Oil Is Spiking on a Widening War. CPI Has to Decide If the Fed Cares: Weekly Market Insights, September 7-11

Thursday's Producer Price Index arrives first, and it typically gets less attention than Friday's CPI, but this year the order matters more than usual. PPI feeds directly into several CPI subcomponents, particularly on the goods side, so a hot or cool PPI print Thursday morning is a genuine preview of what Friday might bring, not just a standalone data point. July's PPI for final demand came in flat month over month, below the roughly 0.2% consensus, with prices up 4.7% over the trailing twelve months. Whether August holds that pattern or accelerates, especially given how much crude has moved since the July reference period ended, is the first real test of whether the recent oil spike is already bleeding into the pricing pipeline.

The detail that makes this week genuinely unusual is what won't happen alongside the data. The FOMC's blackout period began Saturday, September 5, and runs through Thursday, September 17, meaning no Fed governor or regional bank president can publicly discuss monetary policy between now and the meeting itself. In a normal week, a hot or cool inflation print gets contextualized within hours by whichever Fed official happens to be scheduled to speak. This week, PPI and CPI will have to speak entirely for themselves. That raises the odds of an outsized initial market reaction on each release, since there's no scheduled Fed voice available to walk a surprising number back or confirm it before the meeting.

Friday's CPI carries the heavier weight for an obvious reason: it's the last inflation reading the Fed will see before its decision, five days later. The University of Michigan's preliminary September consumer sentiment reading, out the same morning at 10:00 AM, adds a second inflation-expectations data point worth checking against the government's own number. August's final reading came in at 51.7, down 6.3% from July, with one-year inflation expectations at 4.0%, itself down from a preliminary 4.3% but still well above the 3.4% consumers were expecting back in February, before the war began. If September's preliminary reading shows expectations climbing again, that's a household-level echo of the same question CPI is answering at the government level.

Oracle reports fiscal first-quarter results Thursday after the close, with Adobe following the same evening, a rare case of two of the market's most closely watched AI-infrastructure and AI-monetization stories landing within hours of each other.

Oracle's setup is about scale and financing. The company carries a reported $638 billion in remaining performance obligations, its backlog of contracted future cloud revenue, and management has guided for total cloud revenue growth of 58% to 64% in U.S. dollar terms for the quarter. Consensus estimates cluster around $19.1 billion in revenue and roughly $1.74 in non-GAAP earnings per share. The options market is pricing an earnings-day move of close to 10% in either direction, and the stock has fallen more than 20% from its highs earlier this year on investor concern about how much debt-funded capital expenditure the AI buildout actually requires. The question Thursday isn't whether Oracle's backlog is real. It's whether the market believes Oracle can convert that backlog into cash flow fast enough to justify the capital spending it's taking on to build it.

Adobe's test runs on a different axis entirely. The stock has recovered roughly 22% since its last earnings report, when a beat-and-raise quarter still sold off because management held its annual recurring revenue guidance unchanged despite a shift toward AI-first and freemium pricing models. Thursday's report is the market's chance to see whether that pivot is translating into subscription and revenue growth or still mostly reads as a story about future potential. Running Oracle and Adobe as a pair says more than either alone: one is a bet on AI infrastructure spending being justified by future demand, the other is a bet on AI features actually getting monetized inside existing software products. A strong reaction in one and a weak reaction in the other would say something specific about which side of the AI trade the market currently trusts more.

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.

Every name below earns its place by answering "why this stock, this week," not because the chart looks interesting.

Kroger (KR). Kroger hosts its second-quarter earnings call at 8:00 AM Friday, thirty minutes before the CPI report lands. That timing isn't a coincidence worth ignoring. Tyson Foods updated its fiscal 2026 outlook on September 3 to flag additional beef-margin pressure tied to a severe cattle shortage, and Kroger's own gross margin has already been squeezed by exactly that kind of input-cost pressure. Watch the guidance range and any specific commentary on how much of rising beef and grocery costs the company can actually pass through to shoppers, not just the headline earnings-per-share number. That's a real-time, company-level answer to roughly the same question CPI is about to answer at the economy-wide level: how much of this year's cost pressure is landing on the consumer's plate.

Oracle (ORCL) and Adobe (ADBE). Both report Thursday evening, hours apart, for the reasons laid out above. The specific numbers to watch are Oracle's cloud infrastructure revenue growth against its 58% to 64% guided range, and Adobe's subscription revenue and annual recurring revenue trend against the flat guidance that triggered its last selloff.

ConocoPhillips (COP) and Delta Air Lines (DAL). These two sit on opposite sides of the same oil shock, and running them as a pair says more than either alone. ConocoPhillips is a clean, large-cap proxy for higher crude prices flowing straight to the bottom line. Delta carries the mirror image through jet fuel costs, one of its largest single expense lines. If both move in genuinely opposite directions on the same week's Hormuz headlines, that's the market pricing the actual mechanism rather than reacting to noise. Company-specific news at either name, or a sudden de-escalation signal from either side of the conflict, could complicate that read quickly.

AeroVironment (AVAV). AeroVironment isn't reporting earnings this week, but it received a $465 million U.S. Army contract on September 2 under the Enduring-High Energy Laser program, the first production contract for a high-energy laser weapon system in U.S. history according to the company. With an active, escalating conflict driving fresh defense-spending headlines most weeks, watching whether the broader defense complex trades as a genuine beneficiary of the widening Hormuz conflict, rather than just this one contract, is the thing worth tracking.

Gold (spot) versus the 2-year Treasury yield. Gold's roughly 2.4% drop over the past week is a clean, real-time signal of how the market is pricing the dollar and real yields, which makes it worth checking against even for traders who never touch a gold position. If Friday's CPI comes in hot and reinforces the case for a hike, gold likely keeps falling as the 2-year yield pushes further above 4.4%. If core comes in soft and the market reads that as the Fed's excuse to look through the oil shock, gold has room to reverse even with a live war still pushing headline inflation the other way. Watching whether gold and the 2-year yield keep moving in the mirror-image pattern they've held all month, or break from it, is a fast cross-check on which story the market is actually buying.

iShares Russell 2000 ETF (IWM). Small-cap balance sheets carry more variable-rate debt and less pricing power than the mega-caps that dominate the S&P 500, which makes the Russell 2000 the most amplified reader of whatever this week's PPI and CPI data do to rate expectations. The index already showed that sensitivity Friday, rising 0.25% even as the other three major indexes fell, an unusual divergence worth watching for whether it repeats or reverses once real inflation data, rather than a jobs report, is the catalyst.

Also on the radar: the ECB's Thursday rate decision as a global monetary-policy cross-check landing the same morning as U.S. PPI; China's Tuesday-night CPI reading for a read on global demand and deflation pressure; GameStop, Casey's General Stores, Braze, ServiceTitan, and Chewy among Tuesday and Wednesday's smaller earnings; and next week's calendar, where the FOMC decision, quadruple witching, and the Nasdaq-100/S&P 500 quarterly rebalance all land inside the same five trading days, which makes however this week resolves the direct setup for that one.

These are conditional paths built from this week's actual research, not predictions, and none is more likely than the others based on the evidence available as of this cutoff.

Inflation confirms the hike case. Both PPI Thursday and CPI Friday come in at or above the Cleveland Fed's nowcast, with core CPI running at 0.2% month over month or hotter. Confirmation: the 2-year yield pushes through 4.4%, the dollar extends its recent gain, rate-hike odds move above 65% to 70%, and gold extends its slide.

Core cools even as headline gets pushed up by oil. Headline CPI ticks higher on gasoline and energy pass-through, but core comes in soft, in the 0.1% to 0.2% range. Confirmation: Fed funds futures pull hike odds back down despite the higher headline number, long-end yields ease even as energy-linked components stay elevated, and gold stabilizes or bounces.

The AI-capex trade holds through earnings. Oracle and Adobe both clear their bars and hold their post-earnings gains regardless of what the week's macro data show. Confirmation: semiconductor and software-sector proxies move higher specifically on Thursday evening and Friday morning, independent of the CPI reaction that same day.

The war overrides the calendar entirely. Iran's promised escalation materializes as a further disruption to Hormuz shipping, a strike with confirmed casualties, or an expansion to a new front. Confirmation: oil and gold rise together, an unusual combination that signals genuine risk-off positioning, and equities sell off broadly regardless of what CPI or Thursday's earnings showed that same day.

Kroger flags real consumer strain. Kroger's Friday call cites input costs, particularly beef and other grocery categories, that it can't fully pass through, and trims guidance toward the low end of its range. Confirmation: consumer-staples names underperform broadly, and the reaction complicates any reading of Friday's jobs-report strength as proof the consumer is fine.

This week's trap is straightforward: a hot jobs report, an escalating war, and the last inflation data before a live hike decision are all pulling in different directions, and there's no Fed official available all week to help sort out which one the market should believe.

• Watch Thursday's PPI reaction as a genuine preview of Friday's CPI, not just a standalone release. A surprise in either direction on PPI is worth adjusting size around before Friday's number, not after it.
• Check the options-implied move on Oracle and Adobe before sizing anything into Thursday's close. A move that looks small next to either stock's recent realized volatility is itself worth noting given how much both names have swung on recent reports.
• Treat any fresh Hormuz or Iran headline as a standalone catalyst for energy, defense, and airline names all week, even without a fixed release time attached. The past 48 hours alone produced a carrier-missile exchange and three destroyed tankers with no scheduled slot on any calendar.
• Widen stops or reduce size heading into Friday's 8:30 AM CPI release specifically. With the FOMC blackout in effect, no Fed voice can walk back an extreme market reaction before the September 15-16 meeting itself.
• Watch Kroger's 8:00 AM call for guidance language before CPI even lands. A share of the "is the consumer actually fine" question gets answered thirty minutes before the government's own number arrives.
• Look past this week too. Next week stacks the FOMC decision, quadruple witching, and the quarterly index rebalance into a single five-day stretch, and however this week's data resolves becomes the direct setup walking into that one. Scanning for unusual volume around Thursday's and Friday's releases is exactly the kind of real-time flag a tool like Trade Ideas is built to catch as it happens rather than after the fact.