Nvidia Earnings Beat Ends Post-Report Slump on $108B Guidance

A pattern finally broken, but the premium remains

In the run-up to its fiscal second-quarter report in late August, Nvidia had a strange problem. It kept beating expectations, and its stock kept falling anyway. The shares had dropped on the trading day after each of the previous four reports, even though the company met or beat estimates for earnings per share, revenue and forward guidance every time.11 CNBC described an "expectations premium": a year of outstanding profits had raised the bar for what counted as a real beat.11

This time the pattern broke, though it wasn't obvious at first. After an initial dip in extended trading, Nvidia closed up 8.7% the following day, its biggest one-day percentage gain since April 2025.22 CNBC estimated the move added about $440 billion to the company's market value.23 Six weeks later, though, the stock was trading around $226, well below the price targets analysts have set.3 The premium has eased, but it hasn't gone away.

The numbers: a clean beat on every headline line

Fiscal second-quarter revenue came in at $96.22 billion, up 106% from a year earlier. Analysts had expected roughly $92.2 billion.14 Adjusted earnings per share were $2.22. Most consensus figures were $2.09 or $2.10, so the beat was about 5.7%.4 Data center revenue was $89 billion, about 92% of the total.29 Estimates for that segment had sat around $85 billion.25

On a GAAP basis, diluted EPS was $2.46, up 128%, and net income was $59.7 billion.9 Nvidia earns very high margins on enormous sales. MarketBeat put its net margin for the quarter at about 64%.5

The trend over several quarters matters as much as any single figure. Year-over-year revenue growth went from about 73% to 85% and then 106% over three quarters.1 The prior quarter had already been strong: $81.6 billion in revenue against estimates of about $78.4 billion, and adjusted EPS of $1.87 against $1.76.8 Growth is speeding up at a revenue base most companies never reach.

Why guidance, not the quarter, moved the stock

Before the report, analysts and strategists mostly agreed that the quarter itself mattered less than the outlook. Wedbush's Matthew Bryson argued that because Nvidia beats so consistently, investors already assume it will exceed whatever it guides.11 Siebert's Mark Malek said consensus for the third quarter was around $104 billion, while buy-side investors privately expected more, and that the stock would rise or fall on that gap.22 Rosenblatt's Kevin Cassidy said guidance above $105 billion would be an upside surprise and $110 billion would be a blowout.11

Nvidia guided to $108 billion, plus or minus 2%.9 Outlets measured that against different consensus figures: $104.2 billion from LSEG,26 $103.9 billion in another tally,28 and $105.39 billion from Visible Alpha.21 On any of them, the guide landed between Cassidy's "upside" and "blowout" levels. It also assumes no data center compute revenue from China, so any sales there would be extra.9

The longer-range outlook mattered more. CFO Colette Kress said Nvidia expects about 70% revenue growth in fiscal 2028.27 Analysts surveyed by LSEG had expected about 44%.27 CEO Jensen Huang said underlying demand is "much greater than 70%" and that supply is what limits growth.27 Nvidia rarely gives a forecast that far ahead, and it did more than any single quarterly beat to change the debate. Direxion's Jake Behan said the conversation moved from whether AI spending is peaking to how long the buildout can last.24

Where the coverage diverges: the first reaction versus the second

Reports of the stock's reaction differ depending on when they were written. Some accounts described shares slipping after hours, which they attributed to a weaker gross margin outlook.25 Others reported a gain of about 4% later in the same evening session.27 Some of the coverage that described a sell-off was written in the first hour after the release.

The initial dip had a clear cause. Nvidia guided third-quarter gross margin to 74%, down from 75%, and said margins could bottom at 71% to 72% in the January quarter as memory and component costs rise, before settling at 72% to 73% the following year.14 Investing.com's Thomas Monteiro called it the first sequential margin decline of the current cycle.28

My reading is that the full-day move tells you more than the first after-hours reaction. Investors decided that much more revenue at slightly lower margins is a good trade. That holds only if margins actually stop at the floor Nvidia gave. Pressure from memory costs is now the main risk to the profit story, more than demand.

The new ceiling: being sold out

One unusual point from the coverage is that Nvidia's strength may make big surprises harder. Seaport Research's Jay Goldberg asked where upside can come from when a company is already sold out.14 If capacity is mostly spoken for, Nvidia can't easily beat its own guidance by a wide margin, and investors have come to expect wide margins of victory. Huang said the company is working with memory suppliers and securing power, land and data-center capacity, and he pointed to a price increase planned for the first quarter.28

That is a structural form of the expectations premium. Investors aren't doubting demand. Shipments are limited by supply chains Nvidia doesn't fully control, so the size of its beats is limited too.

The financing question

A second concern, mentioned less often, is how Nvidia's growth is being paid for. REX Financial's Bill Birmingham noted that Nvidia disclosed $108.5 billion of maximum guarantee exposure, including a $105 billion structure tied to SB Energy and OpenAI.22 He argued that the AI economy is now limited by access to capital rather than chip supply, and that guarantees make Nvidia's future revenue lower quality because they add default risk.22 Before the report, Morgan Stanley had already pointed to concerns about circular financing.12 JPMorgan, after the prior quarter, flagged $119 billion in supply commitments as a risk if hyperscaler spending slows.20

This is the strongest case against the stock, and the earnings call didn't resolve it. Nvidia now acts partly as a lender to the AI buildout as well as its main supplier, as CNBC put it.11 That role raises the stakes if any large customer has trouble paying.

Buybacks as the next catalyst

Several analysts suggested a way to narrow the gap between results and the share price: return more cash to shareholders. Bank of America reiterated a $350 target and argued that a shift toward Apple-style buybacks could re-rate the shares. It added that returning more cash would lower the risk of overfunding the AI buildout.21 Cassidy at Rosenblatt made the same comparison to Apple, noting that iPhone growth stopped impressing investors but buybacks then supported the stock.11 A later Rosenblatt note keeping its $390 target cited a "massive $235B buyback."2 I take this as a sign that Wall Street sees capital returns, more than further beats, as the next source of share-price gains.

What analysts are watching, and where targets sit

Analysts remain almost unanimously bullish. MarketBeat counts 55 analysts with a consensus Buy rating and an average target of $321.89, ranging from $210 to $515.1 Yahoo Finance shows an average of $328.72.3 Recent reiterations include UBS at $300, Citi at $315 and Cantor Fitzgerald at $350.2 Consensus for the current quarter is about $2.47 in EPS on $108.8 billion in sales.2 That figure is already slightly above the midpoint of Nvidia's guidance, so the bar has risen again.

Valuation explains much of that optimism. Nvidia trades at about 25 times forward earnings.17 That is a modest multiple for a company more than doubling its revenue. A year ago, the consensus target was $212.84.1

For the coming report, the main things to watch are whether gross margins hold in the 72% to 74% range, whether shipments keep pace with guidance despite memory shortages, whether Nvidia formalizes larger capital returns, and whether its financing exposure grows. History offers some reassurance. LPL Financial found that after Nvidia's 15 reports since October 2022, the stock averaged about 21% returns over three months, with gains in 70% of those periods.24

My conclusion is that the August report showed the expectations premium can be overcome, but only with a forecast that changes the long-term outlook, not just a quarterly beat. Nvidia can't produce a 70% growth outlook every quarter. With the stock still well below analyst targets, the next jump will probably depend on margins, supply and cash returns.