As Nvidia gears up for its quarterly earnings release next Thursday, Morgan Stanley anticipates the chipmaker will likely deliver another "beat and raise" quarter, driven by persistent demand strength and improving supply conditions. However, the investment bank cautions that simply surpassing expectations may no longer be sufficient to propel the stock higher, marking a shift from previous quarters.

Morgan Stanley maintains an "Overweight" rating on Nvidia with a $288 price target and continues to list it as a "top pick." The bank projects revenue of $91.1 billion for the July quarter and $102.3 billion for the October quarter, while noting that wafer starts for Blackwell's 4nm process have improved, suggesting there is still upside to current forecasts.

The stronger the earnings, the higher the bar becomes. Nvidia has seen its shares decline the day after earnings for four consecutive quarters, reflecting a shift in investor focus from "can they beat" to "can they sustain the beat." Morgan Stanley believes that medium-to-long-term variables such as market share, gross margins, the revolving capital model, and the ramp of Rubin are the key determinants of whether Nvidia's valuation can expand further, and these are difficult to fully address in a single quarterly report.

Blackwell Demand Remains Robust

Morgan Stanley forecasts Nvidia's July quarter revenue at $91.1 billion and October quarter at $102.3 billion, with corresponding earnings per share of $2.07 and $2.34, roughly in line with consensus estimates of $2.08 and $2.36.

On the surface, Morgan Stanley's earnings estimates do not significantly exceed market consensus. The bank believes the real focus should be on supply-side signals. During its June Asia tour, Morgan Stanley observed that 4nm wafer starts for Blackwell have increased, implying that current supply assumptions in the model may still be conservative. In other words, the market may only see "demand remains strong," but how much Nvidia can actually deliver could be a source of future earnings revisions.

Morgan Stanley therefore expects Nvidia to continue its "beat and raise" pattern. However, since the market has already revised up expectations for the October quarter based on recent supply signals, the largest near-term earnings upside may not be in the next quarter but in the January quarter, when Rubin shipments begin to scale up significantly.

Rubin Shipments Begin, but the Real Impact Comes Next Year

The Rubin product line is progressing as expected. Nvidia has confirmed that Rubin will begin shipping in the third quarter, with Morgan Stanley estimating approximately 150,000 units shipped, corresponding to around $9 billion in revenue, at an average selling price of about $60,000 per unit.

Recently, KYEC, a key testing partner, lowered its full-year guidance and pushed more AI-related revenue to the fourth quarter of 2026 and 2027, raising concerns about Rubin's shipment pace. However, Morgan Stanley's Asia semiconductor analyst Charlie Chan believes this change reflects longer test cycles rather than lower shipment volumes. The bank's updated model maintains the full-year Rubin test volume unchanged, only shifting more weight from the third quarter to the fourth quarter.

More importantly, even if third-quarter Rubin shipments come in below expectations, the impact on Nvidia's overall results would be relatively limited. Morgan Stanley's current estimate of 150,000 units is already significantly below the revised industry estimates, so short-term test cycle changes are insufficient to alter Nvidia's overall revenue trajectory.

Three Key Questions That Will Determine Post-Earnings Stock Direction

For Nvidia, the real challenge has shifted from "can earnings beat expectations" to "can valuation find a new upside narrative." Morgan Stanley identifies margins, revolving capital, and market share as the three most contentious issues for investors, and the upcoming earnings call may still not provide definitive answers.

First, margins. Morgan Stanley expects management to reiterate the FY2027 gross margin guidance of "mid-70s." However, the bank is cautious about the consensus expectation of a recovery in gross margins in the second half of FY2028. Cost pressures from DRAM, front-end wafers, packaging, and substrates are likely to persist, meaning margins could face further downside risk. If Nvidia lowers its margin guidance, it would undoubtedly pressure the stock in the near term, but from a valuation perspective, a margin cut could also serve as a sentiment-clearing event.

Second, the revolving capital model. Jensen Huang has already explained the logic behind the $500 billion cooperation framework with partners through a blog post. Morgan Stanley expects management's comments on the earnings call to remain largely unchanged.

Third, market share. With ASIC competition intensifying and AMD advancing its AI chip business, the market is focusing on whether Nvidia can maintain its leading position. Morgan Stanley expects management to emphasize Rubin's significant improvement in AI factory economics compared to Blackwell. However, since Rubin is still in early production, the market will need more shipment data and customer feedback to assess its true competitiveness against ASIC and AMD.

The Market May Still Be Underestimating 2027

If near-term earnings lack strong catalysts, longer-term growth expectations may be the key to further valuation expansion for Nvidia. Morgan Stanley believes market expectations for Nvidia's 2027 remain too low. The bank projects FY2027 revenue of $393 billion and FY2028 revenue of $598.8 billion, significantly above the consensus estimate of $562.4 billion.

Jensen Huang stated at the GTC conference that the company has $1 trillion in demand visibility for Blackwell and Rubin combined from 2025 to 2027. This figure includes networking business but excludes Groq, standalone CPUs, RTX, and software businesses.

Morgan Stanley notes that current consensus estimates imply combined data center revenue of approximately $1.054 trillion, while its forecast is $1.09 trillion. Considering the 2025 data already includes about $30 billion in Hopper and related networking revenue, the bank believes this forecast does not set an especially high bar.

Meanwhile, inference demand is accelerating, with growth from frontier labs, enterprises, and sovereign customers. Combined with the ramp of new products like Rubin and potential incremental revenue from profit-sharing arrangements, Morgan Stanley believes Nvidia's actual business scale in 2027 could be significantly higher than current market expectations.

Even Strong Earnings May Not Immediately Move the Stock

So next week's earnings report may present a familiar scenario: results beat again, guidance raised again, but the stock may not surge as a result. Four consecutive quarters of post-earnings declines suggest that the market's pricing logic for Nvidia is evolving. As the earnings base continues to rise, "beat and raise" is transitioning from a stock catalyst to a market baseline requirement.

Morgan Stanley remains bullish on Nvidia's long-term fundamentals and believes Blackwell demand, Rubin's ramp, and growing inference demand will continue to support results. But for near-term stock performance, the market is truly waiting to see whether market share can be defended, margins can stabilize, and whether Rubin can prove Nvidia's leadership in the next round of AI infrastructure competition.

In other words, the focus of next week's earnings may not be whether Nvidia can deliver another strong report, but whether that report is strong enough to make investors willing to pay up for higher valuations again.