Investing

NVIDIA Reveals Plan to Reverse Its Stock Slide - Will It Work?

Eric Bleeker

Published: July 28, 2024 2:52 pm

Last Updated: July 28, 2024 11:00 pm

NVIDIA‘s (Nasdaq: NVDA) stock is now down more than 20% from its peak intraday price of $140.76. Several factors are weighing on the company’s share price, but one of the strongest is the belief that the current surge in chip sales is temporary.

It looks like NVIDIA will take that idea head-on in its next earnings report. Let’s look at what NVIDIA is reportedly planning to reveal and whether it could reverse the company’s stock slide.

Need to Know Highlights

• Goldman Sachs is reporting that in NVIDIA’s next earnings call they’ll focus on the ROI customers can earn from their graphics cards.
• One of the greatest threats to NVIDIA is that customers aren’t earning a return on their chips and rather buying to keep up in an “arm’s race.” This could lead to spending falling after Blackwell’s launch as customers pause buying to reflect lower customer use.
• If you’re looking for stocks poised to ride the next big wave of AI spending, make sure to grab a complimentary copy of our brand-new “The Next NVIDIA” report. It features a software stock we’re confident has 10X potential and could be missing form your portfolio if you’ve only bought hardware stocks like NVIDIA.

What NVIDIA Plans to Reveal Next Earnings

Goldman Sachs is reporting that NVIDIA is going to highlight profits made by end customers on its next earnings call. The company has done this in the past, but often in more vague statements around ROI.

Here’s an example from NVIDIA’s most recent call:

“Training and inferencing AI on NVIDIA CUDA is driving meaningful acceleration in cloud rental revenue growth, delivering an immediate and strong return on cloud providers’ investment. For every $1 spent on NVIDIA AI infrastructure, cloud providers have an opportunity to earn $5 in GPU instant hosting revenue over 4 years.”

NVIDIA is currently fighting on two fronts in the GPU market; the first is ‘training.’

In that market, NVIDIA is more dominant. Customers are currently spending vast sums for ever more complex models. Spending in the training market seems a bit more secure as customers aren’t necessarily spending to an ROI figure, but rather in an “arm’s race” to push the capabilities of their leading models further.

The second front NVIDIA is fighting is on ‘inferencing.’

Inferencing happens after models are trained. It’s using the models to actually make predictions. In this market, NVIDIA is less dominant. NVIDIA estimates about 40% of their revenue comes from inferencing, but it’s a market where large cloud providers like Amazon (Nasdaq: AMZN), Alphabet (Nasdaq: GOOGL), and Microsoft (Nasdaq: MSFT) continue developing custom chips primarily intended to be more cost-efficient in inferencing.

There is also a question of just how much inferencing capacity is needed. Some Wall Street estimates say inferencing capacity is already exceeding near-term needs. Another factor is that if a lot of inferencing demand is coming from large cloud companies. This is a part of the market driven far more by ROI. That is to say, large cloud companies have P&Ls that Wall Street watches closely. If they’re losing money because large orders for chips intended for inferencing aren’t being rented by customers, they’ll pull back on orders far more quickly.

When companies issue earnings, its a chance for them to tell their side of the story. So it makes sense that NVIDIA would want to use its earnings to give investors added confidence customers are actually seeing solid returns on their GPU purchases.

Let’s explore a couple ‘case studies’ below where customers are seeing real revenue from NVIDIA’s GPU chips. Then we’ll end with a more cautionary example.

Defending NVIDIA: Meta’s Advertising ROI

It’s estimated that Meta (Nasdaq: META) is NVIDIA’s second-largest customer, trailing only Microsoft. With this level of massive spend, whether or not Meta is seeing positive ROI has been a huge question among investors.

Wall Street researcher Bernstein recently did the math, and their numbers are very positive for NVIDIA investors. Bernstein estimates the following:

1. Meta is projected to spend $14.1 billion in core AI spend in 2024. Mark Zuckerberg has publicly said Meta would have 340,000 of NVIDIA’s H100 GPUs by the end of the year. Meta has issued further statements that they’ll have 600,000 in compute equivalent chips. This would include custom processors from the likes of Broadcom (Nasdaq: AVGO) but gives a sense of their total AI buildout.
2. Bernstein estimates that 50% of content on Instagram and about 40% on Facebook comes from AI algorithms. These algorithms are producing about a 10-11% lift in time spent on apps.
3. This additional time spent in Meta’s apps produces an estimated incremental $16.3 billion in revenue this year.
4. Add up all these figures and Meta would be producing about a 22% ROIC on their investments in artificial intelligence for content ranking. 

It’s worth noting these figures for Meta are estimates. However, this is exactly the kind of concrete example NVIDIA needs to help allay fears around the return on buying their AI chips.

While it’s common for NVIDIA bears to cite a lack of revenue or “killer apps” in the AI space, it’s estimated Meta is generating an incremental $16.3 billion at a 22% ROIC. That’s probably the single best case study on what AI investments can produce!

Defending NVIDIA:  ServiceNow

ServiceNow (NYSE: NOW) reported earnings this week and was up more than 15% the next day. One focus of the company’s earnings call was how AI products and services to customers had driven their revenue beat. Here is a highlight from their earnings call:

“Moving to the topic of AI. Now Assist net new ACV doubled quarter-over-quarter, significantly overachieving our expectations, and it became the fastest-growing new product in the company’s history. We signed 11 Now Assist deals with $1 million-plus NNACV in Q2, two of which were over $5 million. We had impressive wins across industries, from banking, health care and manufacturing, to semis, tech and many others.”

Obviously, a lot from the quotes above won’t make sense unless you actively follow ServiceNow’s business. However, here’s the key point: ServiceNow used specific examples from across industries – automotive, healthcare, semiconductors – to show how generative AI is being adopted broadly across industries. 

This is likely what NVIDIA will attempt to do in their own earnings as well. Show with as much clarity as allowed specifically how customers in a broad series of industries are using their chips and the kinds of returns they’re seeing.

Reason for Concern: Google

Finally, let’s end with what could be a reason for concern from this earnings season. Alphabet issued earnings and all the call received a number of questions about the ROIC they were seeing from their AI infrastructure investments. The commentary from their CEO, Sundar Pichai, was fairly muted:

“Look, I – obviously, we are at an early stage of, what I view, as a very transformative area and in technology where you’re going through these transitions, aggressively investing upfront in a defining category, particularly in an area where in a leverage way cuts across all our core areas of products, including search, YouTube and other services as well as fuels growth in cloud and supports the innovative long-term bets and Other Bets is definitely something for us makes sense to lean in.

I think the one way I think about it is when we go through a curve like this, the risk of under-investing is dramatically greater than the risk of over-investing for us here, even in scenarios where if it turns out that we are over-investing, we clearly – these are infrastructure, which is widely useful for us. They have long useful lives and we can apply it across, and we can work through that. But I think not investing to be at the front here, I think, definitely has much more significant downside.” 

The key point here is that Pichai seems to say, “If we’re not getting great ROIC from customers, we can also use our AI chips for internal projects and decelerate investments.”

That’s a practical way of looking at the market. However, it plays into the point made by bears in the AI market. That is to say that while companies feel like they need to invest to keep up right now, pretty soon large hyperscalers like Google will run the numbers and see that they’ve overinvested. At this point, the market could see a correction that would be very bad for NVIDIA’s share price.

It’s now up to NVIDIA to begin countering this narrative. Pay close attention to what they highlight in their next earnings. It could go a long way in deciding whether NVIDIA’s stock runs in the second half or stays in neutral. 

Get More AI Ideas

The next trend in AI is pushing from clusters of thousands of AI chips to perhaps a million or more. This mega-trend will create new winners which we’ve detailed in our brand-new “The Next NVIDIA report.” 

The report is complimentary, but available for a limited time only. Make sure to grab a copy as breakthroughs in the space are moving at an extremely rapid pace.