Morgan Stanley sets bold new price target on Nvidia stock
Despite the volatility, Morgan Stanley is still standing its ground on Nvidia (NVDA) stock.
The big bank stuck with an Overweight rating on the AI bellwether’s stock, slapping a $250 price target.
The fresh target implies an eye-popping 30% upside from Nvidia’s current price, hovering around $192.
In fact, the bank lays out an even loftier $330 bull case on robust execution and a $150 bear case if growth cools off quicker than expected.
It has clearly been a busy week for Nvidia, marked by China green-lighting first H200 chip imports and encouraging hyperscaler capex cues.
Consequently, since the last time I covered Nvidia stock (on Jan. 22), it’s up nearly 4% with its current market cap at a dizzying $4.68 trillion.
Nevertheless, Nvidia has its work cut out this year, but Morgan Stanley analysts feel that the fundamentals look mostly intact.
However, that conviction has yet to be reflected in the stock market, for a variety of reasons. Morgan Stanley believes Mr. Market is still fixated on the overhangs, which conflict with their latest checks as we enter a new phase of the relentless AI trade.
Vera Rubin is proof point that can reset the narrative
Morgan Stanley feels that Vera Rubin is the clearest way for Nvidia to cut through all the competitive noise in the AI space.
More Tech Stocks:
- Morgan Stanley sets jaw-dropping Micron price target after event
- Nvidia’s China chip problem isn’t what most investors think
- Quantum Computing makes $110 million move nobody saw coming
- Morgan Stanley drops eye-popping Broadcom price target
- Apple analyst sets bold stock target for 2026
The bank expects several AI chipmakers to post 75%+ growth this year, suggesting everyone’s gaining ground amid red-hot demand. Rubin, in Morgan Stanley’s view, is what brings clarity.
For the most part, comparisons based on vendor specifications oversimplify reality. Until systems are running at scale, customers can’t gauge the true cost of ownership.
That’s where Rubin switches up the conversation, with Nvidia delivering entire end-to-end systems, not just individual chips.
As CEO, Jensen Huang put it at this year’s CES event,
Morgan Stanley expects a significantly quicker-than-expected transition from Blackwell, on the back of superior manufacturability.
For instance, board assembly time is expected to drop to nearly five minutes, compared to about two hours previously, paving the way for a much smoother ramp into 2026.
Related: Bank of America delivers sobering stock market take
Vera Rubin versus Blackwell
- Blackwell is Nvidia’s current flagship AI platform, and Vera Rubin is its successor, tailor-made for quicker and more cost-effective AI factories.
- Vera Rubin is expected to cut inference token cost by up to 10 times versus Blackwell.
- For training, Nvidia says Rubin needs four times fewer GPUs than Blackwell, reducing racks, power, and cooling requirements while delivering five times quicker AI computing power.
- Vera Rubin adds “context memory storage” in enabling chatbots to respond faster to longer prompts and conversations.
Source: Nvidia investor relations press release
That explains Nvidia’s monstrous lead in AI chip sales (nearly 85% share), and Rubin will help significantly extend it.
Big Tech is still spending, and the numbers matter for Nvidia
Nvidia’s growth engine is humming, and that is unlikely to change anytime soon, given current hyperscaler capex patterns. Moreover, for it to ‘win’, it just needs those marginal dollars to follow into high-performance computing, and the most recent numbers underscore that trend.
Related: Apple report drops iPhone 18 bombshell
For perspective, CreditSights estimates that the top five hyperscalers will spend nearly $602 billion in 2026, up 36% year over year. Also, Goldman Sachs points to a wave of AI investment of $500 billion-plus this year.
The most recent proof point comes from Microsoft and Meta.
For context, Microsoft just reported a whopping $37.5 billion capex figurefor fiscal Q2 2026, up nearly 66% year-over-year. Nearly two-thirds of that spending is for chips and compute assets
Meta’s results echoed that same trend.
It guided 2026 capex to$115 billion to $135 billion, comfortably above Wall Street expectations of $110 billion and significantly higher than the $72.2 billion it reported in 2025.
For Nvidia, as long as these hyperscalers lean on compute-heavy builds, the tailwind will continue powering their story.
Wall Street’s latest Nvidia price targets versus the current price
- Bank of America: $275 (+42.9%).
- Citi: $270 (+40.2%).
- Wells Fargo: $265 (+37.7%).
- J.P. Morgan: $250 (+29.9%).
- RBC Capital Markets: $240 (+24.7%).
The risks that could reset Nvidia’s valuation
Morgan Stanley remains constructively bullish on Nvidia, but it also lays out a scenario where things could go wrong.
In its bear-case scenario, Nvidia stock could plummet to $150, on the back of sluggish growth, competitive pressures, and valuation compression.
Perhaps the biggest risk is demand normalization.
If AI spending cools faster than expected and the demand/supply imbalance corrects quickly, we could see a meaningful acceleration in growth rates.
Competition is another critical pressure point.
Specifically, the risks of custom silicon (ASICs) or a much stronger ramp-up from AMD will derail Nvidia’s pricing power. Also, with AMD pushing its own full-stack agenda with Helios, Nvidia clearly has its hands full.
Financing optics also matters a ton.
Though Nvidia’s direct credit exposure appears mostly limited, anxiety around client funding structures can weigh down investor sentiment (who are now in show-me mode with the AI trade).
Throw in export restrictions, tariffs, or geopolitical friction, and things could bite even more, calling into question Nvidia’s lofty pricing multiples (currently 42-times forward GAAP earnings).
