TG Therapeutics (Nasdaq: $TGTX) Stock Hub 2026: BRIUMVI At A $1 Billion Run Rate, $750 Million Of Debt And The Subcutaneous Readout That Decides 2027

01 What TG Therapeutics Is In August 2026

TG Therapeutics is no longer a clinical-stage story. In the three months to June 30, 2026 the company
recognised $240.3 million of total revenue, reported net income of $7.8 million
and finished the period with $612.3 million in cash, cash equivalents and investment securities.
Those three numbers come from the same source, the exhibit to the Form 8-K filed on August 3, 2026, and they
describe a company that sells a product, collects the money and keeps some of it.

Almost all of that revenue is one drug sold in one country. BRIUMVI, ublituximab-xiiy, is an
anti-CD20 monoclonal antibody approved in the United States for relapsing forms of multiple sclerosis. United
States net product revenue for BRIUMVI was $227.7 million in the quarter, which the company
describes as approximately a 64 per cent increase over the same period a year earlier. Everything else, meaning
sales to the ex-United States partner and the royalty that comes back from it, added $12.6 million.
That is 5.2 per cent of the quarter.

The concentration cuts both ways and is the first thing to hold on to. A single product growing at this rate
produces operating leverage that few commercial-stage biotechs ever reach: the company has now printed positive
net income in consecutive periods and generated positive operating cash flow in the first half. The same
concentration means that a label change, a reimbursement decision, a competitor launch or a manufacturing
interruption would land on essentially all of the revenue at once. There is no second product to absorb it.

The second thing to hold on to is that the profit is currently being spent on purpose. Revenue grew 70.3 per
cent in the quarter and net income fell 72.4 per cent, from $28.2 million to $7.8 million. That is not a
deterioration in the business. It is the arithmetic of a company running a commercial launch and, at the same
time, paying to manufacture a subcutaneous formulation that has not been approved and standing up a second
manufacturer. Research and development spending tripled year on year, and $54.6 million of that
increase in the quarter is that single manufacturing line.

The company describes itself as a fully integrated commercial-stage biotechnology company focused on B-cell
diseases. The practical translation, as of August 2026, is one approved product carrying the whole income
statement, three separate attempts to widen it, and a $750 million term loan sitting behind all of it.

02 BRIUMVI: The Label, The Dosing And The Safety Section

BRIUMVI is a glycoengineered monoclonal antibody directed at a specific epitope on CD20-expressing B cells.
The removal of certain sugar molecules from the antibody, which the company calls glycoengineering, is intended
to allow efficient B-cell depletion at lower doses. That is the mechanistic claim; what matters commercially is
the label.

The United States indication, in the exact words of the prescribing information in force,
reads: “BRIUMVI is indicated for the treatment of relapsing forms of multiple sclerosis (MS), to include
clinically isolated syndrome, relapsing-remitting disease, and active secondary progressive disease, in
adults.” The application is BLA 761238, approved on December 28, 2022
under standard review, with the United States launch in January 2023.

The European indication is narrower: “Briumvi is indicated for the treatment of adult
patients with relapsing forms of multiple sclerosis (RMS) with active disease defined by clinical or imaging
features.” The European marketing authorisation was granted on May 31, 2023 following a positive CHMP
opinion on May 30, and the marketing authorisation holder in Europe is
not TG Therapeutics — it is Neuraxpharm Pharmaceuticals S.L., the commercial
partner. Anyone reading the European product record will find Neuraxpharm’s name on it, and that is the correct
structure rather than an error.

Dosing, from Table 1 of the label, is the operational fact that shapes the whole pipeline strategy: a first
infusion of 150 mg over four hours, a second infusion of 450 mg two weeks later over
one hour, then 450 mg every 24 weeks over one hour. Two visits to start, then two visits a year.
The three programmes described further down this page — the consolidated single infusion, the subcutaneous
formulation and the autoinjector — all exist to shorten that.

The safety section is not decorative and belongs in any honest description of the asset. In the multiple
sclerosis trials, infusion reactions occurred in 48 per cent of premedicated patients, with
0.6 per cent serious. The overall infection rate was 56 per cent against 54 per cent on teriflunomide, serious
infections 5 per cent against 3 per cent, and there were three infection-related deaths among
BRIUMVI-treated patients. The label carries contraindications in active hepatitis B and in patients with a
history of life-threatening infusion reaction, and warnings covering hepatitis B reactivation, progressive
multifocal leukoencephalopathy, reduction in immunoglobulins and, added in the most recent revisions,
clinically significant liver injury reported in the post-marketing setting for anti-CD20
therapies approved in multiple sclerosis, BRIUMVI included.

Three label supplements have been approved since launch on top of the original approval: two on
August 18, 2025 and one on January 16, 2026. They correspond to changes in
the dosing and administration sections and in the warnings, including hepatitis B reactivation. Exclusivity runs
twelve years from approval, to December 2034, with patents the company states as extending to
2042.

03 ULTIMATE I And II: What The Registrational Data Established, And What It Did Not

Approval rests on two trials, ULTIMATE I and ULTIMATE II, of identical design: randomised,
double-blind, double-dummy, parallel-group, active-comparator studies running 96 weeks in patients with relapsing
multiple sclerosis. Patients received either BRIUMVI intravenously with an oral placebo, or
teriflunomide 14 mg orally once a day with an intravenous placebo. Entry required at least one
relapse in the prior year, or two in the prior two years, or a gadolinium-enhancing lesion in the prior year,
with a baseline EDSS between 0 and 5.5.

ULTIMATE I randomised 549 patients, ULTIMATE II randomised 545. The primary
endpoint in both was the annualised relapse rate.

In ULTIMATE I, the annualised relapse rate was 0.076 on BRIUMVI against 0.188 on
teriflunomide, a 59 per cent relative reduction, p<0.001. In
ULTIMATE II it was 0.091 against 0.178, a 49 per cent relative reduction,
p=0.002. The imaging endpoints were more emphatic still: gadolinium-enhancing T1 lesions per scan
at week 96 fell by 97 per cent in both trials, and new or enlarging T2 lesions by 92 and 90 per cent
respectively, all at p<0.001.

The disability endpoint did not read out positive, and that matters. In the pooled analysis
of both trials, confirmed disability progression at 12 weeks was 5.2 per cent on BRIUMVI against 5.9 per
cent on teriflunomide, a 16 per cent risk reduction with p=0.510. The prescribing
information states it plainly: there was no statistically significant difference in disability progression
between the two arms. The published New England Journal of Medicine analysis gives the same result as a hazard
ratio of 0.84 with a 95 per cent confidence interval from 0.50 to 1.41.

What that means in practice is that BRIUMVI is established on relapse and lesion control, the endpoints on
which the anti-CD20 class competes, and is not established on slowing disability accumulation against an active
comparator. Neither is the class as a whole, on this particular comparison. Anyone describing the drug as
proven to slow disability progression is going beyond the label.

One arithmetic caution for readers who cross-check. The European public assessment report quotes pooled
figures — 0.09 against 0.23 relapses a year — that do not match the per-trial numbers above. That is not a
contradiction but a different aggregation of the same two studies. The per-trial figures from the FDA label or
the journal publication are the ones to use.

04 The Commercial Ramp And What It Is Doing To The Balance Sheet

The commercial line is the part of this company that has actually changed in the last two years. United
States net product revenue for BRIUMVI, quarter by quarter, went from $138.8 million in the
second quarter of 2025 to $227.7 million in the second quarter of 2026. The company stated
approximate figures of $182 million for the fourth quarter of 2025 and about $594 million for the full year
2025, against total global revenue of roughly $616 million for that year.

Two of the intermediate points on that path deserve a caveat, and the chart in this section carries it: the
2025 quarterly figures are company-stated approximations given at an investor presentation, and the first
quarter of 2026 is derived by subtracting the reported second quarter from the reported half year. The
direction is not in question. The decimals between the audited points are.

What the ramp has done to the shape of the business is visible in three places. Accounts receivable stood at
$401.6 million at June 30, up $95.9 million in six months, which on quarterly revenue of $240
million is roughly 150 days of sales outstanding. The company attributes this to large, established, credit-worthy
customers on extended payment terms and recorded no allowance for credit losses at the balance sheet date.
Inventory, current and non-current together, reached $223.5 million, up $80.1 million in the
half. And cost of revenue rose 117.5 per cent against revenue growth of 70.3 per cent.

That last line has a specific explanation rather than a margin problem. Inside cost of revenue sits the
royalty owed to LFB under the licence that underpins ublituximab: $24.4 million in the
quarter and $45.0 million in the half, with $24.3 million still payable at June 30. The rate is
disclosed as escalating from mid-single digits to high-single digits on net sales. As BRIUMVI sales grow, the
royalty grows with them and the rate steps up. Gross margin on total revenue works out at 82.9 per cent for the
quarter, though the company does not publish that as a metric and it should be read as a calculation rather
than a disclosure.

Outside the United States, the model is different and much smaller. TG sells product to
Neuraxpharm, which holds the European marketing authorisation and commercialises there, and
receives a royalty in return. In the quarter that was $8.1 million of product sales and $3.6 million of royalty.
No sell-through figure for the European market is published by TG, so the size of BRIUMVI’s actual European
business is not verifiable from the company’s own disclosures.

05 Guidance Raised Twice, And The Run-Rate Claim

Guidance has been raised twice this year, and the sequence is worth setting out because it is the cleanest
available evidence on how the launch is tracking against the company’s own expectations.

On February 26, 2026, with the full-year 2025 results, TG guided to total global revenue of
approximately $875 to $900 million for 2026 and BRIUMVI United States net product revenue of
approximately $825 to $850 million. On May 6, 2026 those became approximately
$925 million and $885 to $900 million. On August 3, 2026 they
became approximately $950 million and $890 to $905 million.

The quarterly targets were beaten in both quarters for which one was given. The first-quarter target of about
$185 to $190 million of BRIUMVI United States revenue came in at about $194.8 million; the second-quarter target
of about $220 million came in at $227.7 million. No quarterly target was given for the third quarter in the
August release.

The cost side of the same guidance moved in the same direction, which is the part that gets less attention.
Full-year operating expense, defined by the company as research and development plus selling, general and
administrative and excluding non-cash compensation, went from “approximately $350 million” in February
and May to “approximately $350 – $400 million” in August, on top of approximately $100 million of
subcutaneous manufacturing and second-manufacturer start-up costs. Revenue guidance rose by $25 million; the
upper end of the expense guidance rose by up to $50 million.

Beyond 2026 there is no numerical guidance at all. What exists is a statement by the chairman
and chief executive, Michael S. Weiss, in the August 3 release: the second quarter puts the company
“on track to exit 2026 at an approximately $1 billion annualized U.S. BRIUMVI revenue run rate”. That is
an exit run rate, not a 2027 revenue target, and it is a forward-looking statement by management rather than a
reported figure. It is quoted here as such.

06 ENHANCE, The Consolidated Infusion, And Why There Is No PDUFA Date

Three separate programmes exist to make BRIUMVI easier to give. The first has already read out.

ENHANCE is a Phase 3, randomised, double-blind, placebo-controlled trial testing whether the
two-visit start can be collapsed into one. The primary endpoint was bioequivalence of total drug exposure between
the approved regimen — 150 mg on day 1 followed by 450 mg on day 15 — and a
single consolidated 600 mg infusion on day 1, measured as area under the curve to week 16 and
compared as a geometric mean ratio.

On May 27, 2026 the company announced that ENHANCE met its primary endpoint, with a geometric
mean ratio of approximately 1.0 and a 90 per cent confidence interval inside the predefined bioequivalence bounds
of 0.80 to 1.25. Secondary endpoints on safety, B-cell depletion and imaging were consistent with prior studies.
No infusion reactions of grade 3 or higher occurred in either arm, and infusion reactions were
lower than those reported in ULTIMATE I and II, with fewer in the single-infusion arm although the difference
between arms was not statistically distinguishable.

The regulatory status is where care is required. The company’s stated position is a
supplemental BLA filing targeted for the second half of 2026. As of the Form 10-Q filed on
August 5, 2026, no supplemental BLA has been filed: the string does not appear in the filing in
that sense, and no 8-K announces a submission or an acceptance. It follows that
there is no PDUFA date for any TG Therapeutics programme. That is not a gap in the research; it
is a positive finding. Anyone quoting a decision date for the consolidated regimen or for the subcutaneous
formulation is quoting something that does not exist.

A related set of data sits in the same category. In November 2024 the company reported that rapid 30-minute
infusions were well tolerated, with all infusion-related reactions mild and resolving, and that patients already
B-cell depleted on a prior anti-CD20 therapy could switch directly to 450 mg given over one hour without the
150 mg starting dose, with 97 per cent of infusions completed without interruption. Neither schedule is
in the label. Table 1 still specifies four hours for the first infusion. No regulatory filing has been
announced for the 30-minute infusion.

07 Subcutaneous BRIUMVI: The Largest Call Option The Company Owns

The subcutaneous programme is the largest single call option TG Therapeutics owns, and it is also where
the money is currently going.

On June 3, 2026 TG reported Phase 1 results for a high-concentration
400 mg in 2 mL subcutaneous formulation of ublituximab. More than 100 patients were treated in
the study, of whom more than 80 received the subcutaneous route across doses from 50 mg to 400 mg, with over 225
subcutaneous injections given and more than three quarters of them at the full 400 mg. Mean bioavailability was
greater than 60 per cent relative to intravenous administration, with the lower bound of the
95 per cent confidence interval above 55 per cent. Kinetics were linear across the dose range.

Local injection-site reactions occurred in fewer than 5 per cent of patients and systemic
injection-related reactions in about 21 per cent, not dose-dependent, mostly at the first
injection, and resolving in all patients. No severe injection-site reactions and no new safety signals were
reported.

The Phase 3 is NCT07211633, a randomised, open-label, parallel-group non-inferiority study
with three arms: intravenous BRIUMVI on the approved schedule, subcutaneous 400 mg every eight weeks, and
subcutaneous 400 mg every twelve weeks. The primary endpoint is exposure to week 24, with non-inferiority
declared if the lower bound of the 90 per cent confidence interval of the geometric mean ratio exceeds 0.80.
Enrolment is complete. Modelling from the Phase 1 data projects a geometric mean ratio of about
1.21 for the quarterly regimen with a lower bound of 1.15, and about 1.58 for the every-eight-weeks regimen with
a lower bound of 1.50.

Topline is guided for year-end 2026 or the first quarter of 2027. Weiss has spoken of
advancing toward a potential approval in 2028. A device bridging study for an autoinjector,
NCT07503873, has been recruiting since April 2026 with primary completion estimated for
July 2027.

The strategic claim attached to this programme — that a subcutaneous option
“could nearly double the addressable market opportunity for the BRIUMVI franchise” — is a company
statement about an unapproved product and is reported here as that, not as an estimate anyone has verified. What
can be said without stretching is that the competitive target is specific: Kesimpta, ofatumumab,
Novartis, which is the anti-CD20 that patients administer themselves at home. Ocrevus Zunovo, the subcutaneous
ocrelizumab from Genentech, is given by a healthcare professional and sits in a different position.

08 Myasthenia Gravis, Schizophrenia And What Is No Longer In The Pipeline

Two new indications entered the clinic in the space of two months, both using the approved molecule in
diseases where B-cell depletion has a mechanistic rationale.

Myasthenia gravis. In June 2026 TG announced positive topline Phase 1 data for subcutaneous
BRIUMVI in myasthenia gravis and started a randomised Phase 2 that the company describes as potentially
registration-directed. The design is unusual and worth reading closely: intravenous BRIUMVI as
maintenance therapy following induction with efgartigimod, in adults. The trial is
NCT07673744, randomised, double-blind, multicentre, placebo-controlled,
120 patients, started July 30, 2026, with primary completion estimated for
July 1, 2028.

Schizophrenia. In July 2026 the company began a Phase 2 in adults with treatment-resistant
schizophrenia, NCT07680946, 60 patients, started July 3, 2026, primary
completion estimated December 1, 2028. This is an immunological hypothesis applied to a
psychiatric indication. It is early, it is small, and no data exist.

A pediatric programme in relapsing multiple sclerosis is also running: NCT07220252, Phase
2/3, 240 patients, with arms for ublituximab, placebo and fingolimod, primary completion estimated January 2030.

Both of the new adult indications share the same characteristic: the earliest primary completion is
mid-2028. Neither will produce a readout that moves this file in the next twenty-four months.
They matter as evidence of what management intends to do with the cash the commercial product generates, not as
near-term catalysts.

One negative finding belongs here for completeness. Programmes that appear in older descriptions of this
company — TG-1701, TG-1801, umbralisib — do not appear in the 2025 Form 10-K or the second-quarter
2026 Form 10-Q pipeline table. The two clinical records are closed: NCT03671590 terminated, NCT03804996 completed
and NCT04806035 terminated. No anti-CD3 programme exists in either filing. Any source
attributing one to TG Therapeutics is describing something that is not in the company’s own disclosures.

09 Azer-cel: The Only Asset That Is Not Ublituximab

Azer-cel, azercabtagene zapreleucel, is the one asset in the pipeline that is not
ublituximab. It is an allogeneic, off-the-shelf CAR-T directed at CD19, built from donor T cells using the ARCUS
genome-editing technology and designed to avoid graft-versus-host disease.

It is licensed rather than owned outright. In January 2024, through the subsidiary TG Cell Therapy, TG took an
exclusive global licence from Precision BioSciences covering autoimmune disease and all
non-oncology indications. The economics, from the filings: $7.5 million upfront, split as $5.25
million in cash and $2.25 million as an equity investment in Precision; a further $2.5 million
paid in equity on January 7, 2025; another $7.5 million on short-term clinical or time-based
milestones, partly in equity at a preset premium; and up to $288 million of further clinical,
regulatory and commercial milestones, with royalties running from high single digits to low double digits on net
sales. The first of those milestones was reached on February 23, 2026, triggering the $7.5
million payment.

The FDA cleared the IND in August 2024 for progressive forms of multiple sclerosis, and the
first patient was dosed in August 2025. The Phase 1, NCT06680037, is recruiting
with 100 patients planned and primary completion estimated December 2027. A long-term follow-up record,
NCT07577583, is registered but not yet recruiting, with an anticipated start in May 2027.

The near-term item is presentational: TG has committed to
present preliminary Phase 1 azer-cel data in progressive multiple sclerosis in the second half of
2026. That commitment was made in January and repeated on August 3. It is one of only three remaining
2026 pipeline milestones the company lists, the others being the full ENHANCE results and the subcutaneous Phase
3 topline.

Progressive multiple sclerosis is a harder problem than the relapsing form, and the field is littered with
assets that worked on inflammation and did nothing for progression — as the ULTIMATE disability endpoint
demonstrates within this same company’s own file. Preliminary Phase 1 data in a handful of patients will not
settle it either way.

10 The Second Quarter Of 2026, Line By Line

The second quarter of 2026, reported on August 3 and filed in full on August 5, reads as follows.

Revenue. Product revenue net was $235.794 million against $138.843 million a
year earlier. Licence, milestone, royalty and other revenue was $4.541 million against $2.305
million. Total revenue $240.335 million against $141.148 million, up 70.3 per cent.

Costs. Cost of revenue $41.194 million. Research and development
$95.337 million, of which $8.070 million is stock-based compensation. Selling, general and
administrative $82.130 million, of which $19.805 million is stock-based compensation. Total costs
and expenses $218.661 million against $106.305 million.

Below the operating line. Operating income $21.674 million against $34.843
million. Interest expense $16.572 million against $6.716 million, the consequence of replacing a
$250 million loan with a $750 million one. Other income $5.132 million. Net income before taxes $10.234 million,
income tax expense $2.453 million, net income $7.781 million. Earnings per share
$0.05 both basic and diluted, on 141.8 million basic and 157.3 million diluted weighted-average
shares.

The cascade has been recomputed line by line and reconciles: 235.794 plus 4.541 gives 240.335; 41.194 plus
95.337 plus 82.130 gives 218.661; 240.335 less 218.661 gives 21.674; 21.674 less 11.440 of net other expense
gives 10.234; less 2.453 of tax gives 7.781.

Stock-based compensation across both operating lines totalled $27.875 million
in the quarter, equal to 11.6 per cent of revenue, and $47.825 million in the half. It is a real cost of running
this company even though it does not consume cash, and the company’s own operating expense guidance is stated
before it.

Cash generation. For the first half, net cash provided by operating activities was
positive $21.369 million, against negative $21.279 million in the first half of 2025 — a swing of
$42.6 million. Capital expenditure was $0.344 million, so free cash flow for the half was about
$21.0 million. Deriving the second quarter alone by subtracting the first quarter’s disclosed
$17.9 million of cash use gives roughly $39.3 million generated in the second quarter, a
calculation rather than a published figure.

The gap between $27.6 million of half-year net income and $21.4 million of operating cash is working capital:
receivables absorbed $95.9 million and inventory $80.1 million, offset by $99.4 million of growth in payables and
accruals and $47.8 million of non-cash compensation added back.

11 The Blue Owl Loan And A Balance Sheet That Changed Shape

The balance sheet changed shape completely in March 2026, and the change is the most consequential financial
event of the year for this company.

On March 18, 2026 TG drew a $750 million term loan under a First Amendment
to its financing agreement with Blue Owl, using part of the proceeds to repay in full the $250
million initial term loan taken in August 2024. The new facility matures March 18, 2031.
Quarterly amortisation of $37.5 million begins with the quarter ending
March 31, 2030, with a balloon at maturity, and those instalments can be deferred if a total net
leverage ratio falls below an agreed threshold. Pricing starts at 4.75 per cent over the SOFR-linked option, with
a 25 basis point step-down on reaching a leverage threshold, reset quarterly. The loan is secured on
substantially all assets of the company and certain guarantor subsidiaries. An additional
uncommitted facility of up to $250 million sits alongside it. At June 30 the company states it
was in compliance with all financial covenants.

Repaying the old loan cost $9.153 million, recognised in the first quarter as a loss on
extinguishment, and the new financing cost $4.9 million to arrange. The loan is carried at
$745.387 million net of unamortised discount and issuance costs, all non-current.

Set against $612.3 million of cash, cash equivalents and investment securities at June 30,
net debt works out at roughly $137.7 million. Total equity was
$604.083 million, down from $648.020 million at December 31 despite $27.6 million of net income,
because of the buyback described in the next section.

Other capital structure items are small but should not be misread. Legacy convertible notes total
$15.5 million of principal at a 5 per cent coupon, converting at
$1,125 per share — a conversion price that makes them economically irrelevant as dilution,
equivalent to roughly 13,800 shares. Hercules warrants cover 115,042 shares at $17.95 and 50,172 shares at
$14.70, 165,214 in total.

The item that deserves attention is the shelf. TG filed an automatic shelf registration on Form
S-3ASR on August 8, 2025, effective on filing, registering an unlimited amount of debt,
equity or other securities. Nothing has been issued under it as of June 30, 2026, and the previous at-the-market
programme with Cantor Fitzgerald and B. Riley has expired. But an unlimited automatic shelf is the maximum
possible financing flexibility, and it is a structural dilution risk that exists whether or not it is used.

12 A Share Count That Is Shrinking, And The Buyback That Stopped

The share count is going down, not up, which is unusual enough in this sector to state plainly.

Shares outstanding were 155,305,953 at December 31, 2025 and
153,069,053 on the cover of the Form 10-Q dated August 3, 2026 — a reduction of about
2.24 million shares, or 1.4 per cent, in seven months. Shares issued actually rose over
the same period, from 158,849,596 to 159,949,330, on restricted stock grants; the whole of the reduction in the
count comes from repurchase. Weighted-average diluted shares fell from 162.5 million in the first half of 2025 to
158.7 million in the first half of 2026.

The buyback history is specific. A first $100 million programme authorised in August 2024 was completed in
September 2025, retiring 3,502,334 shares at an average of $28.55. A second programme authorised
in September 2025 was increased to $300 million on March 18, 2026, the same day the new loan was
drawn. Under it the company repurchased about $100.0 million at an average of $30.44. Treasury
stock stood at 6,871,546 shares at a cost of $200.2 million at June 30.

All of that second-programme spending happened in the first quarter. The Form 10-Q issuer purchases table
shows zero shares repurchased in April, May and June 2026, with
$200,008,004 still available under the authorisation. The programme has no fixed expiry and
obliges the company to buy nothing. A reader tracking whether management still considers the shares cheap has a
clean signal to watch in the third-quarter filing.

Dilution from equity awards is the offsetting force and is not trivial. At June 30 there were
11,220,924 restricted shares outstanding across two plans, 4,060,156 options,
and 1,949,005 stock tracking units. The tracking units are worth understanding: they can be
settled either in a share or in a cash payment of equal value at the committee’s discretion, so they are
simultaneously a dilution item and a cash exposure that rises with the share price. Unrecognised cost on them is
$84.8 million over a weighted-average 3.6 years, and in the second quarter alone they produced
$15.3 million of expense, of which $8.6 million was pure fair-value remeasurement because the
share price rose.

Insider activity in the last three months contains no open-market purchases and no open-market
sales. Five directors received annual stock tracking unit grants of 8,325 units each on June 11, and
director Laurence N. Charney made three transfers in early August totalling 6,050 shares, recorded under code G
and described in the filings as bona fide gifts to a 501(c)(3) institution. Gifts are not sales and produce no
proceeds. None of the six filings is flagged as made under a Rule 10b5-1 plan.

13 Market Data, Ownership And Retail Sentiment

The market data below was read on August 21, 2026 and changes continuously.

The shares traded around $55.38, up 9.5 per cent on the day, on volume of
about 2.72 million against an average of 2.37 million, a relative volume of roughly 2.5. Market capitalisation
was about $8.48 billion on 153.08 million shares outstanding and a float of 141.64 million.
Institutional ownership stands at 72.99 per cent and insider ownership at 7.47 per cent. The consensus target
carried by the aggregator was $69.71, which is a third-party aggregation rather than a figure this page endorses.

Short interest is 20.70 per cent of the float. On a company that is profitable, cash
generative and raising guidance, one share in five of the tradable float being sold short is a fact that shapes
how this stock behaves on news. It amplifies moves in both directions and it is the single most important thing
to know about the mechanics of the shares as distinct from the business.

The disclosed institutional register, taken only from Schedule 13G filings and the proxy table rather than
from aggregators, shows BlackRock at 20,134,882 shares, Vanguard at 15,596,480
and State Street at 8,100,619 as at the April 14, 2026 proxy date. Chief executive
Michael S. Weiss is listed at 11,544,752 shares, 7.54 per cent, though that figure includes
5,995,707 unvested restricted shares and 1,766,666 shares issuable on options, so beneficial ownership overstates
outright holding by a wide margin.

One trap in that register deserves flagging. Vanguard filed a Schedule 13G/A on March 27, 2026 showing
zero shares. That is not an exit. The proxy explains it as an internal realignment, after which a
new filer, Vanguard Portfolio Management, filed for 8,400,210 shares on April 29. Reading only the amendment
produces the wrong conclusion. What is true is that the position as filed fell from 15.6 million to 8.4 million
shares, and the realignment does not explain that part.

On the day this page was written the retail conversation was dominated by a single theme:
an unconfirmed takeover rumour. Posts referenced a market-chatter alert speculating about an
acquirer with a capitalisation between $20 billion and $40 billion, and discussed what price would be acceptable.
No primary source supports it. There is no company announcement, no Form 8-K, no Schedule 13D and
no statement from TG Therapeutics. Between August 8 and August 21, 2026 the company filed nothing at all with the
Securities and Exchange Commission, and the last item on its news wire is the second-quarter release of August 3.

The distinction matters more than usual here. A 9 per cent move on a $8.5 billion company with a fifth of its
float short, on no filing and no release, is a flow event and a rumour, not a change in the business. The
business changed on August 3, when guidance went up. Nothing verifiable has changed since.

14 The Anti-CD20 Field And Where BRIUMVI Sits In It

Three anti-CD20 antibodies are approved by the FDA for multiple sclerosis besides BRIUMVI, and the company
names them itself.

OCREVUS, ocrelizumab, Genentech, BLA 761053, approved March 28, 2017,
intravenous. OCREVUS ZUNOVO, ocrelizumab with hyaluronidase, Genentech, BLA 761371, approved
September 13, 2024, subcutaneous and administered by a healthcare professional.
KESIMPTA, ofatumumab, Novartis, approved for multiple sclerosis on
August 20, 2020 as supplement 70 to BLA 125326, subcutaneous and self-administered by the
patient.

The 2025 Form 10-K states the position in the company’s own words: BRIUMVI directly competes with
ocrelizumab, the only other approved intravenously administered anti-CD20 antibody, and with the subcutaneous
version of ocrelizumab given by healthcare providers; it also competes with ofatumumab, and ofatumumab
“would represent direct competition for any future self-administered subcutaneous formulation of
ublituximab currently under development”.

That sentence is the strategic map. TG’s subcutaneous programme is not aimed at Ocrevus. It is aimed at
Kesimpta, and the differentiator being tested is dosing frequency: quarterly rather than monthly. Whether that
matters commercially depends on data that will not exist until the end of 2026 at the earliest.

The 10-K also names two emerging classes as potential threats, neither approved in multiple sclerosis:
BTK inhibitors, which the company says could alter treatment paradigms and sequencing in ways
that affect anti-CD20 utilisation, and therapies directed at CD40 ligand, described as being at
varying and often early stages.

Two clarifications for accuracy. Rituximab is not FDA-approved for multiple sclerosis; it is
used off-label in some settings and the company does not list it among approved competitors. And no market share
figure for the anti-CD20 class appears in any primary source examined for this page — the only related statement
on record is a management objective, not a measurement.

15 The Catalyst Map To 2028

The calendar for the next eighteen months has three company-committed items and no regulatory decision dates,
for the reason set out earlier: nothing is under review.

One registry inconsistency should be noted by anyone checking these dates independently. The ClinicalTrials.gov
record for ENHANCE, NCT05877963, still shows the study as recruiting with primary completion in January 2027,
while the company announced completion of randomised-cohort enrolment in October 2025 and positive topline in May
2026. Where the registry and the company disagree here, the company’s filings are the more current source.

16 Risks And Red Flags

Concentration. One product, one country, 95 per cent of revenue. Every operational risk that
attaches to BRIUMVI attaches to the whole income statement at once.

Leverage. $750 million of secured debt against $612.3 million of cash. Interest expense more
than doubled year on year to $16.6 million in the quarter and is now large enough to consume most of operating
income: $21.7 million of operating income against $11.4 million of net other expense. Amortisation of $37.5
million a quarter begins in 2030. The loan is secured on substantially all assets.

Working capital. Receivables at $401.6 million represent roughly 150 days of sales, and
inventory has risen $80.1 million in six months. No credit loss allowance is recorded. If collection slows or
inventory has to be written down, cash generation reverses quickly.

The royalty escalator. The LFB royalty rises from mid-single to high-single digits as sales
grow, so the cost of revenue line gets structurally worse as the product succeeds. It was $24.4 million in the
quarter, 59 per cent of cost of revenue.

Spending ahead of approval. Roughly $100 million of full-year cost is being spent
manufacturing a subcutaneous formulation that has no approval and whose pivotal data do not read out until year
end at the earliest. If that Phase 3 misses, the spending does not come back.

The unlimited shelf. An automatic S-3ASR registering an unlimited amount of securities is in
force. Nothing has been issued under it, and nothing prevents issuance.

Label and safety. Three supplements since approval have added or expanded warnings, most
recently on hepatitis B reactivation in January 2026, and the label now carries post-marketing liver injury
language for the class. Anti-CD20 therapy carries progressive multifocal leukoencephalopathy risk. Three
infection-related deaths occurred in the registrational programme.

Governance signal. At the June 11, 2026 annual meeting all directors were re-elected, but two
drew unusually high withheld votes: Sagar Lonial at 54.85 per cent for and 45.15 per cent withheld, and Kenneth
Hoberman at 65.52 per cent for. One management proposal was not approved. The company has only two executive
officers, and the chairman is also chief executive and president.

Rumour risk, today specifically. A material part of the current share price move rests on an
unverified takeover story with no primary source. Positions taken on that basis are exposed to its being denied
or simply fading.

17 Merlintrader Health Score

The Merlintrader Health Score measures balance-sheet robustness and execution over a twelve to eighteen month
horizon on five weighted pillars. It is not an indication to buy or sell, and it says nothing about what the
shares are worth.

18 Bottom Line

The constructive case. BRIUMVI is growing at roughly 64 per cent year on year in its main
market, guidance has been raised twice in six months, and the company is profitable with positive operating cash
flow. The share count is shrinking. Cash covers most of the debt. Three attempts to widen the franchise — a
one-visit start, a home-administered subcutaneous option and two new indications — are all funded from the
product’s own cash flow rather than from equity issuance. If the subcutaneous Phase 3 reads out non-inferior at
the end of 2026, the company enters 2027 with a differentiated presentation aimed squarely at the only
self-administered competitor in the class.

The sceptical case. Ninety-five per cent of revenue is one molecule in one country, and the
cost of selling it is rising faster than the revenue: cost of revenue up 117 per cent against revenue up 70 per
cent, with a royalty that escalates as sales grow. Net income fell 72 per cent in the quarter. The balance sheet
is now leveraged with $750 million secured on substantially all assets, and interest expense has more than
doubled. Roughly $100 million a year is being spent to manufacture a product that is not approved. The disability
endpoint in the registrational programme did not reach significance, which limits how the drug can be positioned
against the class. And a fifth of the float is short, which means the share price is not currently a clean read
on any of this.

What would settle it. The subcutaneous Phase 3 topline at year end or early 2027 is the
single largest variable, because it determines whether this is a one-presentation intravenous product facing a
self-administered competitor or a company with both. Before that, the third-quarter print will show whether the
BRIUMVI ramp is still beating the company’s own targets, whether receivables keep growing faster than sales, and
whether the buyback resumes. Those three lines in one filing will say more than any amount of speculation about
who might or might not be interested in buying the company.

Related Research On Merlintrader

• Biotech Stock Hubs & Catalyst Index 2026 — every biotech company hub on Merlintrader, each with its own update date.
• Top Ten Biotech Stocks Right Now — the names we are following most closely.
• Free Biotech Catalyst Calendar — PDUFA dates, advisory committees and clinical readouts in one filterable calendar.