The Data Held and the Stock Halved: Outperform Maintained on a Slower sHTG Ramp
Key Takeaways
- The quarter's numbers are a voucher and a shrinking partner base. Fiscal fourth-quarter revenue of $37.8M was a third below a consensus range of $57.0M to $61.2M, because Sarepta and Novartis recognition both fell by more than a quarter and nothing replaced the June Madrigal upfront. The derived loss of ($2.98) per share carries the $215.0M voucher expense, or $1.50 a share. Without it the loss is ($1.48), slightly better than every consensus figure we found.
- The voucher was made redundant 23 days after it closed, and it can now be sold. The FDA admitted the severe-hypertriglyceridemia supplement to its STAR pilot on October 7, which carries the same six-month review the voucher was bought to secure. The supplement went in on November 2, a PDUFA date is due within about 50 days, and the calendar second-quarter 2027 launch target stands. Arrowhead now holds a fully expensed asset of a kind that has sold for $180M to $220M over the past year.
- The stock nearly halved, and its worst day came on a competitor's launch. Shares are 45.0% below our August upgrade close. Arrowhead's own dimer readout cost 18.4% over September 15 and 16. The single worst session, down 15.3% on November 9, was the day the competing APOC3 product's maker roughly halved its 2026 sales guidance on a slow severe-hypertriglyceridemia ramp. Management called that a rival's execution problem. We do not accept the framing, because Arrowhead's own conversion of prescriptions into paid shipments is, in the chief executive's words, "still a choke point."
- The balance sheet is cleaner in structure and thinner in substance. The November refinancing replaced a 15% paid-in-kind facility with a $400M floating-rate loan due 2031 and removed every cash sweep, but retiring the old loan cost $404.4M against a $195.4M carrying value. Pro forma cash of about $1.18B sits against $1.10B of funded debt, and fourth-quarter operating cash burn was $192.4M with no partner receipts. Research spending started to fall a quarter early while selling and administrative expense rose 42.5%.
- Rating: Maintaining Outperform, fair value cut to $35 to $95 (weighted about $70) from $65 to $130 (about $102). We were right about the data and wrong about how quickly a new market would pay for it. At $47.78 the enterprise value is 1.8 to 2.4 times management's $3B to $4B peak framing, against 3.0 to 4.1 times in August, and we think the market now prices the slow ramp more fully than our base case does. Conviction falls to 6 from 7.
Results vs. Consensus
Arrowhead reported its fiscal year, not its quarter. The release and the 10-K contain annual statements only, so every fourth-quarter figure in this section is the fiscal year less the nine months reported in the June 10-Q, both on filed numbers. The company did not publish a fourth-quarter earnings-per-share figure, and the one used here is our derivation on implied weighted shares of about 143.7M. The derivation reconciles to the dollar on every line.
FQ4 2026 Scorecard
| Metric | Actual (derived) | Consensus | Beat/Miss | Magnitude |
|---|---|---|---|---|
| Revenue | $37.794M | $57.0M to $61.2M | Miss | (33.7)% to (38.2)% |
| Collaboration and license revenue | $34.692M | n/a | n/a | (52.4)% QoQ |
| REDEMPLO product revenue | $3.102M | n/a | n/a | +30.1% QoQ |
| Diluted EPS (attributable) | ($2.98) | ($1.52) to ($1.62) | Miss | ($1.36) to ($1.46) |
| Diluted EPS ex-voucher | ($1.48) | ($1.52) to ($1.62) | Beat | +$0.04 to +$0.14 |
| Operating expenses ex-voucher | $259.421M | n/a | Record | +5.7% QoQ |
| Operating loss ex-voucher | ($221.627)M | n/a | n/a | vs. ($170.093)M in FQ3 |
| Cash from operations | ($192.403)M | n/a | n/a | No partner receipts in the quarter |
| Cash and investments | $1,192.983M | n/a | n/a | (25.5)% QoQ like-for-like |
| Forward guidance | Not provided | n/a | n/a | "too early" |
Consensus ranges span three independent providers. Every published quarterly consensus we found excludes the voucher expense: a consensus loss near ($1.55) that included $1.50 a share of voucher would imply an underlying loss of a few cents, which no model carried.
Quality of the miss. The headline loss is a regulatory purchase and the revenue miss is a recognition schedule. Neither tells you much about the operating business on its own, and read together they say something unflattering. The voucher was expensed in full under the research and development standard because Arrowhead told the FDA, before closing, that it intended to use it on its own supplement. That accounting choice is why $215.0M appears as an operating expense rather than an asset, and why the voucher's resale value appears nowhere in the balance sheet. Strip it out and the per-share loss of ($1.48) is a few cents better than consensus. But it is better on a cost base that set another record and a recurring revenue base that fell for the second quarter running. The recurring base has gone from about $63.7M in March to $50.3M in June to $37.8M in September, a 40.7% decline in six months, while operating expenses before the voucher rose from $215.0M to $259.4M.
Year-over-Year Comparison
| $000s except per-share | FQ4 2026 | FQ4 2025 | Change |
|---|---|---|---|
| Product revenue, net | 3,102 | n/a | new |
| Collaboration and license revenue | 34,692 | 256,472 | (86.5)% |
| Total revenue | 37,794 | 256,472 | (85.3)% |
| Cost of product revenue | 1,799 | n/a | new |
| Research and development | 190,465 | 174,687 | +9.0% |
| Selling, general and administrative | 67,157 | 37,679 | +78.2% |
| Acquired priority review voucher | 215,000 | n/a | new |
| Total operating expenses | 474,421 | 212,366 | +123.4% |
| Operating expenses ex-voucher | 259,421 | 212,366 | +22.2% |
| Operating (loss) income | (436,627) | 44,106 | n/a |
| Interest income | 14,368 | 9,053 | +58.7% |
| Interest expense | (22,795) | (21,694) | +5.1% |
| Total other income (expense) | 915 | (7,985) | n/a |
| Income tax (benefit) expense | (6,827) | 20,000 | n/a |
| Net (loss) income including noncontrolling interest | (428,885) | 16,121 | n/a |
| Net (loss) income attributable to noncontrolling interest | (884) | 39,871 | n/a |
| Net loss attributable to Arrowhead | (428,001) | (23,750) | n/a |
| Diluted EPS (derived) | ($2.98) | ($0.18) | n/a |
Both quarters are fiscal year less nine months. The prior-year quarter carried $126.5M of Sarepta revenue, including the first DM1 milestone, and the $130.0M Sanofi upfront recognised through the Greater China entity, which is why its revenue and its noncontrolling-interest line are not comparable to anything since.
The year-over-year table is mostly a record of what last year was. The September 2025 quarter recognised the first $100M Sarepta DM1 milestone and the Greater China upfront, so the 85.3% revenue decline and the swing from operating income to loss measure the end of a licensing windfall rather than a deterioration. The two lines that do carry information are the expense lines. Research and development is up only 9.0% on the year, against 22.1% in the June quarter. Selling, general and administrative expense is up 78.2%, and that is the cost of the commercial organisation the company says is now largely built.
Sequential Comparison
| $000s except per-share | FQ4 2026 | FQ3 2026 | Change |
|---|---|---|---|
| Total revenue | 37,794 | 75,253 | (49.8)% |
| Revenue excluding discrete items | 37,794 | 50,253 | (24.8)% |
| Product revenue, net | 3,102 | 2,385 | +30.1% |
| Research and development | 190,465 | 198,223 | (3.9)% |
| Selling, general and administrative | 67,157 | 47,123 | +42.5% |
| Operating expenses ex-voucher | 259,421 | 245,346 | +5.7% |
| Operating loss ex-voucher | (221,627) | (170,093) | ($51.5)M wider |
| Total other income (expense) | 915 | (8,812) | $9.7M swing |
| Net income (loss) attributable to noncontrolling interest | (884) | 15,364 | $16.2M swing |
| Net loss attributable to Arrowhead | (428,001) | (194,280) | n/a |
| Net loss attributable ex-voucher | (213,001) | (194,280) | ($18.7)M wider |
| Diluted EPS (derived for FQ4) | ($2.98) | ($1.36) | n/a |
| Cash and investments (like-for-like) | 1,192,983 | 1,601,960 | (25.5)% |
| Common shares outstanding | 141,290 | 141,135 | +0.1% |
FQ3 discrete item: the $25.0M Madrigal upfront. FQ3 cash restated to include the $35.1M held at the Greater China entity, which the June release excluded and the September release includes again.
The sequential table explains why a voucher-adjusted earnings line that looks in line still deserves a worse reading than it gets. The operating loss before the voucher widened by $51.5M in a single quarter. Below the line, a $5.3M net gain on the equity-method stake in Bisirna (a $7.6M dilution gain against a $2.3M share of losses) and $4.0M of other income produced $0.9M of net other income, against an $8.8M expense in June. The noncontrolling interest absorbed $0.9M of loss after taking $15.4M of income in June. Those two items are worth about $26.0M of swing in Arrowhead's favour. With a $6.8M tax benefit, they are why the attributable loss ex-voucher widened by only $18.7M while the operating loss widened by $51.5M.
Fiscal Year 2026
| $000s except per-share | FY2026 | FY2025 | Change |
|---|---|---|---|
| Total revenue | 450,817 | 829,448 | (45.6)% |
| of which product revenue, net | 6,603 | n/a | first year |
| Research and development | 739,144 | 607,159 | +21.7% |
| Selling, general and administrative | 202,045 | 123,943 | +63.0% |
| Acquired priority review voucher | 215,000 | n/a | new |
| Total operating expenses | 1,157,988 | 731,102 | +58.4% |
| Operating (loss) income | (707,171) | 98,346 | n/a |
| Net loss attributable to Arrowhead | (724,206) | (1,631) | n/a |
| Diluted EPS | ($5.11) | ($0.01) | n/a |
| Cash from operations | (271,949) | 179,552 | n/a |
Revenue assessment
Revenue of $37.794M breaks into $19.292M from Sarepta, $14.065M from Novartis, $1.335M from Sanofi through the Greater China entity, nothing from Madrigal and $3.102M of REDEMPLO product sales. Every partner line except the smallest fell. In August we told readers to carry Sarepta at the reset $26M level and Novartis near $20M, which produced the roughly $50M September-quarter revenue we wrote into our implied path. Sarepta came in 25.8% below that anchor and Novartis 29.7% below it. Our own estimate was about $12M too high, and consensus was higher still.
"For our Novartis collaboration, we recognized $89 million during the year, nearly all of which relates to the $200 million upfront payment we received in October. As of September 30, 2026, we still carried approximately $113 million of deferred revenue, which will be recognized over time as we fulfill our research and development obligations." — Daniel Apel, Chief Financial Officer
The Novartis decline is the more interesting of the two, because the balance sheet contradicts it. The 10-K classifies $91.6M of the $112.9M remaining Novartis deferral as current, meaning the company expects to recognise it within twelve months. That is about $23M a quarter on an even schedule, against the $14.1M recognised in September. Recognition follows the input method on preclinical work, so a slow quarter of activity produces a slow quarter of revenue and a faster one later. Sarepta has no such buffer: the step-in took the trial cost out of Arrowhead's hands and the reimbursement revenue with it, and the line has now fallen from $41.7M in March to $19.3M in September.
Assessment: the collaboration base is decaying faster than we modelled and product revenue is still too small to offset it. Revenue is now the residue of three agreements in their recognition tails, and it will move with milestone timing until severe hypertriglyceridemia sales arrive. A model should carry roughly $40M to $47M a quarter through calendar 2027 before any new deal or milestone, with Novartis as the line most likely to surprise upward.
Cost assessment
Operating expenses before the voucher reached $259.4M, a record for the third time in four quarters, but the composition moved in the direction we asked for in August. Research and development fell 3.9% to $190.5M, and candidate costs, the outsourced clinical and manufacturing spend, fell 11.4% to $112.2M. In August the chief financial officer dated the step-down in pivotal-study spend to "beginning in fiscal 2027." It began a quarter early. The offset was selling, general and administrative expense, up 42.5% to $67.2M, with salaries up 71.0% to $24.1M and professional and outside services up 35.2% to $31.5M. That is the field force expansion from this summer arriving in the payroll.
"As of today, our commercial organization is largely in place, and we expect only incremental additions as we approach a potential U.S. sHTG launch." — Daniel Apel, Chief Financial Officer
That sentence is the closest thing to an expense ceiling this management team has given in five quarters, and it is the most useful forward statement on the call. It implies that selling, general and administrative expense should run near the September level rather than rise again ahead of launch. Cost of product revenue appeared for the first time at $1.8M for the year. The 10-K warns that product gross margin will "decrease modestly" once inventory manufactured before approval, and expensed as research cost at the time, is used up.
Assessment: we give credit for the research line and withhold it on the total. The voucher-adjusted cost base is still higher than it has ever been, and a commercial organisation sized for over 20,000 prescribers is now being carried against $3.1M of quarterly product sales for at least two more quarters. The "largely in place" statement is the commitment to grade in the transition quarter.
EPS assessment
The derived loss of ($2.98) per share is ($428.0)M on about 143.7M weighted shares, and $1.50 of it is the voucher. As a range check, the simple difference between the reported full-year ($5.11) and nine-month ($2.10) figures is ($3.01). Ex-voucher, ($1.48) is $0.04 to $0.14 better than the three consensus figures we found. The voucher carried no tax benefit, and the $6.8M tax credit in the quarter is the resolution of prior-year estimates on filing the 2025 return, which the 10-K states plainly. Net loss including the noncontrolling interest was ($428.9)M, so the Greater China allocation was not a factor this time.
Interest expense of $22.8M was again almost entirely non-cash: the 10-K records interest paid of nil for the year, against $58.5M of paid-in-kind interest and accretion on the credit facility and $33.0M of non-cash accretion on the royalty liability. That changes from November. The new term loan pays cash interest quarterly at three-month SOFR plus 5.50%, and at a SOFR near 4% that is roughly $9M to $10M a quarter that did not leave the building before.
Revenue Composition
Arrowhead reports one segment. The useful decomposition is by counterparty, because each stream has its own mechanic and its own reason to stop.
| Stream | FQ4 2026 | FQ3 2026 | QoQ | % of FQ4 | Mechanic | Durability |
|---|---|---|---|---|---|---|
| Sarepta | $19.292M | $26.395M | (26.9)% | 51.0% | Ratable recognition of initial consideration plus cost reimbursement and clinical supply | Falling since the step-in; second annual fee of up to $50.0M due over the next twelve months |
| Novartis | $14.065M | $20.232M | (30.5)% | 37.2% | Input-method recognition of the $200M upfront against preclinical work | $112.9M deferred, $91.6M of it current |
| Sanofi / Greater China | $1.335M | $1.241M | +7.6% | 3.5% | Transition services and commercial supply | Small; milestones of up to $255.0M remain |
| Madrigal | $0 | $25.000M | n/a | 0% | Upfront recognised in June | Milestones only from here |
| REDEMPLO product revenue | $3.102M | $2.385M | +30.1% | 8.2% | Net commercial sales, derived as total less collaboration revenue | Growing, still immaterial |
| Total revenue | $37.794M | $75.253M | (49.8)% | 100% | Per the income statement | n/a |
Each FQ4 figure is the 10-K annual amount less the nine-month amount in the June 10-Q. Percentages are rounded and sum to 99.9%.
Sarepta: the largest line and the fastest decline
Sarepta contributed $316.9M for the year, of which $181M was the second DM1 milestone recognised in December. The ongoing stream has fallen in each of the last three quarters as the step-in took the clinical programmes, and their reimbursable cost, into Sarepta's hands. The cash side is better than the revenue side: the 10-K says Arrowhead is eligible for the second instalment of the annual fee, up to $50.0M, over the twelve months from September 30.
Assessment: carry the recognised stream near $19M and falling, and treat the annual fee as a 2027 cash event rather than revenue. The 10-K does not say how the instalment will be recognised.
Novartis: a slow quarter against a full deferral
Novartis recognised $89.0M for the year, $87.1M of it from the upfront and $1.9M from cost reimbursement. The fourth-quarter $14.1M is the lowest of the four, while the current deferral implies about $23M a quarter ahead.
Assessment: the line most likely to beat a model in 2027. We carry $18M to $20M a quarter, below the current-deferral pace, until a quarter confirms the faster schedule.
Product: real growth that is slower than demand
Product revenue of $3.1M was up 30.1% in a quarter when cumulative new-to-brand prescriptions rose 63% and prescribers rose more than 50%. The full-year figure, about $6.6M, was stated aloud by the chief financial officer. We take that as a durable improvement in disclosure.
Assessment: revenue is growing at half the rate of the demand indicators. The gap is the funnel management describes, and it is examined in the next section.
Commercial Performance
REDEMPLO has been on the US market for FCS since November 18, 2025. Management opened this call by committing to report three launch indicators regularly: new-to-brand prescriptions, distinct prescribers and covered lives. In August we wrote that we would believe the launch was readable when a denominator survived two consecutive quarters. The prescriber count has now survived two, and the cumulative prescription count has returned after a quarter's absence. Conversion has not appeared.
| Metric | FQ4 2026 (to Sep 30) | Since quarter-end | FQ3 2026 (as reported in August) | What it tells you |
|---|---|---|---|---|
| Cumulative new-to-brand prescriptions | 950, +63% QoQ | +300, to over 1,250 | Not disclosed (May: >400) | Returned as a committed metric |
| New prescriptions per week | ~28 average in the quarter (derived) | ~50 to 60 | ~20 to 30 | The step-up is a post-quarter phenomenon |
| Distinct prescribers | >640, up more than 50% | >860 by early November | >400 | Survived two consecutive quarters |
| Prescribers with more than one patient | ~27% | n/a | Not disclosed | New |
| US lives with a coverage pathway | >160M | n/a | "favorable policies in place for the most significant payers" | New quantified metric |
| Product revenue | $3.102M, +30.1% QoQ | n/a | $2.385M | Grows at half the rate of prescriptions |
| Implied paid quarterly doses | ~276 at list; ~345 at a 20% gross-to-net; ~394 at 30% | n/a | ~212 / ~265 / ~303 | Our derivation at $11,250 per quarterly dose |
| Prescription-to-paid conversion | Not disclosed | n/a | Not disclosed | Sixth consecutive quarter withheld |
| US wholesale acquisition cost | $45,000 a year | n/a | $45,000 a year | One price across FCS and sHTG |
| Approved and active markets | US, EU (Germany live since August), Canada, Australia, China (partner) | UK authorised in October | US, EU, Canada, China, Australia | England funds only genetically confirmed FCS |
The in-quarter weekly rate is derived: 950 less the ~583 implied at June 30 by the 63% growth figure is ~367 prescriptions over 13 weeks.
Assessment: the demand indicators are genuinely good and the revenue still lags them by a wide margin. If every paid patient took one quarterly dose in the quarter, $3.1M of revenue covers roughly 280 to 390 patients against 950 prescribed, which puts the paid share between about 29% and 41%. A bridge programme that puts patients on drug while coverage is pending, typically before any claim is paid, would explain part of the gap, and management confirmed one exists. The more uncomfortable explanation is arithmetical. Over 1,250 new-to-brand prescriptions inside a year is about a fifth of the roughly 6,500 US FCS patients we have used since initiation. Either the clinically diagnosed population is larger than that estimate, or some prescriptions are going to patients at the boundary of the label, whom payers are entitled to decline. Both readings fit a funnel that fills faster than it pays. The second also fits the post-quarter surge in weekly prescriptions, which arrived as the field force grew and as a competitor's severe-hypertriglyceridemia launch raised awareness of the class.
Key Topics & Management Commentary
Overall Management Tone: Management was the most expansive it has been in five quarters about the long-term company and the most guarded about near-term numbers, declining guidance while describing a "potential blockbuster" and a platform breakthrough it did not think "can be overstated." The posture toward the competitor's slow launch was dismissive rather than analytical. That was the least convincing stretch of the call, because the same remarks conceded Arrowhead's own conversion bottleneck. On the balance sheet management was more forthcoming than in any prior quarter, sizing the voucher's resale range, the refinancing terms and the product revenue unprompted.
1. The voucher that became redundant
Arrowhead closed its $215.0M rare pediatric disease priority review voucher purchase on September 14 and expensed it in full. On October 7, twenty-three days later, the FDA admitted the plozasiran severe-hypertriglyceridemia supplement to its Split Real Time Application Review pilot. STAR supplements qualify for a six-month priority review on their own, and the company withdrew its notice to use the voucher.
"As to the voucher, following the FDA's acceptance of plozasiran into the STAR pilot program in October, it is no longer needed for our planned sHTG supplemental application and is eligible to be sold to a third party. Publicly announced PRV transactions over the last 12 months have ranged from approximately $180 million to $220 million." — Daniel Apel, Chief Financial Officer
The chief financial officer then illustrated a pro forma cash position of about $1.4B if the voucher were sold at prevailing rates, and immediately added that no sale agreement exists and no proceeds are in the reported figures. The refinanced credit agreement expressly permits the company to sell any priority review voucher without lender consent, which the old one did not.
Assessment: the purchase now looks like expensive insurance that the FDA made unnecessary, and most of the premium is recoverable. If the voucher sells at the middle of the stated range, the net cost is about $15M. If it sells at the bottom, about $35M. In August we accepted the strategic logic and called the "greater than 3x return" framing loose. The framing is now moot, and the accounting has created a hidden asset of roughly $1.25 to $1.53 a share that the income statement has already written off.
2. STAR, the supplement and the calendar
The supplement was submitted on November 2. Under STAR the review goal is six months, and the FDA will generally target action at least a month before that date, which points to an action window around April 2027 if the supplement is accepted as filed. Entry into STAR requires the FDA to determine that the clinical evidence indicates the drug may demonstrate substantial improvement on a clinically relevant endpoint over available therapies.
"The FDA granted plozasiran entry into the Split Real-time Application Review, or STAR pilot program for our planned sHTG supplement. STAR provides a 6-month priority review framework and the possibility of action earlier than the formal goal date. We submitted our sNDA on November 2, so we expect to receive our PDUFA date over the next 50 days or so." — Dr. Christopher Anzalone, President and Chief Executive Officer
The 10-K adds the conditions the call omitted: the company must deliver the second part of the split submission within the window the programme contemplates, and the FDA has said acceptance into STAR does not guarantee an expedited review or an approval.
Assessment: STAR entry is the most important regulatory signal of the quarter, and it bears directly on the label debate we have carried since May. The FDA has not decided anything about the pancreatitis claim, but the admission criterion is a substantial improvement on a clinically relevant endpoint. Triglyceride lowering alone, against an approved first mover with the same mechanism, is a harder case to make on that test than a reduction in pancreatitis events. We raise our probability of a pancreatitis claim in the indication statement modestly.
3. The competitor's launch is the market's evidence
On November 9, the maker of TRYNGOLZA cut its 2026 product-sales guidance to about $50M from $100M to $110M, citing a slower-than-anticipated severe-hypertriglyceridemia ramp. It reported $9M of third-quarter sales, its first full quarter in the indication, and about 1,500 prescribers. Its shares fell 21.4% and Arrowhead's fell 15.3% on a day the main biotechnology ETF rose 3.1%. Management was asked whether this was execution or market.
"So broadly, look, what we learned from Ionis doesn't affect our expectation in sHTG at all. I think what we've seen from them are some missteps in communication, some missteps in commercial execution, and we don't expect to have those missteps." — Dr. Christopher Anzalone, President and Chief Executive Officer
The same chief executive, in prepared remarks, said revenue would not tell the full story "until after the first several quarters" of Arrowhead's own launch, and described the step between prescription and paid shipment as the slow part of the process. In August he volunteered that severe hypertriglyceridemia would be "a relatively slow ramp because this is a brand-new market."
Assessment: we do not accept the execution framing, and we do not think management fully does either. The competitor cited only a slower-than-anticipated launch ramp, but the likely cause, building payer criteria and educating physicians in a market that has never treated triglycerides aggressively, is the friction Arrowhead's own FCS launch has shown for a year. Arrowhead will launch second, with quarterly dosing, a pancreatitis data set and pre-approval payer discussions. Those are real advantages in share. None of them makes the market grow faster. We now carry a slower class ramp as a separate bear point, and it is the main reason our fair value falls.
4. FCS demand is accelerating, and conversion is still the choke point
The FCS indicators improved across the board. Cumulative new-to-brand prescriptions reached 950 at quarter-end and over 1,250 by early November, the weekly rate rose to 50 to 60 from 20 to 30, and prescribers passed 860, with about 27% writing for more than one patient.
"We are seeing strong demand for REDEMPLO and continue to work with payers to increase the rate at which prescriptions are converted to paid shipments. This is still a choke point, and there is substantial work to do, but we are off to a good start." — Dr. Christopher Anzalone, President and Chief Executive Officer
The commercial lead was explicit that the counts are written prescriptions, not paid claims, and that moving from one to the other requires benefit verification, office support and prior authorisation. Our derivation in the previous section puts the paid share of prescribed patients at roughly 29% to 41%.
Assessment: this is the first quarter in which the three metrics management chose are both consistent with the prior quarter and committed for the future, which addresses half of our sixth bear point. The other half is the number that converts the metrics into a forecast, and it was withheld for the sixth consecutive quarter. Until it appears, the right way to read this launch is through product revenue, and product revenue grew 30%.
5. The recurring revenue base fell for the second quarter running
The base we can observe has gone from about $63.7M in March, the only clean quarter, to $50.3M in June and $37.8M in September. In August we wrote that the clean base "lasted exactly one quarter." It has now fallen 40.7% in two.
"While we have consistently stated that FCS revenue will have an immaterial impact to our 2026 financial statements, we continue to invest in supporting this FCS launch and preparing for the potential blockbuster opportunity in sHTG." — Daniel Apel, Chief Financial Officer
Assessment: the revenue line is no longer a meaningful offset to the cost base, and it will not become one before the larger launch. That makes the cash line, not the income statement, the right place to judge the next four quarters.
6. Research spending turned, commercial spending did not
Candidate costs fell 11.4% sequentially and total research and development fell 3.9%, a quarter ahead of the fiscal 2027 timing the company set in August. Selling, general and administrative salaries rose 71.0%.
"Our expanded sales team is now largely in place, building relationships and mapping priority accounts in preparation for a potential sHTG launch." — Andy Davis, Senior Vice President and Head of the Global Cardiometabolic Franchise
Assessment: the research commitment was met early and the commercial build is, by management's account, complete. For the first time we can model a cost base that flattens rather than one that rises without limit: roughly $185M to $190M of research and $65M to $70M of selling and administrative expense a quarter through the launch. That remains about $255M a quarter against roughly $45M of revenue.
7. The refinancing: better terms, a bigger bill than the balance sheet showed
On November 2 Arrowhead replaced its 2024 Sixth Street facility with a $400.0M senior secured term loan at three-month SOFR plus 5.50%, maturing November 2031, with no amortisation and no sweep of product, royalty, milestone or partnering proceeds. The old loan carried 15% paid-in-kind interest and a two-times multiple of invested capital if repaid before August 2028.
"The new loan matures on November 2, 2031, has no scheduled principal amortization before maturity and importantly, eliminates any and all required cash sweeps to Sixth Street tied to product sales, royalty, milestones or any existing or new licensing and partnership proceeds." — Daniel Apel, Chief Financial Officer
The 10-K gives the other half. The payoff of the old facility was about $404.4M, including principal and paid-in-kind interest of about $363.8M, against a carrying value of $195.4M at September 30. The $68M discount is measured against the full two-times multiple, not against what the balance sheet showed. Net of the new loan and its 2% fee, the transaction cost about $12.4M of cash.
Assessment: the economics are clearly better and the structure is much better. Removing the sweep means every partner milestone and every dollar of product revenue now stays in the business, which matters more than the coupon. But the carrying value understated the claim on the company by about $209M, and our August net-cash figure of $293.0M inherited that understatement. Paid-in-kind interest is also becoming cash interest of roughly $9M to $10M a quarter.
8. The balance sheet after the voucher and the refinancing
Cash and investments closed at $1,193.0M. On a like-for-like basis, including the Greater China entity's cash in both periods, they fell $409.0M in the quarter, or $194.0M before the voucher. Cash used in operations was $192.4M in a quarter with no partner receipts, which makes it the cleanest burn figure the company has produced. There were no at-the-market sales in the quarter.
"To put our debt position in context, our funded debt now consists principally of the $400 million term loan due in November 2031 and $700 million of 0% convertible notes due in January 2032. In other words, no principal payments are required before late 2031." — Daniel Apel, Chief Financial Officer
Pro forma for the refinancing, cash is about $1,180.6M against $1,100.0M of funded debt, a net position of about $80.6M. Including the $400.4M royalty liability, which we have counted since initiation, the net position is about ($319.8)M. At $195.8M of quarterly operating and capital burn plus the new cash interest, the company has about five quarters before it reaches the facility's $125.0M minimum-cash covenant, and about six with a voucher sale.
Assessment: the maturity wall is gone and the runway is shorter than the headline suggests. Five to six quarters takes the company to the end of 2027 or early 2028, three to four quarters into a calendar second-quarter 2027 launch, which is exactly when a slow ramp would be most visible. The refinancing also opened a basket for additional convertible debt of up to the greater of $1.1B and 10% of market capitalisation. We read that as the next financing instrument being chosen in advance, and at these prices we would rather it be debt than equity.
9. The full SHASTA data set held up
The detailed results were presented at the European Society of Cardiology Congress on August 30, published in the New England Journal of Medicine on November 4 and presented again at the American Heart Association the weekend before the call. Median triglyceride reductions of 79% and 81% held. The journal states the placebo-adjusted reductions as 45 and 47 percentage points, and gives a pooled pancreatitis rate of 1 per 100 patient-years against 5 on placebo across 757 patients.
"In SHASTA-3 and 4, the number of patients needed to treat in order to prevent a single acute pancreatitis event was 9 in patients with TGs greater than 880 without history of pancreatitis and those with TGs greater than 500 and history of pancreatitis. Within that latter group alone, the NNT was only 3." — Dr. Christopher Anzalone, President and Chief Executive Officer
Among patients with a prior pancreatitis history the event rate fell 91%. The widely quoted 100% reduction in the highest-risk subgroup is zero events among 35 treated patients against five among 17 on placebo, which is striking and small. Treatment-emergent adverse event rates were similar in the pooled active and placebo groups. The chief medical officer disclosed one new imbalance: worsening glycemic control in 14.3% of treated patients against 8.7% on placebo, with mean changes in haemoglobin A1C he described as consistent with prior experience. The imaging sub-study showed no statistically significant change in liver fat.
Assessment: the data set survived peer review intact. The glycemic imbalance is the one item a label or a payer could use, in a population where diabetes is common, and it belongs on the watch list rather than in the bear case. The placebo-adjusted figure of 45 to 47 points is the number payers will quote, and it is less flattering than the 80% headline. The mean reduction and the imaging subgroup's size, both of which we asked for in August, appear in neither the release, the 10-K nor the call.
10. The platform delivered twice, and the market sold both
ARO-DIMER-PA, the company's first dual-functional RNAi candidate, reported interim single-dose results on September 15, with mean maximal reductions of 72% for PCSK9 and 88% for APOC3. The stock fell 18.4% across that session and the next. The figures management cited on this call are higher, 89% for PCSK9 and 95% for APOC3 with LDL cholesterol down 65% and ApoB down 66%, from a later presentation across more dose cohorts and follow-up. The same single-dose study produced the immune-related adverse events at 200mg and 400mg that management discussed on the call. On October 14, ARO-MAPT, a subcutaneous candidate designed to reach the brain, reported a 73% mean reduction in cerebrospinal-fluid total tau at day 105, against the 50% to 60% benchmark management set in August.
"At the 12.8 milligram per kilogram regimen administered on day 1 and 15, mean CSF total tau fell 73% by day 105 with a maximum individual reduction of 85%. Day 105 was 90 days after the last dose, and these findings support the potential for maintenance dosing every 3 months or less frequently." — James Hamilton, Chief Medical Officer and Head of R&D
Management is moving quickly into frontotemporal dementia and progressive supranuclear palsy, slowly in Alzheimer's disease, and expects two wholly owned central-nervous-system clinical filings around the turn of the year. ARO-DIMER-PA will seek a first approval on LDL lowering, subject to regulators, and the chief executive sees a reasonable chance that the chosen dose will be below 200mg. On partnering, management said US plozasiran and ARO-MAPT are off the table, and obesity, liver, complement and early central-nervous-system work are open.
Assessment: the platform pillar earned an upgrade on evidence. Delivery across the blood-brain barrier by a skin injection, at a knockdown above the company's own bar, is the most valuable new capability Arrowhead has shown since the liver. Two caveats keep it out of our valuation. The loading regimen is weight-based at 12.8mg per kilogram, a heavy dose by the standards of a 25mg plozasiran injection. And the dual-silencing molecule's tolerability at the doses that produced the best lipid numbers is unresolved. We treat both as option value, as in August, but the options are now worth more than the market is paying for them.
11. Europe: live, and narrower than the label
REDEMPLO became available and reimbursed in Germany in August. The UK medicines regulator authorised it in October for FCS diagnosed by genetic testing or clinical criteria, and England's technology appraisal body recommended NHS funding, but only for genetically confirmed patients whose response to diet and conventional therapy has been inadequate.
"Outside the U.S., Germany and the U.K. are now active commercial markets. In the U.K., the reimbursement authority recently issued final guidance recommending REDEMPLO for genetically confirmed FCS, supporting reimbursement in England and Wales, and we are working to serve eligible patients there now." — Andy Davis, Senior Vice President and Head of the Global Cardiometabolic Franchise
Assessment: the clinical-diagnosis label was the feature we cited in making the FCS label a bull pillar, and England's funding decision did not pay for it. That is a small market and a small revenue effect. It is also the first sign that payers outside the US will draw the line where the competitor's label draws it, which is a fact to remember when management sizes ex-US revenue, which it still declines to do.
Guidance & Outlook
Arrowhead gives no revenue or expense guidance and declined again: "it's just too early," pending a PDUFA date, a label and payer experience. It also changed its fiscal year end to December 31, effective January 1, 2027. The October to December 2026 quarter will be reported as a transition period on a Form 10-Q transition report, and fiscal first quarter 2027 will mean January to March 2027.
"We believe this change will ease a number of administrative burdens associated with the September year-end and ahead of a potential sHTG launch will improve company and investor communication." — Daniel Apel, Chief Financial Officer
Commitments made in August, graded
| Commitment made in August | Status at this print | Verdict |
|---|---|---|
| Detailed SHASTA-3/4 data at ESC on August 30 (we also asked for the mean reduction and the imaging subgroup size) | Presented August 30, published November 4, presented again at AHA. Medians and placebo-adjusted medians given; mean and subgroup size not in the release, 10-K or call | Delivered, partly answered |
| ARO-DIMER-PA and ARO-MAPT first readouts in September | DIMER-PA September 15. MAPT October 14, with 73% mean tau knockdown against a 50% to 60% benchmark | DIMER on time; MAPT two weeks late and above the bar |
| Supplement submitted before the end of 2026 | Submitted November 2, with STAR entry granted October 7 | Delivered early |
| Voucher purchase closes in the fiscal fourth quarter | Closed September 14 and expensed. Made redundant by STAR 23 days later | Delivered, then superseded |
| Research spending moderates from fiscal 2027; where selling and administrative expense settles | Research down 3.9%, candidate costs down 11.4% in FQ4. SG&A up 42.5%; organisation "largely in place" | Research early; first ceiling language on SG&A |
| A product revenue derivation and a conversion denominator that survive two quarters | Product revenue derivable and stated for the year. Prescriber count survived; prescription count returned. Conversion withheld | Partly delivered |
| Whether the balance sheet needs a raise after the voucher | No equity raise and no at-the-market sales in FQ4. Debt refinanced instead | Delivered |
| An explanation of the noncontrolling-interest allocation | The 10-K reports the Greater China entity's fiscal 2026 revenue of $31.6M and net income of $14.5M. FQ4 allocation ($0.9)M | Answered at the annual level |
| German reimbursement launch and EU rollout | Germany live since August. UK authorised; England funds genetically confirmed patients only | Delivered, narrower than hoped |
| Developments in the Ionis patent suit | First Amended Complaint filed October 1 alleging infringement by commercial sales; trial set for July 11, 2028. Not mentioned on the call | Escalated procedurally |
New commitments made on this call
| Item | Commitment | Date |
|---|---|---|
| PDUFA date | Expected from the FDA | Within about 50 days of November 12 |
| sHTG launch | US launch, subject to approval, initially in patients at high risk of pancreatitis | Calendar Q2 2027 |
| Obesity | ARO-ALK7 non-diabetic monotherapy cohorts and type 2 diabetic tirzepatide combination cohorts: body composition by MRI, safety, knockdown, weight | Before end 2026 |
| ARO-DIMER-PA | Multiple-dose data and dose selection, likely below 200mg | First half to mid-2027 |
| ARO-MAPT | Multiple-dose healthy-volunteer data, then Alzheimer's patient data including p-tau; FTD and PSP study designs with regulators now | 2027; patient data second half |
| CNS pipeline | Clinical trial applications for two wholly owned candidates, more entries by end 2027 | Around end 2026 to early 2027 |
| Zodasiran YOSEMITE | Study completion, data, then filing | Mid-2027; data second half 2027; NDA by end 2027 |
| Commercial organisation | Largely in place; only incremental additions | Through launch |
| Launch metrics | New-to-brand prescriptions, distinct prescribers and covered lives reported regularly | Ongoing |
| Voucher | Available for sale; no agreement | Open |
| Fiscal calendar | December year end; transition report for October to December 2026 | From January 1, 2027 |
Implied path for the transition quarter (October to December 2026), our estimate: revenue of $40M to $50M (Sarepta near $19M, Novartis $14M to $23M, product near $4M, Sanofi near $1M), operating expenses near $255M, and an operating cash burn near $195M plus the $12.4M net cost of the refinancing and about two months of cash interest. That puts year-end cash and investments near $980M before any voucher sale or partner receipt. The transition-quarter income statement will also have to account for the roughly $209M by which the old facility's payoff exceeded its carrying value. The 10-K does not say whether that will be treated as an extinguishment loss or a modification.
Street at: no meaningful consensus exists for a transition quarter that most models will not have carried as a separate period.
Guidance style: dates, not numbers, and the dates keep landing. Every clinical and regulatory commitment from August was met or beaten except a two-week slip on ARO-MAPT. The record on financial forward statements improved this quarter: the research step-down arrived early.
Analyst Q&A Highlights
Fourteen analysts asked questions. Unlike August, the session engaged with commercial reality: the opening question was the competitor's guidance cut, and the commercial questions that followed covered launch definition, guidance, payer access, pricing, bridge programmes and the size of the prior-pancreatitis population. Nobody asked about the refinancing payoff, the voucher accounting, the collaboration revenue decline or the cost base. Half the questions went to the pipeline.
Whether the competitor's slow launch is about them or about the market
The opening question asked directly whether the competing product's guidance cut reflected company execution or a market that will take longer to develop, and what the headwinds could be. Management answered with conviction and little evidence, attributing the cut to communication and commercial missteps it does not expect to repeat, then citing its own prescription growth.
Q: "Given that Ionis has set expectations for their early sHTG launch to be softer than expected, do you view that as a company-specific execution issue or evidence that the sHTG market may take longer to develop than anticipated?"
— Maurice Raycroft, Jefferies
A: "So broadly, look, what we learned from Ionis doesn't affect our expectation in sHTG at all. I think what we've seen from them are some missteps in communication, some missteps in commercial execution, and we don't expect to have those missteps."
— Dr. Christopher Anzalone, President and Chief Executive Officer
Assessment: the question asked for headwinds and the answer offered none. The rest of the reply conceded the "middle of the funnel" as the slow part, which is a market-wide problem rather than an execution one. The answer is a statement of intent, and the next three quarters of the competitor's sales are the better test of it.
What "rapid launch" means if it does not mean revenue
A follow-up pressed on the apparent contradiction between promising a "rapid launch" and saying revenue would not tell the story for several quarters, and asked whether rapid was defined by prescriptions rather than sales. The commercial lead confirmed the three leading indicators as the yardstick, and management added that it was not walking away from revenue.
Q: "And maybe because you guys highlighted communication and kind of expectation setting, is your definition of a rapid launch on sort of the metric that you guys outlined like NBRx as opposed to revenue. So I just wanted to perhaps clarify that so everybody's level set."
— Jason Gerberry, Bank of America
A: "As we have mentioned at the outset of this call, the 3 leading indicators that we will be tracking and looking to consistently report to you and others are the 3 metrics you mentioned at the outset of the call, that's new-to-brand prescriptions, which is largely synonymous with new patients." The other two indicators he named were breadth of prescribers and payer coverage.
— Andy Davis, Senior Vice President and Head of the Global Cardiometabolic Franchise
Assessment: the answer is the clearest statement yet of how management wants the launch judged, and it is a yardstick chosen because it will look good early. We will use it, alongside derived product revenue, because only the second number converts into value.
When guidance arrives
Asked whether guidance might come next year, management said it was too early, pending a PDUFA date, a label and experience with payers.
Q: "Just wanted to get your thoughts on potentially providing guidance. What's the current thinking there? Is that something you might share next year at some point?"
— Michael Ulz, Morgan Stanley
A: "Yes. Unfortunately, it's just too early to provide guidance. We will be happy to do that when we can. We have just filed our sNDA. We don't have a PDUFA date yet, much less a label."
— Dr. Christopher Anzalone, President and Chief Executive Officer
Assessment: a reasonable refusal for revenue, and a missed opportunity for expenses. A company that has just said its commercial build is complete could have framed a cost run rate, and that was the guidance investors most needed after a third record quarter of spending in one fiscal year.
Whether the price survives the competitor's payer friction
The pricing question asked whether $45,000 is the right level given early friction in the competitor's launch. The commercial lead said current payer policies for the competing product follow its label, that pre-approval payer conversations have produced no pushback on price, and that early friction is procedural rather than price-related.
Q: "On the payer side, do you think you're at the right pricing range given some of the initial friction that's happening with the Tryngolza launch?"
— Prakhar Agrawal, Cantor
A: "We do believe we have the right pricing range here for REDEMPLO, and there are 2 kind of leading indicators that give us confidence. The first would be the currently published policies related to Tryngolza and sHTG, which do appear to have policy criteria consistent with label."
— Andy Davis, Senior Vice President and Head of the Global Cardiometabolic Franchise
Assessment: policies that track the label are good news for access and say nothing about speed. The answer supports the price and leaves the ramp question where it was. The label wording, including whether pancreatitis appears in the indication, will set the payer criteria.
How many patients start on bridge supply before they become revenue
A question about the competitor's bridge programme asked whether REDEMPLO follows the same playbook. Management confirmed it runs a quick-start bridge and a patient assistance programme in FCS and will do the same in severe hypertriglyceridemia.
Q: "Recently we've learned that the Quick Start program that Ionis has adopted is essentially a bridge program before patients get covered. So wanted to get your thoughts on if REDEMPLO follows a similar playbook here for FCS, given that many of the patients may be first-to-therapy and not yet approved."
— Jiayi Yuan, RBC Capital Markets
A: "So like any rare disease launch, we do have a quick start bridge program ourselves and a patient assistance program so patients can start therapy while coverage is being established. And so we would expect a quick start program to be in place for SHTG as well, should we be -- should we indicated for that next year."
— Andy Davis, Senior Vice President and Head of the Global Cardiometabolic Franchise
Assessment: the most informative exchange for modelling, and the least discussed. A bridge programme means some patients counted in the prescription metrics are likely on therapy before any claim is paid, which would explain part of the gap between 63% prescription growth and 30% revenue growth. The share of patients on bridge supply was not given. It is the next denominator to ask for.
How large the highest-need population is
Asked how many US patients have had a prior pancreatitis event and how long broad uptake beyond them will take, the commercial lead gave a population figure for the first time and conceded the broader market will take years.
Q: "Just curious what your latest thinking is around how many patients in the U.S. have had a prior pancreatitis event, and just in general, I guess kind of what you think the timeline and path is to really getting broad uptake outside those patients and kind of more of the primary prevention setting."
— James Condulis, Stifel
A: "We estimate there are around 200,000 patients in the U.S. who have a prior history of AP due to high triglycerides."
— Andy Davis, Senior Vice President and Head of the Global Cardiometabolic Franchise
Assessment: 200,000 secondary-prevention patients at $45,000 gross is a $9B list-price pool, and it is where the number needed to treat is three. If the first two years of the launch concentrate there, as management intends, a slow market can still produce a meaningful revenue line. The same answer said the high-risk segment is about a third of the broader severe-hypertriglyceridemia population, and that penetrating the broader population "will take a little bit more time" and depends on payer policies.
Whether the dual-silencing molecule's side effects limit its dose
A question on ARO-DIMER-PA's safety asked how long the injection-site reactions and fevers lasted and what they imply for repeat dosing. The chief medical officer confirmed immune-related events at the two highest single doses and said dose selection awaits the full multiple-dose data. The chief executive then said the eventual dose is reasonably likely to be below 200mg.
Q: "I think there were some injection reactions in pyrexia. So wanted to understand how long these events lasted? Are these going to be single dose effect and implications for the MAD dosing?"
— Prakhar Agrawal, Cantor
A: "Like you said, what was presented in the SAD component of the study at some of the higher dose levels, 200, particularly 200 and 400 milligrams, we saw some immune-related AEs, specifically pyrexia, headaches, some ISRs."
— James Hamilton, Chief Medical Officer and Head of R&D
Assessment: this is the tolerability question the stock's September decline left open. The best lipid numbers came from the doses with the tolerability signal, and the company is now signalling a lower dose. That would trade some efficacy for cleaner safety, and the multiple-dose data will show how much.
What it takes to move the brain-delivery programme forward
The last question asked which patient data would connect tau lowering to disease biology and how the three indications would be sequenced. Management said knockdown plus safety is enough to start Phase II studies in the primary tauopathies, with Alzheimer's patient biomarkers and tau imaging to follow.
Q: "I'm wondering what the next patient data set is that can connect the CSF total tau lowering to an effect on disease biology, and what result would justify advancing toward a registrational program?"
— Patrick Trucchio, H.C. Wainwright
A: "Based on the knockdown data we have from healthy volunteers, we think that showing the knockdown and the safety is sufficient to at least get Phase II studies up and running in those indications."
— James Hamilton, Chief Medical Officer and Head of R&D
Assessment: a sensible sequencing. MAPT mutations cause frontotemporal dementia directly, so knockdown is close to a mechanistic endpoint there in a way it is not in Alzheimer's disease. The chief executive summarised it as fast in the primary tauopathies and slow in Alzheimer's, which is the right order for a company with five to six quarters of cash.
What They're NOT Saying
- How the refinancing will be booked. The old facility's payoff of about $404.4M exceeded its $195.4M carrying value by about $209M. The call presented the transaction as a $68M discount, which it is against the contractual multiple. Whether the transition quarter carries a loss of that order was not discussed.
- The new convertible-debt basket. The refinanced agreement permits additional convertible indebtedness up to the greater of $1.1B and 10% of market capitalisation, against no capacity under the old one. It appears only in the 10-K.
- Prescription-to-paid conversion, for the sixth quarter. Also missing: the share of patients on bridge supply and the gross-to-net. The three committed metrics are all upstream of payment.
- Why the weekly prescription rate doubled after the quarter ended. The quarter averaged about 28 a week. The 50 to 60 figure is a post-quarter rate, and no reason was given for the step beyond field-force reach. Over 1,250 prescriptions in a year is about a fifth of the FCS population we have assumed, and the label boundary was not discussed.
- The patent suit. The competitor filed an amended complaint on October 1 alleging that commercial sales of plozasiran infringe its patent, and a trial is set for July 11, 2028. The suit was not mentioned on a call that attributed the same competitor's launch problems to its missteps.
- Why Novartis recognition fell 30.5%. This happened while $91.6M of the deferral sits in current liabilities, implying a faster schedule ahead.
- The narrower English funding decision. England's appraisal funds only genetically confirmed FCS, despite a UK label that includes clinical diagnosis. The commercial lead stated the restriction without addressing its significance.
- Product gross margin. The 10-K expects it to decrease once pre-approval inventory, already expensed as research, is used up. The call did not mention it.
- The mean triglyceride reduction and the imaging subgroup size. Both were asked for in August. Neither appears in the release, the 10-K or the call.
- An operating expense run rate for 2027. "Largely in place" is the nearest thing to it.
Market Reaction
- Pre-print setup: the shares closed at $47.15 on November 12, down 4.5% on the session and 45.7% from our August report's $86.91 close. They were down 29.0% year to date from $66.39, down 18.9% over thirty days from $58.13, and up 13.8% over twelve months from $41.42. The 52-week closing range was $38.09 to $89.59, the high being the August 4 pre-print close. The S&P 500 was up 10.9% year to date.
- The path since August: about 7.2% was lost into mid-September, through the detailed congress data, to $80.67 on September 14. The interim ARO-DIMER-PA release then took 18.4% off in two sessions, to $65.79 on 4.4M and 5.1M shares. A slower 11.6% decline to $58.16 on November 6 followed, including a flat response to the October 14 ARO-MAPT data (up 0.4% on 4.2M shares) and a 6.8% fall in the November 5 market selloff. On November 9 came a 15.3% drop to $49.24 on 8.1M shares, as the competitor fell 21.4% and the main biotechnology ETF rose 3.1%.
- After-hours move: down about 2.4% after the release on November 12.
- Reaction session (November 13): the shares opened at $47.68, up 1.1%, traded between $44.52 and $48.29, and closed at $47.78, up 1.3% or $0.63. The S&P 500 fell 0.4%, so the relative move was 1.7 points.
- Volume: 3.1M shares against a 2.0M thirty-day average, 1.6 times normal.
- Peers: the competitor rose 6.4% the same session to $29.06. The main biotechnology ETF fell 0.4%.
The print itself barely moved the stock, and that is the right reading of it. The release confirmed facts the market already had: the voucher purchase, STAR entry, the supplement, the refinancing and every clinical readout had been announced before November 12. What was new was a collaboration revenue line weaker than expected and a set of FCS demand indicators stronger than expected. The intraday low, 5.6% below the prior close, was bought back to a modest gain on above-average volume. We read that as holders who had already absorbed the competitor's guidance cut finding nothing in the release to make it worse.
The more important story is the three months before the print, and it does not support a simple narrative of disappointment with Arrowhead. The two largest declines followed a platform readout that was positive on efficacy and a competitor's guidance cut. Neither was a negative fact about plozasiran. The detailed pivotal data, which carried our largest residual risk in August, cost 3.0% on the first session after the congress presentation and was recovered within three sessions. What the stock has repriced is the timing of the commercial opportunity, and the speed with which the market reacted to the competitor's guidance cut says that timing was carrying most of the August valuation.
The setup has inverted. In August the stock entered the print at its 52-week high and needed the quarter to add to a priced-in win. In November it entered 47.4% below that high, 23.8% above its 52-week low, and at an enterprise value of about $7.19B at the reaction close. That is 2.4 times the low end of management's $3B to $4B peak framing for severe hypertriglyceridemia, against 3.0 to 4.1 times three months ago. The multiple has compressed by more than the outlook has.
Street Perspective
Debate: is the severe-hypertriglyceridemia market slow for everyone, or only for the first mover?
Bull view: the first entrant carried the cost of educating physicians and writing payer criteria, and the second entrant inherits both. Payer policies already track the competitor's label, Arrowhead is negotiating before approval, and it brings quarterly dosing and a pancreatitis data set the first mover did not have at launch. The FCS indicators, with prescribers up 34% in five weeks, show the field force is reaching the right physicians.
Bear view: one quarter of real-world evidence says this market builds slowly, and the company launching second has described the same choke point in its own orphan launch for a year. Two companies each claiming more than $3B of peak sales in a market that produced $9M in its first full quarter is a sign of shared optimism, not shared evidence. A second entrant also splits the patients who do convert.
Our take: the bears have the better argument on timing and the bulls on share. We now assume the market develops at roughly the pace the first mover is seeing and that Arrowhead takes a larger share of it than its arrival order implies. That is a later peak, not a lower one, and it is the main reason our fair value falls by a third.
Debate: does the balance sheet need equity before sHTG revenue arrives?
Bull view: no principal falls due before late 2031, every cash sweep has gone, the voucher can be sold for about $200M, the Sarepta annual fee and further milestones are due, and the cost base has stopped rising. On those terms the company reaches a launch that is about six months away with its cash largely intact.
Bear view: the quarter consumed $194M before the voucher with no partner receipts, cash interest starts now, and the runway to the covenant floor is about five quarters. The refinancing created room for $1.1B of new convertible debt, and the stock has halved, so the next financing will dilute more than the last one did.
Our take: a financing in the next eighteen months is more likely than not, and the refinancing has made it more likely to be convertible debt than common equity. The bear case is right about the need and wrong about the instrument. We carry a 2027 convertible in the base case and an equity raise below $40 in the bear case.
Debate: is the platform in the price?
Bull view: Arrowhead has now shown liver, lung, muscle, fat and brain delivery and a molecule that silences two genes, and the central-nervous-system result in particular is a capability few companies have. At a $7.2B enterprise value, little of that is being paid for after subtracting plozasiran.
Bear view: the market sold both readouts for good reasons. The dual-silencing molecule's best numbers came with a tolerability signal, the brain programme needs a heavy weight-based dose, and both are years from revenue at a company with a short runway and no appetite to partner its best central-nervous-system asset.
Our take: the platform is worth more than zero and less than the bulls claim. Its value right now is as partnering currency, and management has ruled the most valuable piece off the table. We still exclude it from the fair value and treat a central-nervous-system partnership on the early programmes as the likeliest non-dilutive financing.
Model Update Needed
| Item | Prior assumption (August) | Suggested change | Reason |
|---|---|---|---|
| Quarterly revenue base | ~$50M, upward slope from product only | $40M to $47M through calendar 2027 | Sarepta $19.3M and Novartis $14.1M in FQ4, both more than a quarter below the August anchors |
| Sarepta | $26M a quarter | ~$19M and declining; annual fee of up to $50M as 2027 cash | Step-in continues to remove reimbursable cost |
| Novartis | ~$20M a quarter | $18M to $20M, with upside to ~$23M | $91.6M current deferral against $14.1M recognised in FQ4 |
| REDEMPLO product revenue (FCS) | $2.4M, growing | $3.1M, growing ~25% to 30% a quarter | Revenue grows at half the rate of prescriptions; bridge supply and payer criteria |
| Research and development | $198M, moderating from fiscal 2027 | $185M to $190M a quarter | Candidate costs down 11.4% in FQ4 |
| Selling, general and administrative | $47M, rising through 2027 | $65M to $70M, broadly flat | Organisation "largely in place, only incremental additions" |
| Priority review voucher | $215.0M outflow, capitalised | Expensed in FQ4; $180M to $220M resale as a contingent asset | ASC 730 treatment; STAR made it redundant |
| Credit facility | $181.4M carrying value, paid-in-kind interest | $400.0M principal from November 2, SOFR + 5.50% cash, due 2031 | Refinancing; payoff $404.4M |
| Net position | $293.0M net cash on carrying values | About $80.6M net of funded debt; about ($319.8)M including the royalty liability | Payoff value replaces carrying value; voucher; FQ4 burn |
| sHTG approval and launch | Calendar Q2 2027 on voucher priority review | Unchanged, via STAR; action window around April 2027 | Supplement filed November 2 |
| sHTG revenue ramp | Education-paced, as management describes | Paced to the first mover's observed launch; meaningful revenue from 2028 | The competitor's $9M first full quarter and halved guidance |
| Peak sHTG revenue | $3.0B base, $4.0B bull | Unchanged, reached two to three years later | Management reiterated $3B to $4B "over time" in US sales |
| Financing | None modelled before label | 2027 convertible in the base case | Five to six quarters of runway; new $1.1B convertible basket |
| Fiscal calendar | September year end | December year end; October to December 2026 transition period | Board decision of November 9 |
| Diluted share count | 143.4M weighted, 151.6M fully loaded | ~143.7M weighted, ~151.8M fully loaded | 141.290M outstanding, 2.468M pre-funded warrants, ~8.0M convertible shares |
| Contingency | Ionis suit unquantified | Unchanged; trial July 11, 2028 | Amended complaint now targets commercial sales |
Valuation impact: at the $47.78 reaction close, 141.290M shares plus 2.468M pre-funded warrants give an economic market capitalisation of about $6.87B. Add $700.0M of convertible principal, the $400.0M new term loan and the $400.4M royalty liability, and subtract about $1,180.6M of pro forma cash. The enterprise value is about $7.19B, or 2.4 times $3.0B of peak revenue and 1.8 times $4.0B.
Fair-value range: $35 to $95, probability-weighted about $70, from $65 to $130 weighted about $102. We have lowered the multiple of peak revenue each scenario earns to reflect a longer time to peak, and reflected the refinancing payoff in net debt. We also moved weight from the bull case to the base case: bull 25% from 30%, base 50% from 45%. The voucher is added back at about $200M.
| Scenario | Value | Probability | Implied EV / peak sales | Conditions |
|---|---|---|---|---|
| Bull | $95 | 25% | 3.4x on $4.0B | Approval around April 2027 with pancreatitis in the indication; the ramp beats the first mover's on dosing and payer preparation; voucher sold near $200M; financing is a convertible at a premium |
| Base | $75 | 50% | 3.6x on $3.0B | Approval on schedule with pancreatitis data in the clinical section; class-paced ramp concentrated in secondary prevention; peak reached two to three years later than we assumed in August; a 2027 convertible |
| Bear | $35 | 25% | 2.6x on $2.0B | Approval slips or the label carries a glycemic warning; payer criteria restrict use to prior-pancreatitis patients; an equity raise below $40 before the ramp is visible |
The probability-weighted value of about $70 is 46.5% above the $47.78 close. The weighted value fell by about a third and the price fell by about 45%, so the gap between them is wider than in August. We do not take that as a sign that the market is simply wrong. A good part of the gap is the market charging for a financing we also expect, and for a ramp it has watched in real time.
Thesis Scorecard Post-Earnings
| Thesis Point | Status | Notes |
|---|---|---|
| Bull 1: The severe-hypertriglyceridemia expansion is the whole case, and it is de-risked | Confirmed | Peer-reviewed data held: pancreatitis rate 1 vs 5 per 100 patient-years, number needed to treat 3 in secondary prevention. STAR entry, supplement filed November 2, PDUFA date within about 50 days, launch Q2 2027. The regulatory path is clearer; the commercial ramp moves to Bear 9. ON TRACK, unchanged |
| Bull 2: The TRiM platform extends beyond the liver | Confirmed | ARO-MAPT 73% mean CSF tau knockdown against a 50% to 60% benchmark, which the company calls the first clinical validation of siRNA delivery across the blood-brain barrier. ARO-DIMER-PA validated dual silencing, with a tolerability signal at the highest doses. EARLY EVIDENCE to ON TRACK |
| Bull 3: Serial partnering funds the pipeline | Neutral | No new deal in the quarter. Sarepta annual fee of up to $50M due; Novartis deferral $112.9M. US plozasiran and ARO-MAPT ruled out for partnering; obesity, liver, complement and early CNS open. ON TRACK, unchanged |
| Bull 4: The FCS label is best-in-class and transferable | Confirmed, with a caveat | Germany live, UK authorised, over 160M US lives with a coverage pathway, prescribers over 860. England's funding covers genetically confirmed patients only, so it does not fund the clinical-diagnosis label. ON TRACK, unchanged |
| Bear 1: The cost base | Confirmed, turning | Operating expenses ex-voucher a record $259.4M, up 5.7%. Research down 3.9% a quarter early; SG&A up 42.5% with the organisation now "largely in place." MATERIALIZING, unchanged; first evidence of a ceiling |
| Bear 2: Powering of the pancreatitis endpoint | Challenged | The pooled result held in peer review. New watch item: worsening glycemic control 14.3% vs 8.7%. Mean reduction and imaging subgroup size still undisclosed. CONTAINED, unchanged |
| Bear 3: Revenue is milestone timing, not a business | Confirmed, escalating | Revenue $37.8M, a third below consensus; recurring base down 24.8% QoQ and 40.7% in two quarters. Product revenue is 8.2% of the total. MATERIALIZING, unchanged |
| Bear 4: Funding | Confirmed | Like-for-like cash down $409.0M ($194.0M before the voucher). Refinancing payoff exceeded carrying value by about $209M; net about $80.6M on funded debt, about ($319.8)M with the royalty liability. Five to six quarters to the covenant floor. Offsets: no maturities before late 2031, no sweeps, a saleable voucher. EMERGING, unchanged |
| Bear 5: Valuation and positioning | Challenged | Down 45.0% since our upgrade. EV of about $7.19B is 1.8x to 2.4x the peak framing against 3.0x to 4.1x in August. The valuation concern has reversed into the case for the rating. CONTAINED, unchanged |
| Bear 6: Prescription-to-revenue conversion is unexplained | Confirmed, improving | Three launch metrics now committed and consistent across quarters; prescriber count survived two quarters. Conversion withheld for the sixth quarter; product revenue up 30.1% against prescriptions up 63%; a bridge programme confirmed. MATERIALIZING, unchanged |
| Bear 7: Material items disclosed only in the filing | Partly challenged | Improved: voucher resale range, product revenue, Novartis deferral and refinancing terms stated aloud. Filing-only: the $209M payoff gap, the $1.1B convertible basket, the amended Ionis complaint and trial date, the expected gross-margin decline. EMERGING, unchanged |
| Bear 8: The Greater China vehicle is being diluted and drained | Challenged | The 10-K finally explains the allocation: the entity earned $14.5M on $31.6M of revenue in fiscal 2026. FQ4 allocation ($0.9)M; ownership stable at 56.38%. EMERGING to CONTAINED |
| Bear 9 (new): The sHTG market builds slowly for the whole class | New | The first mover halved its 2026 guidance to about $50M after $9M in its first full sHTG quarter with about 1,500 prescribers. Arrowhead's own FCS conversion is "still a choke point." Management declined the read-across and declined guidance. EMERGING |
Overall: the thesis is weaker on timing and stronger on substance. Every scientific and regulatory question we carried in August was answered in Arrowhead's favour or left unchanged: the pivotal data survived peer review, the FDA put the supplement on its fastest track, and the brain-delivery platform beat its own benchmark. What changed is the evidence on how quickly a new market pays for a good drug. It came from a competitor, and it caused the stock's worst single session. One new bear point, Bear 9, now carries more of our valuation than any of the old ones. The balance sheet improved in structure and deteriorated in substance in the same quarter. Product revenue is growing, but at half the rate of the demand management asks us to watch.
Action: buy, at a smaller position than we argued for in August. We upgraded at $86.91 and the stock is at $47.78. The error was not in the data, which held, but in how much a slow commercial ramp would cost a stock priced for a fast one. The ramp is now the market's central assumption. At about 2.4 times the low end of management's peak framing, with approval odds that STAR entry has raised, a saleable $200M voucher and five to six quarters of cash, we think the price has moved further than the outlook. We would revisit the rating on three things: a PDUFA date later than the second quarter of 2027, a transition-quarter product revenue figure that fails to grow, or an equity raise below $40 before approval.