CRSP — Q2 2026 Financial Report Analysis

Overview

CRISPR Therapeutics' net loss fell to $91.2 million in the second quarter of 2026, from $208.5 million a year earlier. Most of that improvement is one item dropping out, not a change in the business: last year's quarter carried a $96.3 million charge for licensing drug rights from Sirius Therapeutics. Reported revenue rose to $10.2 million from $0.9 million, but $10.0 million of it is a single upfront payment from a license deal the company itself calls "immaterial".

The quarter's real operating news is Casgevy, the sickle-cell and beta-thalassemia gene therapy CRISPR co-owns with Vertex Pharmaceuticals. Casgevy sales reached $76 million, up 78% from the first quarter and 151% from a year ago. Those sales are booked by Vertex, not CRISPR, so they never appear in CRISPR's revenue line. They show up indirectly, by shrinking the net cost CRISPR pays Vertex for its 40% share of the program. That cost fell to $40.3 million from $45.2 million. The program is still losing money for CRISPR, just less of it.

The balance sheet is the other story. A $600 million convertible bond sold in March lifted cash and investments to $2.36 billion, up from $1.98 billion at the end of 2025. That pays for a pipeline that now includes three in vivo (inside-the-body) liver-editing programs in human trials.

At a glance

• $76 million in Casgevy sales (+151% year on year): the approved product is growing quickly, but CRISPR's 40% share still cost it $40.3 million net this quarter, so the program has not yet turned profitable for CRISPR.
• $91.2 million net loss, down from $208.5 million: about $94 million of the $117 million improvement comes from a one-off licensing charge in last year's quarter that did not repeat.
• $2.36 billion of cash and investments: at the first half's pace of cash use ($192.4 million in six months), that is several years of funding. Part of it is borrowed, from a $600 million convertible bond due 2031.

Key figures

Percentage changes on revenue and profit are marked n/m (not meaningful) where the base is tiny or both periods are losses. Operating margin (the share of revenue left after operating costs) is not shown: with $10 million of revenue against $127 million of costs, it would be a large negative number that says nothing useful.

Takeaway: Casgevy's growth is real, with sales up 151% year on year. But it still reaches CRISPR only as a smaller bill, $40.3 million this quarter against $45.2 million a year ago. It will not become income until the program as a whole turns profitable. Even then, $221.8 million of costs CRISPR deferred in 2022–2024 must first be repaid to Vertex out of those profits. For the next several years, CRISPR's value rests on its pipeline and its $2.36 billion cash pile far more than on Casgevy's earnings.

How Casgevy money actually flows to CRISPR

Casgevy is an ex vivo therapy. A patient's own blood stem cells are taken out, edited with CRISPR/Cas9 to switch back on production of fetal hemoglobin (the oxygen-carrying protein babies make before birth), then returned to the patient. Vertex leads development, manufacturing and sales. Under their amended agreement, the two companies split the program's profits and losses 60/40, with Vertex taking 60%.

That means Casgevy's sales never show up as CRISPR revenue. CRISPR's 40% share of program costs, minus its 40% share of program revenue, lands in one line: collaboration expense, net. The filing says the drop to $40.3 million from $45.2 million was "primarily attributable to an increase in our share of CASGEVY revenue." For the first half, the line fell to $86.2 million from $102.7 million, which implies $45.9 million in the first quarter and $40.3 million in the second. The quarterly bill is shrinking as sales ramp, but it is still a cost.

Two contract details make the eventual payoff smaller than the sales growth suggests:

• $221.8 million of deferred costs. In 2022, 2023 and 2024, CRISPR used an option to postpone part of its share of Casgevy spending. Those amounts are payable to Vertex only out of future Casgevy profits, capped at a set amount per year. None has been recorded as a liability yet, because repayment isn't yet probable. The practical effect is that CRISPR's first slices of Casgevy profit are already spoken for.
• A $43.0 million priority review voucher payment. If Vertex uses or sells the FDA voucher it received for Casgevy before the program has a profitable calendar year, it owes CRISPR $43.0 million (or 42% of the sale proceeds). This hasn't been recognized either.

On the commercial side, the filing and press release point to more patients able to get the therapy. Casgevy is approved in 39 countries. Germany agreed reimbursement for patients 12 and older in May. In July, the FDA approved Casgevy for children as young as 2. The company says that approval makes roughly 5,500 more US patients potentially eligible, and it came 53 days after filing. Submissions for children aged 5 to 11 have also been filed in Saudi Arabia and the UK.

Spending and the pipeline

Operating expenses fell to $127.5 million from $230.2 million. Excluding the acquired-R&D charge, they fell to $125.0 million from $134.0 million. The pieces:

• R&D: $67.2 million vs $69.9 million. Employee costs fell by $5.2 million to $11.7 million, and facility costs fell by $1.6 million. License and sublicense fees rose by $4.4 million to $5.1 million. Spending on outside research (trials, contract labs) was flat at $20.8 million.
• Acquired in-process R&D: $2.5 million vs $96.3 million. Last year's charge was the cost of entering the Sirius siRNA deal: $25.0 million in cash plus 1,842,105 CRISPR shares issued at $38.00 each (about $70.0 million).
• G&A: $17.6 million vs $18.9 million, on lower employee costs, including stock-based pay.

Where that money is going, per the 10-Q and earnings release:

• In vivo liver editing. These use a lipid nanoparticle, a tiny fat bubble that carries the gene-editing machinery into liver cells after an infusion, with no cell removal needed.
  • CTX310 turns off ANGPTL3, a gene that raises blood fats. It is in a Phase 1b trial focused on severe hypertriglyceridemia (very high blood triglycerides) and refractory high cholesterol.
  • CTX340 targets angiotensinogen, a protein that drives blood pressure. It received FDA clearance and started a Phase 1 trial in hard-to-treat hypertension.
  • CTX460 treats alpha-1 antitrypsin deficiency. It is the first candidate from CRISPR's own "SyNTase" gene-correction platform and also started Phase 1.
  • CTX321, a reworked Lp(a) program, is in pre-trial studies.
• siRNA. These drugs silence a gene temporarily rather than editing it. CTX611, a long-acting Factor XI blocker meant to prevent blood clots with less bleeding risk, is in a Phase 2 trial in knee-replacement patients. It is co-developed 50/50 with Sirius.
• Cell therapy. Zugo-cel (formerly CTX112) is an off-the-shelf CAR-T, an immune cell engineered to hunt down specific targets, here B cells. It is in Phase 1 trials for lupus, systemic sclerosis, myositis and two autoimmune blood disorders, a new trial in autoimmune neurological diseases, and a Phase 1/2 trial in B-cell cancers.

What the headline numbers hide

• The loss improvement is mostly a comparison effect. Net loss narrowed by $117.4 million. $93.8 million of that is last year's Sirius licensing charge not recurring. Revenue added another $9.3 million, nearly all of it the one-time $10.0 million upfront payment. Strip out both, and lower collaboration costs, R&D and G&A together improved the result by about $9.0 million (our arithmetic from the filing's line items). Most of the remainder came from higher other income.
• "Revenue up more than tenfold" means nothing here. Collaboration revenue was $10.0 million this quarter and $11.0 million for the half. The 10-Q says it came from "an upfront payment" on "an immaterial license and collaboration agreement". The filing doesn't name the partner, and nothing indicates the payment will repeat.
• Loss per share fell faster than the loss, partly because there are more shares. Weighted diluted shares rose 10.8%, to 96.5 million from 87.1 million. That reflects shares issued since then, including the 1,842,105 issued to Sirius and sales through CRISPR's at-the-market programs (selling stock gradually into the market). No shares were sold in the first half of 2026, and $557.2 million of capacity remains under the current program.
• The new bond adds potential dilution and a small interest cost. The $600 million of convertible notes due March 2031 pay 1.7308% a year. The coupon was raised from 1.125% so that a 35% Swiss withholding tax on interest doesn't hit bondholders. Holders can convert at about $76.56 per share, and up to 11.36 million shares could be issued. Interest and amortization cost $3.3 million in the quarter. Because the cash came from debt, shareholders' equity fell to $1,746.6 million from $1,921.8 million at year-end, even as cash rose.
• Cash use was a bit lower than the accounting loss. Operating cash outflow for the first half was $192.4 million, against a $214.1 million net loss. That gap is normal once you add back about $34 million of stock-based pay, a non-cash cost. Cash use rose from $167.8 million a year earlier, largely on timing: the first half of 2025 included a $25.0 million milestone payment received from Vertex.
• Other income is partly market noise. Other income of $27.1 million (vs $22.1 million) includes interest earned on the investment portfolio. It also includes gains or losses on corporate equity stakes CRISPR holds, which drove the $5.0 million increase. That piece can swing either way next quarter.

Outlook

CRISPR gives no revenue or earnings guidance. Its financial commitment is that existing cash will fund operations and capital spending for "at least the next 24 months", before any further partner payments. That is a conservative floor. At the first half's operating cash use of about $96 million a quarter, $2.36 billion would last considerably longer. The cost base will rise, though, if the in vivo and zugo-cel programs move into larger trials.

Milestones the company has flagged for the second half of 2026:

• A durability update on CTX310's Phase 1a trial, scheduled for presentation at the European Society of Cardiology Congress on August 28, and a Phase 1b update later in the year.
• A Phase 2 update for CTX611.
• Clinical updates on zugo-cel in both autoimmune disease and B-cell cancers. A combination study with Lilly's pirtobrutinib in aggressive lymphomas has also started.
• Updates on the Lp(a) program and on CRISPR's option to pick up to two more siRNA targets with Sirius.

Our read: the main financial metric to watch is the collaboration expense line. If Casgevy keeps growing at anything like this year's pace, that line keeps shrinking toward zero, and at that point the program stops costing CRISPR cash. Profit sharing won't start right away, because the $221.8 million of deferred costs comes first. Before then, the stock's direction will be set mostly by the CTX310, CTX611 and zugo-cel data readouts listed above. The next 10-Q, for the third quarter, is likely in early November, based on the May 4 and August 3 filing dates this year.

This analysis is based on CRISPR Therapeutics' Form 10-Q for the quarter ended June 30, 2026, and its earnings release of August 3, 2026.