In This Article:
This is The Takeaway from today's Morning Brief, which you can sign up to receive in your inbox every morning along with:
-
The chart of the day
-
What we're watching
-
What we're reading
-
Economic data releases and earnings
The dust has settled on the latest Nvidia (NVDA) earnings week frenzy.
And I think it's important to take stock of where things stand for the world's most important stock (sorry, Apple (AAPL)). Why? Because you should be thinking about whether the pullback in Nvidia is a buying opportunity or the start of a deeper sell-off as expectations are reset.
We know Nvidia's margins in the first half of the year will be below their usual robust levels as Blackwell AI chips ramp up. I would argue the Street knew this ahead of the results, so they got flustered over nothing.
On Nvidia's earnings call, execs sought to push back on the bears, who have put forth a narrative that there will be a digestion period for AI investments by hyperscalers such as Amazon (AMZN) and that Nvidia's margins may have peaked.
"Once our Blackwell fully rounds, we can improve our cost and our gross margin," Nvidia CFO Colette Kress said. "So, we expect to probably be in the mid-70s later this year."
We also know that, fundamentally, Nvidia's business is strong and likely to stay strong.
Fourth quarter revenue rose 12% sequentially and 78% from the prior year. Data center sales more than doubled from the prior year. Earnings handily beat analyst estimates.
"We're going to have to continue to scale as demand is quite high, and customers are anxious and impatient to get their Blackwell systems," Nvidia founder and CEO Jensen Huang said.
Listen: Nvidia could be unstoppable
Nowhere in the company's 2025 guidance or commentary from Huang did I sense that AMD (AMD) is taking Nvidia's market share; ditto custom chips from Amazon. I heard no hint that hyperscalers are sending AI chips back to Nvidia or have stopped fawning over Jensen to get more of these chips at any cost.
Put together, I would argue what we heard from Nvidia in terms of demand and margins was all well known going into the results. So, the sell-off could prove to be an overreaction, a function of investors aiming to model out mixed first quarter guidance for the next two years.
But there are a couple of things we don't yet know about Nvidia that warrant greater attention. These play into the long-term bull thesis.
For starters, there's Huang's point about DeepSeek's R1 requiring 100x more compute resources compared to pre-training models due to inference time scaling. Look, most of us have no clue what this even means. But the casual observer could read it as the market has it strong on DeepSeek, and there could be a lot of upside to Nvidia estimates as DeepSeek and other reasoning models gain hold.