A 19% earnings beat with 3% less profit: which Merck does $145 believe?

A 19% beat that earned 3% less than in 2025

Merck & Co. ($MRK) posted a quarter that looked like a blowout: non-GAAP EPS of $2.49 against a $2.09 estimate, revenue of $18.09 billion against $17.33 billion expected, and a 3.8% jump in the shares the next day. The same release shows non-GAAP EPS fell 3% from $2.58 in the quarter to September 2025, and GAAP EPS fell 12% to $2.04. The beat measures how low the bar was, not how much the company earned.

Sales grew 5%, but about $550 million of the $809 million increase came from U.S. customers buying early - a wholesaler stocking up before a price change is the pharma equivalent of a supermarket pulling December sales into November. Strip that out and growth was roughly 1.5%. Original IV Keytruda fell 5% to $7.77 billion while the new under-the-skin Qlex version added $830 million; Bridion lost 67% and Januvia/Janumet 39% to generics.

The tax line did the rest. The non-GAAP tax rate for the quarter was 10.9%, and the full-year rate guide was cut to 32-33% from 35-36%. That is most of why the EPS range rose to $2.77-2.82 from $2.66-2.76 while the operating expense guide went up, not down.

2027 starts with a $2 billion generic hole and $250 million of paybacks

On the 29 October 2026 call, management described 2027 as a year of investment: moderate revenue growth, more than $2 billion of headwinds from generic Bridion, the Januvia family and Adempas, higher expenses, higher interest cost and a higher tax rate. Of the $550 million timing benefit, about $300 million reverses in the quarter to December 2026 and about $250 million - a one-off CDC purchase of Gardasil - reverses in 2027. Keytruda growth is moderating ahead of key U.S. patents that begin expiring in 2028, with price pressure in Germany already visible.

The balance sheet is carrying the pipeline bet. The 10-Q for the quarter to June 2026 shows $8.78 billion paid in cash for Cidara and $5.84 billion for Terns, leaving $6.85 billion of cash against $53.9 billion of debt. In the first half of 2026 the company generated $9.29 billion of operating cash flow, spent $1.88 billion on capex, paid $4.25 billion in dividends and bought back $1.63 billion of stock, with about $3 billion of buybacks planned for the full year.

Fair value $145.82: why the model treats the pipeline as already paid for

At $144.68 the stock sits 0.8% below the FPI fair value of $145.82 - fairly valued, with no margin either way. The 145.5x trailing P/E is an accounting artifact: $14.7 billion of acquired-research charges produced a nine-month GAAP loss of $510 million, so trailing earnings are close to zero. Add the one-time charges back to the 2026 guide and underlying EPS lands at roughly $8.82-8.87, which puts the stock near 16x - below the industry median of 21.3x.

The problem is that underlying EPS in 2025 was $9.18 before $0.20 of similar charges, so the core business earns about 4% less while the share price is up 37.4% in 2026 and 72.5% over twelve months. That is pure multiple expansion: P/S of 5.3x is the highest since 2022 and above the 3.6x industry median, and EV/EBITDA of 31.0x compares with 14.0x for peers.

The DCF half of the model is driven by cash the business produces today and how fast it grows. A free cash flow yield of 4.50% (industry 4.42%) and a dividend yield of 2.35% are respectable; a three-year revenue growth rate of 3.1%, ROIC of 5.4% against an industry 10.5%, and roughly $47 billion of net debt are what keep the number from rising further. The 30% peer weight sees a company priced like a grower on sales and earning like a mature one on margins (net margin 3.6% versus 14.4% for the industry).

Street vs model: a $165 median against $145.82, a 13% gap with a date on it

Wall Street rates the stock a buy - 26 buy, 11 hold, 1 sell. The median target is $165, the mean $160.47, the range $105 to $180. After the results, Guggenheim went to $171 and KeyBanc to $165 on 30 October 2026, Cantor Fitzgerald to $155 on 2 November 2026, and Barclays had set $180 on 26 October 2026. Eight targets in the last month average $160.13.

The median sits 13.2% above the FPI fair value, the mean 10.0% above. The gap has a calendar: the Street is paying for the sac-TMT decision due by 15 June 2027, intismeran's 42% reduction in melanoma recurrence risk and Winrevair growing 93% - revenue that arrives, if it arrives, in 2027 and beyond. The model's 50% DCF weight anchors on cash flows the company shows now; it cannot pay for a 2030 drug in 2026. Insiders have not been paying either: 28 sales worth $57.0 million and zero purchases in twelve months.

Three numbers that move $145 before 2 February 2027

The next results arrive on 2 February 2027 with a $1.77 EPS estimate and $16.32 billion of revenue expected. Signals to watch:

• Q4 revenue above $16.32 billion despite the $300 million timing unwind would show demand carried the quarter, not stocking. A print under $1.77 would put the 2027 outlook in play.
• 2027 EPS guidance against roughly $8.85. A guide above the 2026 underlying number would mean the $2 billion generic hole is being filled; a guide below it would confirm the investment-year framing and strip the 16x argument of its growth premium.
• Price levels: the 50-day average at $146.66 and the fair value at $145.82 sit on top of the current price; the 200-day average at $127.08 is 12% below; the 52-week high is $156.92. Qlex at 9.6% of combined Keytruda sales is the gauge of whether the 2028 patent hill can be climbed before it arrives.