Every assumption is ours and stated; change any of them and the answer moves.

Path A: pay the tax

The tax on a $500,000 gain at 20% plus the 3.8% net investment income tax is $119,000, leaving $381,000 to invest. Compounded at 8% for ten years that becomes $822,550. Selling in year ten realises a gain of $441,550 and a further $105,089 of tax, leaving $717,461.

Path B: roll it into a fund

The full $500,000 goes into a Qualified Opportunity Fund in 2027. Under the OZ 2.0 rolling deferral the gain is recognised in the taxable year containing the fifth anniversary, and the 10% basis step-up means only $450,000 is taxable: $107,100, payable from outside cash. That is $11,900 less than paying immediately, and five years later. To compare terminal wealth honestly, carry that payment forward at the same 8% from the five-year mark: it costs $157,365 of year-ten wealth. The model assumes the fund interest is worth at least the deferred gain at that point, which on these fee and growth assumptions it is at every gross return shown below. If a fund fell far enough that it was not, the §1400Z-2(b)(2)(A) lesser-of rule would cut the year-five tax to that lower value. It is the statute’s only piece of downside protection, and it pays out precisely in the outcomes where everything else has gone wrong.

Now the fund. At a 12% gross asset-level return — a normal development pro forma, not a conservative one — the representative fee schedule delivers $1,343,470 before the promote, of which $52,802 goes to the sponsor above the 8% preferred, leaving the investor $1,290,669. The ten-year election makes all of that appreciation free of federal tax, including the depreciation that would otherwise be recaptured. Subtract the deferred tax carried forward and the investor holds $1,133,304, against $717,461 for paying the tax: an advantage of $415,842, or 58%.

What the tax break is actually worth

That $415,842 conflates two different things: a good development and a tax break. Separate them by running the identical 12% deal, with the identical fees, outside a zone and after tax. The investor starts with $381,000 instead of $500,000, pays the same fees, and pays 23.8% on the exit gain, finishing with $840,097. The Opportunity Zone structure is therefore worth $293,207 on this deal and the rest is the deal itself.

Where it breaks

Run the gross return down and the advantage collapses faster than intuition suggests, because the fee drag and the year-five tax are fixed while the exclusion scales with the outcome. At a 10% gross return the investor finishes with $953,877, an advantage of $236,416. At 8% gross — an entirely plausible outcome for a development completed into a flat rental market — the finish is $771,818 and the advantage is $54,357, less than five points of the original gain. The break-even is a net investor return of about 5.75% a year, which given the fee stack requires a gross asset-level return of about 7.4%. Below that the tax-free exit is worth less than paying the tax in 2027 and owning an index fund.

The failures in this asset class are not mostly market failures. They are structural, and each of them can take the tax benefit away while leaving you holding the real estate, which is the worst of both outcomes.

The fund fails its own test

A Qualified Opportunity Fund whose two testing dates average below 90% qualified assets pays a penalty on the shortfall unless it shows reasonable cause, and a sustained or wilful failure can decertify it. Decertification is not a fund-level inconvenience; it removes the deferral and the exclusion for every investor at once. The same applies below the fund, where a QOZB that drifts outside the 70% tangible-property or 50% gross-income tests taints the fund’s holding. You cannot audit this yourself; you can ask for the fund’s testing history, its Form 8996 filings and the identity of the accountant who signs off, and you can decline funds that treat the question as a formality.

The clock beats the building

The exclusion needs ten years of holding, and a development that runs long, refinances into trouble or is sold by a lender in year seven forfeits it. Construction schedules slip; the exclusion does not. Ask what happens to your ten-year clock in the sponsor’s downside cases — a workout, a recapitalisation, a partial asset sale — and get the answer in the operating agreement rather than on a call.

The year-five tax arrives without cash

Under the rolling deferral the tax on the original gain falls due in the fifth year while a development fund is typically still building or leasing up. Investors in the first round met the same event on December 31, 2026 as a fixed date and many were unprepared. The lesser-of ceiling limits the size of the bill to the fund interest’s fair market value, but only a fund that has lost money gets that relief, and it arrives as a smaller bill rather than as cash. Reserve the money at the outset, outside the fund, in something liquid, and treat any sponsor projection that assumes a distribution will cover it as a projection.

The sponsor is the whole investment

There is no exchange, no daily mark, no redemption and, in most funds, no independent valuation. What you have is a private-placement memorandum, a sponsor and a decade. The long tail of single-asset funds is where a first-time sponsor with one deal and a tax pitch operates, and a deal that could not be financed on its merits is exactly the deal that gets wrapped in an incentive and sold to someone chasing a deadline.

The map moves under you

Designations last ten years, and the second-round map drawn in 2026 for 2027 onward is not the first-round map; the 2018 tracts expire on December 31, 2028. A fund raising in one cycle to deploy in the next carries designation risk that no amount of real-estate diligence addresses. Notice 2026-40 softens the compliance edge of this, since Treasury intends to let existing funds keep treating expired first-round tracts as zones for certain tests through 2047. It does not change the exit market for a building in a tract nobody wanted to designate twice, and until the intention is written into a regulation it is an intention.

The state does not follow

Federal exclusion is not state exclusion. In non-conforming states the deferred gain is taxed on the ordinary schedule and the appreciation is taxed at exit, so an investor in such a state is paying full state tax on both legs while accepting a ten-year federal lock-up. California is the significant case; New York, New Jersey and Massachusetts are commonly listed with it. Check conformity before you model anything.

Political risk survives permanence

A tax expenditure that has produced this much adverse journalism — the marina, the designated tract containing a well-connected investor’s land, tax-free capital in casinos, crypto data centres and luxury apartments — carries political risk. Permanence in the 2025 Act reduces it substantially without eliminating it, and the new §6039K and §6039L reporting will produce, for the first time, a public record of exactly who received what.

The questions below are the ones that separate a real Opportunity Zone investment from a tax pitch, and they are asked in this order because each one can end the conversation.

1. Verify the tract, not the neighbourhood. Get the parcel’s census tract number and check it against Treasury’s published designation list for the round that governs your investment. Designation is binary and it is the whole benefit.
2. Read the fee table against a ten-year hold, in dollars. Acquisition, annual asset management, development, financing, disposition, property management and the promote, totalled on your cheque over ten years and compared with the 1.6-point annual drag in the costs section.
3. Ask what the sponsor has finished. Not what it has raised — what it has built, leased, sold and returned, with dates and realised net returns. Most sponsors here have nothing to show, and a completed record outside the OZ world beats three OZ deals under construction.
4. Underwrite the deal as if the incentive did not exist. Rents, absorption, construction cost, debt terms, exit cap rate. If it does not stand up at a 12% gross return without the tax break, the tax break is covering a hole.
5. Get the compliance apparatus in writing. Who tests the 90%, on what dates, who audits it, who prepares the §6039K reporting, and what happens if a test fails.
6. Model the fifth year in cash. Your deferred gain, your rate, your state, the date the tax is due, and where the money comes from.
7. Read the exit provisions. How and when the fund sells, whether the sponsor may extend, whether an investor may transfer, and what a transfer does to the holding period. The ten-year condition makes these terms matter more here than in any other private real-estate wrapper.
8. Price the alternative honestly. Your own break-even, computed as the worked example computes it, on your rate and your realistic alternative return rather than the sponsor’s.

The sequence below assumes you have, or expect, a realised capital gain and are deciding what to do with it. It is written for the second-round rules that govern investments from January 1, 2027.

1. Establish the gain and the clock. The exact amount, the character and the date of sale, because the 180-day window runs from it. A gain whose window crosses into a new calendar year is also a choice of regime: under Notice 2026-40, a pre-2027 gain invested on or after January 1, 2027 is deferred under the new rules, so the investment date decides which law you get. Nothing else can be decided until this is fixed.
2. Compute your own break-even before you take a meeting. Your marginal federal and state rate, your realistic alternative return, and the minimum gross asset-level return a fund must therefore produce. The worked example’s 7.4% is ours; yours will differ.
3. Decide whether you want the real estate at all. An OZ investment is a ten-year illiquid development position in American multifamily housing. If that is not an exposure you want, the tax break is not a reason to acquire it.
4. Choose a lane: standard or rural. A Qualified Rural Opportunity Fund carries a 30% step-up rather than 10% and a 50% rather than 100% substantial-improvement test, a real advantage in renovation deals; it also means a thinner exit market. Require a wider expected return for it, not a narrower one.
5. Shortlist on sponsor, then on deal. Three or four sponsors with completed projects, then the specific asset. Ignore the fund that arrives with a deadline attached.
6. Run the diligence list, in the order above.
7. Reserve the year-five tax separately, in cash or short Treasuries, sized to your rate on 90% of the deferred gain, on the day you invest. That is the ceiling: if the fund is worth less than the deferred gain when the inclusion date arrives, the lesser-of rule reduces the bill, and you will have over-reserved rather than under-reserved.
8. Diversify by tract and by sponsor if the cheque allows. Two funds in different metros under different sponsors is materially safer than one.
9. Write the exit date down. The month your ten years is complete, the election you have to make, and the accountant who will make it.

Each reading below is given at the level it stood on the date shown, so a reader in a later year can see at a glance what has moved and what it would mean.

The second-round map

Rev. Proc. 2026-14 published 25,332 eligible tracts, 8,334 of them entirely rural, with governors’ nominations due September 28, 2026 (October 28 with the single available extension) and certification expected around November 28, 2026 for a January 1, 2027 effective date. Watch how heavily states use the rural allowance and how many tracts in the top income quartile of the eligible pool are nominated: a repeat of the first round’s pattern is the strongest predictor of a repeat of its distributional results.

The transition rules in writing

Notice 2026-40 of June 18, 2026 states an intention rather than a rule: that funds and businesses will be allowed to treat expired first-round tracts as Opportunity Zones for certain tangible-property, active-business, gross-income and intangibles tests through December 31, 2047. Watch for proposed or final regulations carrying that safe harbour. Until they appear, a fund holding first-round tract property past December 31, 2028 is relying on published intent, and any sponsor that describes the point as settled is overstating it.

Fundraising after the reset

Novogradac-tracked funds raised $850.8 million in the first quarter of 2026 and $264.0 million in the second, the second-lowest quarter on record, on a cumulative $43.88 billion. A first full year under OZ 2.0 that does not clear roughly $3 billion of tracked equity would say the permanent programme has a smaller natural constituency than its first-round peak, which matters because fund size drives fee levels and sponsor quality.

The first realised exits

The first-round cohort reaches its ten-year mark from 2028 onward, and the first credible fund-level realised net returns should appear then. Watch whether sponsors publish them. A market that reaches 2030 with no public realised-return series has told you something about the results.

Treasury’s next update

Coyne and Johnson’s Working Paper 128 (June 2026) put qualified OZ property at $112 billion across roughly 12,800 funds through 2024. What it does not measure is the operating-business share; that figure comes from Novogradac’s tracking, reported by the Center for American Progress at under 2%, with the Tax Policy Center at under 3%. Treasury’s next update is the cleanest available measure of whether OZ 2.0 changed the mix; a business share still under 5% would confirm that the incentive is a housing subsidy whatever the statute says.

The reporting regime

Sections 6039K and 6039L take effect with the new programme, backed by the §6726 penalty. The first published aggregate from that data will be the first time anyone can see, contemporaneously rather than six years late, where the money went. Watch for the first Treasury or IRS release drawn from it.

The residential market itself

Our stored Zillow top-tier US level was 715,426 in July 2026 against 714,222 in December 2024, and our Parcl Labs price-per-square-foot series fell 2.25% between June 30 and September 7, 2026. Because roughly three-quarters of identified fund investment includes housing, a national residential market that stays flat is a market in which the ten-year exclusion applies to very little appreciation, and the whole calculation in the worked example shifts toward paying the tax.