Wall Street’s AI Paradox: Why Has NVIDIA’s Stock Flatlined as Hyperscaler Spend Explodes?
-
Nvidia (NVDA) shares dropped another 2.23% on Friday and are now negative for the year despite massive spending plans from NVIDIA’s largest customers.
-
NVIDIA is curerntly trading for slightly north of 23X forward earnings.
-
Wall Street is worried about competition: Broadcom AI revenue surged 74% to $6.2B while AMD Data Center hit $5.4B as competition intensifies.
-
A recent study identified one single habit that doubled Americans’ retirement savings and moved retirement from dream, to reality. Read more here.
Here’s the paradox keeping Wall Street analysts up at night: Tech giants are pouring $700 billion into AI infrastructure, yet Nvidia (NASDAQ:NVDA) stock has gone cold. Bloomberg’s headline on Friday captured it perfectly: “Nvidia Shares Go Cold Even as Big Tech Spending on AI Balloons.” The stock that was supposed to be the ultimate AI beneficiary is seeing limited gains while its customers announce record spending sprees.
The numbers tell a story that seems disconnected from reality. Nvidia just posted year-to-date returns of negative 1.98% through Friday. Over the past month, the stock has been essentially flat while Meta Platforms (NASDAQ:META), Alphabet (NASDAQ:GOOGL), and Amazon (NASDAQ:AMZN) announced AI infrastructure buildouts that would make a small country jealous.
Prediction markets tell the same story. Polymarket traders are pricing in just a 35% probability that Nvidia closes February above $190. The crowd expects the stock to stay range-bound in the $185-$190 zone, not exactly the behavior you’d expect from a company sitting at the center of a $700 billion spending boom.
If you’re looking for reasons for weakness, there are plenty that investors could point to.
Insider selling has been consistent. CFO Colette Kress dumped approximately 164,000 shares across 60 transactions over three months. EVP Ajay Puri liquidated 438,973 shares worth roughly $80.8 million. Director Mark Stevens sold 572,500 shares for about $103.7 million. Yet, I wouldn’t describe this level of (normally) planned selling as atypical for a company of NVIDIA’s size.
The competitive threats are real. When there’s a prize the size of being the ‘brains’ of the AI build out, competitiors will emerge and hyperscalers will do everything in their power to diversify.
Broadcom (NASDAQ:AVGO) just reported AI semiconductor revenue growing 74% year-over-year to approximately $6.2 billion in Q4, with guidance projecting $8.2 billion in Q1 as hyperscalers invest in custom chips. Since Q4, Broadcom executives have said their business continues accelerating. Broadcom’s design services are essential for Google’s TPUs, and third-party data shows the majority of AI processors Google deployed last year were TPUs rather than NVIDIA’s GPUs.
Alphabet’s CTO casually mentioning their “real secret weapon” is co-designing hardware, models, and infrastructure in-house. When your biggest customers are also becoming your competitors, that’s margin pressure waiting to happen.
Advanced Micro Devices (NASDAQ:AMD) saw its Data Center segment hit $5.4 billion in Q4, up 39% year-over-year, driven by what CEO Lisa Su called “rapid scaling of data center AI franchise.” Su upped the long-term growth rate for the company’s data center products.
We were providing live commentary and analysis on Arista’s earnings on Thursday afternoon when the company’s CEO said this on their conference call:
“A year ago, it was pretty much 99% NVIDIA, right? Today, when we look at our deployments, we see about 20%, maybe a little more, 20%-25%, where AMD is becoming the preferred accelerator of choice.
And in those scenarios, Arista is clearly preferred because they’re building best-of-breed building blocks for the NIC, for the network, for the IO, and they want open standards as opposed to a full-on vertical stack from one vendor. So you’re right to point out that, AMD, and in particular.”
Arista only sees a part of the market, but the comments – and fears of NVIDIA losing market share – were enough to push down shares 2.23% in yesterday’s trading.
And yet, despite all this, NVIDIA seems more resilient than you’d imagine. A quarter ago, Wall Street was projecting $6.81 in Fiscal 2027 (the year NVIDIA is now in) sales. Today, that number has risen to $7.74.
Given how over-consensus each of the major hyperscalers’ capex was, would it shock you if earnings ended up being over $9 per share this year?
That’s likely what’s currently ‘holding’ NVIDIA’s shares back. Currently, they trade for about 23X forward earnings. A next wave of upward earnings revisions could easily lead to a higher multiple as it becomes clear how much growth still remains in the upcoming years.
For example, if NVIDIA’s next report on February 25th kicked off a wave of upward revisions and we found ourselves headed into summer with Wall Street closer to a $9 estimate on EPS in Fiscal 2027, if the shares re-rated to 25X Fiscal 2027 earnings, NVIDIA shares would trade for closer to $225 per share.
That would represent about 23% upside from where NVIDIA trades today. Wall Street estimates are currently even above this number. UBS just raised its price target to $245, citing strong GPU production expectations for 2026. The Street’s price target is currently $253.88. It seems we’re just waiting for that next round of upward revisions to kick off the next movement in share price.
Most Americans drastically underestimate how much they need to retire and overestimate how prepared they are. But data shows that people with one habit have more than double the savings of those who don’t.
And no, it’s got nothing to do with increasing your income, savings, clipping coupons, or even cutting back on your lifestyle. It’s much more straightforward (and powerful) than any of that. Frankly, it’s shocking more people don’t adopt the habit given how easy it is.