EUR 850.00 For Business Accounts Only

Morningstar | Maintaining Roche's FVE After Solid 3Q; Shares Continue to Incorporate Bearish Pipeline Projection

Roche reported third-quarter sales that slightly beat consensus expectations but were in line with our estimates, and we're adjusting our fair value estimate slightly to $42/CHF 333 to account for minor sales forecast and exchange rate changes. The third quarter was Roche's second consecutive quarter with 7% constant-currency top-line growth, driven by solid performance in both pharmaceutical (7%) and diagnostics (6%) segments. On biosimilars, we're encouraged by increased clarity on the timing of U.S. biosimilar launches (likely Rituxan in the first half of 2019, followed by Herceptin and Avastin in the second half of 2019) as well as the current rate of Herceptin's decline in Europe (which has been slowed by branded subcutaneous Herceptin popularity, at least temporarily). We've also slightly increased our long-term sales estimates for Rituxan, Herceptin, and Avastin in China, as dramatic price cuts are now being annualized and volume growth is starting to become apparent.

There is a flurry of conferences for the remainder of the year, and we're looking forward to details surrounding Tecentriq's positive data in triple-negative breast cancer at the annual meeting of the European Society for Medical Oncology on Oct. 22 (we model $1 billion peak Tecentriq sales in this indication), Kadcyla's positive data in early breast cancer at the San Antonio breast cancer symposium in December, and more details surrounding Roche's CD20 bispecifics in lymphoma at the annual meeting of the American Society of Hematology in December. With the recent Hemlibra approval in the broader hemophilia A market in the U.S. and upcoming likely approvals for Roche's new flu drug baloxavir marboxil and extended approvals for Venclexta (AML) and Tecentriq/Avastin (first-line lung cancer) in December, we continue to model five-year growth ahead of consensus and view shares as undervalued. The firm maintains a stable, wide moat, supported by novel drug development and innovative diagnostics.

On biosimilars, Teva and Celltrion look poised to launch their Truxima (biosimilar Rituxan) in the U.S. in early 2019, with Food and Drug Administration approval expected next month. Our estimate of a roughly 20% U.S. Rituxan decline in 2019 factors in both Gazyva's uptake in follicular lymphoma as well as Truxima's launch, although we note that Truxima will not launch with a complete label (the label will focus on follicular lymphoma) and other biosimilar entrants are likely delayed (Novartis received a complete response letter in May).

We're maintaining our Tecentriq metastatic triple-negative breast cancer forecast ahead of next week's data. Roche noted that the benefit of Tecentriq is largely in the PD-L1 positive population, but this still likely supports uptake in the 40%-50% of patients with this form of the disease. Roche is also expanding its clinical program in TNBC to earlier stages of TNBC, with neoadjuvant data expected in 2019 and an adjuvant trial now running. Competition prevents us from further raising our market share assumptions, as Keytruda is poised to report data in the neoadjuvant setting later this year and in the metastatic setting in 2019, making it the closest competitor.

We've slightly increased our sales assumptions for breast cancer drug Kadcyla based on the Katherine study, which showed better disease-free survival than Herceptin when used as an adjuvant therapy in patients who failed to show a complete response with neoadjuvant treatment. We had very bearish expectations for Kadcyla following its failure to differentiate from Herceptin in the first-line metastatic setting in the Marianne study, so we've slightly increased our near-term estimates and now see Kadcyla sales peaking at almost $1.4 billion in 2022 ahead of biosimilar competition. Data in 2019 for Kadcyla with Perjeta in the adjuvant setting from the Kaitlin study could support larger mid-term sales growth for the drug.