Phase Space AI

Financial Model Notes

Halozyme Therapeutics [HALO]

Halozyme Therapeutics [HALO] — Financial Model Notes

2026-07-29. Every figure below is traceable to a named primary filing. [HOUSE] marks a house estimate; [GUIDANCE] marks management guidance. Nothing else is asserted.


1. Data lineage

Item Source document Accession Filed
FY2025 / FY2024 / FY2023 income statement 10-K, R5.htm 0001159036-26-000024 2026-02-17
Revenue disaggregation FY2023–FY2025 10-K, R54.htm same same
Customer and geographic concentration 10-K, R40.htm same same
Revenue narrative, remaining performance obligations 10-K, R55.htm same same
Royalty revenue by product FY2024/FY2025 10-K MD&A same same
Royalty term and patent expiry language 10-K, revenue-recognition policy and IP section same same
Merck litigation, German injunction, USPTO PGRs 10-K, legal and risk sections same same
Q1-2026 income statement, balance sheet 10-Q, R4.htm, R2.htm 0001159036-26-000068 2026-05-11
Q1-2026 revenue disaggregation 10-Q, R46.htm, R47.htm same same
Business combinations (Elektrofi, Surf Bio) 10-Q, R38.htm; 10-K R46.htm both both
Debt terms 10-Q, R59.htm; 10-K R67.htm both both
Share count 10-Q cover, R1.htm 0001159036-26-000068 2026-05-11
FY2025 adjusted-EBITDA reconciliation 8-K Ex-99.1 0001159036-26-000023 2026-02-17
FY2026 guidance 8-K Ex-99.1 0001159036-26-000023 2026-02-17
FY2026 guidance reiteration; $1bn buyback 8-K Ex-99.1 0001159036-26-000067 2026-05-11
Cash flow, SBC XBRL companyfacts, us-gaap taxonomy max filed 2026-05-11
Prices, volume, volatility Alpaca /v2/stocks/bars, adjustment=all 2026-07-29
Options chain Alpaca /v2/options/contracts + /v1beta1/options/snapshots 2026-07-29

Lineage note. HALO's XBRL companyfacts is current (max filed 2026-05-11) — it contains the Q1-2026 10-Q. This contrasts with ALNY, whose feed stops at 2026-04-06. Both were checked; the difference is issuer level, which is what makes staleness dangerous to assume away.

Taxonomy warning encountered and respected. The peer-multiples tool reported for HALO: "stale taxonomies present and MUST NOT be read: srt ends 2021-12-31. Fresh: dei, us-gaap." All figures here are read from us-gaap and dei only.


2. Revenue build — three lines, and the lumpy sub-lines broken out

2.1 Annual, $m

FY Royalties Product sales, net Collaborative agreements Total Total YoY
2021 443.3
2022 360.5 (Antares acquired May-2022) 660.1 +48.9%
2023 447.9 300.9 80.5 829.3 +25.6%
2024 571.0 303.5 140.8 1,015.3 +22.4%
2025 867.8 376.4 152.3 1,396.6 +37.6%
2025 mix 62.1% 27.0% 10.9% 100%

FY2022's line-level split beyond royalties was not retrieved; only total $660.1m and royalties $360.5m are used. FY2021 total $443.3m is from the FY2022 release.

2.2 Product sales, decomposed — this is where the leading indicator lives

$m FY2023 FY2024 FY2025 Q1-2025 Q1-2026 Q1 YoY
Proprietary product sales (Hylenex, XYOSTED) 130.8 166.6 194.6 36.3 41.9 +15%
Bulk rHuPH20 sales 115.4 86.3 133.0 27.0 73.5 +172%
Device partnered product sales 54.6 50.5 48.8 14.7 15.1 +3%
Total product sales, net 300.9 303.5 376.4 78.0 130.4 +67%

Modelling note: bulk rHuPH20 is the forward indicator. It is enzyme sold to partners for manufacturing and is ordered ahead of partner production. It nearly tripled year on year in Q1-2026 while HALO's own proprietary products grew 15%. The Q1-2026 10-Q discloses $265.8m of remaining performance obligations on unfulfilled product purchase orders, expected to be delivered by end-2027 (FY2025 10-K: $289.5m). This is the single most under-followed number in the name.

2.3 Collaborative agreements, decomposed — the lumpy component, isolated

$m FY2023 FY2024 FY2025 Q1-2025 Q1-2026
Upfront licence and target-nomination fees 2.0 27.0 18.5 0.2 0.0
Event-based development and regulatory milestones 69.0 72.5 47.0 0.0 0.0
Sales-based milestones 0.0 30.0 70.0 15.0 0.0
Sub-total: LUMPY 71.0 129.5 135.5 15.2 0.0
Device licensing and development revenue (recurring) 9.5 11.3 16.9 3.4 5.6
Total collaborative 80.5 140.8 152.3 18.6 5.6
Lumpy as % of total revenue 8.6% 12.8% 9.7% 5.7% 0.0%

TTM lumpy through 2026-03-31 = $120.3m = 8.0% of TTM revenue. Recurring TTM = $1,508.5m − $120.3m = $1,388.2m, which is the reverse-DCF base.

2.4 Royalty revenue by product, $m — the concentration answer

Product Partner FY2024 FY2025 Change % of FY2025 royalties % of FY2025 total revenue
DARZALEX Janssen 374.8 482.7 +29% 55.6% 34.6%
VYVGART Hytrulo argenx 28.9 157.2 +444% 18.1% 11.3%
Phesgo Roche 70.1 105.6 +51% 12.2% 7.6%
Other various 97.2 122.3 +26% 14.1% 8.8%
Total 571.0 867.8 +52% 100% 62.1%

Customer concentration, % of total revenue (10-K; counterparties are lettered, not named):

FY2023 FY2024 FY2025
Partner A 44% 41% 41%
Partner B 19% 17% 17%
Partner C 7% 9% 16%
Partner D 10% 8% 6%

The 10-K separately states Janssen and Roche = 69% of accounts receivable.

Geographic revenue, $m: US 587.2 → 690.5 → 823.0; Switzerland 149.0 → 212.4 → 276.4; Belgium 58.4 → 84.0 → 217.5 (+159%); Japan 15.1 → 18.9 → 46.4; all other foreign 19.6 → 9.5 → 33.3.


3. Cost structure and the operating-margin correction

$m FY2023 FY2024 FY2025 Q1-2025 Q1-2026
Cost of sales 192.4 159.4 228.8 48.4 79.2
Amortisation of intangibles 73.8 71.0 76.7 17.8 29.5
Research and development 76.4 79.0 81.5 14.8 25.6
Selling, general and administrative 149.2 154.3 207.1 42.4 57.9
Impairment of intangible asset 0.0 0.0 48.7 0.0 0.0
Acquired in-process R&D expense 0.0 0.0 284.9 0.0 0.0
Total operating expenses 491.7 463.8 927.6 123.3 192.2
Operating income 337.6 551.5 469.0 141.5 184.5
GAAP operating margin 40.7% 54.3% 33.6% 53.4% 49.0%
Clean operating margin (ex IPR&D and impairment) 40.7% 54.3% 57.5% 53.4% 49.0%

Total gross margin (revenue less cost of sales) = 83.6% FY2025, 79.0% Q1-2026. The decline is mix: bulk rHuPH20 carries a lower margin than royalties, and it is the fastest-growing line.

The 33.6% figure the screen used as a terminal margin is a one-year artefact. FY2024 was 54.3%, FY2023 was 40.7%, and Q1-2026 is 49.0%. Only FY2025 carries the $284.9m IPR&D and $48.7m impairment. Any model that takes a trailing GAAP margin off FY2025 without restating it will misprice this name.


4. The adjusted-EBITDA reconciliation, as filed — model this exactly

$m FY2025 FY2024
GAAP net income 316.9 444.1
less investment and other income, net (21.5) (24.4)
plus interest expense 18.1 18.1
plus income tax expense 150.0 113.0
plus depreciation and amortisation 88.1 81.3
EBITDA 551.6 632.2
plus transaction costs for business combinations (Elektrofi) 14.6
plus intellectual-property litigation costs (Merck) 16.7
plus severance and share-based-compensation acceleration (Elektrofi) 24.6
plus impairment of intangible asset 48.7
plus other one-time items 1.4
Adjusted EBITDA (as filed) 657.6 632.2
Acquired IPR&D expense (Surf Bio) — NOT added back (284.9)
Adjusted EBITDA, restated for the IPR&D charge [HOUSE] 942.5 632.2
Growth rate As filed Restated
FY2024 → FY2025 adjusted EBITDA +4.0% +49.1%
FY2025 → FY2026 guidance ($1,125–1,205m) +71% to +83% +19.4% to +27.9%

Modelling instruction: underwrite +19% to +28%. The +71–83% headline in the Q1-2026 release is arithmetic against a base carrying a one-off $284.9m acquisition charge that the company's own adjusted-EBITDA definition — which explicitly excludes impairments, transaction costs, severance and litigation — declines to exclude. The same distortion runs through the non-GAAP diluted EPS growth of +87% to +99%.

Q1-2026 for reference: EBITDA $218.3m; adjusted EBITDA $229.5m (vs $162.0m in Q1-2025, +41.7%); GAAP diluted EPS $1.22; non-GAAP diluted EPS $1.60 (vs $1.11, +44.1%). Note that the clean quarter-on-quarter growth rates (+42%) match revenue growth (+42%), not the +71–83% annual headline.


5. Balance sheet, 2026-03-31

$m 2026-03-31 2025-12-31
Cash and cash equivalents 309.7 133.8
Marketable securities 8.9 9.0
Restricted cash 2.3 2.6
Total cash and investments 320.9 145.4
Accounts receivable, net and contract assets 458.0 441.3
Inventories (current) 155.5 176.5
Goodwill 582.3 580.4
Intangible assets, net 952.0 981.5
Total assets 2,672.7 2,525.3
Current portion of long-term debt, net 208.7 0.0
Long-term debt, net 1,935.9 2,142.6
Total liabilities 2,453.1 2,476.5
Total stockholders' equity 219.6 48.8
Shares issued and outstanding 118,523k 117,782k

Convertible debt, gross principal $2,179.6m:

Series Principal Coupon Conversion price Maturity Status at $83.62
2027 $209.6m 0.25% $77.17 2027-03-01 in the money; reclassified current
2028 $470.0m 1.00% $56.02 2028 deep in the money
2031 $750.0m 0.00% $87.20 2031 just out of the money
2032 $750.0m 0.875% $87.20 2032 just out of the money

Capped calls were purchased on the 2031/2032 issues (cap price $136.78, $106.8m and $104.0m paid) and on the 2028 issue (cap $75.4075). A $750m revolving credit facility exists under the Credit Agreement.

Future debt maturities and interest: 2026 $11.8m; 2027 $221.1m; 2028 $481.3m; 2029 $6.6m; 2030 $6.6m; thereafter $1,513.1m.

Scheduled intangible amortisation: 2026 $118.0m; 2027 $118.0m; 2028 $118.0m; 2029 $83.3m; 2030 $60.6m; thereafter $483.4m.


6. Cash flow

$m FY2022 FY2023 FY2024 FY2025
Operating cash flow 240.1 388.6 479.1 651.6
Capital expenditure 4.8 15.3 10.7 7.0
Free cash flow 235.3 373.3 468.4 644.6
Stock-based compensation 24.4 36.6 43.4 51.6
SBC as % of revenue 3.7% 4.4% 4.3% 3.7%

Accruals check: (316.9 − 651.6) / ((2,525.3 + 2,063.5) / 2) = −14.6%. Cash comfortably exceeds earnings. PASS.

FCF yield: $644.6m / $9,918m market cap = 6.5%; / $12,099m EV = 5.3%.

Capital intensity is negligible — $7.0m of capex on $1,396.6m of revenue (0.5%). This is a royalty business and the model should treat it as one: incremental revenue converts to cash at close to the incremental margin.

Share count is declining. Weighted diluted: 134.2m (FY2023) → 129.4m (FY2024) → 123.9m (FY2025) → 122.9m (Q1-2026). The $1bn authorisation to 2028-12-31, with ≥$400m projected in 2026, continues this.


7. FY2026 — [GUIDANCE]

Line Guidance vs FY2025 actual
Total revenue $1,710 – 1,810m $1,396.6m → +22% to +30%
Royalty revenue $1,130 – 1,170m $867.8m → +30% to +35%
Adjusted EBITDA $1,125 – 1,205m $657.6m → +71% to +83% as filed; +19% to +28% restated
Non-GAAP diluted EPS $7.75 – 8.25 +87% to +99% as filed; distorted the same way
Included investment ~$60m of new Hypercon and Surf Bio spend
Stated expectation ENHANZE royalty revenue exceeds $1 billion

Q1-2026 delivered $376.7m, about 21% of the $1,760m midpoint — the ramp is second-half weighted and untested. That is invalidation condition #4 in HALO_Trade_Construction.md.


8. FY2027 — [HOUSE], isolated so it can be replaced

No guidance and no consensus exists (Alpha Vantage quota exhausted — see disclosed limitations).

Line FY2027 [HOUSE] Reasoning
Royalties $1,300m +12% on the FY2026 midpoint. DARZALEX indication expansion; VYVGART all-serotype gMG plus ocular MG; RYBREVANT FASPRO full year. Deliberately decelerating from +30–35%, because the US rHuPH20 patent expires in 2027 and the step-down timing is undisclosed
Product sales $600m Bulk rHuPH20 momentum plus the $265.8m of disclosed unfulfilled purchase orders deliverable through 2027
Collaborative $100m Below FY2025's $152.3m. Vertex/Oruka/GSK upfronts and target nominations, no large sales-based milestone assumed
Total ~$2,000m +14% on the FY2026 midpoint
Non-GAAP diluted EPS [HOUSE] ~$9.75 Operating leverage on $2.0bn plus ~4% share-count reduction from the funded buyback

TTM revenue reportable twelve months from now (through Q2-2027) ≈ (1,760 + 2,000) / 2 = $1,880m. This is the denominator of the 12-month target and the single most replaceable input in the model.


9. The unmodellable input — stated, not estimated

The magnitude of the post-2027 rHuPH20 royalty rate reduction is not disclosed anywhere in HALO's filings.

What is disclosed: - Royalties run at "an on average mid-single digit percent rate" of partner sales. - The rate is "reduced" — not terminated — in a country upon expiry of HALO's rHuPH20 patents, where no collaboration-specific patent covers the product. - "Janssen's patents covering DARZALEX SC do not impact the timing for this royalty reduction." - US rHuPH20 expires 2027; European 2029; additional patents valid into 2029. - DARZALEX is 34.6% of FY2025 total revenue; US revenue is 58.9% of total.

What is not disclosed and is not estimated here: the reduced rate, the products it applies to individually, and whether any collaboration-specific patents cover DARZALEX SC in any given country.

Modelling instruction: do not put a number on this. The correct treatment, used in HALO_Valuation.md, is to discount the exit multiple with the argument stated in the open (15.0x EBIT against a 27.4x growth-matched anchor) and to identify the flip point (8.0x EBIT) so a book can express its own view. A specific-looking percentage would be a fabricated number in the highest-leverage position in the model.


10. Model verification and known weaknesses

Verified: - Revenue components sum to the reported totals: 867.8 + 376.4 + 152.3 = 1,396.6 ✓ (FY2025); 240.7 + 130.4 + 5.6 = 376.7 ✓ (Q1-2026). - Product sub-components sum: 194.6 + 133.0 + 48.8 = 376.4 ✓; 41.9 + 73.5 + 15.1 = 130.4 ✓. - Collaborative sub-components sum: 18.5 + 47.0 + 70.0 + 16.9 = 152.3 ✓ (rounding). - Royalty-by-product sums to the royalty line: 482.7 + 157.2 + 105.6 + 122.3 = 867.8 ✓. - Adjusted-EBITDA reconciliation ties from GAAP net income to $657.6m ✓, and the implied FY2025 base from the guidance growth rate ($1,125m / 1.71 = $658m) independently confirms it ✓. - Operating margin arithmetic: 469.0 / 1,396.6 = 33.58% ✓ (the screen's 33.6%). - Market cap reconciles to the screen: 83.62 × 118.613m = $9.92bn vs screen $9.9bn ✓. - Share count cross-checked: 10-Q cover 118,613k (2026-04-30); balance sheet 118,523k issued and outstanding (2026-03-31); GAAP diluted weighted 122,875k (Q1-2026). - Demonstrated CAGR reproduces the screen: (1,396.6 / 660.1)^(1/3) − 1 = 28.36% vs screen 28.4% ✓.

Known weaknesses, stated: 1. The royalty step-down magnitude is unmodellable (§9). This is not a gap more work would close. 2. FY2022 line-level revenue split was not retrieved. The recurring-versus-lumpy decomposition therefore runs FY2023→FY2025 (two years). The three-year window is used only for total and royalty-only CAGRs, where the components are known. 3. Partner A/B/C/D are not named in the 10-K. The inference that Partner A is Janssen is reasonable from the geographic and product data but is not asserted as fact anywhere in this memo. 4. No consensus. FY2027 is a house estimate. 5. The EV/TTM-sales history uses annual capital-structure snapshots, not daily. HALO's share count fell from 147m to 123m and its net debt rose from $1.1bn to $1.8bn over the window, so this matters more here than for a stable-structure name. The current-day figure (8.02x) is exact. 6. Convertible treatment is conservative (GAAP diluted shares plus full debt at carrying value) and slightly overstates EV if the in-the-money 2027/2028 series convert. 7. No regressed beta. WACC of 10% is the framework default, named as such. 8. Terminal margin of 50% is a judgement, sensitivity-tested from 33.6% to 55% rather than defended.