U.S. stocks slid on Tuesday, November 17, 2026, and technology led the way down. According to midday market reporting carried by MarketScreener, the Nasdaq Composite fell 1.1% to 24,254.1, the S&P 500 dropped 1.0% to 7,338.8, and the Dow Jones Industrial Average lost 1.3% to 50,927.2. Every sector except materials traded lower, and NVIDIA shares weighed on sentiment before the company’s Nvidia earnings release after the closing bell.

The timing explains why so many traders kept one eye on a single calendar entry. NVIDIA’s investor-relations page lists “NVIDIA 3rd Quarter FY27 Financial Results” for 5:00 p.m. EST today. We previewed the setup in our earlier look at report day, and this analysis picks up where that preview stopped: what the midday tape says, what it cannot say, and why an AI-model audience should care about a chipmaker’s quarterly numbers.

One caution before the details. The reporting available at midday covers only the stretch before the release. Nothing in it describes revenue, earnings, guidance, margins, or product announcements, and this article does not guess at them. Where we discuss scenarios, we label them as analysis, not as reported fact.

Add Shattered once in the Google app and our stories show up in your news suggestions.

The Midday Numbers: Nasdaq 24,254.1, S&P 500 7,338.8, Dow 50,927.2

The headline figures are simple to state. At midday on November 17, the three main U.S. benchmarks all traded in the red, and the Dow showed the deepest percentage decline of the three. The Nasdaq, which carries the heaviest technology weighting, fell 1.1%. The S&P 500 sat between the two at 1.0%.

A separate midday report, also dated November 17, captured an earlier and milder snapshot. In that version, NVIDIA shares were down 1.2%, the Dow was down 366.22 points, or 0.71%, to 51,243.38, the S&P 500 was down 29.15 points, or 0.39%, to 7,387.52, and the Nasdaq Composite was down 81.65 points, or 0.33%, to 24,428.11. The table below lines the two snapshots up so readers can see the spread.

The “level gap” column is our own subtraction, not a reported figure. It shows that between the two snapshots the Dow sat roughly 316 points lower, the S&P 500 about 49 points lower, and the Nasdaq about 174 points lower. All three gaps point the same direction, which fits a market that was still sliding as the session wore on.

Why Two Midday Reports Show Different Declines

Readers who spot the mismatch between a 1.1% Nasdaq drop and a 0.33% drop deserve a straight answer. The supplied reporting does not say what time each snapshot was taken, so we cannot pin the difference on a specific minute. The most plausible reading is timing. Index levels tick every few seconds, and “midday” can describe any moment from late morning into early afternoon.

Timing, not contradiction

Two reports that disagree on percentages but agree on direction are not in conflict. Both show all three indexes lower. Both put NVIDIA in the red. The size of the drop is what changed, and a market that is falling into an event tends to produce exactly this pattern, with early losses small and later losses larger as positioning gets trimmed.

How we handle the conflict

We follow a conservative rule here. When sources differ, we present both, we label them as separate snapshots, and we avoid claiming one is the final number. Closing levels will settle the question, and neither snapshot should be treated as the day’s official result. Anyone building a chart, a trading note, or a research memo from this story should pull the official close instead of reusing an intraday print.

Technology Leads Declines, Materials Alone Holds Up

The sector detail matters more than the index headline. According to the midday report, technology led sector declines, and every sector except materials fell. That is a broad pullback, not a narrow rotation. When ten of eleven groups fall together, the cause usually sits above individual company news, in positioning, in rates, or in a shared worry that hits everything at once.

The reporting offers no specific cause for the breadth of the decline, and we will not supply one. The one link the coverage draws is the investor focus on NVIDIA’s pending results. That link is plausible because the chipmaker has become a bellwether for the whole AI build-out, but plausible is not the same as proven. A market can fall for several reasons on the same day, and a single earnings date rarely explains an entire session.

Materials as the lone holdout is a small detail with a possible message. Materials stocks carry little direct exposure to AI chip demand, so their relative calm fits a story in which the pressure concentrated in growth and technology names. We stop short of saying that is what happened. The coverage does not draw that conclusion, so it stays a reading, not a finding.

Nvidia Earnings at 5:00 p.m. EST: The Calendar Fact

The most solid fact in the whole story is the schedule. NVIDIA’s official event listing identifies the release as the third-quarter fiscal 2027 results and sets it for November 17, 2026, at 5:00 p.m. EST. The release lands after the regular session ends, so the market’s verdict arrives in after-hours trading and in the Wednesday, November 18 session, not in today’s close.

That sequencing shapes how to read today’s decline. Investors who sold on Tuesday acted without knowing the numbers. Some trimmed exposure to cut risk. Others may have taken profits in names that ran ahead of the report. A few may have hedged with options. None of that tells us how the quarter went, only how cautious holders felt before finding out.

We tracked a similar pre-report mood recently, when Arm and Marvell slid in an AI chip selloff before the report. Pressure on semiconductor names has built across the week. Today’s broad decline fits that pattern, though the supplied reporting does not link the two directly.

What Is Confirmed and What Is Not

Pre-earnings coverage invites over-reading, so a clean ledger helps. The table separates what the supplied reporting confirms from what it leaves open.

The last row deserves a word. Many earnings-eve stories lean on analyst commentary to fill space. The coverage behind this one contains no on-record quotations from named speakers, so this article carries none. We would rather publish a shorter list of verified facts than dress up a thin record with invented color.

How Pre-Earnings Trading Usually Works

Trading ahead of a major report follows a recognizable rhythm. Volume often thins in the final hour before the release as desks avoid taking new risk. Implied volatility in options markets tends to climb into the event, then collapse once the numbers print. Large holders sometimes cut position size to stay within risk limits, and that selling can look like a bearish verdict when it is really housekeeping.

A stock that carries a heavy index weight pulls the whole tape with it. When shares of a mega-cap technology name drift lower on no company-specific news, passive funds and index-tracking products feel the effect, and the Nasdaq Composite, where large technology names dominate, feels it most. That mechanical link is one reason the Nasdaq’s decline looked larger than the S&P 500’s in the first snapshot, although the Dow’s decline was larger still, so weighting alone cannot tell the whole story.

Readers who want a framework for the numbers themselves can use our guide on how to read NVIDIA earnings, which walks through customer concentration and the line items that tend to move the stock.

Another practical point concerns reading the first headline after the release. Automated news feeds often publish a revenue or earnings-per-share figure within seconds, and the first reaction in after-hours trading can reverse once management starts talking on the call. Careful readers wait for the full release, the guidance range, and the commentary before drawing a conclusion. The first minute of a reaction is often the least reliable one.

Why Nvidia Sets the Tone for AI Models

This story sits in our AI and machine learning coverage for a reason. Frontier model labs, open-weight projects, and the startups building on top of them all depend on GPU supply. Training runs, inference capacity, and the price per token that developers see on an API invoice trace back, in part, to what accelerators cost and how many can be delivered.

That chain means NVIDIA’s quarterly report works as a read on the whole AI compute market. Management commentary on demand, supply constraints, and customer plans shapes how labs budget for the next generation of models. A confident outlook tends to support aggressive model-scaling plans. A cautious one can slow them. Neither outcome is known as of this writing, which is exactly why the pre-report mood has been so tense.

Developers feel this indirectly. Rate limits, context-window pricing, and the availability of newer model tiers all reflect compute economics several steps upstream. When chip supply loosens, providers can cut prices or expand access. When it tightens, they ration. Teams planning AI spending for 2027 will read the report for hints about which way that dial is turning.

Market Impact: Who Feels a Beat, a Miss, or a Shrug

We cannot say how the market will react, but we can map who has exposure. Direct exposure sits with NVIDIA holders and the funds that carry it as a top position. Second-order exposure runs through chip designers, memory makers, networking vendors, and equipment suppliers that sell into the same data-center build-out. Third-order exposure reaches cloud providers and AI labs, whose spending plans depend on component availability.

The sentiment loop also runs the other way. A weak reaction in the chipmaker can drag down names that depend on its customers, and a strong reaction can lift them. We covered how the AI selloff resumed before earnings, a reminder that expectations had already cooled before today’s session opened.

For broad-market investors, the concern is concentration. When a handful of AI-linked stocks account for a large share of index gains, a stumble in one can move the entire benchmark. Today’s decline, with technology in front, shows that concentration at work, even if the reporting stops short of crediting any single cause.

Individual investors face a similar choice. Selling into a nervous session locks in a loss before the facts arrive, while holding through the report accepts the risk of a gap in either direction. Neither path is free of cost, and neither is a reaction to today’s decline alone. Long-term holders usually ask whether the thesis behind the position changed, not whether the index dropped on a Tuesday.

Competitive Picture: Where Rivals Fit in the AI Compute Chain

NVIDIA does not operate alone, and the competitive frame shapes how analysts read its report. AMD sells accelerators aimed at the same data-center customers. Cloud providers design in-house chips to cut dependence on outside suppliers. Specialist designers and networking firms supply the pieces that make large clusters work. Each of these players gains or loses mindshare depending on what NVIDIA says about demand.

Our earlier piece on how a leadership change tests NVIDIA’s AI chip share lays out the rivals in more detail. The short version is that the supplied reporting for today’s story names no competitor and makes no comparison, so we treat this section as background, not as news from the session.

What matters for model builders is optionality. Labs that can run across more than one hardware family hold more negotiating power on price and delivery. That is a structural point, not a claim about this quarter, and it explains why every AI-chip report draws attention from companies that never buy a single NVIDIA part.

Historical Context: Earnings Eves and Index Moves

NVIDIA’s reports have grown into set-piece events for the entire market, and the days before them often carry a nervous undertone. The supplied reporting gives no historical figures for past reactions, so we avoid quoting specific percentages from earlier quarters. What we can say is structural. As AI spending moved from experiment to budget line, one company’s quarterly call became a proxy for the sector’s health.

That shift changed the pre-report routine. Years ago, a chipmaker’s results mattered mainly to chip investors. Now software firms, cloud providers, utilities that serve data centers, and even retailers watch the call for signals. A pre-earnings dip in technology is therefore not unusual, and a 1% move across the major indexes on a nervous day does not by itself point to a crisis.

Context also comes from the week’s wider mood. Our look at U.S. futures falling ahead of NVIDIA and a major AI IPO showed the market juggling several AI-linked catalysts at once. Today’s decline lands inside that busy stretch, so reading it as a verdict on one company would oversimplify.

Five Predictions for the Next 48 Hours

These are analyst-style judgments from our desk, not reported facts and not investment advice. Each is conditional, because the results are unknown as we write.

  • Volatility peaks at the open on November 18. After-hours trading will set the tone, and the regular session usually digests it within the first hour. Expect the largest swings there.
  • Guidance draws more attention than the quarter itself. Investors usually trade the forward outlook, so commentary on demand and supply will likely outweigh backward-looking figures.
  • Technology stays the swing sector. With materials the only group that held up today, the next large move in the indexes will probably follow technology in either direction.
  • AI-chip peers move in sympathy. Designers, memory makers, and equipment suppliers tend to follow the lead chipmaker after a report, so expect correlated moves.
  • Model labs and cloud providers cite the call. Within days, AI companies will point to the supply commentary when they discuss capacity and pricing plans.

Each prediction could fail. A surprise in either direction, or a macro event unrelated to AI, could override all five. We will update our coverage once the results are public and verifiable.

What AI Teams and Developers Should Watch

Signals that touch your roadmap

Engineers do not need to trade stocks to benefit from reading the report. Three themes matter most for people who build on AI models. First, any commentary on supply and delivery times, since that affects how quickly providers can add capacity. Second, remarks on inference demand, which hint at how crowded shared model endpoints may get. Third, signals about next-generation hardware timing, which shape when new model tiers can launch.

What not to infer from a midday dip

A falling index on report day tells a developer nothing about API pricing next month. Price changes follow contracts, capacity plans, and competition, not one afternoon of trading. Teams should wait for the actual results and for provider announcements before changing budgets or migrating workloads.

For more background on AI models, our AI and machine learning section collects current coverage of model releases, benchmarks, and the infrastructure behind them.

Risks to This Reading

Several things could make today’s analysis look dated within hours. The most obvious is the report itself. A result far from expectations in either direction would swamp the pre-report mood. The second is data revision, since intraday index figures change by the second, and the two snapshots already differ. The third is cause. The coverage ties investor focus to the pending results, but it offers no proof that the report drove the broad decline, so another factor may have contributed.

We also flag a limit on our sourcing. The supplied reporting includes no named speakers and no direct quotations, and it gives no information about what NVIDIA management plans to say. Readers who want a more complete picture should wait for the company’s published materials and for the official market close.

Looking past the earnings date, the broader question for the AI sector is durability. Spending commitments, power availability, and customer financing all affect how long the build-out can keep its current pace. None of those topics appear in the supplied reporting for November 17, so we flag them as open questions. They will matter for model builders long after this week’s price action fades from memory.

The Takeaway

The takeaway: on November 17, 2026, U.S. stocks fell at midday, technology led, only materials held up, and the market waited for NVIDIA’s 5:00 p.m. EST report. Two separate midday snapshots put the Nasdaq down between 0.33% and 1.1%, the S&P 500 down between 0.39% and 1.0%, and the Dow down between 0.71% and 1.3%. What happens next depends on numbers that did not exist when the reports behind this story were written.

Frequently Asked Questions

When does NVIDIA report its fiscal third-quarter FY27 results?

NVIDIA’s investor-relations listing schedules “NVIDIA 3rd Quarter FY27 Financial Results” for November 17, 2026, at 5:00 p.m. EST, after the U.S. market close.

How far did U.S. stock indexes fall on November 17, 2026?

At midday, one report showed the Nasdaq Composite down 1.1% to 24,254.1, the S&P 500 down 1.0% to 7,338.8, and the Dow down 1.3% to 50,927.2. A separate midday report showed smaller declines of 0.33%, 0.39%, and 0.71%. The difference most likely reflects when each snapshot was taken.

Which sectors fell?

Technology led the declines, and every sector except materials fell, according to the midday report.

Did NVIDIA’s pending report cause the selloff?

The reporting says NVIDIA shares weighed on sentiment ahead of the results, but it names no single cause for the broad decline. We treat the link as plausible and unproven.

How much did NVIDIA shares fall?

In a separate November 17 midday report, NVIDIA shares were down 1.2%. The other report describes the shares as weighing on sentiment without giving a percentage.

What are the results themselves?

Unknown at the time of writing. The supplied reporting covers only the period before the release, so revenue, earnings, guidance, margins, and product news remain unconfirmed.

Why does an AI-model audience care about a chipmaker’s earnings?

Model training and inference run on accelerators, so supply, demand, and pricing signals from the leading chipmaker shape how labs plan capacity and how providers price access. That influence is indirect and slow, so a single report rarely changes API pricing overnight.