Two Wars, a Trade Fight, and CPI Still Has to Land Clean: Weekly Market Insights, September 7-11

Read-through is what happens when the market prices one company's stock off a different company's news, on the assumption that a similar drug or business will succeed or fail the same way. Novartis's pelacarsen, an Lp(a)-lowering drug developed with Ionis Pharmaceuticals ($IONS), missed the primary endpoint of its Phase 3 outcomes trial. Here's the distinction that matters: it successfully lowered the target biomarker, Lp(a) itself, but that didn't translate into fewer actual heart attacks, strokes, or cardiovascular deaths compared with placebo, in a trial of more than 8,300 patients.

BMO Capital Markets downgraded Amgen ($AMGN) the same day, and Amgen's own drug didn't fail. Amgen has a similar-mechanism Lp(a) drug, olpasiran, that hasn't even read out yet. The market priced in a failure that hasn't happened, purely because the mechanism looks similar.

Separately, Novartis disclosed a second trial miss the same week. Tuesday's market commentary tied that second miss to further pressure on Dyne Therapeutics ($DYN) and Sarepta Therapeutics ($SRPT), on the same read-through logic. That second claim comes from a single source and wasn't independently confirmed for this edition, so treat it as a lead worth watching rather than a settled fact.

Ionis takes the most direct hit of the group. Under its 2019 licensing deal, it stood to earn tiered royalties in the mid-teens to low-20% rangeRangeA range (or trading range) is a price zone where a stock oscillates between a defined support floor and resistance ceiling without making meaningful new highs or lows — price is consolidating rather than trending, and the dominant trade is Read definition → plus $650 million in milestones once pelacarsen reached approval. Most of that is gone now.

Google announced a 13 billion euro investment in Finnish AI infrastructure this week, its largest-ever European investment, including new data centers and an expansion of its existing site in Hamina. You won't find Nokia ($NOK) named anywhere in Google's own release. It's been a supplier on that Hamina buildout since 2023, alongside the Finnish telecoms DNA and Elisa, and it separately launched its own AI network-operations platform on Google Cloud this month. Nokia traded up close to 5% on the news.

Qualcomm ($QCOM) and Amazon Web Services announced a multi-generational AI infrastructure partnership on September 7, and Qualcomm shares rose more than 7% within a day, then kept climbing as the week went on. The mechanism behind it is unusual, so it's worth slowing down on.

Amazon received a warrant for up to 25 million Qualcomm shares at $161.26 each. Only 3.75 million of those vested immediately. The rest vest in tranches, tied to Amazon actually placing orders, up to a maximum of $60 billion through 2036. That $60 billion is a ceiling for the warrant to fully vest, not revenue Qualcomm has already booked, and that's a distinction most of the week's coverage skipped entirely.

One more thing worth noticing: Amazon's own stock barely moved on the news. That's informative on its own. The market is pricing this as a Qualcomm story, not an Amazon one.

Compass Pathways ($CMPS) closed up 5% Tuesday on a specific, dated cluster of events, not a vague mood. There's a new US doctor-education campaign for its psilocybin therapy, COMP360. Management is presenting at two Cantor healthcare-conference sessions this week. And the real anchor: the FDA is holding its first-ever public hearing on the therapeutic use of psychedelic drugs in supervised settings, scheduled for Monday, September 14.

Compass's rolling NDA submission is on track to complete in the fourth quarter of 2026, helped along by a Commissioner's National Priority Voucher that shortens the review window. Final approval would still need the DEA to reschedule psilocybin off Schedule I, so this isn't a done deal by any stretch. Separately, a chunk of the ticker's social chatter this week recycled Eli Lilly's July acquisition of a different, earlier-stage competitor as though it were new, and speculated Compass could be next. That specific claim has no confirmed backing. It's retail speculation, not a signal, and worth knowing the difference.

Pharvaris announced a late-stage trial win for its own oral hereditary-angioedema drug, cutting monthly attacks 83% against placebo. Biocryst ($BCRX) already sells an approved, revenue-generating drug for the same disease, so the market read Pharvaris's success as a direct competitive threat rather than background noise. Biocryst posted its worst day in over six months on it. Not everyone agreed that was the right read: TD Cowen initiated Biocryst at Buy with a $30 price target, roughly triple where it was trading, the same day.

Oracle ($ORCL) reports fiscal first-quarter results Thursday after the close, with Adobe ($ADBE) following hours later. It's still the clearest read available on which side of the AI trade the market currently trusts. Oracle carries a reported $638 billion in remaining performance obligations, future revenue it already has under contract but hasn't recognized yet, and has guided to 58-64% cloud revenue growth for the quarter. Consensus sits around $19.1 billion in revenue and $1.74 in non-GAAP earnings per share, and the options market is pricing a roughly 10% move on the report, in either direction. The stock is down more than 20% from its highs this year on a fair question: does that backlog convert to cash flow fast enough to justify the debt-funded spending behind it?

Adobe has recovered about 22% since its last report, when a beat-and-raise quarter still sold off because management held annual recurring revenue guidance flat despite an AI-first pricing pivot. Thursday is the market's chance to see whether that pivot is finally showing up in the numbers instead of just the narrative. For more on how the AI-capex debate has been building all year, this guide's July breakdown of the AI infrastructure spending question is still the fuller reference.

Kroger's ($KR) second-quarter call Friday at 8:00 AM lands 30 minutes before CPI, and that timing isn't a coincidence worth ignoring. Tyson Foods flagged worsening beef-margin pressure from a cattle shortage on September 3, and Kroger's own margins have already absorbed a version of that same input-cost squeeze. How much of that Kroger says it can pass through to shoppers, versus how much it has to eat itself, is a company-level answer to the same question CPI's food-at-home component measures nationally, half an hour later.

Oil and the widening conflict. WTI is trading near $94 and Brent near $98, both closing in on $100 for the first time in over a month, after Brent touched $99.46 intraday Tuesday. The weekend's US strikes on three Iranian oil tankers, a response to Iranian missiles fired at a US carrier and destroyer, were compounded Tuesday by a genuinely different escalation.

Saudi Arabia said Houthi attacks halted operations at several of its energy facilities. That's a second front, not a continuation of the first. Saudi energy infrastructure, not just Iranian shipping, is now a live target.

Goldman Sachs raised its oil forecasts this week: Brent and WTI to $85 and $80 for December 2026, and $80 and $75 for 2027. The bank's bear case, not its base case, has Brent above $120 in 2027 if Gulf output stays roughly 4 million barrels a day below prewar levels. Worth knowing that range exists even if it isn't the expected outcome.

Separately, reports that Iran and Oman are close to a deal on a "temporary safe route" through the Strait of Hormuz have been circulating. But Trump told Axios over the weekend the US is "only semi-negotiating" and plans to keep leaning on its naval blockade rather than pursue further strikes, while Iran's Foreign Ministry says it won't fully reopen the strait until that blockade lifts. Those are two very different versions of the same situation, which means any oil-price relief tied to that reported deal might be pricing in an agreement that doesn't actually resolve anything yet.

The trade war's stock-level exposure. Beyond the policy question above, Trump's specific naming of Gulfstream as the domestic alternative to Bombardier puts General Dynamics ($GD) in a real, if genuinely uncertain, position. Uncertain because the FAA, not the White House, actually certifies aircraft, and it's unclear how a unilateral sales ban would even get implemented. Bombardier itself pushed back by noting it has 3,500 US employees, 2,800 US suppliers across 47 states, and more than $2.5 billion in annual US spending, which complicates any clean "domestic winner" story even if the threat gets followed through on.

Gold versus the 2-year yield. Gold fell to around $4,420 an ounce last week, down roughly 2.4%, as the stronger dollar and higher real yields that come with rising hike odds pressured the metal even with an active war underway. That's a good reminder that gold trades on rate expectations as much as safe-haven demand, maybe more so lately. If Friday's CPI reinforces the hike case, gold likely keeps falling as the 2-year yield (last at 4.377%, its highest since January 2025) pushes higher. If core cools and the Fed's look-through argument gains ground instead, gold has room to reverse.

Next week stacks the FOMC decision (September 15-16), quadruple witching, and the Nasdaq-100/S&P 500 quarterly rebalance into the same five-day span. However this week's PPI, CPI, and earnings resolve becomes the direct setup for that stretch, and there's not much time to digest between the two weeks.

iShares Russell 2000 ETF ($IWM). Small-cap balance sheets carry more variable-rate debt than the S&P 500's largest names, and less pricing power to offset it. That makes the Russell the most amplified reader of whatever this week's inflation data does to rate expectations. It already diverged from the other major indexes Friday, rising 0.25% while they fell. Worth watching whether that repeats once real inflation data, not a jobs report, is the thing driving the reaction.

Dyne Therapeutics ($DYN) and Sarepta Therapeutics ($SRPT). Named in Tuesday market commentary as further pressured by Novartis's disclosed second trial miss, on the same read-through logic hitting Amgen. It sits here in the radar rather than the priority list because the underlying claim wasn't independently confirmed for this edition.

Also worth knowing about: the ECB's Thursday decision, a global policy cross-check landing the same morning as US PPI; China's Tuesday-night CPI, a global demand read; and Chewy's Wednesday earnings, a narrower consumer-discretionary data point.