MRK — Q2 2026 Financial Report Analysis

A $5.7 billion deal charge turns a 5% sales-growth quarter into a loss

Merck's second quarter of 2026 is two stories that point in opposite directions. The commercial business grew: sales of $16,607 million were up 5.1% on the same quarter a year earlier (4% excluding currency movements), with Keytruda, Winrevair, Welireg and Animal Health all adding volume. The reported bottom line went the other way: a net loss of $1,335 million, against a $4,427 million profit a year ago.

Almost all of that swing is one accounting entry. In May 2026 Merck bought Terns Pharmaceuticals for $6.8 billion, and because the deal was essentially the purchase of a single unapproved drug candidate — MK-4208, an oral leukemia treatment in Phase 1/2 trials — accounting rules required Merck to expense nearly the whole price immediately rather than carry it on the balance sheet as an asset. The filing puts it plainly: Merck "recorded a charge of $5.7 billion to Research and development expenses (which primarily represented acquired in-process research and development [IPR&D] with no alternative future use) in the second quarter." That is $2.31 per share, and it is the entire difference between a loss and a profit.

It is not a cash charge in the usual sense either — the money left in the quarter, but it shows up as an investment, not as an operating cost. Cash from operations in the first half was $9,288 million, up 60% from $5,793 million a year earlier.

Key figures

Operating margin — the share of revenue left after the costs of making and selling the products and of research, but before interest and tax — is calculated here from the income statement: sales less cost of sales, SG&A, R&D and restructuring costs. "Ex-FX" means excluding the effect of changes in currency exchange rates, which flatter or depress reported growth for a company that sells in dozens of currencies but reports in dollars. The last row strips the $5.7 billion Terns charge out of 2026 and the $200 million Hengrui licensing payment out of 2025.

Takeaway: The loss is an artifact of how a single deal is booked, not of the business deteriorating — but the quarter still contains a real problem underneath it. Gross profit was lower than a year ago in absolute dollars ($12,212M vs $12,249M) despite $801M more revenue, because cost of sales rose 23.6%. Merck grew the top line and got nothing through to gross profit for it.

What actually happened to profitability

Strip out Terns and Merck's underlying operating margin was roughly 30.8% this quarter against about 32.9% a year earlier — a two-point decline that has nothing to do with acquisition accounting. Two things caused it.

Cost of sales rose $838 million, or 23.6%, on $801 million of extra revenue. Roughly $491 million of that increase is acquisition-related: charges booked to cost of sales rose to $1,067 million from $576 million, mostly amortization of intangible assets — the gradual writing-down of the value assigned to drugs Merck bought rather than invented, which flows through cost of sales even though no extra manufacturing happened. Merck's own explanation for the gross-margin drop is "higher amortization of intangible assets and inventory write-downs." On Merck's non-GAAP basis, which removes those acquisition charges, gross margin still slipped to 81.1% from 82.2%, so about a point of the decline is genuine — inventory written off, not accounting.

SG&A grew faster than sales, up 10% to $2,904 million, "primarily due to higher administrative costs (including investments in IT), as well as higher promotional costs in support of product launches." Launch spending on Winrevair, Capvaxive, Ohtuvayre and now Lipfendra is real money going out ahead of the revenue it is meant to produce.

Partly offsetting both: restructuring costs fell to $151 million from $560 million.

The tax line is worth a note because it looks bizarre. Merck recorded a $654 million tax charge on a pretax loss of $683 million — an effective rate of -95.9%. The reason is that the Terns charge earns no tax deduction: the filing attributes 108.9 percentage points of that rate to the charge "for which no tax benefit was recorded." So the deal costs $5.7 billion pretax and $5.7 billion after tax.

The product picture: broadening, but still Keytruda

Keytruda (plus its new under-the-skin version, Keytruda Qlex) was $8,366 million, up 5.2%, and still 50.4% of everything Merck sells. Growth came from "strong global uptake in earlier-stage indications, including triple-negative breast cancer (TNBC), cervical cancer, head and neck cancer and bladder cancer, as well as higher global demand in metastatic indications, including urothelial cancer." Two qualifications belong on that number. First, Keytruda Qlex contributed $463 million — that is largely existing patients switching from the infused version, not new demand, though the switch is strategically the point (see below). Second, for the six-month period the filing flags "an approximate $250 million favorable impact due to the timing of wholesaler purchases" — a stocking effect that will not repeat, so year-to-date Keytruda growth of 8% overstates the underlying rate.

Beyond Keytruda, the launch portfolio is doing the work:

And the declines:

Three of those four declines deserve to be read differently from each other. Januvia is a finished story — an off-patent diabetes franchise in permanent generic decline, and the drag on it will shrink simply because the base keeps getting smaller. Vaxneuvance's 35% fall is mostly a comparison problem: the year-ago quarter was inflated by U.S. public-sector purchasing, so the drop exaggerates the underlying deterioration, though Merck does concede real "competitive pressure" — notably from its own Capvaxive, which grew $55 million while Vaxneuvance lost $81 million. Gardasil's +4% is the one genuinely encouraging reversal, after a period of weakness driven by China; the gain here came from Asia Pacific and Europe, and European tender timing means some of it is a shift between quarters rather than new demand.

The real question: what Merck is buying, and why

Merck has committed roughly $16 billion of cash to two asset acquisitions in six months — $9.2 billion for Cidara in January (a $9.0 billion IPR&D charge, taken in Q1) and $6.8 billion for Terns in May ($5.7 billion charged here). Both were structured as asset acquisitions rather than business combinations because a single drug candidate "accounted for substantially all of the fair value of the gross assets acquired," which is why the cost hits the income statement all at once instead of being spread over years. It also bought Verona (Ohtuvayre) last October and, in July, Targan for about $650 million to expand Animal Health's poultry business.

The financing is visible on the balance sheet. Cash and equivalents fell to $6,849 million from $14,565 million at the end of 2025, and long-term debt rose to $51,081 million from $46,750 million, including $6.0 billion of senior unsecured notes issued in May at coupons from 4.30% to 5.85% out to 2056. That is why "other (income) expense, net" swung to a $99 million expense from $7 million of income, which Merck attributes "primarily to higher net interest expense."

The strategic logic is Keytruda. The drug is half of Merck's revenue and its patents begin running out around the end of this decade. The filing's own disclosure of the erosion so far is modest — a biosimilar copy launched in Argentina in 2025, "further launches in smaller international markets during 2026," and the company "anticipates the impact of biosimilar erosion to Keytruda sales will be immaterial in 2026." But that is precisely the window in which a replacement has to be bought or built, and the spending pattern says Merck has chosen to buy. Keytruda Qlex is the other half of the defence: converting patients to a subcutaneous injection covered by separate, later-expiring formulation patents makes the franchise harder for an infused biosimilar to take. $463 million in the quarter is the first real evidence that conversion is happening.

Guidance: the headline cut hides a small raise

Merck raised and narrowed its full-year 2026 sales outlook to $66.3–67.3 billion, from $65.8–67.0 billion, which includes "a positive impact from foreign exchange of approximately 1% at mid-July 2026 exchange rates" — so perhaps half a point of the raise is currency rather than volume.

The EPS guidance looks alarming and is not. Non-GAAP EPS is now guided to $2.66–2.76, down from $5.04–5.16. The entire reduction is the Terns charge, which "was not previously included in the outlook": $2.31 per share for the write-off plus about $0.12 per share of costs to finance the deal and advance MK-4208, for $2.43 in total. Prior midpoint $5.10, minus $2.43, is $2.67 — the new midpoint is $2.71. Underlying guidance went up about $0.04, helped by roughly $0.15 per share of currency benefit.

The gross-margin guide coming down a point, from approximately 82% to approximately 81%, is the line to watch: it confirms that this quarter's margin pressure is expected to persist rather than reverse, and it is guided independently of the deal charges.

View on trajectory

Add the $5.93 per share of Cidara and Terns charges back to the $2.71 midpoint and Merck is guiding to roughly $8.64 of underlying full-year EPS, against $8.98 in 2025 — and 2025 itself carried $0.20 of business-development charges. So even setting the acquisitions entirely aside, and with about $0.15 of currency help, Merck expects to earn less per share this year than last, on 5%-ish sales growth. The gap is exactly what this quarter showed: rising amortization from bought-in products, inventory write-downs, launch spending growing faster than launch revenue, and higher interest on the debt used to fund the deals.

That is a reasonable trade if MK-1406 (the long-acting flu preventive from Cidara, now in Phase 3) and MK-4208 turn into products, and if the newer launches keep compounding — Winrevair at +75%, Welireg at +67% and Capvaxive at +42% are growing from bases now large enough to matter, and the FDA approval of Lipfendra (enlicitide), described as "the first and only once-daily oral PCSK9 inhibitor" for lowering LDL cholesterol, adds a potentially large primary-care product to a portfolio that has been narrowing toward oncology. The honest summary is that Merck is paying real, immediate, non-deductible money today for pipeline that may or may not replace Keytruda later this decade, while the current portfolio's profitability slowly erodes. The quarter neither proves nor disproves that bet; it just makes the size of it visible.

Source: Merck & Co., Inc. Form 10-Q for the quarterly period ended June 30, 2026 (accession 0000310158-26-000212, filed August 7, 2026), and the second-quarter 2026 earnings release filed as Exhibit 99.1 to the Form 8-K dated August 4, 2026.