The July earnings cluster is done, but Q3 2026 earnings season is not close to over. Retail heavyweights report in a five-day window from August 18 to 20, NVIDIA and Salesforce report the same evening on August 26, and the AI-infrastructure trio of Marvell, Broadcom, and Oracle carries the season into mid-September, right past a Fed rate decision. Adobe, Costco, and Micron close things out before the quarter ends September 30. For options traders, the calendar itself is still the edge: implied volatility inflates on a schedule and crushes on a schedule, and knowing those dates in advance is worth more than any single trade idea.
- Retail opens the next leg of the season August 18-20 (HD, TGT, LOW, WMT, TJX); NVIDIA and Salesforce report the same evening, August 26.
- Three AI-infrastructure names, Marvell (Aug 27), Broadcom (Sept 2), and Oracle (Sept 14), report within about two and a half weeks of each other and tend to move together on AI-capex sentiment.
- A Fed rate decision lands September 16, between the Adobe/Oracle software cluster and the Costco/Micron close of the quarter.
- Pre-earnings IV typically expands 15-30% in the 5-10 trading days before a report, then crushes 30-50% immediately after the print. That pattern has not changed.
- Dates below were checked against company investor-relations pages and earnings-date trackers as of August 11, 2026. A few (marked below) are tracker consensus rather than official company confirmation, so verify before positioning.
Where the Season Stands: July’s Cluster Already Happened
If you were positioning around the July banks-to-Mag-7 stretch, that part of the calendar has already played out, and this month brought a second wave behind it. Since the start of August alone, traderc.com has covered pre-earnings setups and recaps on AMD, Duolingo, Rocket Lab, Super Micro, AST SpaceMobile, and CoreWeave. That is useful context even if you missed those prints: it shows the expected-move framework below applied in real time, name by name, all in the same three-week stretch. If you want to see how a stock actually trades in the days after a report rather than just on the print itself, the post-earnings drift piece is worth a read before the next batch below.
The Rest of the Q3 2026 Earnings Calendar
A naming note that trips people up every quarter: reports landing in August and September mostly cover companies’ fiscal Q2 2026 results (or Q3/Q1 for companies on non-calendar fiscal years), but the reports themselves land in calendar Q3, which is why the trade press calls all of it “Q3 earnings season.” Report dates, not quarter labels, are what determine your options expiration and IV timeline.
Here is the slate of high-liquidity options names still to report through the end of Q3 2026 (September 30):
| Date | Company (Ticker) | Session | Expected-move tier | Notes |
|---|---|---|---|---|
| Tue, Aug 18 | Home Depot (HD) | Before open | Low (2-4%) | Unofficial start of the retail leg; a read on housing-linked consumer spend. |
| Wed, Aug 19 | Target (TGT), Lowe’s (LOW) | Before open | Low-mid (3-5%) | Two more consumer-discretionary reads the same morning. |
| Thu, Aug 20 | Walmart (WMT), TJX Companies | Before open | Low (2-4%) | Closes out the retail cluster. WMT is the single broadest read on US consumer spending of the whole quarter. |
| Wed, Aug 26 | NVIDIA (NVDA), Salesforce (CRM) | After close | High (6-9% NVDA, mid 4-6% CRM) | The single most-watched print of the quarter, NVDA is effectively the AI-capex barometer for every semis and software name reporting after it. CRM reports the same evening. |
| Thu, Aug 27 | Marvell Technology (MRVL) | After close | High (7-10%) | Smaller float and a history of outsized post-earnings swings; trades on the same AI-networking narrative as NVDA a day earlier. |
| Wed, Sept 2 | Broadcom (AVGO) | After close | Mid-high (5-7%) | Custom AI silicon and networking; company-confirmed date. Often trades on hyperscaler capex commentary from the two prior weeks. |
| Thu, Sept 10 | Adobe (ADBE) | After close (tracker consensus, not yet company-confirmed) | Mid (4-6%) | First look at enterprise software demand and AI-feature monetization ahead of Oracle. |
| Tue-Wed, Sept 15-16 | FOMC meeting (not earnings) | Statement + dot plot, 2:00 PM ET Wed | Index-level event | A rate decision lands in the middle of the software/AI-infrastructure stretch. See how FOMC decisions move options prices. |
| Mon, Sept 14 | Oracle (ORCL) | After close (tracker consensus, not yet company-confirmed) | Mid-high (5-8%) | Cloud infrastructure backlog is the number that has moved this stock hardest the last few quarters; lands the day before the FOMC meeting. |
| Thu, Sept 24 | Costco (COST) | After close | Low-mid (3-5%) | Fiscal Q4 close-out; membership fee and renewal-rate commentary matters more than same-store sales here. |
| Tue, Sept 29 | Micron (MU) | After close | High (6-9%) | Closes the quarter; memory pricing is a direct read-through on the same AI-datacenter demand NVDA, MRVL, and AVGO already priced in earlier in September. |
Two tracker-sourced dates above (Adobe, Oracle) had not been individually reconfirmed by the companies themselves as of this writing, though they were consistent across multiple independent trackers. Everything else on the table is a company-confirmed date. Recheck both companies’ investor-relations pages in the days before positioning; report dates do move.
The IV Run-Up Playbook
Earnings options trading is really two separate trades: the inflation phase before the print and the crush at the print. Pre-earnings implied volatility typically expands 15-30% in the 5-10 trading days before a report, then gives back 30-50% immediately after the numbers hit. Neither phase requires predicting the direction of the stock.
10 to 5 days out: the accumulation window
IV starts creeping higher as the report date enters the front expiration cycle. This is the window where long-premium structures (calendars, diagonals, long straddles you intend to exit before the print) are cheapest relative to what they will be worth at peak IV. If you buy premium here, the plan is usually to sell it back before the announcement, not to hold through it.
5 days to the day before: peak inflation
The steepest IV climb usually happens in the final week. Short-premium traders who want to sell the inflated front expiration get the most credit in the last one to two sessions before the report. The trade-off is obvious: you collect peak premium precisely because you are now committed through the binary event.
Print day: the crush
The front-month IV collapse happens at the first tradable moment after the news: the opening bell for a before-open reporter like Home Depot, the next morning for an after-close reporter like NVIDIA. The stock’s actual move versus the expected move decides who wins. Our TSLA and IBM case study walks through two real prints where the crush, not the direction, decided the P&L.
To measure the expected move yourself, take the at-the-money straddle price in the expiration immediately after the report and divide by the stock price. For a deeper walkthrough of the mechanics, start with how to trade options around earnings.
Strategy Selection by Volatility Tier
The most common mistake this stretch of the calendar invites is running the same structure on every name. A strategy sized for a 3% Home Depot move behaves completely differently on a stock pricing a 9% swing. Match the structure to the tier:
| Volatility tier | Typical names (rest of Q3) | Expected move | Structures that fit |
|---|---|---|---|
| Low IV | WMT, HD, TGT, COST, TJX | 2-5% | Debit spreads and narrow iron condors. Premium is thin, so wide short-premium structures collect little for the risk. |
| Mid IV | LOW, CRM, AVGO, ADBE | 4-7% | Iron condors at the expected-move wings; calendars into the print for premium buyers. |
| High IV | NVDA, MRVL, ORCL, MU | 6-10% | Credit spreads and defined-risk short premium (iron condors, short strangles with wings). Rich premium compensates for the wider distribution, but only defined-risk structures cap the tail. |
A hypothetical illustration of the tier logic: a trader expecting a quiet retail print might place an illustrative iron condor on Walmart with short strikes just outside the 3% expected move, risking a defined amount to collect the elevated pre-earnings credit. On NVIDIA, that same trader would need strikes roughly two to three times as far from the stock price to sit outside the expected move, and would still want wings on the position because AI-infrastructure names regularly blow through their straddle in either direction. Both examples are illustrative only, not recommendations.
Retail prints also have their own personality: they cluster on consecutive mornings, they gap at the open rather than overnight, and the read-through from one name (same-store sales, tariff cost commentary, consumer trade-down) often moves the next day’s reporter before it even opens. The AI-infrastructure names behave the opposite way, correlated on capex sentiment rather than on shared operating metrics, which is the subject of the next section.
The AI-Capex Correlation Trap: Aug 26 to Sept 14
The single most important risk-management note for the rest of the quarter is not one earnings cluster week, it is a loosely connected chain running from NVIDIA’s report on August 26 through Oracle’s on September 14. NVDA, MRVL, AVGO, and ORCL are not the same business (GPUs, custom networking silicon, custom AI ASICs, and cloud infrastructure backlog, respectively), but all four trade heavily on the market’s read of AI-capex durability, and a disappointing data point from any one of them has repeatedly moved the others in premarket trading this year.
Holding short-premium positions on more than one or two of these names in the same three-week window is not diversification. It is one large bet that hyperscaler AI spending stays on the trajectory the market is currently pricing, spread across four separate binary events. Staggering exposure, something like never having more than two AI-infrastructure earnings positions open at once, keeps a single disappointing print from resetting sentiment across a book of correlated names. The Fed’s rate decision landing September 16, two days after Oracle, adds a second, unrelated source of index-level volatility right in the middle of the stretch. See how FOMC decisions move options prices for how that interacts with single-name event risk.
One more pattern worth respecting through this stretch: a strong headline beat does not guarantee a positive reaction, especially in a crowded, high-expectation name. We have documented the beat-and-fall pattern repeatedly this year, and NVDA, MRVL, and ORCL all currently carry the kind of elevated expectations that make the pattern more likely, not less.
What the Rest of the Season Tells You
The macro read stacks in a useful order again this leg. Retail reports first (Aug 18-20) and speaks directly to the health of the US consumer heading into the holiday-prep quarter. NVIDIA, Marvell, and Broadcom (Aug 26 to Sept 2) speak to whether hyperscaler AI-infrastructure spending is holding up or starting to moderate. Adobe and Oracle (Sept 10 and 14) give the enterprise-software read, including whether AI features are actually converting to paid seats and cloud consumption, not just headline capex. Costco and Micron close the quarter (Sept 24 and 29), the former on consumer resilience, the latter as a direct read-through on the same AI-datacenter memory demand the earlier names already priced. Options traders who track which of these names beat and still fell, versus beat and ran, get an early sentiment signal that raw EPS tables do not show.
Bottom Line
The rest of Q3 2026 runs retail first (Aug 18-20), NVIDIA and Salesforce together on Aug 26, Marvell the next day, Broadcom on Sept 2, then Adobe and Oracle bracket a Fed meeting from Sept 10 to 16, and Costco and Micron close the quarter Sept 24 and 29. Match your structure to each name’s volatility tier, treat the IV run-up and the crush as separate trades, and treat NVDA through ORCL as one correlated AI-capex risk rather than four independent bets.
FAQ
Q: Why is it called Q3 2026 earnings season if most of these are Q2 results?
A: The season is named for when the reports happen (the third calendar quarter), not the period they cover. Most of the names in the table above are reporting fiscal Q2 2026 results; a few with non-calendar fiscal years (Oracle, Broadcom, Costco) are reporting a different fiscal quarter that still lands in September. Traders anchor on report dates rather than quarter labels for exactly this reason.
Q: How do I calculate the expected move before one of these reports?
A: Take the price of the at-the-money straddle in the expiration immediately following the report date and divide it by the stock price. A $28 straddle on a $400 stock implies roughly a 7% expected move, in NVIDIA’s typical range. Most broker platforms also display this directly on the options chain.
Q: Why treat NVIDIA, Marvell, Broadcom, and Oracle as correlated instead of four separate trades?
A: They sell different things, GPUs, networking silicon, custom AI chips, and cloud infrastructure, but the market currently prices all four heavily on the same underlying question: is hyperscaler AI-capex spending sustainable. A disappointing data point from the first name to report has repeatedly moved the others before they even opened this year. Sizing them as one correlated bet rather than four independent ones is a risk-management choice, not a prediction about any one print.
Q: Should I hold positions through the September 15-16 FOMC meeting?
A: That is a personal risk decision, but understand what you are holding: index-level event risk landing between the Adobe and Oracle reports. Many premium sellers reduce position count or size that week specifically because single-name and macro event risk are stacking in the same few sessions.
Q: Do these report dates change?
A: Yes, occasionally. Dates above were checked against company investor-relations pages and multiple earnings trackers as of August 11, 2026; two (Adobe, Oracle) were tracker consensus rather than individually company-confirmed at that time. Confirm the date on the company’s investor-relations page before opening any position tied to it.
Keep learning: see how this same framework played out on AMD and CoreWeave earlier this earnings season, or start with the core guide to trading options around earnings.