Twelve Cannabis Companies, One Tax Code: The 280E Scorecard
🎙 Listen — Q&A podcast
A cross-company analysis of the largest pure-play U.S. multi-state operators — FY2025. The nine coverage names were analyzed by Claude (Anthropic's Claude Opus) from SEC XBRL filings through the RoboSystems knowledge graph and validated one company at a time; three more (Curaleaf, Cresco, Glass House) are added from their FY2025 public releases, flagged † and pending independent validation.
Coverage — $TRLV · $GTBIF · $VRNOF · $AAWH · $VREOF · $JUSHF · $TSNDF · $MRMD · $PLNH. Comparison set — $CURLF · $CRLBF · $GLASF. Sector benchmark — $MSOS.
Twelve companies. $6.8 billion in revenue. $3.4 billion in gross profit — a blended margin near 50%, the profile of a healthy consumer-products industry.
A combined pre-tax loss of roughly $260 million.
And a $786 million income-tax bill.
Eleven of the twelve paid income tax while posting a net loss. The lone company that turned a GAAP profit — Green Thumb Industries — paid $147 million in tax, more than its entire $138 million of operating income. This is not a story about twelve badly run companies. Most of them generate positive operating cash flow. It is a story about one line of the U.S. tax code — Section 280E — and what it does to a margin-healthy industry that the federal government still classifies alongside heroin.
For the nine RoboSystems coverage names, every figure below was pulled by Claude Opus from each company's SEC XBRL filings through the knowledge graph, then re-validated statement by statement against each filer's 10-K. Curaleaf, Cresco, and Glass House (marked †) are not yet in the graph for FY2025; their figures come from each company's FY2025 earnings release and are presented as a comparison set pending independent validation. Everything else is what cross-company financial analysis looks like when the filings are a queryable database instead of a stack of PDFs.
Over What 280E does. Section 280E bars any business "trafficking" a Schedule I substance from deducting ordinary operating expenses — rent, payroll, marketing, interest. Only cost of goods sold is deductible. The result: cannabis operators are taxed on gross profit, not net income, which produces large tax bills on top of pre-tax losses. A normal company pays ~21% of pre-tax income. As the scorecard shows, these twelve paid 14%–31% of gross profit — whether or not they made money.
The methodology, briefly
- Graph-sourced and re-validated — the nine coverage names (SEC XBRL, FY2025, period ending 12/31/2025): revenue, gross profit, operating income, income-tax expense, net income, cash, total assets, goodwill, stockholders' equity, unrecognized tax benefits, and the cash-flow statement — operating cash flow and cash income taxes actually paid. Pulled by Claude Opus from each filer's 10-K via the RoboSystems graph, then checked company-by-company (income statement, balance sheet, cash flow). The most-recent disputed-tax reserve also reflects Q1 FY2026 (period ending 3/31/2026) where the 10-Q tags it.
- Web-sourced, pending validation — the three added names (†): Curaleaf, Cresco, and Glass House file as 40-F foreign private issuers and are not loaded in the graph for FY2025. Their figures come from each company's FY2025 earnings release (Curaleaf and Cresco SEC EX-99.7; Glass House 3/24/2026 release); the disputed-tax-reserve estimates for Curaleaf (
$439M) and Cresco ($172M) are compiled from public reporting. Treat these as a comparison set to be independently validated. - Filing/management-reported: adjusted EBITDA (a non-GAAP measure), net debt including leases, debt-maturity schedules, and the medical-vs-adult-use revenue mix. Enterprise values are from the SSC comparables set (priced 5/29/2026).
- Scenario ranges reflect a re-rating toward consumer-staples multiples (CPG ~10–14×, Altria ~8–9×, Constellation ~12×) if 280E is lifted. They are not price targets and not advice.
Table A — The operators: this is a real industry
| Company (Ticker) | Revenue | Gross margin | Adj. EBITDA | Adj. EBITDA margin |
|---|---|---|---|---|
| Trulieve (TRLV) | $1,181M | 60.2% | $427M | 36% |
| Curaleaf (CURLF) † | $1,268M | 50.0% | $275M | 21.7% |
| Green Thumb (GTBIF) | $1,175M | 48.9% | $348M | 30% |
| Verano (VRNOF) | $822M | 50.3% | $229M | 28% |
| Cresco (CRLBF) † | $656M | 49.5% | $157M | 24.0% |
| Ascend Wellness (AAWH) | $501M | 33.9% | $117M | 23% |
| Vireo Growth (VREOF) | $269M | 47.3% | ~$33M Q1¹ | 31% |
| Jushi (JUSHF) | $263M | 43.4% | $50M | 19% |
| TerrAscend (TSNDF) | $261M | 52.3% | $68M | 26% |
| Glass House (GLASF) † | $182M | 42.0% | $17M | 9% |
| MariMed (MRMD) | $160M | 36.2% | $17M | 11% |
| Planet 13 (PLNH) | $103M | 38.6% | ~breakeven² | — |
| TOTAL / blended | $6,840M | 49.4% | — | — |
† Web-sourced from the company's FY2025 earnings release; pending independent validation (not in the SEC XBRL graph for FY2025). ¹ Vireo is mid-transformation (a roll-up); FY2025 trailing figures understate it — Q1 FY2026 adj. EBITDA was $32.7M at 30.8%, with the run-rate higher pro-forma for the pending Fluent acquisition. ² Planet 13 posts a modest operating loss and does not tag a clean consolidated operating-income line in its XBRL — confirmed in the graph.
Eight of twelve run gross margins between 42% and 60%. Ten of twelve are adjusted-EBITDA positive, most at double-digit margins. By every operating measure above the tax line, this is a normal — in places, an excellent — consumer industry.
Table B — The 280E squeeze: taxed on gross profit, not income
| Company | Income-tax expense | Tax as % of gross profit | Net income | Disputed tax reserve³ |
|---|---|---|---|---|
| Trulieve | $208M | 29% | −$116M | $668M |
| Curaleaf † | $124M | 20% | −$228M | ~$439M |
| Green Thumb | $147M | 26% | +$114M | $203M |
| Verano | $92M | 22% | −$258M | $474M |
| Cresco † | $45M | 14% | −$140M | ~$172M |
| Ascend Wellness | $51M | 30% | −$118M | $204M |
| Vireo Growth | $28M | 22% | −$68M | $120M |
| Jushi | $35M | 31% | −$69M | $177M |
| TerrAscend | $29M | 22% | −$81M | $177M |
| Glass House † | $12M | 15% | −$29M | ~none disclosed |
| MariMed | $4M | 6% | −$14M | $27M⁴ |
| Planet 13 | $12M | 29% | −$64M | $56M |
| TOTAL | $786M | ~25% | −$1,071M | ~$2,695M |
† Web-sourced, pending validation. ³ Disputed tax reserve = unrecognized tax benefits (UTB): disputed taxes the operators have booked as reserves while contesting that 280E applies — taxes claimed but not paid, pending litigation. For the nine, graph-sourced from each 10-K (12/31/2025); Curaleaf/Cresco from public reporting; Glass House discloses no UTB and shows income-tax-payable of $0. ⁴ MariMed books most of its 280E exposure as a $27M accrued income-tax payable (current liability) rather than as a UTB — the same overhang under a different line.
This is the heart of it. A normal corporation pays roughly 21% of pre-tax income. These twelve paid 25% of gross profit — and the group had no pre-tax income to speak of. The $786M tax bill landed on a combined pre-tax loss of ~$260M.
Three things the data surfaces that the headlines miss:
- They booked $786 million in tax — and the nine we can see in cash paid only about $127 million of it. The cash-flow statements for the nine graph-validated names show what each actually wrote to the government: Trulieve paid $1.5 million in cash against a $208 million expense; Planet 13 paid essentially nothing; Vireo paid $1 million against a $28 million expense (its current provision was ~$42 million). Across those nine, cash income taxes paid totaled roughly $127 million — barely a fifth of their $606 million expense. The rest was accrued, not paid — most of it piling onto the disputed-tax reserves below. (The three added names don't break out cash taxes in their releases.) There are two reasons cash lags expense: the leveraged names run a contested-position gambit (deduct on the return, reserve the disputed tax, dare the IRS to litigate), while Green Thumb's gap is plain deferred-tax timing ($30 million cash; $144 million non-cash). The cleanest cash payer is Verano — $78 million.
- The reserves are enormous and nearly universal. The twelve carry ~$2.7 billion of disputed taxes — booked but contested rather than paid. For the smaller names, the reserve exceeds the entire enterprise value (see Table C-2). Glass House is the telling exception: a lower-margin biomass/wholesale model leaves it with a far smaller 280E footprint and no disclosed reserve at all.
- Green Thumb is not the exception people think. It carries a $203M reserve, in line with Ascend and Curaleaf. The thing that sets Green Thumb apart isn't restraint on the tax fight — it's that it is profitable enough to absorb the tax at all. Everyone is contesting 280E. Only Green Thumb is winning the operating business underneath it.
Table C-1 — Survival: who can wait out the catalyst
The catalyst (below) is a when, not an if for some and a maybe for others. Survival is about who can carry the tax and the debt long enough to get there. Nine graph-sourced (12/31/2025); three † from FY2025 releases.
| Company | Cash | Stockholders' equity | Net debt (incl. leases)⁵ | Debt-maturity pressure |
|---|---|---|---|---|
| Green Thumb | $274M | $1,911M | net cash | None — $200M facility undrawn |
| Curaleaf † | $102M | $756M | ~$447M | $500M 11.5% notes, 2029 |
| Trulieve | $256M | $1,143M | ~net cash | 10.5% notes, 2030 |
| Verano | $83M | $704M | ~$335M | Refinanced to 2029–30 |
| Cresco † | $91M | $253M | ~$240M | $325M term, 12.5%, 2030 |
| TerrAscend | $37M | $99M | ~$200M | ~$220M balloon, 2028 |
| Vireo Growth | $138M | $307M | ~$260M | $117M due 2028 |
| Glass House † | $23M | $175M⁶ | ~$46M | Secured facility, 2030 |
| MariMed | $9M | $50M | ~$80M | None material (~75% post-2030) |
| Planet 13 | $5M | $51M | minimal funded debt | Credit line roll, ~mid-2026 |
| Ascend Wellness | $86M | −$47M | ~$520M | $300M @ 12.75%, 2029 |
| Jushi | $24M | −$115M | ~$260M | $160M term @ 12.5%, 2029 |
⁵ Net debt is management/filing-reported and includes finance and operating leases, so it exceeds the graph's tagged long-term-debt line for several names. ⁶ Glass House equity includes ~$92.5M of mezzanine preferred; common-type equity is ~$82M.
Ascend and Jushi carry negative book equity — technically insolvent on paper, still operating, because the red ink is a tax artifact, not an operating collapse. Green Thumb and Trulieve sit on net cash and ~$3 billion of combined equity: they win whether the catalyst comes next quarter or in three years. The squeezed middle — TerrAscend's 2028 balloon, Planet 13's 2026 credit-line roll — is where timing risk actually bites. Across the nine we can see in detail, eight of nine ran positive operating cash flow in FY2025 (only Planet 13 was negative, at −$14M). The losses are on the tax line, not in the stores.
Table C-2 — The overhang: disputed tax reserve as a share of enterprise value
The single sharpest way to see the back-tax risk: take each operator's disputed 280E reserve (the most recent filing — Q1 FY2026 where the 10-Q tags it, else the FY2025 10-K) and divide it by enterprise value. (Reserves still compound, so the four names that retag it in Q1 — Trulieve, Ascend, Jushi, TerrAscend — show a higher figure here than in Table B's 12/31/2025 column.) For several names the unpaid, contested tax is a large fraction of what the whole enterprise is worth — and for Planet 13 it exceeds it.
| Company | Disputed tax reserve | Enterprise value⁷ | Reserve ÷ EV |
|---|---|---|---|
| Planet 13 | $56M | $49M | 114% |
| Verano | $474M | $766M | 62% |
| Jushi | $186M | $363M | 51% |
| TerrAscend | $188M | $471M | 40% |
| Ascend Wellness | $219M | $636M | 34% |
| Trulieve | $696M | $2,043M | 34% |
| MariMed | $27M⁴ | $113M | 24%–27% |
| Cresco † | ~$172M | $856M | 20% |
| Vireo Growth | $120M | $982M | 12% |
| Curaleaf † | ~$439M | $3,763M | 12% |
| Green Thumb | $203M | $1,807M | 11% |
| Glass House † | ~none | $1,294M | ~0% |
⁷ Enterprise value from the SSC comparables set, priced 5/29/2026. Reserves: Trulieve, Ascend, Jushi, TerrAscend at Q1 FY2026 (3/31/2026); MariMed = accrued income-tax payable (~$30M Q1); others at FY2025 (12/31/2025). † Curaleaf/Cresco web-sourced and Glass House discloses none — pending validation.
The pattern is the whole risk in one column. Planet 13's contested reserve is bigger than its entire enterprise value; Verano's, Jushi's, and TerrAscend's are a third to two-thirds of theirs. If the operators' 280E position prevails, these reserves age out harmlessly — or even reverse into gains. If the IRS prevails, they become real, interest-bearing liabilities sitting in front of the equity. The names at the top of this table have the most to gain and the most to lose from how the tax fight resolves; Green Thumb, Curaleaf, Vireo, and Glass House sit lightest.
The catalyst: a tax switch, half-flipped
The entire sector is a leveraged bet on one variable — whether, and how broadly, cannabis moves off Schedule I. As of late June 2026 the switch is half-flipped, and that asymmetry is the whole investable distinction between these names.
- Stage 1 — DONE (April 2026): A partial order moved state-licensed medical marijuana to Schedule III, lifting 280E on medical income only, prospectively for the full 2026 tax year. Adult-use stays Schedule I. The first visible benefit shows up in Q2 FY2026 results (Q1 closed before the effective date).
- Stage 2 — CONTESTED (now): A DEA hearing on moving all marijuana to Schedule III — the bigger prize that lifts 280E from adult-use too — opened June 29, 2026 and runs to no later than July 15. The roster was stacked with rescheduling opponents, and a parallel D.C. Circuit emergency stay could even vacate the medical relief. As of this writing there is no outcome; a broad final rule, if it comes, would still need months to take effect.
A subtlety worth holding onto: a broad Schedule III order would lift 280E from adult-use without federally legalizing recreational sales — because 280E keys off the drug's schedule, not the legality of the sale. The tax prize can arrive even though the product stays federally unauthorized.
The medical-vs-adult-use mix is what splits the trade: medical-tilted operators (Trulieve, Jushi, Verano, parts of TerrAscend and Planet 13) capture Stage-1 relief now; adult-use-heavy names (Green Thumb, Ascend, MariMed, Cresco, Glass House) need the broad order. The sector has already priced some of this — the $MSOS index sits near a 2026 high, up roughly 104% over the year into the hearing — but the billion-dollar names still trade at ~1–3× revenue and ~3–14× EBITDA, a fraction of the consumer-staples businesses they resemble above the tax line. The market is paying for medical relief and pricing out almost everything else.
The tiers
Tier 1 — The Survivors. Green Thumb (GTBIF), Trulieve (TRLV). Net cash, real GAAP economics, the cleanest balance sheets. Green Thumb is the only consistently profitable major and a buyer, not a target; its catalyst (adult-use relief) simply hasn't arrived — its most recent quarter still carried a 76% effective tax rate. Trulieve is the cleanest expression of the near-term trade — 60% margins, medical-heavy, already restructured to capture the medical fix, and the first plant-touching operator on the NYSE. They win on either timeline. → Green Thumb coverage · Trulieve coverage
Tier 2 — The Operating Machines. Curaleaf (CURLF) †, Verano (VRNOF), Cresco (CRLBF) †, TerrAscend (TSNDF), Jushi (JUSHF). Genuine scale and EBITDA at high gross margins, with free cash flow suppressed almost entirely by tax and interest. Curaleaf is the biggest operator of all ($1.27B revenue, $275M EBITDA) and an international consolidator, but levered and impairment-scarred; Cresco runs a No.-1-brand CPG model and has formally taken the position 280E doesn't apply. The medical-tilted names here capture Stage-1 relief now; all carry heavy balance-sheet beta to the broad order — and refinancing-timing risk (TerrAscend's 2028 wall). TerrAscend is the named defendant in the first 280E clawback suit (below). → Verano · TerrAscend · Jushi · Curaleaf and Cresco are added comparison names (no individual RoboSystems report yet).
Tier 3 — The High-Beta Options. Ascend Wellness (AAWH), Vireo Growth (VREOF), MariMed (MRMD), Planet 13 (PLNH), Glass House (GLASF) †. Smaller, more levered, or more idiosyncratic — the most binary expressions of the thesis. Ascend is a good retail operator inside a punishing capital structure (negative equity, a 2029 debt wall); Vireo is the fastest grower and most aggressive consolidator (and paid just $1M of cash tax on a $28M expense); MariMed is a 25-quarter positive-EBITDA brand house with a back-tax overhang bigger than its market cap; Planet 13 is priced as a survival bet where the catalyst is simply removing a cash drain; Glass House is a different animal — a low-cost California cultivation play with the lightest 280E footprint in the group, an operational/CA-market story more than a tax story. → Ascend · Vireo · MariMed · Planet 13 · Glass House is an added comparison name (no individual RoboSystems report yet).
The risks that bind the whole group
- The IRS is litigating the gambit. That ~$2.7B of disputed reserves assumes the operators' position prevails. The IRS sued TerrAscend in May 2026 to claw back an $8.3M refund — the first known 280E clawback — and is examining multiple filers; aggregate industry back-280E exposure is reported near $1.6 billion. If the position fails, those reserves — the ~$480M of nine-name tax not paid in cash, plus everything accrued before — become real, interest-bearing liabilities. Table C-2 shows who is most exposed.
- The broad order may not come. The DEA hearing is contested and opponent-heavy; a D.C. Circuit stay could even unwind the medical relief. Stage 2 is optionality, not a base case.
- A nine-MSO antitrust overhang. A February 2026 state "cannabis cartel" price-fixing suit names nine operators — a live overhang for the group.
- Impairment and dilution scars. Several names carry boom-cycle goodwill/intangible write-downs (Cresco took a $105M impairment in FY2025; Curaleaf and Verano have multi-year write-down histories); the roll-ups (Vireo) dilute relentlessly.
Bottom line
Strip out one line of the tax code and these twelve companies earn roughly $3.4 billion of gross profit a year at consumer-staples margins, with most already generating positive operating cash. Leave it in, and the same twelve post a collective net loss north of a billion dollars, book a $786 million tax bill, and carry ~$2.7 billion of disputed taxes — for one name, more than the whole enterprise is worth. That gap — between the operating business and the after-tax business — is the cannabis trade. Tier 1 owns the business; Tier 2 owns the operating leverage to the tax; Tier 3 owns the option on the tax itself. Which one you want depends entirely on your read of a switch that, as of today, is half-flipped and being fought over in two federal venues at once. The data says these are real businesses hiding behind a punitive tax rate. What they're worth depends on the policy — and that, not the operations, is the variable. You decide.
This analysis was built with Claude + RoboSystems — Claude Opus did the company-by-company analysis on RoboSystems' open, structured access to every public company's SEC filings. The nine RoboSystems coverage names were pulled and re-validated directly from SEC XBRL; Curaleaf, Cresco, and Glass House (†) were added from their FY2025 public releases and are flagged pending independent validation. You can set up the same tools in about five minutes at robosystems.ai. New customers get 50% off your first month with code CANNABIS50.
Full coverage on the nine RoboSystems names — financials, catalyst scenarios, and re-rating math: Trulieve · Green Thumb · Verano · Ascend Wellness · Vireo Growth · Jushi · TerrAscend · MariMed · Planet 13
Curaleaf, Cresco, and Glass House are included here as a comparison set; individual RoboSystems coverage reports for them are not yet published.
Data: the nine coverage names' FY2025 figures were pulled from SEC XBRL via the RoboSystems knowledge graph and re-validated statement-by-statement; Curaleaf, Cresco, and Glass House (†) are web-sourced from FY2025 company releases and pending independent validation. Enterprise values from the SSC comparables set (5/29/2026). Disclaimer: This is a fact-based reference report. Implied values are scenario ranges, not price targets. Nothing here is investment advice.