A look at Nvidia's skyrocketing stock and what's next

Jay Goldberg

Posts: 75   +1
Staff

A few weeks ago, I received a last minute invitation to participate in a livestream panel discussing Nvidia. At the time, there was a wild conspiracy making the rounds on Twitter about the company's results. This panel was clearly going to explore those in depth, and if nothing else, I was intrigued by who was behind the wild rumors and how they benefited from them.

Editor's Note:
Guest author Jonathan Goldberg is the founder of D2D Advisory, a multi-functional consulting firm. Jonathan has developed growth strategies and alliances for companies in the mobile, networking, gaming, and software industries.

So I attended the panel, listened (often in shock at what I heard), and pushed back where I could. The whole experience was a bit unsettling, staring into the mouth of the Internet rumor machine does not give one faith in humanity.

We are not going to dig into that theory, nor are we going to link to any part of it. For the most part, those rumors seems to have subsided and we do not want to add any oxygen to the attention machine. Nonetheless, I wanted to lay out my views on Nvidia, to set the record straight on how I view the company and its current outlook.

As everyone knows, Nvidia's stock is on a tear right now, up 200+% this year alone, solidifying its position as the best performing semiconductor stock probably of all time. The stock has been driven by two consecutive earnings reports delivering blow-out numbers.

For a company of Nvidia's size to grow earnings this much in such a short time is probably something most of us will only see once in our careers. Really strong numbers. This series of sudden surprises is probably what attracted the attention of the rumor mill, with so much attention focused on the company taking a contrary position is going to generate an audience online.

... to be clear, Nvidia's results are driven by real demand.

However, to be clear, Nvidia's results are driven by real demand. The software world is scrambling to catch up with the potential offered by transformer-based AI models. There is a huge change taking place in the landscape and no one wants to be left behind. As I have noted in the past, Nvidia has a lock on the market for AI training semis right now, a position that is unlikely to change any time soon. So all that interest in AI translates into big business for the company.

Moreover, the company has spent the past few years beefing up their offerings for the cloud. I wrote about this about 18 months ago, when their analyst day presentations made it clear that Nvidia was laying claim to dominance in the data center. This means they are selling a lot more than just GPUs for the cloud – they also have CPUs, networking chips and a host of systems tying everything together. They also have deployed some surprisingly robust software offerings, something which challenges all their semis peers.

All of which is to say that they are very well positioned right now, riding the wave of AI, gaining share and garnering those massive earnings.

Will it last? Here I need to break the question into two time frames.

Over the next year or so, the company has all that wind in its sails, but that will not last forever. Right now the company is struggling to meet demand, but at some point next year, supply will likely catch up. No company can keep up the pace that Nvidia is on, and semis are cyclical, so inevitably their earnings will pause. Couple that with the fact the Nvidia has never been terribly good at forecasting quarterly demand, or cared to get better at it, and yes eventually the stock will take a hit.

Given the super-premium multiple the company is trading at currently, any miss will see the stock trashed. Over the past ten years, Nvidia's stock fallen over 5% in a day 53 times. As much as the 30-year view of their stock is up and to the right, zoom in on any two to three year period on that chart and it looks more like a giant zig zag.

All of this makes it a terrible stock for retail investors. Professional investors are spending immense resources right now to get a finger on the pulse of Nvidia's progress, which gives them a big edge versus retail investors intrigued by what they may see or hear on Twitter – tread carefully – and to be clear, I do not own any Nvidia shares, and joking aside, I will not be buying any.

Over the long term, there is still a lot to like about Nvidia. Yes, interest in AI has gotten out of hand. And yes, there is a lot of hyperbole taking place in the segment. Precisely forecasting demand for AI semis is very challenging with a lot of moving parts – consumer adoption, software models, killer apps (or lack thereof), etc.

Nonetheless, the advances in compute that transformers, LLMs and other AI models offer are now clear, and these capabilities will get woven into the fabric of our digital lives no matter what. There is nothing on the horizon that looks set to challenge Nvidia's position in that any time soon. They may not be able to continue an unending series of monster earnings surprises, but they will participate in a major way.

Permalink to story.

 
What's next is nothing good no matter how you look at it.

But if I had to throw in a few guesses (From the likely to the more unlikely) I'd say that they'd stay the course by only accelerating and pushing harder in the same directions.

For gamer this means a couple things: They don't really have to worry about gaming revenue anymore, but they also have enough money that they don't have to concede it to AMD or someone else altogether. This means they'll stay on the same track but push even further: they'll keep charging even more for even less VRAM, far more closed ecosystem AI features and just enough tech to clearly outclass AMD Radeon division but giving gamers enough to make cards actually compelling.

So I expect Nvidia to bump up prices another 15-20% next gen so 450 USD for the next 5060ti and 1800 on the top end for the 5090 probably still keeping the exact same VRAM numbers.

Also they'll introduce DLSS 4 to go even heavier on tenser core frame generation so they can get away with minuscule gains on Rasterization performance and pretty much require most modern games to bend the knee, kiss the ring and implement DLSS 4 so they can sell most gamers on at least another 500 USD to get something that can run DLSS 4.

I also expect a version of ChatGPT and other AI asset creation for voice overs and even texture and model work and not to mention actual code for the game engine to be heavily publicized: Use our AI stuff and conveniently, with one of our supremely expensive 'AI' oriented cards to create your games far more easily from the dialgue to the assets to the coding.

In fact they might find a way to tie this AI created game to just work better, if not exclusively, with DLSS 4 so again you have to either pay for a 50X0 card or be stuck on 720p low detail to even run the thing because it depends so much on their tensor cores now.
 
Games running at 720p low unless ran with something called dlss4 (which is not even mentioned by any source, even the fakest ones), where are those theories coming from ? not even mlid would write something that silly.
 
Last edited:
Nvidia's stock has an amazing soar between Jan-Jun 2023, then it got stagnating until Aug.
Since Aug 31, 2023 it went nowhere but down, down and down. The bubble is bursting already.

This article is late. WAY late.
 
Nvidia's stock has an amazing soar between Jan-Jun 2023, then it got stagnating until Aug.
Since Aug 31, 2023 it went nowhere but down, down and down. The bubble is bursting already.

This article is late. WAY late.
I'm sure this article is based on long term evaluation.
What you said is true, but it is very short term.

I bought a load of tech stock at the beginning of the pandemic, and it included a ton of Nvidia stock I bought in early 2020 for $67 a share. So from that viewpoint things couldn't look much better.

With the way they stick in to their customers, and the fact that the biggest stick seems to be coming as they switch focus to AI, anyone that unloads Nvidia now is a fool.
 
Great article. I believe the bulls want the stock to rally unfortunately the bears are also equally opportunistic. Is ai a bubble? No but I believe it will go through cycles of saturation points like the one we are in; where the market saturated all orders and demand hysteria is netting somewhere rumored to be 10x roi. I don't hold any Nvidia stock either unfortunately lol, but if I had the opportunity I would buy on the dip. Another question does AMD have the potential for a Nvidia bull like rally? AMD is trading at 30% off their 52 week high but is almost double their 52 week low.
Nvidia is definitely one of those growth stocks you want in your portfolio long term besides a good index imo.
 
Now that gaming GPU's are no longer their top priority. Be prepared to pay even more for their next flagship card featuring DLSS 4 with 2x ray tracing performance while being 300% more expensive.
 
Now that gaming GPU's are no longer their top priority. Be prepared to pay even more for their next flagship card featuring DLSS 4 with 2x ray tracing performance while being 300% more expensive.
I am predicting the 4090 to be a 4 to 5 year card for some and that we might be paying for the same performance per dollar in q1 2025 with natural efficiency deltas and software sets that the culprit of an ai might bring to the table.. I always thought that Nvidia might just using the gaming portion of its business going forward as a clutch or an insurance policy just in case it needs to get over the performance targets, in order to rally its stock further. From the gaming revenue shortfall, I recall there was almost 5 billion dollars worth of revenue Nvidia can potentially use for a rainy day. Why saturate all your opportunities at once when you can fall back on the gaming side to potentially have a guaranteed bull position until the next cycle. If gamers are lucky we might get see some price fluctuations from the 300% price jump that you are predicting but it will probably be like the 20% off msrp @$1000 for the 4080 drop in price for split second to test the waters for the lucky few and then retest those bottom fluctuations in pricing to increase revenue to get over the target figure.
 
People can still opt not to buy new expensive GPUs and by that force Nvidia to lower the prices. It's just that people don't even think of that as an option anymore.
 
It’s great to have an opinion but it’s better to have “facts”. If you listen to TSMC’s earnings call 2023 Q2, TSMC says …

“Okay. Gokul, this is C. C. Wei. Let me answer your question. For the AI, right now, we see very strong demand, yes. For the front-end part, we don't have any problem to support. But for the back end, the advanced packaging side, especially for the CoWoS, we do have some very tight capacity to -- very hard to fulfill 100% of what customers needed. So we are working with customers for the short term to help them to fulfill the demand, but we are increasing our capacity as quickly as possible. And we expect these tightness somewhat be released in next year, probably towards the end of next year. But in between, we're still working closely with our customers to support their growth.”

What does this mean? CoWoS Service (Chip on Wafer on Substrate) is in “limited supply”. That means TSMC’s customers, like NVDIA must decide their priorities. Do they make high margin AI chips & boards or lower margin Gaming chips & boards? If they choose the later, why should they take a margin hit?

As you can see, in this environment, how are gaming board prices gonna come down? They can’t. Not until ALL the bottlenecks gating supply are alleviated AND demand is less that supply. When will that happen? Well, according to TSMC above, next year, maybe. Probably 2025.

Now, you might ask, can another company step in and offer this service, CoWoS? The answer is “No”. Why? Well, for another company to offer CoWoS they would have to buy the unfinished product from TSMC, perform the service, then sell it to the board assemblers. If they “**** it up”, they have to eat the cost of the chip. CoWoS becomes a “lower margin” business for anybody other than TSMC. For TSMC, they’re making the big money, “making chips” and any additional services they offer “adds profit”. If any step in the process affects yield, they just charge more per unit, they don’t “eat it”. From a profit point of view it only makes sense for TSMC to perform the CoWoS service. (Now sure, the most profit is what NVDA does, designing the chips.)

I would keep listening to TSMC for news.

Now, as far as an investment reco, if the environment is “gated supply with massive demand”, is that the best time to be a seller?

ObiwanBill
 
Great article. I believe the bulls want the stock to rally unfortunately the bears are also equally opportunistic. Is ai a bubble? No but I believe it will go through cycles of saturation points like the one we are in; where the market saturated all orders and demand hysteria is netting somewhere rumored to be 10x roi. I don't hold any Nvidia stock either unfortunately lol, but if I had the opportunity I would buy on the dip. Another question does AMD have the potential for a Nvidia bull like rally? AMD is trading at 30% off their 52 week high but is almost double their 52 week low.
Nvidia is definitely one of those growth stocks you want in your portfolio long term besides a good index imo.
The time to sell Nvidia is now. QT + Higher Interest rates will mean the markets will do a massive correction over the next 2 years.
 
Great article. I believe the bulls want the stock to rally unfortunately the bears are also equally opportunistic. Is ai a bubble? No but I believe it will go through cycles of saturation points like the one we are in; where the market saturated all orders and demand hysteria is netting somewhere rumored to be 10x roi. I don't hold any Nvidia stock either unfortunately lol, but if I had the opportunity I would buy on the dip. Another question does AMD have the potential for a Nvidia bull like rally? AMD is trading at 30% off their 52 week high but is almost double their 52 week low.
Nvidia is definitely one of those growth stocks you want in your portfolio long term besides a good index imo.

I remember when I purchased a few hundred shares of AMD @ $1.62. I thought selling at $15 was a good idea at the time. :/
 
People can still opt not to buy new expensive GPUs and by that force Nvidia to lower the prices. It's just that people don't even think of that as an option anymore.

Its already happening

The amount of people that buy too end cards is like 2% and always has been
 
Honestly, this whole focus on AI in the "cloud" for me is the same as rellying on drugs consumption to grown your economy. You might get large amounts of money quickly and in the end your population gets sick and dies.
 
Last edited:
Back