Weekly Market Insights August 10-14, 2026: A Weak Jobs Report Bought the Fed Room to Wait. CPI and a Wall of AI Earnings Decide If It Holds.

Every name below reports this week and can move on its own merits. Together, they touch nearly every layer of the AI infrastructure stack, which makes the group more useful as a set than any single report is on its own. DayTradingToolkit's guide to building a day-trading watchlist is worth a look if eight names in one week feels like too many to hold in your head at once; the point of a watchlist is to remove decisions, not add them.

Rocket Lab (RKLB): the one report with nothing to do with the AI-cloud argument. Rocket Lab reports after Monday's close, with a call at 5:00 p.m. ET, the same day this article publishes. The Zacks consensus calls for a loss of 3 cents per share, a 70% year-over-year improvement, on revenue near $231.6 million, up more than 60%. The stock has gained over 46% since its last report. Watch progress on the Neutron rocket program, which has faced execution-risk questions from analysts, and any update on the $397 million Space Force contract the company secured this summer. A clean report here says something about launch-sector demand independent of whatever the AI trade does the rest of the week.

Lumentum (LITE): the optics name with the highest bar in the group. Lumentum reports fiscal fourth-quarter results Tuesday after the close, call at 5:00 p.m. ET. Analysts expect $2.99 per share, up more than 700% from 32 cents a year ago, on revenue near $988.6 million, more than double the year-ago quarter. Growth like that comes from cloud and AI demand for optical components: narrow-linewidth laser shipments were already up more than 120% year over year last quarter, and management has said several product lines are effectively sold out for now. The number worth watching isn't the beat itself. It's whether 1.6-terabit transceiver shipments are ramping on schedule, since a company already growing this fast has little room for a soft answer on the next leg.

Super Micro (SMCI): the report that already told you part of the answer. Super Micro reports Tuesday as well, call also at 5:00 p.m. ET. A preliminary update in July flagged fourth-quarter revenue landing near the low end of its $11.0 billion to $12.5 billion guidance range, but gross margin coming in at 15% to 17%, well above the 8.2% to 8.4% originally guided, on more than $60 billion in new orders for the quarter. The report itself should mostly confirm those numbers. What it can't confirm in advance is the other half of the Super Micro story: the company is operating under an independent board investigation tied to a March DOJ indictment alleging former employees, including a co-founder, illegally diverted roughly $2.5 billion in AI servers to China. Super Micro itself hasn't been charged, but expect analysts to spend real time on the call asking whether the legal overhang has changed customer or supplier behavior, not just whether the margin number holds.

CoreWeave (CRWV): the most direct test of the AI-infrastructure argument all week. CoreWeave closes out Tuesday's earnings window with a call at 5:00 p.m. ET. CoreWeave doesn't build chips or components, it rents out the AI compute those chips go into, which makes it about as pure a read on AI-demand pricing and utilization as this list has. The stock has underperformed recently on industry overbuild and capacity-delay concerns, worries some analysts have called overblown given what they describe as a roughly four-to-one industry-wide supply-demand imbalance. Watch whether Tuesday's reaction spreads into Wednesday's Nebius report or stays contained to CoreWeave alone. If both AI-cloud names move the same direction for the same underlying reason, that's a sector signal. If they diverge, the market is drawing a line between specific companies rather than the theme.

Nebius (NBIS): the earnings report Michael Burry picked a fight with. Nebius releases results before Wednesday's open, call at 8:00 a.m. ET, thirty minutes ahead of CPI the same morning. Analysts expect revenue near $576 million to $583 million and continued progress toward a $7 billion to $9 billion full-year exit ARR target already more than half contracted. Under normal circumstances that would be the whole story. It isn't this time. On August 6, Michael Burry disclosed a direct equity short in Nebius at $211.77, choosing a straight short over options because, in his words, the puts were priced too expensively to bother with. His stated thesis: AI infrastructure companies may be depreciating GPUs too slowly relative to how fast that hardware actually loses value, which would mean today's reported profits look better than they really are. Nebius shares came under renewed pressure in the days around the disclosure, adding to a rough stretch for the stock since late July. Wednesday's report won't settle an accounting debate like that in one sitting, but capital-expenditure pacing, the timeline for its new Nvidia Vera Rubin hardware to start generating billable revenue, and any commentary on lease obligations will all get read through Burry's thesis whether management addresses it directly or not.

Cisco (CSCO): the legacy name still writing the AI-buildout story. Cisco reports fiscal fourth-quarter results Wednesday after the close. The stock recently broke out above a level it hadn't touched since the dot-com era, on the strength of networking demand tied to AI data-center buildouts. Analysts are looking for networking-segment growth to accelerate to around 26% annually, even as overall product growth is expected to decelerate from last quarter's decade-best pace, with persistently high component costs expected to keep gross margin capped near 66%. Cisco isn't a pure AI-infrastructure story the way CoreWeave or Nebius is, which is exactly what makes it useful this week: a strong number here says the AI buildout is showing up in demand for the unglamorous plumbing underneath it, not just the headline compute names.

Coherent (COHR): the second photonics name, telling a different story than Lumentum's. Coherent releases fiscal fourth-quarter and full-year results Wednesday after the New York Stock Exchange closes, webcast at 4:30 p.m. ET, with consensus near $1.62 per share. Where Lumentum's story is almost entirely about optical growth velocity, Coherent's is about diversification: alongside its networking and datacenter optics business, the company runs a silicon-carbide semiconductor unit that recently locked in $1 billion in joint investment from Denso and Mitsubishi Electric, plus a strategic optics partnership with Nvidia announced earlier this year. That mix means Coherent's report can move on a different axis than Lumentum's even though both sell into the same AI-networking buildout. Watch whether the market treats them as one trade or two.

Applied Materials (AMAT): the week's biggest number, arriving last. Applied Materials closes out the earnings calendar Thursday after the close, the same session as PPI and jobless claims. The stock has roughly doubled this year, and the options market has been pricing a post-earnings move north of 10%, larger than anything else on this week's calendar. As the largest supplier of chip-fabrication equipment to foundries like TSMC and Intel and memory makers like Micron and SK Hynix, Applied Materials sits one level removed from the AI trade the rest of this list sits inside of: it doesn't sell compute or optics, it sells the machines that build the chips those products depend on. A confident outlook here reads as evidence that semiconductor capital spending still has room to run. A cautious one, arriving after CoreWeave, Nebius, and both photonics names have already reported, lands as the week's final word on whether AI-capex skepticism spread to the equipment layer or stayed contained to the cloud names carrying the newer debt.

The watchlist above explains what individual companies can reveal. The board below helps determine whether a reaction is broad, company-specific, or driven by something outside any single earnings call.

Oracle is worth watching even though it doesn't report this week. Burry's August 6 disclosure paired the new Nebius short with a fresh Oracle short at $144.63, built on similar leverage-and-depreciation logic, and Oracle carries its own separate overhang: an S&P Global downgrade of its long-term credit rating to BBB- on July 9, one notch above speculative grade, citing Oracle's roughly $638 billion contract backlog and its heavy concentration in OpenAI, which S&P estimated accounts for close to half of that total. If Nebius's Wednesday report reassures the market on capital discipline and Oracle rallies in sympathy, that suggests the market is treating Burry's thesis as company-specific. If both stay pressured regardless of what Nebius actually reports, the thesis itself is what's trading, not the fundamentals underneath either company.

Memory is the sector to watch for continuation rather than reversal. Micron and Sandisk both slid in the days after Sandisk's Wednesday report, and Citi trimmed its Micron price target around the same time on concerns that memory pricing could peak next year rather than keep climbing. Nothing about CPI or PPI directly confirms or denies that, but a memory sector that keeps sliding through a week of otherwise-positive AI-infrastructure earnings would say the market has drawn a specific line around pricing risk, not around enthusiasm for AI hardware broadly.

Several of this week's names, Rocket Lab aside, share the same underlying macro driver even though they sell entirely different products. DayTradingToolkit's guide to correlation risk is worth a look before treating four or five separate earnings reactions as four or five independent trades.

Friday's jobs report bought the market a specific, narrow kind of relief: permission to believe the Fed doesn't need to hike in September. That relief rests on an assumption that hasn't been tested yet, that a one-month contraction in payrolls means the same thing for policy as a genuine slowdown. This week is where the assumption gets tested twice: once by the inflation data, and once by whether AI-infrastructure earnings can survive the same scrutiny that just took down AMD and Sandisk despite both companies beating estimates.

Start with the more mechanical branch. If Wednesday's CPI and Thursday's PPI come in at or below consensus, roughly 0.2% to 0.3% on the core monthly reads, the disinflation story stays intact, and the Fed's July hold looks less like the start of an argument and more like a reasonable pause. In that world, this week's earnings reports get judged on their own terms rather than as referendums on interest-rate policy, and a rally that's already broadened past megacap tech, Friday's session included gains across SPY, QQQ, and the Russell 2000, has room to keep including small caps rather than narrowing back to a handful of names.

The harder branch is a hot print. Markets went into the weekend pricing something in the neighborhood of a mid-40s percentage chance of a September hike, down sharply from the high 50s before Friday's jobs number. A CPI or PPI surprise to the upside would push that back toward where it sat a week ago, at the worst possible moment for this week's earnings calendar: companies like Nebius and CoreWeave are asking investors to be patient with heavy, multi-year capital spending on the promise of future AI revenue, a story that gets harder to tell at a higher discount rate. Rate-sensitive growth names don't need bad news to fall in that scenario. They just need the market to start pricing their future cash flows differently.

Then there's the branch that has nothing to do with the Fed at all. Layer last week's pattern back in. AMD beat and fell 7%. Sandisk beat and fell into Thursday. Palantir beat and rose 12%, but only because the guidance raise was large enough to change the multi-year story, not just the trailing quarter. This week's eight reports face the same test regardless of what CPI and PPI do. A cooperative inflation print doesn't automatically rescue an earnings call that doesn't clear the bar on capital-expenditure pacing, customer concentration, or forward demand visibility, the exact questions Burry's Nebius and Oracle shorts are built around. It's entirely possible for the macro data to come in fine and for one or two of this week's AI-infrastructure names to fall anyway, for the same reason AMD and Sandisk did.

The wildcard sits outside both the calendar and the earnings reports.

None of these branches are mutually exclusive, and the honest answer is that this week probably delivers pieces of more than one. The evidence worth tracking in real time: whether Wednesday's 10-year yield reaction to CPI agrees with the equity reaction, whether Tuesday's and Wednesday's AI-infrastructure earnings get judged on guidance rather than the trailing quarter, and whether oil headlines stay quiet enough to let the inflation data do the talking on its own.

Tuesday's secondary data matters for context, not as a standalone trigger. ADP's private payrolls read and existing home sales are useful cross-checks against Friday's jobs report and the rate backdrop, but neither is likely to move the tape the way Wednesday's CPI or Thursday's PPI will.

Chart-only setups without a scheduled catalyst don't earn a spot on the primary list this week just because a ticker has been active. This is a calendar-heavy week; let the calendar do the filtering rather than chasing a chart that happens to be moving.

Leveraged single-stock products tied to any of this week's earnings names carry their own compounding and liquidity considerations on top of whatever the underlying stock does. DayTradingToolkit's guide to single-stock leveraged ETFs covers the mechanics if one of these products exists for a name on this week's list. The underlying company should stay the primary object of analysis either way.

Friday's Michigan sentiment print is confirmation data at best. By the time it lands at 10 a.m., CPI, PPI, and retail sales will have already done most of the week's work.

For a broader framework on separating scheduled-event windows that deserve preparation from the ones that only add noise, DayTradingToolkit's guide to how economic reports move the market and its companion CPI trading playbook cover the mechanics in more depth than this article has room for.