The Fulcrum Day: PCE Ran Hot, and Nvidia Reports Tonight Into a Market Priced for Perfection

1. Macro Pulse

The setup: a genuinely two-part fulcrum day, exactly as flagged Monday and again yesterday. The morning belongs to PCE, already released and modestly hotter than hoped on the headline; the evening belongs to Nvidia, the single most consequential earnings report remaining this quarter. Futures are softer but not sharply so, consistent with a market absorbing a mixed data point rather than a clean shock. [CNBC, 08/26]

PCE's headline miss is real but narrow, and the core reading is the more important number. July's headline Personal Consumption Expenditures price index rose 0.2% month over month and 3.7% year over year, both a tenth of a point above the 0.1% and 3.6% Dow Jones consensus. Core PCE, which strips out food and energy and is the figure the Fed weighs most heavily, matched expectations precisely at 0.2% month over month and 3.3% year over year, extending a streak of 65 consecutive months above the Fed's 2% long-run target. Goldman Sachs had estimated stock-valuation effects alone could add roughly 0.11 percentage points to core PCE this month, a reminder that elevated equity markets are themselves now a modest, mechanical contributor to the inflation figures the Fed is watching. [CNBC/KuCoin, 08/26]

The market's reaction has been proportionate rather than alarmed. S&P 500 futures fell about 0.1%, Nasdaq 100 futures declined roughly 0.4%, and Dow futures actually ticked up slightly. September rate-hike odds moved to approximately 40%, a modest increase from the high-30s range flagged in yesterday's brief, consistent with a data point that complicates the rate-relief case at the margin without overturning it. A pending revision to the PCE's calculation methodology, due by late September and affecting categories including computer hardware and legal services, adds a layer of interpretive uncertainty to how much weight to place on today's specific print. [KuCoin/BigGo, 08/26]

Nvidia reports after today's close, and nearly every preview gathered this morning agrees on the same framing: the headline beat is close to assumed. The company's own guidance calls for revenue of approximately $91.0B, plus or minus 2%, implying a range of roughly $89.2B to $92.8B. Analyst consensus sits only marginally above that midpoint, at $91.7B to $92.2B in revenue and around $2.08 to $2.09 in earnings per share, against $46.74B in revenue and $1.05 EPS in the same quarter last year, implying revenue growth near 95% to 97%. Wedbush's Matt Bryson expects Nvidia to beat both lines and issue upbeat Q3 guidance on strong hyperscale spending, citing what he calls the best supply position in the industry, where component and material access, not end demand, is now the binding constraint on shipments. [Kiplinger/InvestingLive, 08/26]

What actually moves the stock tonight is narrower than the headline number. TIKR's framing is precise: Nvidia's guidance has assumed zero data-center compute revenue from China for several consecutive quarters, so any signal that business is returning represents pure, unpriced upside, while the Q3 guide itself is the other genuine swing factor. CFO Colette Kress confirmed in June that the next-generation Vera Rubin platform is in full production and set to ramp in fiscal Q3, one of the few concrete catalysts that could lift calendar-2027 estimates independent of tonight's Q2 print. Kress also offered a direct rebuttal to the capex-durability worry this brief has tracked since the OpenAI-Anthropic divergence in mid-August, noting that already-deployed Hopper systems are earning more money today than on their first day in service, because Nvidia continues improving the hardware through software, a genuine counter to the fear that customers might simply upgrade rather than expand. [TIKR, 08/26]

History argues for caution regardless of tonight's actual numbers. Nvidia beat its Q1 fiscal 2027 revenue estimate this year, posting 85% year-over-year growth, and the stock still fell almost 5% the following week. The company's margin of beating both Wall Street consensus and its own guidance has narrowed over time, from beating the midpoint of its guidance range by 22.8% in fiscal Q2 2024 to roughly 5.5% above consensus in the most recent quarter. Shares closed Monday at $208.48, down 2.91% that day alone, even after the business grew revenue 85% the prior quarter, a market capitalization near $5.2T pricing in outcomes that leave little room for anything short of genuine outperformance on the questions that matter most. [TIKR/Motley Fool, 08/26]

2. Economic Calendar

Marvell Technology also reports this week, having rallied 4.8% Tuesday ahead of its own turn, a useful bellwether alongside Nvidia for the broader semiconductor demand picture. [Kiplinger, 08/26]

[ANALYSIS] The sequencing today into tomorrow into Friday is worth holding in mind deliberately. Nvidia's results tonight will substantially shape the mood markets carry into Jackson Hole's opening tomorrow, and Warsh's own remarks Friday will be delivered into whatever market environment tonight's report and tomorrow's initial symposium commentary create. A weak Nvidia report combined with a hawkish Warsh keynote would compound into a genuinely difficult end to the week; a strong Nvidia report and a measured Warsh would do the opposite. These are still separate events with separate inputs, but their proximity means each one's market impact will be read partly through the lens of what came immediately before it.

3. Earnings Spotlight

Tonight's Nvidia report is this quarter's defining earnings event, and it lands into a market that has spent this earnings season applying increasingly granular scrutiny to every beat.

4. Pre-Market Movers: Upside

5. Pre-Market Movers: Downside

6. Sector Rotation Radar

[ANALYSIS] Today's sector positioning is almost entirely a function of what has not happened yet rather than what has. The PCE print is already known and the market's reaction to it has been modest; the far larger, still-unresolved variable is tonight's Nvidia report, and a secondary, related one is how that report's reception colors the opening of Jackson Hole tomorrow. Treating today's relatively calm session as informative about anything beyond today itself would be a mistake, this is a market waiting, not a market that has resolved its major open questions.

7. Today’s Session Playbook

1. The regular session is the undercard tonight is the main event. Today's PCE print has already been digested with a modest, proportionate reaction. Position with tonight's 4:20 PM Nvidia release and the subsequent earnings call explicitly in mind.
2. Top story: Nvidia needs more than a beat. With consensus already assuming roughly 95% to 97% revenue growth, the headline number is close to a formality by every account gathered this morning. The Q3 guide and any China signal are what will actually move the stock, and history argues a beat alone offers no guarantee of a positive reaction.
3. S&P 500 levels: today's modest pre-market pullback keeps the index near its recent range. The far more meaningful test of support and resistance arrives with tomorrow's open, once tonight's Nvidia reaction and the market's read on it are fully known.
4. Event risk: a two-part day, and then Jackson Hole opens tomorrow. PCE this morning was the smaller of today's two catalysts. Tonight's Nvidia report is larger, and tomorrow's start to Jackson Hole, building toward Friday's Warsh keynote, is larger still in terms of how the week ultimately resolves.
5. Under-the-radar: Kress's Hopper-monetization comment. Her point that already-deployed chips earn more today than on day one, through continuous software improvement, is a specific, credible rebuttal to the capex-durability question that has dogged this trade since the OpenAI-Anthropic divergence. Listen for whether this theme resurfaces on tonight's call with any new supporting detail.

This brief has spent the better part of two months tracking a single organizing principle: this market prices the gap between results and expectations, not results in isolation, and that scrutiny has grown more granular with nearly every passing week. Cisco beat on every growth metric and fell on a margin miss. Applied Materials beat revenue, EPS, and guidance and fell on an operating-income shortfall. Walmart beat on an adjusted basis and fell nearly 10% once the market found the GAAP profit decline underneath. Tonight, that same discipline meets its highest, most consequential test of the entire season. Nvidia is expected to grow revenue somewhere between 95% and 97% year over year, a growth rate that would be treated as an extraordinary outlier at almost any other company in the market, and by nearly every account gathered this morning, that number alone is close to assumed. What actually decides tonight's reaction is narrower and harder: the Q3 guide, and whether China, entirely excluded from guidance for multiple consecutive quarters, offers any sign of returning.

The historical pattern this brief surfaced this morning is worth sitting with honestly. Nvidia beat its own Q1 revenue guidance this year by a wide margin, posted 85% year-over-year growth, and the stock still fell nearly 5% over the following week. The company's own track record of beating consensus and guidance has narrowed over time, from a 22.8% beat-the-midpoint margin two years ago to roughly 5.5% most recently. That narrowing is not itself alarming, a company approaching a trillion-dollar-plus quarterly run rate cannot sustain the same percentage surprises indefinitely, but it does mean the room for tonight's report to clear an already-extraordinary bar by an extraordinary margin has genuinely shrunk. This is the delta rule operating exactly as this brief has described it all season: not a verdict on whether Nvidia's business is good, which by every available measure it clearly is, but a verdict on whether tonight's specific numbers exceed what a market capitalization near $5.2T has already priced in.

The counter-argument, and it is a genuine one, sits in Kress's own comments from June. Her point that already-deployed Hopper systems earn more today than on their first day in service, through continuous software-driven improvement, is a direct, credible answer to the fear that AI-infrastructure customers might simply upgrade existing hardware rather than commit to genuinely new capital expenditure, the same fear that has shadowed every major AI-capex story this brief has tracked since Alphabet's reaction in July. If that dynamic shows up clearly in tonight's numbers or commentary, it strengthens the case that Nvidia's revenue is durable and compounding rather than merely large. Combined with the Vera Rubin ramp confirmed for fiscal Q3, there is a credible, evidence-based bull case sitting alongside the more cautious historical pattern.

The honest position heading into tonight is that both the optimistic and the cautious readings are well-supported by real evidence, and neither should be treated as settled until the actual numbers and, more importantly, the actual Q3 guide and China commentary are known. What is not in doubt is the stakes. Tonight's report will meaningfully shape the mood markets carry into tomorrow's opening of Jackson Hole and, by extension, the backdrop Chair Warsh speaks into Friday. Two of this week's four flagged catalysts resolve within the next 18 hours. The discipline this brief committed to at the start of the week, judging each catalyst on its own specific terms rather than forcing a single narrative onto all of them, matters more tonight than at any other point this month.