Bank of America resets Nvidia stock forecast

Nvidia (NVDA) stock has gained about 32% over the past year, at the time of writing, Tuesday afternoon, Feb. 17, according to Yahoo Finance. Meanwhile, the SPDR S&P 500 index (SPY) is up about 11% in the same period.

It is impressive that Nvidia has outpaced the S&P 500 by more than 20%, but even more so when we look at other members of the Magnificent 7:

• Meta is 13% down in the same period.

• Amazon is down 12%.

• Microsoft is down almost 3%.

• Apple is only up a little more than 7%.

• Tesla is up 14%.

The only exception, a Mag 7 member that fared better than Nvidia, is Google, which is 62% up in the same period.

Hyperscalers have announced huge capital expenditures (capex) for this year. I’ve covered Amazon earnings and capex outlook here, and Meta’s earnings and capex outlook here. Nvidia will release its Q4 earnings on February 25.

Bank of America analyst Vivek Arya and his team updated their cloud capex outlook, AI data center total addressable market estimates, and revised forecast for Nvidia ahead of earnings.

To refresh your memory on how Nvidia Q3 earnings looked, read my article “Goldman Sachs, JPMorgan and BofA drop verdicts on Nvidia earnings.”

• Revenue is expected to be $65.0 billion, plus or minus 2%.

• GAAP gross margins are expected to be 74.8% plus or minus 0.5%.

• GAAP operating expenses are expected to be approximately $6.7 billion.

Arya and his team believe that capital expenditures for 2026 and 2027 will reach $748 billion and $869 billion, respectively. These numbers represent year-over-year growth of 56% and 16%, respectively.

Given this exceptionally strong 2026 cloud capex outlook, the team has now updated its 2030 AI data center systems total addressable market estimate to approximately $1.4 trillion, up from the prior estimate of $1.2 trillion.

Related: BofA revamps Applied Materials stock price target after CEO message

Analysts expect the overall data center systems total addressable market to accelerate in 2026, growing by 64% YoY, with AI systems growing even faster at approximately 100% YoY, driven by new AI accelerator deployments.

The team expects that AI accelerator vendors will generally pass on their rising HBM/DDR costs to customers, maintaining margins as scale and system sales increase.

Analysts said that even after excluding Google due to its reliance on its own TPUs, 2026 cloud capex is expected to increase approximately 45% YoY, to $180 billion.

Assuming two-thirds of this spending will go toward servers, analysts estimate an increase of approximately $120 billion in potential compute chip sales this year, which is in line with their new Nvidia outlook of a $110 billion increase in data center compute sales.

The team added that the remaining $10 billion to $20 billion of sales will be split across AMD and non-TPU ASICs.

Related: History of Nvidia: Company timeline and facts

Analysts raised their sales estimates for Nvidia for fiscal years 2027, 2028, and 2029 by +7%/+2%/+2% to $342.33 billion/$422.75 billion/$496.3 billion, and raised EPS estimates by +8%/+3%/+3% to $8/$9.98/$11.94.

The team said that Google’s Gemini 3 Pro AI model, which was trained on Google’s custom TPU v7 Ironwood chips co-designed with Broadcom, may be the best AI model for now, but that it may soon change with Blackwell-trained LLMs.

More AI Stocks:

They noted that all GPU-based models on the market today were trained on older-generation Nvidia Hopper GPUs, except Grok-4. However, even Grok-4 was mostly trained on older Nvidia chips, as xAI’s Collosus cluster, on which it was trained, consists of 150,000 H100, 50,000 H200, and only 30,000 Blackwell chips, B200/GB200.

The team expects any Blackwell-majority-trained LLMs to “rewrite the reasoning/coding benchmarks” once broadly available in early 2026.

In a research note shared with me, Arya reiterated a buy rating for Nvidia stock and the target price of $275, based on 28 multiple his estimate for price-to-earnings ratio excluding cash for calendar year 2027, which is within Nvidia’s historical forward year price-to-earnings range of 25 to 56.

• Weakness in consumer-driven gaming market

• Competition with major public firms

• Larger-than-expected impact from restrictions on compute shipments to China

• Lumpy and unpredictable sales in new enterprise, data center, and autos
  markets

• Potential for decelerating capital returns

• Enhanced government scrutiny of Nvidia’s dominant market position in AI
  chips

Related: Morgan Stanley flags $45B hidden cybersecurity opportunity

This story was originally published by TheStreet on Feb 17, 2026, where it first appeared in the Investing section. Add TheStreet as a Preferred Source by clicking here.