Biomea Fusion (Nasdaq: $BMEA) Stock Hub 2026: COVALENT-211 Fully Enrolled, $35.2 Million Of Cash And A Going-Concern Statement

01 What Biomea Fusion Is In August 2026

Biomea Fusion is a clinical-stage company in San Carlos, California, with one molecule that matters,
one early asset behind it, forty-one employees at the last count and thirty-five million dollars in the
bank. It has never sold a product, never recorded revenue, and at June 30, 2026 carried an accumulated
deficit of $469.8 million. Its shares trade on Nasdaq under $BMEA and closed at $1.37 on August 21, 2026,
against $30.00 a share paid by institutional investors in an April 2023 offering.

The molecule is icovamenib, formerly BMF-219, an oral covalent inhibitor of menin. The company is
testing it in type 2 diabetes, in type 1 diabetes and, through an academic platform trial in England, in
obesity in combination with semaglutide. The second asset is BMF-650, an oral small-molecule GLP-1
receptor agonist in a Phase I study in overweight and obese volunteers. Everything else that Biomea once
had, including the oncology programmes the company was founded on, has been terminated or shelved.

Three things define the position in August 2026, and they pull in different directions. The clinical
programme has never been better organised: two Phase II trials in the two patient groups that responded
best in earlier work, one of them fully enrolled as of August 24, 2026, with primary endpoint data due in
the first and second quarters of 2027. The cost base has been cut by roughly two thirds. And the balance
sheet is the thinnest it has been since the 2021 listing, thin enough that the quarterly report filed on
August 5, 2026 states in plain language that there is substantial doubt about the company’s ability to
continue as a going concern.

What follows is what the filings and the company’s own releases say, with the date attached to every
figure, including the places where the company has described the same clinical result with different
numbers in different releases.

02 Menin, Beta Cells And Why The Mechanism Is Unusual

Menin is a nuclear protein encoded by the MEN1 gene. In oncology it has become a well-known
target because it acts as a scaffold for the MLL fusion proteins that drive certain acute leukaemias, and
several menin inhibitors have been developed for that purpose. Biomea’s thesis is unrelated to leukaemia.
It rests on menin’s second role, as a brake on the turnover and growth of pancreatic beta cells, the cells
that make insulin.

Both forms of diabetes end in the same place: too few working beta cells. In type 1 the loss is driven
by autoimmune destruction; in type 2 by metabolic stress. Every widely used therapy manages the
consequence. Insulin replaces what the pancreas no longer makes. Metformin reduces hepatic glucose
output. GLP-1 receptor agonists amplify the insulin signal that remains and suppress appetite. None of
them rebuilds the cell population that has been lost, and none of them is designed to be stopped.

Biomea’s proposition, as described in its own materials, is that partial and reversible inhibition of
menin allows a patient’s remaining beta cells to proliferate and recover function, so that a defined
course of treatment produces an effect that persists after dosing ends. The company calls icovamenib a
potential beta-cell restorative therapy and, in its October 6, 2025 release, described it as a candidate
first non-chronic therapy for type 2 diabetes.

If the mechanism works as described, the commercial shape of the product is unusual: twelve weeks of an
oral tablet, then nothing, then a durable improvement in glycaemic control. If it does not, the same shape
is a liability, because a treatment that is stopped after twelve weeks has no second chance to show an
effect. The clinical evidence for the proposition is discussed in the sections that follow, and the honest
summary is that it is real, it is prospectively defined, and it rests on very small numbers of patients.

03 The Pivot: From Oncology To Diabetes, And What Was Shut Down

Biomea was built as an oncology company. Its first two clinical programmes put BMF-219, the same
molecule now called icovamenib, into blood cancers and solid tumours: COVALENT-101 in acute leukaemia,
lymphoma and myeloma, registered as NCT05153330 and opened in January 2022, and COVALENT-102 in non-small
cell lung, pancreatic and colorectal cancer, registered as NCT05631574 and opened in January 2023. A
second molecule, BMF-500, an oral covalent FLT3 inhibitor, entered a Phase I trial in acute leukaemia in
July 2023 under NCT05918692.

Both BMF-219 oncology trials now carry the status Terminated on ClinicalTrials.gov, checked on
August 24, 2026. COVALENT-101 enrolled 55 patients across 41 sites; COVALENT-102 enrolled 13 across 23
sites. BMF-500 is listed as active but no longer recruiting, with 35 patients enrolled and a primary
completion date of April 2026, and it does not appear in any company release of 2026. The corporate
description has followed the retreat: Biomea now presents itself, in the same sentence in every release
since 2025, as a clinical-stage diabetes and obesity company.

The pivot was not a strategic flourish. It followed the collapse of the oncology thesis and a cash
position that could no longer fund parallel programmes, and it was executed alongside a reduction in
headcount that shows up in every cost line of the last six quarters. Concentration of this kind removes
the possibility of a second shot. There is one mechanism, two indications, and one early asset that has
not yet produced a clinical efficacy readout.

04 The June 2024 Clinical Hold And The Damage It Did

In June 2024 the United States Food and Drug Administration placed a clinical hold on Biomea’s
diabetes trials. The hold was lifted in September 2024. The company’s annual report for 2025, filed on
March 24, 2026, describes the consequence for the type 1 diabetes study in a single sentence: the hold
had a profound impact on the COVALENT-112 trial in type 1 diabetes, where over 90% of the targeted
patient population was not able to complete dosing due to the clinical hold.

The damage to the type 2 trial was different in kind. COVALENT-111, registered as NCT05731544, carried a registered enrollment of 443
participants, healthy adults included, of whom 267 had received at least one dose of icovamenib and were
counted in the modified intent-to-treat population. When the company reported the study, it did not
analyse all 267. The October 6, 2025 release states that the topline efficacy
analysis covered the 163 patients who had completed at least 80 per cent of their planned dosing before
the hold, without other significant protocol deviations, and who were on one or more antihyperglycaemic
agents at baseline.

The trading history attached to those dates is unambiguous. The shares reached $20.21 intraday on
February 1, 2024 and closed at $11.27 on June 6, 2024. On June 7, 2024, the session in which the market
learned of the hold, they closed at $4.14, a fall of 63 per cent in a single day on 11.9 million shares
against a normal turnover under a million. By June 30, 2026 the closing price the company used to value
its own warrants was $1.47. The hold is not the only reason, but it is the event that separates the two
regimes, and it explains why a company running two Phase II trials in 2026 is doing so on total assets of
$36.9 million.

05 COVALENT-111: The Data Everything Rests On

COVALENT-111 is the study everything else is built on. It was a double-blind, randomised,
placebo-controlled trial in adults diagnosed with type 2 diabetes within the previous seven years, with
HbA1c between 7.0 and 10.5 per cent and body mass index between 25 and 40. Three regimens were tested,
all at 100 mg: Arm A once daily for eight weeks, Arm B once daily for twelve weeks, Arm C once daily for
eight weeks followed by twice daily for four. Outcomes were evaluated by diabetes phenotype using the
Ahlqvist algorithm, prespecified in the statistical analysis plan before unblinding.

The 52-week results released on October 6, 2025 reported two findings. In severely insulin-deficient
patients who received twelve weeks of treatment, Arms B and C combined, icovamenib produced a 1.2 per cent
reduction in HbA1c sustained through Week 52, with p=0.01. That analysis covered ten patients. The
strongest single arm was Arm B, six patients, with a mean reduction of 1.5 per cent and the same p value.
In the separate subgroup of patients already on GLP-1-based therapy who had not reached their glycaemic
targets at entry, eleven patients across all arms, eight or twelve weeks of icovamenib produced a
1.3 per cent reduction sustained through Week 52, with p=0.05.

Two features of that readout sit uncomfortably together. The effect was measured nine months after
the last dose, which is the entire point of the mechanism and is not something the diabetes field sees
often. And the groups are ten patients and eleven patients. Professor Ralph DeFronzo of the University of
Texas Health Science Center, quoted in the same release, called the durability remarkable and said it
suggested a restoration of beta-cell function. A prespecified subgroup of six patients with p=0.01 is a
hypothesis with a number attached to it, not a demonstrated treatment effect, and the two Phase II trials
now running exist precisely to convert one into the other.

06 The Same Result, Two Sets Of Numbers

The same results have been described with different numbers in different company releases, and anyone
comparing them side by side will find the discrepancy.

The October 6, 2025 release carries, in its headline bullets, a 1.8 per cent placebo-adjusted mean
reduction in HbA1c for the severely insulin-deficient population in Arm B, and a 1.8 per cent
placebo-adjusted reduction for the GLP-1 subgroup with all arms combined. The body of the same release
gives 1.5 per cent for Arm B and 1.3 per cent for the GLP-1 subgroup. The gap between the two sets is the
placebo adjustment, which the bullets state and the body text does not.

The release of August 24, 2026 restates the findings again, and in this version the severely
insulin-deficient population reaches up to a placebo-adjusted 1.5% mean HbA1c reduction at Week 52
(p=0.01) after a twelve-week course, while the GLP-1 subgroup reaches up to a placebo-adjusted 1.8%
mean HbA1c reduction at Week 52 (p=0.05) after eight or twelve weeks. The p values are unchanged from
October 2025, and the GLP-1 figure is the same 1.8 per cent given in the earlier bullets.

The severely insulin-deficient figure is the one that moved, and it moved by changing category rather
than by changing size. In October 2025 the placebo-adjusted headline for that population was 1.8 per cent
and 1.5 per cent was the unadjusted mean reduction reported in the body for Arm B. In August 2026 the
figure 1.5 per cent appears as the placebo-adjusted result. The same number is presented under two
different definitions ten months apart.

Nothing here suggests the underlying data changed, and the most likely explanation is a different cut
of the same analysis carried under a shorter label. The practical consequence for a reader is that the
number quoted for icovamenib’s effect depends on which release is being read, and that any figure taken
from a secondary source without its original release date is unreliable. Every figure quoted above carries
the release it came from.

07 August 24, 2026: COVALENT-211 Is Fully Enrolled

On August 24, 2026, before the market opened, Biomea announced that enrollment in COVALENT-211 was
complete. The shares traded up more than twenty per cent on the day on roughly eight million shares
against a thirty-session average of about 766,000, a Finviz Elite reading taken while the session was
still running.

COVALENT-211, registered as NCT07502495, is a randomised, double-blind, placebo-controlled Phase II
trial in adults with insulin-deficient type 2 diabetes who are not reaching glycaemic targets on standard
antihyperglycaemic therapy. Eligible patients had to be on a stable dose of one to three therapies for at
least three months, with HbA1c between 7.5 and 10.5 per cent and a body mass index of 32 or below. The
trial enrolled 64 participants across 18 clinical sites, randomised two to one to icovamenib 100 mg once
daily or placebo for twelve weeks on top of stable background therapy, followed by a forty-week
off-treatment period. The primary endpoint is assessed at Week 26, with secondary endpoints followed to
Week 52. The company expects topline Week 26 results in the first quarter of 2027.

The design is a direct transcription of the COVALENT-111 subgroup analysis into a trial built to test
it: the same 100 mg dose, the same twelve-week course, the same patient phenotype, the same off-treatment
follow-up that produced the durability claim. The dosing was informed by the COVALENT-121 food-effect
study, NCT07254286, which enrolled 60 healthy volunteers and completed in January 2026.

The trial’s own registry entry had not caught up. Read on August 24, 2026, NCT07502495 still showed a
status of recruiting and an estimated enrollment of 60 participants, last updated August 5, 2026, against
the 64 enrolled and the completion declared in the release of the same morning. Registry records lag
company announcements as a matter of routine, and the figures used above are the company’s.

Completing enrollment is an operational milestone, not a clinical one. It fixes the timetable and
removes the risk that the trial slips for lack of patients, which for a company with $35.2 million in the
bank is not a trivial risk to remove. It says nothing whatever about the result.

08 COVALENT-212 And The GLP-1 Non-Responders

The second Phase II trial addresses a larger commercial question. COVALENT-212, registered as
NCT07502508, enrolls adults with type 2 diabetes who are not reaching glycaemic targets despite GLP-1-based
therapy. Participants must have been on a stable GLP-1-based regimen for at least three months and may
take up to two additional background therapies, metformin and an SGLT2 inhibitor. HbA1c must be between
7.0 and 10.5 per cent and body mass index between 25 and 45. As in COVALENT-211, icovamenib is added on
top of the existing regimen, which continues throughout.

The trial opened on April 30, 2026, is listed as recruiting across 17 sites, and the company said on
August 24, 2026 that it expects enrollment to complete before year end, with Week 26 topline data in the
second quarter of 2027.

The population it targets is the one that gives the programme its commercial logic. Biomea cites
clinical research indicating that between twenty and forty per cent of people with type 2 diabetes treated
with GLP-1-based therapies do not reach adequate glucose control, defined as HbA1c below seven per cent,
and frequently progress to insulin. GLP-1 agonists are among the most widely prescribed drugs in the
world. An oral add-on that rescues the non-responders, without competing with the incumbent, would be a
different kind of asset from one that has to displace semaglutide or tirzepatide.

That is the argument, and the evidence supporting it at present is eleven patients in a prespecified
subgroup of a trial interrupted by a clinical hold, plus a mechanism that is coherent on paper. The
readout that tests it is due in the second quarter of 2027, one quarter after the company’s own stated
cash runway ends.

09 Type 1 Diabetes: Three Patients And A Terminated Registration

Type 1 diabetes is where the mechanism would matter most and where the evidence is thinnest.
COVALENT-112, registered as NCT06152042, was a Phase II randomised, double-blind trial of icovamenib
against placebo in adults with type 1 diabetes, with the primary endpoint set at mean change from baseline
in stimulated C-peptide area under the curve at Week 26. C-peptide is released in equal amounts to insulin
by working beta cells and is the standard way of measuring how much endogenous insulin production is left.

The trial is the one the June 2024 clinical hold hit hardest, with more than ninety per cent of the
target population unable to complete dosing. On ClinicalTrials.gov, checked on August 24, 2026, its status
is Terminated, with 37 participants enrolled across 11 locations and a primary completion date of
May 20, 2025.

The company nonetheless analysed the patients who did complete treatment and presented 52-week
follow-up data at the American Diabetes Association’s 86th Scientific Sessions in June 2026. As summarised
in the second-quarter release of August 5, 2026: a 52 per cent increase from baseline in mean C-peptide
area under the curve at Week 12 in patients diagnosed within nought to three years who received 200 mg,
with the effect largely preserved through Week 52, about a seven per cent decline from baseline, after the
twelve-week dosing period ended. That group is three patients. Preservation of C-peptide was also reported
in patients diagnosed between three and fifteen years earlier, nine patients. Icovamenib was described as
generally well tolerated across all dosing arms through Week 52.

Biomea has said it is planning a Phase II trial in recently diagnosed type 1 patients, those within
three years of diagnosis, in collaboration with four United States academic centres: the Barbara Davis
Center for Diabetes, the Joslin Diabetes Center, the diabetes division of the University of Texas Health
Science Center at San Antonio, and the University of Miami Diabetes Research Institute. The design would
evaluate extended dosing of six to twelve months at 200 mg and a possible combination with an
immunosuppressive agent. On August 5, 2026 the company expected the trial to begin in the second half of
the year. No initiation had been announced as of August 24, 2026, and the funding for it is not visible on
the current balance sheet.

10 BMF-650, The Oral GLP-1 And The Nearest Readout

BMF-650 is the only asset in the company that is not icovamenib. It is an oral, next-generation,
small-molecule GLP-1 receptor agonist for obesity, and its Phase I trial, called GLP-131 and registered as
NCT07223216, is a randomised, double-blind, placebo-controlled single and multiple ascending dose study in
otherwise healthy overweight or obese participants. The registry lists 80 participants at a single site,
a start date of October 23, 2025 and a status of active, not recruiting, checked on August 24, 2026.

On June 5, 2026 the company announced it had expanded the trial to evaluate a rapid one-step titration.
Titration is the practical bottleneck of the oral GLP-1 class: tolerability, mainly nausea and vomiting,
governs how fast a patient can reach an effective dose, and a shorter titration is a genuine point of
differentiation if the tolerability holds.

The company has guided to initial 28-day clinical weight reduction data from GLP-131 in the third
quarter of 2026, a guidance repeated in the release of August 5, 2026. That window closes on
September 30, 2026, which makes it the nearest company-guided event on the calendar. It is a window, not a
confirmed date: Biomea has not published a fixed day, and the registry’s own primary completion date for
the study reads May 2026, earlier than the guided data window.

That readout has narrow boundaries. Twenty-eight days of dosing in
healthy overweight volunteers produces a weight-change number and a tolerability profile. It does not
produce a competitive position against oral semaglutide or orforglipron, both of which are far further
advanced with far larger datasets. For a company with this balance sheet, its function is closer to a
financing event than to a clinical one: a good number widens the range of terms on which capital can be
raised.

11 OPAL, Leicester And The Semaglutide Combination

The third clinical strand is not run by Biomea and does not cost it a trial budget. On August 13, 2026
the company announced that the first participant had been dosed in a newly activated arm of OPAL, an
adaptive platform trial registered as ISRCTN10203365 and run by the University of Leicester and the
Leicester Diabetes Centre with the NIHR Biomedical Research Centre Leicester, led by Professor Dame
Melanie Davies and Professor Thomas Yates.

The arm evaluates icovamenib at 100 mg once daily for twelve weeks in combination with low-dose
semaglutide for twenty-four weeks, against low-dose semaglutide alone, in people without type 2 diabetes
who are overweight with a weight-related complication or obese. The company describes the arm as randomised and double-blind
and designed to enroll 64 participants one to one, with the primary endpoint assessed at Week 24; the
ISRCTN registration describes the OPAL platform itself as an open-label adaptive randomised trial, so the
blinding applies to the arm as the company has described it, not to the platform. The
endpoints go beyond weight: physical function, body composition, muscle health and metabolic markers.

The scientific rationale Biomea gives is that in preclinical work icovamenib enhanced semaglutide’s
effect, with weight reduction driven by fat loss and preservation of lean mass, and that it upregulates
GLP-1 expression, supports myogenesis and shifts adipose tissue metabolism towards energy expenditure.
Muscle loss under GLP-1 therapy is one of the live commercial questions in obesity treatment, and every
large developer in the field is working on it.

For an investor the relevant feature is the cost structure. An academic platform trial gives Biomea a
readout in the largest metabolic market in the world without funding a trial it cannot currently afford.
The corresponding limitation is that the company does not control the timetable, the analysis or the
communication of the result, and preclinical enhancement of semaglutide in animals has a long history of
not reproducing in people.

12 The Second Quarter Of 2026 In Numbers

Biomea reported its second quarter on August 5, 2026. There is no revenue line, and there never has
been. The figures below come from the condensed statement of operations filed with the quarterly report on
Form 10-Q for the period ended June 30, 2026.

Two lines carry most of the information. The net loss of $8.3 million is the smallest the company has
reported since before the 2024 hold, and $4.1 million of the improvement is not operating: it is the
revaluation of the warrant liability, a non-cash gain that arises because the share price fell. A company
whose reported loss shrinks when its stock falls is telling the reader something about its capital
structure, not about its business.

The other line is the share count. Weighted average shares went from 40.6 million in the second quarter
of 2025 to 72.4 million in the second quarter of 2026, up 78 per cent in a year. The loss per share
improved from $(0.51) to $(0.12). Holding the share count of a year ago and applying this quarter’s loss
gives $(0.21), so roughly a fifth of the improvement disappears once the extra shares are removed from the
calculation; run the other way round, this quarter’s share count against last year’s loss gives $(0.29),
and more than half of it does.

Cash used in operating activities was $21.697 million in the first half of 2026, against $44.879 million
in the first half of 2025. Financing brought in $159 thousand over the six months, entirely from option
exercises and employee stock purchases. No shares were sold under the at-the-market programme in the first
half of 2026.

13 How A $144 Million Year Became A $25.5 Million Half Year

The cost reduction is the part of the story management has executed cleanly, and the size of it is
easy to underestimate. Full-year 2025 operating expenses were $83.512 million, against $144.070 million in
2024 and $126.135 million in 2023. The first half of 2026 ran at $25.540 million, an annualised rate close
to $51 million.

Research and development, the dominant line, fell from $33.8 million in the first quarter of 2024 to
$9.1 million in each of the first two quarters of 2026. The company attributes the year-on-year decline of
$7.4 million in the second quarter to $1.5 million less on preclinical and exploratory programmes,
$1.3 million less on consultants and professional services, $1.0 million less on manufacturing,
$2.5 million less on personnel including stock compensation, and $1.7 million less on facilities, offset by
$0.6 million more on clinical activities. That last item is the only one moving in the direction of the
pipeline.

Headcount is the mechanism. The annual report records 41 full-time employees at December 31, 2025, of
whom 27 were in research and development. Stock-based compensation, which follows headcount and grant
levels, fell from $2.573 million in the second quarter of 2025 to $1.587 million in the second quarter of
2026. Facilities costs fell as laboratory space was given up; the company recorded a $0.510 million gain on
the sale of laboratory equipment in the first quarter of 2026.

The result is a company spending roughly $12.8 million a quarter to run two Phase II trials, a Phase I
trial and a corporate structure. That is efficient. It is also close to the floor: the trials themselves
are the largest remaining item, and cutting further would mean cutting the reason the company exists.

14 The Balance Sheet And The Going-Concern Statement

At June 30, 2026 Biomea held $35.193 million in cash, cash equivalents and restricted cash, against
$56.181 million at December 31, 2025. Total assets were $36.873 million, total liabilities $24.736 million
and stockholders’ equity $12.137 million. Additional paid-in capital stood at $481.921 million and the
accumulated deficit at $469.791 million. There is no financial debt.

The largest liability is not a payable. The warrant liability of $14.3 million is the fair value, under
a Black-Scholes measurement, of the common warrants issued with the June and October 2025 offerings. It is
remeasured at every balance sheet date, it will never be settled in cash, and it either converts into
equity if the warrants are exercised or disappears if they expire unexercised. Stripping it out, the
company owes roughly $10.5 million to trade creditors, landlords and employees, which is also the whole of
its current liabilities.

The going-concern language is the part that governs everything else. The second-quarter release states
that cash runway is projected into the second quarter of 2027. The quarterly report filed the same day
states that management believes that the existing financial resources are not sufficient to continue
operating activities for at least twelve months past the issuance date of the financial statements, and
concludes that there is substantial doubt about the Company’s ability to continue as a going
concern.

The two statements are consistent, and together they define the problem precisely. Twelve months from
August 5, 2026 is August 2027. A runway that reaches the second quarter of 2027 falls short of that by a
quarter or two, which is the technical trigger for the disclosure. The practical consequence is that
Biomea must raise capital before the COVALENT-211 readout produces the result that would determine what
that capital costs, or raise it immediately afterwards on whatever terms the result creates.

15 Dilution: From $30.00 In 2023 To Warrants Struck At $2.50

The dilution history is short, and reading it in order explains the current share price better than any
clinical argument.

Shares outstanding were 72,483,852 at June 30, 2026 and 72,299,440 at December 31, 2025, so 2026 itself
has been quiet: the movement of 184,412 shares in six months is option exercises and the employee purchase
plan. The dilution happened in 2025, at $2.00 and $2.05 a share, roughly one fifteenth of the April 2023
price.

The instruments that have not yet become shares are set out in Note 8 to the quarterly report. At
June 30, 2026 there were 13,184,249 stock options outstanding, 797,000 shares of restricted stock subject
to vesting, 38,499 shares issuable under the employee stock purchase plan and 37,024,389 common warrants:
51,044,137 in total, or a further 70 per cent on top of the shares in issue.

Two further pressures sit behind that figure. The 2021 incentive plan carries an evergreen provision
that increases the pool automatically each January by up to five per cent of the shares outstanding; on
January 1, 2026 it added 3,614,972 shares. And on August 11, 2025 the board repriced eligible options held
by non-executive employees and service providers to an exercise price of $4.50, which is itself now more
than three times the market price.

16 The Warrant Overhang And December 20, 2026

The warrant structure sets both a level and a date, and the two do not line up with the clinical
calendar.

Two tranches are outstanding, both struck at $2.50. The first, 23,000,000 warrants issued on
June 20, 2025, expires on December 20, 2026. The second, 14,024,389 warrants issued on October 8, 2025,
expires on October 8, 2028. Against the last confirmed close of $1.37 on August 21, 2026, both are well out
of the money; the company’s own Black-Scholes input for the share price at June 30, 2026 was $1.47.

If every warrant were exercised, Biomea would receive $92.6 million in cash and issue 37.0 million
shares. That is more money than the company has raised in the last two years combined, and it is the
cleanest financing path available to it. It requires the shares to trade meaningfully above $2.50, which
means it requires a clinical result, and the first Phase II result is not due until the first quarter of
2027 while the larger tranche expires on December 20, 2026.

The sequence therefore has a specific shape. Between now and December 20, 2026 the only company-guided
event that could plausibly move the shares above the strike is the BMF-650 28-day weight data in the third
quarter of 2026. If it does not, 23 million warrants expire worthless, the associated part of the warrant
liability is released into income as a non-cash gain, and the financing that carries the company to the
COVALENT-211 readout has to come from somewhere else: the at-the-market programme, a fresh offering, or a
partnership.

Warrant strikes are not price targets and nothing about the structure suggests where the shares will
trade. The point is narrower and mechanical: a large, dated block of potential capital sits just above the
current price, and its expiry date falls before the readout the company is being valued on.

17 Management, Board And Governance

Michael J.M. Hitchcock, who signs as Mick Hitchcock, has been interim chief executive of Biomea Fusion
since March 25, 2025, when the company announced that founder Thomas Butler was stepping down. Hitchcock
was already a member of the board. Erdtmann continues as chief operating officer and president, and
Thorsten Kirschberg was described as executive vice president of research in the company’s release of
August 13, 2026.

Seventeen months is a long time to be an interim chief executive, and the word still appears in the
company’s own filings: the segment note in the quarterly report of August 5, 2026 refers to the
Company’s interim chief executive officer as the chief operating decision maker. A permanent
appointment would normally precede, not follow, a pivotal data readout and the financing built around it.

There is no chief financial officer. The same quarterly report records the departures of both the
former chief executive and the former chief financial officer during 2025, and its signature page is
signed by the chief operating officer in the capacity of principal financial and accounting officer.
The document also names him twice in two different ways, as Ramses Erdtmann in the risk factors and as
Rainer Erdtmann on the signature page. For a company that has to negotiate a financing before its Phase II
readout, the absence of a permanent chief executive and of any chief financial officer is a structural
feature rather than a detail.

On governance, one episode is on the record. In March 2025 Biomea notified Nasdaq’s Listing
Qualifications Department that it was not in compliance with Listing Rule 5605(c)(2)(A), which requires an
audit committee of three independent directors, because its audit committee had two members. Compliance
was regained in July 2025 with the appointment of Julianne Averill to the board and the committee. The
matter is closed, and the company’s own risk factors note that continued compliance with Nasdaq
requirements is not assured.

Executive officers, directors and their affiliates beneficially owned approximately six per cent of the
outstanding voting stock at June 30, 2026, according to the same quarterly report; Finviz Elite put insider
ownership at 6.76 per cent on August 24, 2026, on a different definition and a different date.
Institutional ownership was 40.70 per cent on the same Finviz reading. Schedule 13G filings on the register at
that date include Heights Capital Management and CVI Investments, filed February 17, 2026, Blue Owl
Capital Holdings, filed February 12, 2026, and Janus Henderson Group, filed December 8, 2025. The first
two are the kind of holders that typically arrive with a structured financing rather than a long-term
position.

18 Market Snapshot And Analyst Coverage

The market data below were taken from Finviz Elite on August 24, 2026 while the session was still
running. The last confirmed close is the August 21 one; the August 24 figures are intraday and will not be
final until the close.

The company’s investor relations site lists eight covering analysts as of August 24, 2026: Yigal
Nochomovitz at Citi, Jonathan Wolleben at Citizens, Joseph Pantginis at H.C. Wainwright, Roger Song at
Jefferies, Anupam Rama at J.P. Morgan, Leland Gershell at Oppenheimer, Edward Tenthoff at Piper Sandler and
Michael King at Rodman & Renshaw. That is heavy coverage for a company of this size, and it is a
residue of what Biomea was in 2023 rather than a reflection of what it is now.

The aggregate target of $6.71 sits about five times above the last close. Wide gaps of that kind are
normal in binary-outcome biotech, where published targets are usually probability-weighted models of a
successful programme rather than forecasts of the next few months. The individual price targets and the
dates on which they were last revised do not appear on the company’s site, and the aggregate figure above
is Finviz’s own calculation.

The short position is the other structural feature. At 14.82 per cent of a 67.6 million share float and
a short ratio above twelve, the position is large relative to daily turnover, which is what produces days
like August 24 on news that changes no clinical fact.

19 Retail Sentiment On Stocktwits

Stocktwits is where the retail conversation about $BMEA happens, and the readings below are taken from
the platform’s public data on August 24, 2026. They are opinions posted by traders and non-professional
investors, not analyst research, and they are recorded here as a measure of positioning and mood rather
than as evidence about the company.

The pattern in the month to August 24 is easy to describe. The community score sat in the low forties
through the second-quarter release on August 5, which the stream largely ignored, dipped to 32 on
August 11, crossed the midpoint on August 14 and climbed steadily from there, reaching the top of the
platform’s scale on August 21 and staying there through the COVALENT-211 announcement. The rise began
before the news, on no company statement.

The recurring topics in the stream on August 24 were the short position and the $2.50 warrant strike,
with several posts treating the strike as a target level. Neither is a company statement, and the warrant
strike in particular is a term of a 2025 financing, not an estimate of value.

20 The Catalyst Map

Only one item on this map is a date. Everything else is a company-guided window, and windows move.

Read in order, the calendar has an awkward property: the two events that determine whether icovamenib
works arrive after the money runs out, and the one financing instrument already in place expires before
them. The BMF-650 window in the third quarter of 2026 is the only company-guided event that falls inside
both the runway and the life of the larger warrant tranche.

Quarterly reporting is the other recurring item. The third-quarter results would normally be published
in the first half of November 2026, on the pattern of the last four quarters, and no date had been
announced as of August 24, 2026.

21 Risks And Red Flags

The going-concern statement is not a formality. The quarterly report of August 5, 2026
states that existing resources are not sufficient for twelve months and that substantial doubt exists.
Biomea must raise capital, and the two results that would set its price arrive in 2027.

The evidence base is very small. The durability claim that supports both Phase II
trials rests on ten patients in one prespecified subgroup and eleven in another, from a trial whose dosing
was interrupted by a clinical hold and whose efficacy analysis covers 163 of 267 dosed patients. Small
prespecified subgroups with attractive p values reproduce less often than they fail to.

The same result has been published with different figures. The October 6, 2025 and
August 24, 2026 releases give different headline HbA1c reductions for the same two populations, as set out
above. This is a communication issue rather than a data issue, but it complicates any comparison with
competing programmes.

Dilution is structural, not hypothetical. 51.0 million potential new shares against
72.5 million outstanding, an at-the-market programme with room on it, an evergreen option pool that adds
up to five per cent a year automatically, and two 2025 financings priced at $2.00 and $2.05 that
established what the market pays for this asset when it needs money.

The type 1 programme is registered as terminated. COVALENT-112 carries the status
Terminated on ClinicalTrials.gov. The 52-week data presented in June 2026 come from three patients in the
lead group and nine in the second. The replacement trial has been described but not started, and it is not
funded on the current balance sheet.

The competitive field in obesity is brutal. BMF-650 is a Phase I oral GLP-1 agonist in
a class where two very large companies have late-stage or approved oral products and an established
manufacturing base. A 28-day weight number in healthy volunteers does not change that.

Leadership has been interim for seventeen months, the company has terminated the
oncology programmes it was founded on, and a Nasdaq audit committee deficiency was reported and cured in
2025.

Nothing is approved. No product, no revenue, no partnership with an upfront payment
disclosed. A $469.8 million accumulated deficit has bought a mechanism, a dataset of a few dozen patients,
and two trials that will report in 2027.

22 Merlintrader Health Score

The Merlintrader Health Score is a one-to-five reading of how robust a company looks over the next
twelve to eighteen months, built on five weighted pillars. It is a description of financial and
operational fragility, not a view on the share price, and it is not a recommendation to buy, sell or hold
anything.

Weighted, that gives a Merlintrader Health Score of 2.2 out of 5 as of August 24, 2026.
The shape is the mirror image of a mature company: the catalyst pillar is near the top of the scale and
both financial pillars are at the bottom. That combination describes an asset whose outcome is decided by
events rather than by its own resilience, and where the financing has to be solved before those events
arrive.

23 Scenarios

These are descriptions of paths the filings and the calendar make possible. They are not forecasts,
they carry no probabilities, and they are not advice.

The path that works. BMF-650 produces a competitive 28-day weight number with clean
tolerability in the third quarter of 2026. The shares move above $2.50, some part of the 23 million
warrants expiring in December is exercised, and the company takes in cash without a discounted offering.
That extends the runway past the COVALENT-211 readout in the first quarter of 2027, which then reports a
Week 26 primary endpoint consistent with the small-subgroup durability seen in COVALENT-111. In that
sequence Biomea becomes a partnering candidate with a mechanism nobody else in diabetes has, and the
financing question changes shape entirely.

The path that grinds. The BMF-650 data are unremarkable or slip past September 30.
The warrants expire unexercised on December 20, 2026. The company funds itself through the at-the-market
programme and a further offering at a price set by a share below $1.50, adding tens of millions of new
shares before the readout. COVALENT-211 reports something ambiguous in the first quarter of 2027, a
directional effect that does not reach the endpoint, and the company continues at a lower price with a
larger share count and the same going-concern language.

The path that ends it. COVALENT-211 misses at Week 26, or the financing fails before
it. On a mechanism whose entire premise is durability after dosing stops, a clean miss in a properly
powered, placebo-controlled trial removes the reason the company exists. In a business with no revenue, no
approved product and $12.1 million of equity, that outcome is not partial.

The path nobody models. A partner takes the type 2 programme before the readout, on
terms that reflect the balance sheet rather than the science. Pharmaceutical companies with metabolic
franchises have paid for mechanisms at this stage before, and a company that has to finance in 2026 is not
in a position to hold out for a good price.

24 Bottom Line

Biomea Fusion is a single-mechanism company with a genuinely unusual idea and almost no financial
margin for error. The idea is that twelve weeks of an oral menin inhibitor can restore enough beta-cell
function to improve glucose control for months after dosing stops. The evidence for it is a prespecified
subgroup of ten patients, another of eleven, and a type 1 signal in three patients, all from trials
disrupted by a 2024 clinical hold. The trials designed to test it properly are enrolled or nearly enrolled
and will report in the first and second quarters of 2027.

Everything financial argues the other way. Cash of $35.2 million at June 30, 2026, equity of
$12.1 million, an operating burn of $21.7 million in the half year, a going-concern statement in the
company’s own quarterly report, 51.0 million potential new shares against 72.5 million outstanding, and
two 2025 financings that priced the equity at around $2.00 when the company needed money. Management has
cut costs hard and well, and the cut is what makes 2027 reachable at all, but there is very little left to
cut that is not the pipeline itself.

The August 24, 2026 announcement removed enrollment risk from COVALENT-211 and fixed its timetable.
That is worth having. It is also, on its own, an operational milestone that a stock up twenty per cent on
the day has priced as something more. The two questions that decide this company, whether icovamenib works
in a properly powered trial and whether Biomea can fund itself until the answer arrives, are both still
open, and the second one has to be settled first.

Related Research On Merlintrader

• Biotech Stock Hubs — the full
  Merlintrader index of biotech and healthcare stock hubs, with the date each one was last verified.
• Weekly Market Pulse — the week ahead
  across the market, with the biotech catalysts, the earnings calendar and the watchlist.
• Top Ten Biotech Stocks Right Now
  — the names Merlintrader is watching most closely, updated regularly.
• Free Biotech Catalyst Calendar — PDUFA
  dates, advisory committee meetings and clinical readouts across the sector.