NVIDIA Earnings Countdown: Surpassing Expectations Is Almost Certain, But Wall Street Is Most Concerned About These Five Issues

Author: Long Yue

Source: Wall Street Insights

During Nvidia’s earnings season, the most important thing is no longer the numbers themselves.

On May 18, a team led by Bank of America Securities analyst Vivek Arya released a preview report ahead of Nvidia’s Q1 earnings, which will be announced after the market close on Wednesday, May 20, Eastern Time.

Based on Nvidia’s historical pattern over the past ten quarters, actual revenue has averaged 7% to 8% above management guidance. Management had previously issued an F1Q27 revenue guidance of $78.0 billion; based on that, actual revenue is likely to land in the $83.0 billion to $84.0 billion range, while current market consensus is only $78.7 billion.

In other words, “beating expectations” is practically a sure thing. But analysts believe that after the earnings are released, the five issues below are what will truly move the market.

Cash Returns: Can Nvidia’s “stinginess” be changed?

This is the topic the report devotes the most ink to, and also the core reason they believe Nvidia’s valuation has been discounted long term.

Nvidia is currently the largest company by market capitalization in the S&P 500, accounting for as much as 8.3% of the index’s weight—exceeding Apple (peak 7.9%) and Microsoft (peak 7.2%) at their respective historical highs. The problem, however, is that Nvidia’s shareholder returns do not match its scale.

The data is straightforward: from 2022 to 2025, Nvidia’s free cash flow return rate (dividends + buybacks) averaged only 47%, while industry peers averaged 80%; even Nvidia’s own average over the earlier decade was 80%.

At the same time, Nvidia’s current dividend yield is just 0.02%, versus a peer average of 0.89%. In equity income funds, Nvidia is held by only 16% of funds, while Microsoft is held by 57% and Apple by 32%.

Where did the money go? Analysts point out that Nvidia has put a large amount of capital into its ecosystem—OpenAI, Anthropic, and technology partners. Externally, these investments have drawn considerable controversy, with some arguing they are “circular financing,” meaning Nvidia lends money to customers, who then use that money to buy Nvidia chips.

How big is the valuation discount? The data shows that Nvidia’s expected P/E ratios for 2026 and 2027 are 26x and 19x, while the average for the other members of the “Mag-7” is 49x and 42x—the discount is close to 50%.

More specifically, analysts forecast that Nvidia’s total free cash flow over 2026+2027 combined will exceed $430 billion, higher than the roughly $375 billion combined for Apple and Microsoft. Yet Nvidia’s market cap is about $5.46 trillion, which is about 28% lower than the combined $7.5 trillion for Apple and Microsoft.

Analysts believe that if Nvidia increases dividends and buybacks, it could attract more long-term funds that prefer income, narrow the valuation discount, and also put to rest concerns about “circular financing.” They list this as a “potential catalyst in the second half of the year.”

Vera Rubin: When will the next-generation chips arrive?

Nvidia’s current flagship products are the Blackwell series. What the market is focused on is: when will the next-generation Vera Rubin platform officially ramp into volume?

The firm’s view is the second half of 2026. Vera Rubin (codename R200) uses TSMC’s 3-nanometer process and shares the “Oberon” rack architecture with Blackwell Ultra, so the product transition should be relatively smooth, with limited expected impact on gross margin.

Looking further ahead, Vera Rubin Ultra (codename VR300) will be launched in the second half of 2027, adopting a brand-new “Kyber” rack architecture, and high-bandwidth memory (HBM) will take an even larger share of costs.

The market also wants to hear Nvidia’s latest remarks about the “trillion-dollar revenue forecast” on the earnings call. Nvidia previously provided an outlook for cumulative revenue of $1 trillion for 2025-2027, but contributions from LPU (language processing unit) rack(s), CPU(s), and Vera Rubin Ultra have not yet been included—will this time include an update?

Gross Margin: Can the 75% line hold?

Gross margin is one of the core supports for Nvidia’s valuation.

The analysts’ judgment: in the short term, because Vera Rubin continues to use Blackwell’s rack architecture, gross margin during the product transition period should remain relatively stable. But over the medium to long term, rising HBM memory cost share is a persistent source of pressure.

Market consensus shows Nvidia’s gross margin will fluctuate in the 74% to 75% range, which this firm does not dispute, but it emphasizes that any gross margin performance that comes in above expectations will be a positive catalyst.

How will the AI accelerator market size forecast be updated?

Bank of America previously provided a “trillion-dollar” forecast framework for Nvidia’s AI market from 2025 to 2027. In this earnings report, the market is watching whether Nvidia will update this forecast, especially by incorporating three new growth drivers that were previously not included:

LPU (language processing unit) racks

Vera CPU (Nvidia’s in-house server CPU)

Vera Rubin Ultra

The firm expects that by 2030, the overall AI accelerator market size will reach about $1.17 trillion, and Nvidia will maintain roughly a 68% to 70% market share.

More specifically, Nvidia’s AI accelerator revenue is expected to grow from $102.2 billion in 2024 to $800 billion in 2030. Over the same period, AMD is expected to rise from $5.0 billion to $80.1 billion, and Broadcom from $9.3 billion to $181.9 billion.

Have the threats from Google TPU and CPU competition been exaggerated?

A recent claim circulating in the market is that as AI enters the “Agentic AI” era, the importance of CPUs will surpass GPUs, thereby threatening Nvidia’s moat.

The firm explicitly disagrees and provides two reasons:

First, Nvidia’s in-house “Vera CPU” will disclose new progress at the upcoming Computex conference, and its competitive strength in the independent CPU market should not be underestimated.

Second, in the already large-scale deployment of Blackwell and TPU clusters, the CPU-to-GPU mix is already 1:2, which does not align with the narrative that “agentic AI needs more CPUs.”

The conclusion is: although the CPU market is large, there are many competitors (with strong rivals in both x86 and ARM architectures), and Nvidia’s dominance in the GPU/AI accelerator space is unlikely to be shaken in the short term. By 2030, Nvidia is expected to maintain about 70% of revenue share in the total addressable AI market of more than $1.7 trillion.

Valuation: The “tech darling” at five-tenths off

Finally, we return to valuation. The report uses a set of data to directly point out the contradictions in Nvidia’s current valuation.

Based on CY26/27 expected P/E ratios, Nvidia trades at 26x/19x, while “Tech Seven” (Mag-7) averages 49x/42x—Nvidia’s discount is nearly 50%.

On an EV/FCF basis (enterprise value/free cash flow), Nvidia is at 28x/20x, while the Mag-7 average is 83x/59x—an over 66% discount.

Based on PEG (price/earnings to growth ratio), Nvidia is at 0.41x, while the Mag-7 average is 2.61x, and the S&P 500 overall is above 1.3x.

Bank of America maintains a “Buy” rating and a target price of $320, based on CY27’s expected P/E of 28x (excluding cash), placing it in the middle-lower portion of Nvidia’s historical valuation range of 25x to 56x.

LAYOUT REFERENCE (source): total_lines=87, non_empty_lines=44, blank_lines=43