Spyre Therapeutics (Nasdaq: $SYRE) Stock Hub 2026: Three Readouts Down, Three To Go, And A Positive Trial The Company Chose Not To Advance

01 What Spyre Is In August 2026

Spyre Therapeutics, Inc. trades on the Nasdaq Global Select Market under $SYRE and is headquartered at
221 Crescent Street, Building 23, Suite 105, Waltham, Massachusetts 02453. It is a Delaware corporation with
Commission file number 001-37722 and employer identification number 46-4312787, and it reports as a single operating
segment. At June 30, 2026 it held $1,145.3 million of cash, cash equivalents and marketable securities, carried total
liabilities of $71.9 million with no financial borrowings, and had an accumulated deficit of $1.23 billion. The share
count on the cover of its most recent quarterly report is 88,173,762 common shares as of July 28, 2026.

The company has no approved product, no revenue line and, on the evidence of its own filings, no marketing
application pending with any regulator. What
it has is four investigational antibodies, three combinations built out of them, two Phase 2 trials that have been
reading out through 2026, a third that has just been initiated, and a balance sheet large enough that none of those
readouts is a financing event. That combination is unusual. Most companies with this much clinical news flow are
managing a cash position; Spyre is managing a sequence.

The sequence has a name inside the company: six proof-of-concept Phase 2 readouts in 2026, described in the
second-quarter release as remaining on track. Three of the six have now happened. SPY001 read out on April 13 and
was well received. SPY002 read out on June 15 and was well received. SPY072 in rheumatoid arthritis read out on
August 25 and was not: the company said the magnitude of effect did not meet its internal bar to prioritise
advancement of SPY072 as a monotherapy in that indication, and the shares fell more than twelve per cent the
following session from a close of $107.36.

Three remain. SPY003 in ulcerative colitis is expected in September 2026, and the psoriatic arthritis and axial
spondyloarthritis sub-studies of the same basket trial that produced the August disappointment are expected in the
fourth quarter. Every figure below is traced to the filing or release it came from and carries its date,
because on a company whose story changes four times a year the date attached to a number is part of the number.

02 From Aeglea To Spyre: What The Pivot Left On The Balance Sheet

Spyre did not grow into what it is. It was assembled. The legal entity was formed in Delaware on December 16, 2013
as Aeglea BioTherapeutics Holdings, LLC, converted to a corporation on March 10, 2015, and spent its first public
years as Aeglea BioTherapeutics, a rare metabolic disease company whose lead asset was pegzilarginase. On June 22,
2023 that public company acquired the assets of a privately held biotechnology company called Spyre Therapeutics,
Inc., and on November 27, 2023 it completed the rebranding and took the name.

Two features of that history still sit on the balance sheet and still matter to anyone reading the accounts.

The first is the contingent value right. In connection with the asset acquisition a non-transferable CVR was
distributed to holders of record at the close of business on July 3, 2023, entitling them to cash payments from
proceeds related to the disposition or monetisation of the legacy assets. At June 30, 2026 the CVR liability stood at
$31.35 million current and $7.27 million non-current, $38.62 million in total, which is more than half of everything
the company owes. In the same quarter Spyre determined that its remaining obligations under the CVR agreement no
longer qualify as derivatives, because all contingencies have been resolved, and moved to measuring the liability by
discounting the estimated future payments at its weighted average cost of capital.

The second is that the legacy asset is still generating money. Pegzilarginase was sold to Immedica Pharma AB in
July 2023 with milestone consideration attached. In the first quarter of 2026 Spyre booked a $30.0 million gain on
achieved milestones, and in the second quarter a further $40.0 million, specifically for Immedica’s sale of a priority
review voucher. That is $70.0 million over the half, credited inside operating expenses as a negative line, which is why the reported operating loss for the half is $87.3 million against $157.3 million of gross operating
spend. A reader who takes the loss at face value and infers the underlying spend from it will understate it by
seventy million dollars.

The money does not stay in the company either. Under the CVR agreement, milestone payments received by
July 3, 2026, net of allowable expenses and adjustments, are distributed to the legacy holders and reduce the CVR
liability. The first half already contains a $22.9 million payment to CVR holders inside financing activities, and
the fair value of the liability rose by $34.9 million over the same period, a movement that runs through other
expense and accounts for most of the $35.0 million of net other expense reported for the half. So the $70.0 million
credited above the operating line is largely matched by a charge below it and by cash leaving the company. The CVR
term ended on July 3, 2026, which closes the window on both.

The practical consequence of the 2023 pivot is that Spyre has no institutional history in immunology to appeal to.
The management team, the antibodies, the trial design and the capital were all put in place inside three years.
Track record here means what has been executed since 2023, not what the ticker did before.

03 Paragon, Parapyre And The May 29 Amendment To The SPY003 Licence

Every antibody in the pipeline came from the same place. Paragon Therapeutics, Inc. is an antibody discovery
company backed by Fairmount, the fund that is also Spyre’s largest holder, and it is a related party in Spyre’s
filings. SPY001, SPY002, SPY072 and SPY003 were each licensed from Paragon under separate agreements, with option
exercises, milestone payments and sublicensing fees running through the related-party note. In the six months to
June 30, 2026 Spyre paid Paragon $3.0 million of milestone payments and recorded $3.1 million of related-party
research and development expense; at June 30 no milestone payments were outstanding and payable.

Paragon also holds equity exposure through Parapyre Holding LLC. At June 30, 2026 there were 1,532,591 warrants
previously issued to Parapyre outstanding, at a weighted-average exercise price of $22.49.

On May 29, 2026 the two parties amended the SPY003 licence. The terms of that amendment set out what Spyre is
allowed to do with its IL-23 antibody. Before the amendment the Field
was limited to inflammatory bowel disease. The amendment expands the Field to encompass all therapeutic,
prophylactic, palliative and diagnostic uses, subject to a restriction the filing calls the Monotherapy Dosing
Restriction: Spyre will not dose a human patient in a clinical trial of SPY003 for an indication outside inflammatory
bowel disease as part of a combination until June 1, 2028, or as a monotherapy until June 1, 2030.

There is a clause inside the clause. If Spyre, or a licensee of Paragon’s retained rights, consummates a material
transaction, including a change of control, then any remaining restrictions outside the original Field survive only
until June 1, 2028 and then terminate. In plain terms, a takeover shortens the leash on SPY003 by two years. That is
a contractual detail rather than an intention, and Spyre has never said it is for sale, but it is the kind of term
that gets negotiated when both sides have thought about the possibility.

04 The Platform: Half-Life Extension And The Coformulation Thesis

The technical claim underneath everything Spyre does is narrow and repeatable. Take an antibody against a target
that has already been validated in the clinic by someone else. Engineer it for extended half-life. Formulate it at
high concentration so it can be delivered subcutaneously. Then dose it far less often than the incumbent, and, where
the biology supports it, co-formulate two of them into one injection.

The four antibodies map onto three validated targets and one new one. SPY001 targets α4β7, the integrin
that Takeda’s vedolizumab has been hitting commercially for a decade. SPY002 and SPY072 are two distinct antibodies
against TL1A, an upstream cytokine implicated in chronic inflammation and fibrosis, and the target that has drawn the
largest immunology deals of recent years. SPY003 targets the p19 subunit of interleukin-23, the pathway behind
risankizumab and ustekinumab. SPY007, disclosed on August 10, 2026, targets interleukin-17A and 17F and is still
preclinical.

The dosing ambition attached to this engineering is stated in the company’s own forward-looking language: a
potential quarterly to twice-yearly subcutaneous maintenance profile, written in the filings as Q3M to Q6M. In the
June release the SKYLINE study lead described SPY002’s target as a Q3-6M subcutaneous maintenance profile. That is
the differentiation claim. It is not a claim about curing anything; it is a claim about how often a patient with a
chronic condition has to inject.

The combinations are where the thesis becomes ambitious. SPY120 is SPY001 with SPY002, α4β7 plus TL1A.
SPY130 is SPY001 with SPY003, α4β7 plus IL-23. SPY230 is SPY002 with SPY003, TL1A plus IL-23. SPY772,
announced on August 10, 2026, is SPY072 with SPY007, TL1A plus IL-17A/F. The company has published preclinical work
for SPY120 showing that combined inhibition of TL1A and α4β7 outperformed either monotherapy in mouse
models of colitis, and that the pharmacokinetic profiles of SPY001 and SPY002 were similar in non-human primates
whether dosed alone or together. Similar preclinical enhancement has been described for SPY130 and SPY230.
Preclinical is preclinical: no combination has produced a human efficacy result, and the first ones are not expected
before 2027.

05 SPY001: The April Readout, Number By Number

On April 13, 2026 Spyre furnished a Form 8-K reporting initial twelve-week induction topline data from Part A of
the Phase 2 SKYLINE trial for SPY001 in moderately-to-severely active ulcerative colitis. The primary endpoint is
change from baseline in the Robarts Histopathology Index at Week 12, a scored measure of how much inflammation a
pathologist can still see in a biopsy.

SPY001 reduced that index by 9.2 points, with a p-value below 0.0001. Clinical remission by modified Mayo Score
was 40 per cent. Endoscopic improvement was 51 per cent. The change in modified Mayo Score from baseline was
-3.7 points. The safety table covered 43 dosed participants: six subjects with any adverse event, that is 14 per
cent; one severe grade three or higher event, 2 per cent; zero drug-related adverse events; zero discontinuations for
adverse events; one serious adverse event, deemed not drug-related; zero drug-related serious adverse events; zero
adverse events of special interest; zero deaths. The single serious event is described in a footnote as chest pain in
a 68-year-old man with coronary artery disease, angina, type 2 diabetes, hypertension and hypercholesterolaemia, who
was ruled out for myocardial infarction. The most common adverse event occurring in at least two patients was back
pain.

The same filing carried two operational announcements that mattered nearly as much. Recruitment for Part A was
closed, and enrolment was open for Part B, which includes three monotherapy cohorts and three combination cohorts
randomised against a shared placebo. And the remaining Part A readouts were dated: SPY002 mid-2026, SPY003 in the
third quarter, Part B induction data across all cohorts in 2027.

Three days later, on April 16, 2026, the company closed an underwritten public offering of 7,475,000 shares at
$62.00 per share, including full exercise of the underwriters’ option, for gross proceeds of approximately
$463 million and net proceeds of $435.2 million. The sequencing is the point: data announced on April 13, underwriting agreement
signed on April 14, over-allotment exercised in full on April 15, closing on April 16. That is a company that had the shelf ready.

06 SPY002: The June Readout, Number By Number

On June 15, 2026 the second Part A cohort read out, and it read out well. SPY002, the anti-TL1A antibody, reduced
the Robarts Histopathology Index by 10.7 points at Week 12, again with a p-value below 0.0001. Clinical remission was
33 per cent, endoscopic improvement was 42 per cent, and the change in modified Mayo Score was -3.7, identical to
SPY001.

The June release did something the April one did not: it described the population. Thirty-five per cent of
participants were advanced-therapy exposed. Mean disease duration was 7.0 years. Mean baseline Robarts score was
16.9 with a standard deviation of 8.5, mean baseline modified Mayo Score was 6.9 with a standard deviation of 1.0,
and 56 per cent had a baseline endoscopy sub-score of 3, which is the most severe grade. That is a moderately sick
population with a third of it having already failed advanced therapy, and it is useful context for the response
rates.

Safety in the SPY002 cohort was noisier than in SPY001, from a larger group. Across 48 participants there were
twenty subjects with treatment-emergent adverse events, 41.7 per cent; two severe grade three or higher events,
4.2 per cent; three drug-related adverse events, 6.3 per cent, being one case each of nausea, hypertension and
arthralgia; two adverse events leading to drug discontinuation, 4.2 per cent; two serious adverse events, 4.2 per
cent, both deemed not drug-related; zero drug-related serious adverse events; zero adverse events of special
interest; zero deaths. The two serious events were a hospitalisation for exacerbation of ulcerative colitis and a
hospitalisation for worsening heart failure in a subject with a history of heart failure and atrial fibrillation, who
was subsequently diagnosed with worsening aortic stenosis.

The company’s framing was that SPY002 delivered Phase 2 results within one year of Phase 1 results. That is an
operational claim, and it is verifiable: the Phase 1 study of SPY002-091 in healthy participants, NCT06672718, began on
November 26, 2024. Whether the efficacy is best in class is not verifiable from an open-label single-arm cohort, and
the release describes the numbers as among the highest reported in ulcerative colitis rather than as superior to any
named comparator.

07 How SKYLINE Is Actually Built: Part A Against Part B

SKYLINE is registered as NCT07012395, a Phase 2 platform trial with a planned enrolment of 645 participants across
267 listed locations. It started on May 27, 2025 and its primary completion date is given as June 2027. Its full
registered title describes it as a platform trial to assess long-acting antibodies as single agents and in
combinations for moderately to severely active ulcerative colitis. A separate long-term extension study,
NCT07652294, opened on June 8, 2026 with the same 645-participant frame and a primary completion date of
March 2, 2029.

The trial has two halves and they are not the same kind of experiment. That distinction governs how every number
published so far can be read.

Part A is an open-label assessment of the safety and preliminary efficacy of a single dose level of each
investigational monotherapy. There is no placebo arm, no randomisation and no blinding. Every participant knows what
they received and so does every investigator. The 40 per cent clinical remission rate for SPY001 and the 33 per cent
for SPY002 are therefore not treatment effects; they are response rates in treated patients, with no control to
subtract. Ulcerative colitis is an indication in which randomised induction studies
routinely record a material placebo response, so the missing subtraction is not a technicality. The size of that
response in these particular cohorts is unknown, because Part A did not measure it.

Part B is the randomised, placebo-controlled assessment. It evaluates monotherapies at two dose levels and the
three combinations against a shared placebo, and it is designed to deliver dose-ranging data on the monotherapies,
proof-of-concept for the combinations, and contribution-of-components analysis, which is the test of whether a
combination does more than its parts. Induction data across all cohorts is expected in 2027. Enrolment into Part B
opened in April 2026 and the company reported in August that it is currently enrolling.

So the entire Spyre efficacy story in inflammatory bowel disease, as of today, rests on uncontrolled data. The
first controlled read from SKYLINE is more than a year away. The only controlled read the company has published on
any antibody is the one that came out on August 25 in a different disease, and it is the one that disappointed.

08 SPY003 And The September Readout

SPY003 is the third monotherapy and the last piece of the Part A set. It is a highly potent and selective
investigational monoclonal antibody targeting the p19 subunit of interleukin-23, engineered with the same half-life
extension technology and formulated at high concentration for infrequent subcutaneous maintenance dosing. Its Phase 1
single-dose study in healthy participants, NCT06873724, enrolled 59 subjects, started on March 10, 2025 and carries a
primary completion date of November 18, 2026.

The readout timing has tightened as the year has gone on. In April the company said third quarter of 2026. In the
June release the study lead said Part A data for SPY003 would come in the third quarter, and described it as the
result that, if successful, would complete proof-of-concept for the investigational monotherapy components. The
second-quarter release of August 4 narrowed it to September, and the catalyst table furnished with the
August 25 Form 8-K says September 2026 outright.

What that readout can and cannot settle follows directly from the previous section. It will be an open-label,
uncontrolled cohort, on the same design as SPY001 and SPY002, so a strong number will not prove superiority over
risankizumab or any other IL-23 inhibitor. What it can do is complete or break the platform claim. Spyre’s entire
argument for combinations is that it owns three optimised monotherapy components. Two of the three have produced
numbers the market liked. If the third does not, the combination programme that Part B is built around loses one of
its two legs in SPY130 and SPY230, and the story narrows to α4β7 plus TL1A.

It also arrives within weeks of a disappointment, into a share price that has given back some of the year’s gain,
with short interest at 14.88 per cent of a 79.01 million share float. Those are the conditions in which a share price
tends to move by more than the content of a readout on its own would explain, in either direction.

09 SKYWAY And The August 25 Rheumatoid Arthritis Result

SKYWAY is registered as NCT07148414, a Phase 2 randomised, placebo-controlled basket trial of SPY002-072, the
compound the company calls SPY072, in adults with moderately to severely active rheumatologic disease. Planned
enrolment is 285 participants across 64 locations; it carries a start date of August 21, 2025 on the registry, against
the company’s own description of initiation in September 2025, and a primary completion date of October 31, 2026. It has three sub-studies, one each in rheumatoid arthritis, psoriatic arthritis and axial
spondyloarthritis, and the company reported in the second-quarter release that all three completed enrolment on or
ahead of schedule, with all three over-enrolling.

The rheumatoid arthritis sub-study is double-blind and placebo-controlled, tests two dose levels of SPY072 at
Week 12 with open-label follow-up to Week 36, and has as its primary endpoint the change from baseline to Week 12 in
the Disease Activity Score in 28 joints with C-reactive protein. The key secondary endpoint is the proportion of
patients achieving an ACR20 response at Week 12.

On August 25, 2026 Spyre reported the result. Both doses demonstrated statistically significant benefits compared
with placebo on one or more of the primary, key secondary and exploratory endpoints. On the primary endpoint the low
dose produced a change in DAS28-CRP of -1.9 against -1.3 for placebo, with p below 0.05; the high dose produced -1.5,
which is not marked as significant. On ACR20 the high dose reached 63 per cent against 43 per cent for placebo, with
a nominal p below 0.05. On ACR50 the low dose reached 38 per cent against 19 per cent, again nominal. ACR70 was
13 per cent for the high dose, 4 per cent for the low dose and 2 per cent for placebo. Arms were 48, 48 and 47
participants. Results were described as generally comparable between advanced-therapy-naive and
advanced-therapy-experienced sub-groups.

Pharmacology was clean. Both doses achieved target drug concentrations and provided complete and durable
suppression of free TL1A through Week 12, which the company reads as complete target engagement. Safety was clean
too: adverse event rates were 27 per cent on active against 36 per cent on placebo, generally mild or moderate; one
serious treatment-emergent adverse event occurred on each arm, neither deemed drug-related; the one death in the
study occurred in a participant receiving placebo; the most common treatment-emergent adverse events were infections
and infestations, in 14 per cent of SPY072 participants against 15 per cent on placebo.

And then the sentence that moved the stock: the results provide proof-of-mechanism for TL1A in rheumatoid
arthritis and support its potential in other autoimmune diseases and as a combination component, but did not meet the
company’s internal bar to prioritise advancement of SPY072 as a monotherapy in rheumatoid arthritis.

10 What An Internal Bar Means, And What It Does Not

The August 25 result contains three statements of different kinds, and the market reaction blended them.

The first is statistical. SPY072 hit its pre-specified primary endpoint at the low dose, with a real p-value below
0.05 against a real placebo arm. That is a positive trial on its own terms. Two further values in the table, ACR20 on the high dose
and ACR50 on the low dose, are marked nominal, which means the p-value was not adjusted for multiplicity and cannot
be read as a confirmatory finding. The remaining values, including the whole ACR70 row, carry no marker at all.

The second is about dose response, and it is the part the company did not comment on. In a well-behaved dataset
the higher dose does better. Here the low dose owns the primary endpoint and ACR50, while the high dose owns ACR20
and ACR70. Both doses achieved complete and durable suppression of free TL1A through Week 12, which means the target
was fully engaged at both levels and the difference cannot be explained by insufficient exposure. When target
engagement is complete at the lower dose, the higher dose has nothing left to add, and the scatter across endpoints
is what a ceiling on the mechanism looks like. That is the biological reading of the table, and it is less
comfortable than the statistical one.

The third is commercial, and it is the company’s own. An internal bar is not a regulatory standard and not a
statistical threshold. It is a judgement about whether a product could compete, and in rheumatoid arthritis the competition is TNF inhibitors, JAK inhibitors and IL-6 blockade, all
long established and with decades of prescriber habit behind them. A 63 per cent ACR20 with a nominal p-value against
a 43 per cent placebo did not clear the standard Spyre set for itself, and the company said so rather than starting a
Phase 3 on it.

What the result does not do is invalidate TL1A in inflammatory bowel disease. SPY002 is a different antibody in a
different disease with a different mechanism of benefit, and its Part A numbers stand where they were. Nor does it
end SPY072: the company reads the clean safety profile and the demonstrated activity as increased conviction that
long-acting TL1A antibodies work as combination components, which is the role it has now assigned SPY072 in
SPY772.

It does, however, remove the only near-term path by which SPY072 could have become a product on its own. Every
remaining route for that molecule now runs through a combination, and combinations read out in 2027 and 2028.

11 What Is Left Of SKYWAY: Psoriatic Arthritis And axSpA

Two of the six 2026 readouts still belong to SKYWAY. The psoriatic arthritis sub-study is double-blind and
placebo-controlled, tests a single dose level of SPY072 at Week 16 with open-label follow-up through Week 40. The
axial spondyloarthritis sub-study has the same shape. Both are expected in the fourth quarter of 2026, a date the
company repeated in the catalyst table furnished with the August 25 filing, after the rheumatoid arthritis result was
already known.

Three details make these harder to handicap than they look.

They test a single dose, not two. Whatever the rheumatoid arthritis data suggest about a plateau in the mechanism,
these sub-studies cannot explore it; they will produce one number per indication against placebo.

They read at Week 16, not Week 12. That is four extra weeks of exposure and a different endpoint window from the
sub-study that has already reported, so the results are not directly stackable against the rheumatoid arthritis
table.

And the internal bar is now visible. Before August 25 a reader could not know what standard Spyre applies to its
own data. Now there is a worked example, and it was applied to a trial that was statistically positive. That cuts
both ways: it raises the probability that a mediocre psoriatic arthritis number is called mediocre by the company
rather than dressed up, and it lowers the value of a result that merely beats placebo.

The registered primary completion date for the whole SKYWAY trial is October 31, 2026, which is consistent with
fourth-quarter topline data and is the only date in this section that comes from a registry rather than from the
company.

12 SPY772, SPY007 And The SKYLIGHT Trial

On August 10, 2026 Spyre announced the launch of SPY772, a new combination programme comprising SPY072, the
long-acting anti-TL1A antibody, and SPY007, a novel long-acting anti-IL-17A/F antibody in preclinical development.
Alongside it the company announced the initiation of SKYLIGHT, a randomised Phase 2 proof-of-concept trial evaluating
dual inhibition of TL1A and IL-17A/F in hidradenitis suppurativa.

The trial design contains a wrinkle that is easy to miss and changes what the result will mean. SKYLIGHT does not
test SPY772. It is registered as NCT07766005, expects approximately 150 adults with moderate-to-severe hidradenitis
suppurativa, and randomises participants to receive background bimekizumab therapy plus either SPY072 or placebo. In
other words the IL-17A/F half of the combination is supplied by an approved third-party drug, not by SPY007, which is
still preclinical. The primary endpoint is the proportion of patients achieving a HiSCR75 response at Week 16,
meaning at least a 75 per cent reduction from baseline in total abscess and inflammatory nodule count with no
increase in abscess or draining-tunnel count. HiSCR50 and treatment-emergent adverse events through 52 weeks are
secondary. The registry lists a September 2026 start, twelve United States sites, a primary completion date of
October 2027 and a study completion date of June 2028.

The clinical development lead described the logic without ambiguity: SKYLIGHT will evaluate whether adding SPY072
to bimekizumab can improve outcomes beyond bimekizumab alone, and positive results would support advancement of the
long-acting SPY772 coformulation. So SKYLIGHT is a mechanism test that de-risks a product that does not yet exist,
using a competitor’s approved antibody as the second component. If it works, Spyre still has to develop SPY007
through preclinical work and its own first-in-human study before SPY772 becomes a candidate. Topline is expected late
2027 or early 2028.

The August 25 filing added one more thing to the SPY007 story by omission: SPY007 does not appear in the catalyst
table at all. It is a preclinical asset with no disclosed timeline of its own.

13 The Second Quarter In Numbers

Spyre reported second-quarter 2026 results on August 4, 2026, alongside the quarterly report on Form 10-Q for the
period ended June 30.

Research and development expense was $65.5 million against $40.1 million in the second quarter of 2025, an
increase the company attributes to higher manufacturing and clinical trial expenses and higher headcount costs.
General and administrative expense was $16.1 million against $11.8 million, driven by headcount. Against those the
company credited a $40.0 million gain on sale of in-process research and development for a milestone achieved on the
2023 pegzilarginase disposal, specifically Immedica’s sale of a priority review voucher, which is booked as a
negative operating expense and reduces total operating expenses to $41.6 million.

Below the operating line, interest income was $10.1 million against $5.9 million a year earlier, which is what a
billion dollars in treasuries and commercial paper produces. Other expense net was $4.6 million. Total other income
was $5.4 million. Net loss for the quarter was $36.2 million, essentially flat against $36.7 million in the second
quarter of 2025, and loss per common share was $0.36 on 84,729,435 weighted-average shares.

That flat net loss is an accounting artefact. Gross operating spend rose from
$51.9 million to $81.6 million year on year, fifty-seven per cent, while the reported loss did not move, because the
milestone gain and the larger interest income filled the gap. For the half, research and development was
$125.9 million against $81.8 million, general and administrative was $31.4 million against $23.7 million, the gain
was $70.0 million against $10.0 million, and the net loss was $105.2 million against $81.5 million. Loss per common
share for the half was $1.09, the same as the prior year, on a share count that grew by more than twenty million.

The line to watch in future quarters is the gain. It is milestone-driven, it relates to an asset the company no
longer owns, and there is no disclosure suggesting it recurs at this scale. At June 30, 2026 a further $3.0 million
remained outstanding and due from a previously achieved milestone.

14 Cash, Burn And The Runway Into The Second Half Of 2029

At June 30, 2026 Spyre held $162.0 million of cash and cash equivalents and $983.3 million of marketable
securities, $1,145.3 million in total. Total assets were $1,173.0 million, all of it current. Total liabilities were
$71.9 million, of which $38.6 million is the contingent value right liability and $5.7 million accounts payable.
There is no debt. Total stockholders’ equity was $1,101.1 million against $715.2 million at December 31, 2025.

Net cash used in operating activities was $127.2 million for the half, against $87.6 million for the first half of
2025. Split by quarter that is $57.4 million in the first and $69.9 million in the second, a sequential increase of
twenty-two per cent that tracks the trial expansion.

The company’s stated runway is into the second half of 2029. That statement first appeared with the first-quarter
results on May 5, 2026, when it replaced the previous guidance of into the second half of 2028, and it was repeated
on August 4. The upgrade is entirely explained by the April offering.

The arithmetic behind that statement runs as follows. At the second-quarter burn rate of $69.9 million,
$1,145.3 million lasts about sixteen quarters, which reaches into 2030. At a rate that keeps growing at the pace of
the last four quarters it lasts considerably less. Neither number is the company’s forecast, and the company’s own
projection sits between them, which is the normal place for it to sit. What matters for a reader is that this is one
of the few clinical-stage companies of its size where the next three readouts do not have to be financed, and where a
bad readout does not immediately become a financing problem.

The offsetting fact is that Spyre has never behaved like a company that waits until it needs money. It raised
$316 million in October 2025 and $463 million in April 2026, both into strength, and on June 26, 2026 it put a
$500 million at-the-market programme in place that was entirely undrawn at quarter end. Companies do not sign
$500 million sales agreements they intend to leave unused.

15 The Capital Structure, Instrument By Instrument

At June 30, 2026 there were 88,064,133 common shares outstanding, against 78,189,811 at December 31, 2025, an
increase of 12.6 per cent in six months. The cover of the quarterly report gives 88,173,762 as of July 28, 2026.
Authorised common stock is 400,000,000 shares, so authorisation is not a constraint.

Series A non-voting convertible preferred stock: 1,086,341 authorised, 346,045 issued and outstanding at both
June 30, 2026 and December 31, 2025, carried at $146.4 million. Those 346,045 shares are convertible into 13,841,800
common shares, a ratio of forty to one. The Series A has no liquidation preference, no voting rights in the ordinary
course, and dividend rights only on an as-converted basis if common dividends are paid; it does carry protective
provisions over amendments to its own terms and over fundamental transactions while at least thirty per cent of the
original issue remains outstanding.

Series B non-voting convertible preferred stock: 271,625 authorised, none outstanding at June 30, 2026 against
16,667 at December 31, 2025. The last block converted on June 23, 2026 into 666,680 common shares at the option of
the holder. Across the life of the instrument, since stockholder approval of the conversion proposal, all 271,625
Series B shares have converted into 10,865,000 common shares, the same forty-to-one ratio.

Warrants: 1,532,591 Parapyre warrants outstanding at a weighted-average exercise price of $22.49. Equity plans: at
June 30, 2026 the 2016 Plan had 15,084,463 shares available for future issuance, of which 6,876,693 were subject to
outstanding option and restricted unit awards, and the 2018 inducement plan had 6,481,956 available, of which
5,981,072 were subject to outstanding awards. Outstanding awards therefore total 12,857,765 shares, and the two plans
together hold a further 8,708,654 shares that are authorised but not yet granted. The 2016 Plan also carries an
evergreen provision that automatically adds five per cent of issued and outstanding shares, including shares
underlying pre-funded warrants and non-voting convertible preferred, every January 1 unless the board approves
less.

Stockholders also approved an amended and restated 2016 Employee Stock Purchase Plan at the annual meeting on
May 27, 2026, by 66,885,864 votes to 45,542.

16 Two Point One Billion Dollars, And What It Was Raised At

Since inception and through June 30, 2026 Spyre has raised approximately $2.1 billion of gross proceeds from the
sale and issuance of convertible preferred stock and common stock, pre-funded warrants, grant proceeds and the
licensing of pegzilarginase rights in Europe and parts of the Middle East. Accumulated deficit at the same date was
$1.23 billion. Those two figures do not net to the cash balance: the deficit includes non-cash charges and the CVR
dividend recorded in 2023, so the $1,145.3 million on the balance sheet is not the difference between them.

The equity raises since the pivot, in the company’s own words in the risk factors, run as follows. In December
2023 it sold 6,000,000 common shares and 150,000 Series B preferred shares in a private placement for gross proceeds
of approximately $180 million. In March 2024 it sold 121,625 Series B preferred shares in a private placement for
gross proceeds of approximately $180 million. In October 2025 it sold 17,094,594 common shares in an underwritten
public offering for gross proceeds of approximately $316 million. In April 2026 it sold 7,475,000 common shares in an
underwritten public offering for gross proceeds of approximately $463 million.

The at-the-market history is smaller and more revealing about behaviour. The 2024 sales agreement with TD
Securities allowed up to $200.0 million. Under it the company sold 445,668 shares during 2025 for net proceeds of
$14.8 million and 296,396 shares during the first half of 2026 for net proceeds of $9.7 million, all of the latter in
the first quarter. That agreement was terminated on June 26, 2026 and replaced the same day with a 2026 sales
agreement for up to $500.0 million, of which the entire $500.0 million remained available at June 30, 2026. The
replacement went in alongside an automatic shelf registration statement filed on the same date.

One inconsistency appears in the company’s own documents. The April offering’s net
proceeds are given as $435.3 million in the first-quarter release of May 5 and as $435.2 million in the
second-quarter release and the quarterly report. The difference is a rounding of one hundred thousand dollars on a
figure of four hundred and thirty-five million, and it does not change anything, but the later filing is the one to
use.

17 June 23: Fairmount Converts, Then Sells 4.68 Million Shares

The single largest capital-markets event of the summer was not a company action. On June 23, 2026 Fairmount
Healthcare Fund II L.P. converted its remaining 16,667 Series B preferred shares into 666,680 common shares and, the
same day, sold 4,684,781 common shares at $85.31. The Form 4 reports the position in common stock held indirectly
through Fund II as zero after the transaction.

At $85.31 a share that sale raised roughly four hundred million dollars, and it was executed ten weeks after the April offering was priced at $62.00, and nine weeks after that
offering closed.

Fairmount did not exit. The Schedule 13D amendment filed the same day reports Fairmount Funds Management LLC and
Fairmount Healthcare Fund II each beneficially owning 8,835,440 shares, or 9.15 per cent, consisting entirely of
common stock issuable on conversion of 220,886 Series A preferred shares, and explicitly excluding the Series A
shares held above a 9.99 per cent beneficial-ownership limitation. The percentage is calculated on 96,572,019 shares
outstanding as of June 23, 2026, a denominator the filing builds from 86,841,253 shares reported in the May 10-Q,
the 666,680 conversion shares, 228,646 shares underlying near-term exercisable options, and the 8,835,440 conversion
shares themselves.

Two things follow. Fairmount converted the liquid instrument and sold it while retaining the illiquid one, which
is a rebalancing rather than a verdict, and the blocker means its stated 9.15 per cent understates its economic
exposure. But a founding holder selling four hundred million dollars of stock at $85.31, against the $62.00 at
which the company itself had sold shares ten weeks earlier, is a fact, and it is one of the reasons insider
transactions read -35.58 per cent on Finviz’s rolling measure.

18 Ownership, The Board Seat And The Annual Meeting

The annual meeting was held on May 27, 2026. Stockholders elected three Class I directors to serve until the 2029
annual meeting: Mark McKenna with 65,608,709 votes for and 1,329,489 withheld, Cameron Turtle with 66,176,245 for and
761,953 withheld, and Laurie Stelzer with 56,954,664 for and 9,983,534 withheld. Broker non-votes were 1,500,346 on
each. The advisory vote on executive compensation passed with 62,075,459 for and 4,760,373 against. KPMG LLP was
ratified as auditor for the year ending December 31, 2026 with 68,430,566 for and 670 against.

The Stelzer vote is the one that stands out. Nearly ten million withheld votes against fifty-seven million for is
a fifteen per cent withhold rate, against roughly one per cent for Turtle and two per cent for McKenna. Withheld
votes in an uncontested election carry no legal consequence, but the dispersion says something about how at least one
large holder or proxy adviser viewed that seat.

Effective the same day, Peter Harwin resigned from the board. The Form 8-K states the resignation was not the
result of any disagreement with the company on any matter relating to its operations, policies or practices, and that
the board was reduced from eight directors to seven. Harwin is a Fairmount co-founder, and the Form 4 filed by
Fairmount in June still lists 406,038 shares held by Peter Harwin and the same number by Tomas Kiselak as indirect
positions.

Institutional ownership on Finviz’s measure reads 100.82 per cent of the float, which is a mechanical artefact of
a float of 79.01 million shares against a share count that includes convertible preferred; it should be read as
saying that essentially the whole tradeable share count is in institutional hands, not as a literal percentage.
Insider ownership reads 10.39 per cent. The largest disclosed holders in the 2026 proxy were Fairmount Healthcare
Fund II at 9.99 per cent, FMR LLC at 9.71 per cent and BlackRock at 5.05 per cent, with executive officers and
directors as a group at 15.69 per cent. Two Schedule 13G amendments were filed on August 14, 2026.

19 Management

Five names carry the story in the company’s own filings and releases.

Cameron Turtle, DPhil, Chief Executive Officer. He signs the Form 8-K filings, he was re-elected to the
board on May 27, 2026 with the smallest withhold rate of the three nominees, and at 2,467,741 shares, or 3.07 per cent, he holds the largest personal
stake among the executive officers in the 2026 proxy. Two directors, Peter Harwin and Tomas Kiselak, are each shown
at 8,787,215 shares and 10.60 per cent, but those are the Fairmount holdings attributed to them as managing members,
and both disclaim beneficial ownership except for any pecuniary interest. His public framing after the August result was to
separate the mechanism from the indication: the data do not lead the company to prioritise SPY072 as a monotherapy in
rheumatoid arthritis, but the safety profile alongside activity in inflammatory bowel disease, hidradenitis
suppurativa and now rheumatoid arthritis increases conviction in TL1A as a combination component. One qualification
belongs with that quotation: Spyre has no efficacy data of its own in hidradenitis suppurativa. SKYLIGHT had not
begun enrolling when the statement was made and its topline is not expected before late 2027.

Sheldon Sloan, M.D., Chief Medical Officer. He fronted the SPY772 announcement on August 10 and made the
strategic argument in one sentence: combination therapy has improved efficacy without apparent safety downsides in
multiple autoimmune conditions, and long-acting coformulations against validated targets look increasingly likely to
deliver indication-leading profiles. That is the whole company thesis stated by the person responsible for proving
it.

Deanna Nguyen, M.D., SVP of Clinical Development and SKYLINE study lead. She delivered the SPY002
commentary in June, including the Q3-6M maintenance target and the statement that optimised monotherapy components
are the foundation for potentially best-in-class combinations.

Josh Friedman, M.D., Ph.D., SVP of Clinical Development and SKYLIGHT study lead. A second study lead named
in August, which is the clearest signal in the disclosure that the clinical organisation has been built out to run
more than one programme at a time.

Scott Burrows, Chief Financial Officer. He signs the quarterly report on Form 10-Q as principal financial
officer and principal accounting officer, and the 2026 proxy lists him as a named executive officer who joined in
September 2023, having previously been chief financial officer of Arcutis Biotherapeutics. He is named in no press
release between April and August 2026, which is why the clinical voices carry the public communication.

The investor relations contact on every release is Eric McIntyre, SVP of Finance and Investor Relations. On the board, the three Class I directors elected in May are Mark McKenna, Cameron Turtle and
Laurie Stelzer; Peter Harwin resigned the same day, reducing the board from eight to seven.

20 Market Snapshot And Analyst Coverage

Market data below was read on August 26, 2026 from Finviz Elite, during the session that followed the SKYWAY
rheumatoid arthritis release. It changes continuously and is a snapshot, not a valuation.

Three readings stand out. Short interest at 14.88 per cent of a 79.01 million share float, with a short
ratio of 9.51 days to cover, is high for a company with a billion dollars of cash and no debt: it is a bet on the
readouts, not on solvency. Beta of 3.02 and an average true range of $6.01 on a $94 share describe an instrument that
moves three times as much as the index and roughly six per cent on an ordinary day. And performance of plus 451 per
cent over twelve months and plus 548 per cent from the 52-week low places most of the record set out above
inside the share price already paid.

The aggregate sell-side target of $123.92 and the recommendation reading of 1.20 on Finviz’s one-to-five scale
both refer to a period that includes days before the session of August 26, 2026, and neither is shown as
re-cut for the August 25 result in the data available. A target is a forecast, not a fact, and it is reported here as a market reading rather than as a conclusion.

21 Retail Sentiment On Stocktwits

The $SYRE stream on Stocktwits had 2,190 watchers when it was read on August 26, 2026, with message volume flagged
as extremely high and the platform’s own sentiment reading bearish at a score of 34. Volume in the fifteen minutes
around the reading was 2,828 messages against a one-day baseline of 559 and a one-month baseline of 288, which is the
signature of a news day rather than a trend.

The content divided cleanly into three groups. The first repeated the release: the effect size did not meet the
company’s target, several accounts pasted the ACR table verbatim. The second treated the miss as terminal, with the
capitalisation used as the argument, one post asking what an $8 to $9 billion market value is doing on a company with
no commercial product. The third made the argument the company itself makes, that rheumatoid arthritis was never the
core, that the absence of a clean dose-response is odd but unsurprising given how TNF inhibitors perform in the
indication, and that SKYLINE and SKYLIGHT combination readouts are what matter.

These are posts by retail traders and automated accounts, not by professional analysts, and none of it is
evidence about the drug. It is useful only as a reading of who is holding and how loudly. On that measure the notable
thing is that the bear case on the stream is about valuation rather than about science, which is a different argument
from the one the company answered on August 25.

22 The Catalyst Map

Everything below comes from the catalyst table Spyre furnished with its Form 8-K on August 25, 2026, cross-checked
against ClinicalTrials.gov registry dates. Company timings are expectations, not commitments, and none of these is a
regulatory date because Spyre has no product in registration.

The shape of that map matters as much as its content. Three dated events inside 2026, all of them proof-of-concept
rather than registrational; then a gap of roughly a year; then the randomised readout the whole platform argument
depends on. Between the fourth quarter of 2026 and the Part B data in 2027 there is currently no scheduled clinical
catalyst at all.

23 Risks And Red Flags

Every efficacy number in inflammatory bowel disease is uncontrolled. Part A of SKYLINE is open-label with
no placebo arm. The 40 per cent and 33 per cent remission rates cannot be converted into treatment effects, and in
ulcerative colitis placebo response in randomised induction studies is not small. The first controlled read is in
2027.

The one controlled read the company has published disappointed. SPY072 hit its primary endpoint in
rheumatoid arthritis and was still not prioritised. That is the only calibration available between a statistically
positive Spyre trial and a commercially interesting one.

There is no clean dose response in the SKYWAY-RA table. The low dose leads on the primary endpoint and
ACR50, the high dose on ACR20 and ACR70, with complete target suppression at both. Whatever explains that is not
exposure.

The combination thesis has no human data behind it. SPY120, SPY130, SPY230 and SPY772 rest on mouse colitis
models and non-human primate pharmacokinetics. Combining two biologics complicates dose selection, safety
interpretation, manufacturing, regulatory strategy and payer acceptance, and none of that has been tested in this
programme.

The valuation carries the whole map. A market value of roughly $8.29 billion sits on a company with no
approved product, no revenue and no filing pending, after a twelve-month gain of 451 per cent. Sound science and a demanding valuation are separate propositions, and the session that followed
August 25 showed how much of the second is embedded in the price.

Dilution is authorised and undrawn. The share count rose 12.6 per cent in six months. A $500 million
at-the-market programme signed on June 26, 2026 was entirely unused at June 30, an automatic shelf is effective, and
13,841,800 common shares sit behind the Series A preferred. Nothing in the balance sheet forces a raise; nothing in
the company’s history suggests it waits until one is forced.

The largest holder sold four hundred million dollars of stock in June. Fairmount converted and sold
4,684,781 shares at $85.31 on June 23, 2026, while keeping its Series A position. A rebalancing, but a large one, and
it is reflected in the negative insider transaction reading.

Concentration and competition. Four antibodies against three validated targets, all licensed from a single
related-party discovery company. The targets are validated precisely because Takeda, AbbVie, Johnson & Johnson,
Merck and others got there first and are still spending; being second on a proven target means competing on dosing
frequency and price, which is a commercial fight rather than a scientific one.

The pegzilarginase gains are not the business. $70.0 million credited over the first half flatters the reported loss, relates to an asset sold in 2023,
is largely offset by a $34.9 million increase in the CVR liability below the operating line, and has already produced
a $22.9 million cash payment to the legacy CVR holders. Operating cash consumption for the half was $127.2
million.

24 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 from five weighted pillars: balance sheet and runway 30 per cent, catalysts 30 per cent,
dilution 20 per cent, liquidity 10 per cent, execution 10 per cent. It is a measure of fragility, not a view on the
share price, and it is not a buy or sell indication.

Weighted result: 4.2 out of 5. The score is high because the balance sheet removes the failure mode that
kills most clinical-stage companies, and it is not higher because everything that would justify the current
capitalisation is still unproven in a controlled setting. A company can score well on fragility and still be
expensive; those are separate questions and this score answers only the first.

25 Scenarios

These are descriptions of paths, not forecasts, and no probability is attached to any of them. Each names the
future observations that would move a reader from one to another.

26 Bottom Line

Spyre in late August 2026 is a company with more money than most of its peers will ever raise, a clinical
organisation that has hit its own dates, four antibodies against targets other people validated, and a valuation that
already assumes the combination thesis works.

The August 25 rheumatoid arthritis result did not break anything structural. It was a positive trial on its primary endpoint that the company declined to advance, and it disclosed
the decision rather than the alternative of a Phase 3 built on it. But it did two things. It showed what the company means by an internal bar, which sets the
standard the two remaining SKYWAY readouts will be judged against. And it demonstrated, in the only placebo-controlled
dataset Spyre has published, that statistical significance and commercial viability are not the same distance apart
in this pipeline as an open-label response rate makes them look.

Everything that carries the current capitalisation is still uncontrolled. The 9.2 and 10.7 point histology
reductions, the 40 and 33 per cent remission rates, the 51 and 42 per cent endoscopic improvements are all
single-arm. They are good numbers and they came from a real trial, but the experiment that tests them against a
placebo does not report until 2027, and the experiment that tests whether two of these antibodies together are worth
more than one reports at the same time.

Between now and then there are three dated events: SPY003 in September, and two rheumatic sub-studies in the
fourth quarter. After that the calendar is empty until Part B. The balance sheet means none of that has to be
financed, which is a genuine and unusual advantage. What the balance sheet cannot do is answer the question the
market has already answered in the price.

Every figure above was verified on August 26, 2026 against the filing or release named in the sources below,
and each carries the date of the document it came from.

Related Research On Merlintrader

• Vera Therapeutics ($VERA) Stock Hub — another immunology company whose value rests on a single mechanism moving through randomised trials.
• Travere Therapeutics ($TVTX) Stock Hub — what the same kind of biology looks like once it has passed a controlled endpoint and reached the market.
• Arcutis Biotherapeutics ($ARQT) Stock Hub — inflammation in dermatology, including the commercial fight that follows a validated target.
• Aurinia Pharmaceuticals ($AUPH) Stock Hub — autoimmune disease with an approved product and a cash position, for the contrast.
• Top Ten Biotech Stocks Right Now — the names we are following most closely, updated regularly.