As the global AI frenzy meets increasing market volatility, Nvidia has once again found itself in the spotlight.

Recently, the chip giant hosted its much-anticipated GTC conference, where CEO Jensen Huang unveiled new AI chips. Yet despite the buzz, Nvidia’s stock slid, raising questions about whether investor enthusiasm for AI is cooling.

Meanwhile, several members of the “Magnificent Seven,” including Amazon and Tesla, have seen sharp selloffs in recent weeks, prompting deeper reflection on tech valuations.

To better understand what’s driving these shifts, we invited Simon Erickson, founder of 7investing to join us today.

Southern Finance: Nvidia’s GTC keynote was highly anticipated, yet that day Nvidia’s stock still dropped. Over the past year, Nvidia has gone from being the hottest AI stock to facing skepticism about its long-term growth. Do you think this shift in sentiment is temporary, or does it signal deeper concerns about AI investments?

Simon Erickson: Well, there's certainly plenty of long-term growth to be had out there. And if you don't take my word for it, look at the capital expenditure budgets that we have coming up for the large cloud data centers, right? You've got Microsoft coming out there and saying they're going to spend $80 billion. You've got Amazon saying they're going to spend $100 billion. And you've got Alphabet saying they're going to spend $75 billion with the capital expenditures. And it's going to go right into the data center, right? It's going to be the processors that are using NVIDIA's A100s and the H100s for the training, the inference for machine learning.

So there's no doubt that this is a long-term trend, that it's got some serious dollars behind it. NVIDIA specifically has been almost dominant when it comes to the GPUs that are sold into the AI accelerators of the data center. I think we might see a bit more competition in the next coming years.

I think that AMD is already very interested and shown their interest in joining the GPU side of the processor space for the data center. But also new companies like Cerebras is going to IPO here very soon. There's a lot of these large tech companies that are developing their own AI chips as well.

They're just much more efficient when it comes to the computing that you want to do. You want to do as many operations as you possibly can per second using the least amount of power as you possibly can. And that's why you see Amazon developing Inferentia and Trainium.

You see Microsoft coming out with its own new chips. You've seen Facebook and Tesla and everybody else developing their own AI-specific chips. Why? Because they're spending $75 billion or more every year on CapEx for the data center.

So big money to be had here. NVIDIA has been by far the dominant share, 90% plus of the GPUs that go to the data center thus far. I think they're going to be just fine going forward, but there probably will be a little bit more competition.

Southern Finance: You talk about competition. How do you view the challenge that DeepSeek has brought to NVIDIA?

Simon Erickson: I think that in general, this is necessary, right? It's kind of like Moore's law has made chips more efficient as you could fit more processors or more transistors onto a processor. You've got to be more efficient now at the platform level rather than just the computing transistors on the chips itself.

So I think that perhaps it's a bit overblown, at least in my opinion, that DeepSeek is bad for NVIDIA. In fact, if you can get more efficient with processing and do more computing, that probably is good for the long-term trend of more efficient AI processors and models in the long run.

Southern Finance: You also talk about lots of Magnificent 7 stocks like Amazon, Apple, and these stocks have experienced significant sell-off over the past few weeks. What factors do you think have contributed to this sell-off? And do you think these companies will regain their premium?

Simon Erickson：First of all, it was absolutely necessary for these stocks to sell off. It was impossible for them to keep the same premium valuations that they had so far. When we're talking about the Magnificent 7, we're talking about the seven largest American companies in the S&P index.

It's Meta, Microsoft, Amazon, Apple, NVIDIA, and Alphabet and Tesla. And together, they were 33% of the S&P 500 index. Of course, the S&P 500 is a weighted index based on market cap, but the top seven of those 500 companies were a third of the overall index.

We're talking about trillion, multi-trillion-dollar companies now. And when you look at the cyclically adjusted PE ratio of the index itself and then of just of those seven companies, you can just see it was just getting ridiculously overvalued and expensive. As a whole, the cyclically adjusted PE ratio, which is kind of a normalized representation of earnings, was sneaking up at 35 times price earnings since the election of November.

That's way too expensive. It's only been higher than that two times in the last two decades. The first was the stock market euphoria of 2021.

That's when this index was selling at about 38 times. And then, of course, during the dot-com bubble of the year 2000, it was selling at 43 times. So when you saw the index as a whole, the S&P 500 selling at 35, 36 times trailing earnings, as a whole, it was just getting too expensive.

But then let's look at the Magnificent 7. Those seven companies themselves just recently a PE ratio of 66 times. Way too high. NVIDIA is at 66 times earnings.

Tesla at one point was 200 times earnings. You almost cannot justify investing in those companies at such a premium valuation. A lot of money was going to this because there was a lot of confidence in it, a lot of optimism in them.

The valuation was simply getting too high. This is a very necessary reversion back to the mean. I'm relieved as an investor to see it selling back off in the last couple of months.

And personally, I think this is an opportunity to get out of those largest tech companies that got very expensive since November and get into some of the other companies, either in the S&P 500 or even outside of the S&P 500. There's a lot of undervalued and overlooked tech companies out there that just aren't getting the same fanfare as the Magnificent 7.

Southern Finance: So you think the Magnificent 7 are overvalued and right now the correction is actually very necessary.

Simon Erickson :100% agree.

Southern Finance: Tesla has seen the biggest percentage decline, down around 50% in just a few weeks. How much of this is due to macroeconomic concerns, and how much is company-specific?

Simon Erickson: It’s company-specific, but I put the disclaimer out there that Tesla for a long time was very arguably overvalued, at least compared to other automakers.

And that's because Tesla isn't really an automaker. It's an AI company that's happening to sell cars. It's then going on and incorporating full self-driving software that's selling for $100 a month into the cars.

And Elon has, of course, talked about the Model 2 introduction, which you'd be using for a robo-taxi. That would position it more to compete against companies like Uber or Didi or Lyft, autonomous transportation companies, rather than just a carmaker that collects money one-time up front. So Tesla had always had a premium valuation.

We're starting to see a little bit of that sparkle wear off in the last couple of months. It's undeniable that Elon Musk is distracted, and he has a whole lot of things going on. But let's talk about Tesla, the company, a little bit here.

First of all, sales are falling in Europe. The company's combined output of the Model S and the Model X, which are very heavily sold into Europe, combined production was down 33% quarter over quarter in the fourth quarter to less than 10,000 vehicles made in the fourth quarter, down from 14,000 the previous quarter. So they're making less cars because demand simply isn't there for them in Europe.

And then on the other side of things, they aren't selling so well in China as well. The Model 3 itself actually had a nearly $4,000 reduction in its average global selling price during just the last quarter, down about $32,800 U.S. dollars globally, up from more than $36,000 in the previous quarter. So you see the impact of things like this 0% financing, rebates, tax incentives, things like this just to spark demand when the cars are not selling completely organically out there right now.

We know that China is a great growth market for Tesla, but it's not without its own competition, whether that be from BYD, which has very fast charging models that it's now introduced now. And then, of course, NIO, which is very competitive on the high end of the premium segment out there. It's not like Tesla is the only game in town.

It's going to have to, to value, to have its valuation make sense. It cannot just be a car company. It has to succeed with Robotaxi or at least its AI ambitions if you want to invest in Tesla, even at its lower valuation today.

Southern Finance: How long do you think Tesla will be like experiencing this kind of high volatility? What is the time that you think Tesla would regain its premium?

Simon Erickson: So for me, I've kind of broken the company into three sectors. The first being kind of the auto and energy sector, which is where we are today. The second is, can they get an adoption of kind of a large enough, enough vehicles on the road to justify the full self-driving software piece where you've got that collecting income stream every single month from the autonomous driving? And then third, the Robotaxi network itself.

I personally think, that the car company, Tesla as a car company, is really only worth about $130 per share. That's based on a valuation model that I spent an entire month working on, but it's just not that lucrative of a business. You want to get into that higher margin business, which has been buying the NVIDIA chips.

You know, the reason they're putting so many GPUs into their neural networks or their data centers is because they want to sell the software. They want to give everybody a free month of self-driving, get people convinced that this is really awesome. They want to stick with it and then are willing to pay Tesla another $100 each and every month for that.

And then if that works out for them, it's probably worth around $250, maybe up to $300 per share, kind of where the stock is priced at today. But the real bang for the buck on Tesla is going to be if it's an autonomous vehicle company, where now it's got the autonomous vehicles out there. People are buying the cars, but they're using them as moneymakers.

You're buying a Tesla not just to drive yourself around, but you're sending it out to be your robotic taxi cab, giving other people rides, and then you're collecting the income stream from that. And Tesla is taking a take rate of perhaps 20%, with no human drivers, fully autonomous driving. There's a light at the end of the tunnel of everything they've been working on, on autonomy.

We've seen very, very little of it right now, other than people enjoying the full self-driving, other than the autopilot that it has incorporated in the cars. There's a reason they're putting all the hardware into the vehicles today. It's because that's going to be the driver of the company's valuation.

And so to answer your question of does it deserve its premium valuation much longer, you've got to make a decision on this. Is it overvalued right now as a car company? Because it's not going to succeed with the robo-taxi network or full self-driving. Or is it completely undervalued because it's actually an AI company and Elon's going to pull more rabbits out of the hat in the future? If it does, it's probably worth closer to $750 per share, meaning Tesla is perhaps one third of its true valuation if it succeeds in those endeavors.

Southern Finance: So Chinese concept stocks have outperformed in recent months while the U.S. stock market has experienced some selloffs and volatility. Besides all the well-known tech giants, what are some under the radar Chinese sectors that deserve specific attention from investors?

Simon Erickson:  One sector to the part of your question that I think is interesting is the biotech space in China.

China has kind of reformed a lot of its own FDA, the Chinese regulator that's going to look over drug development of new pharmaceutical and biopharmaceutical drugs in the last couple of years. And the reform will be finished by 2027, which is two years out from now. But the goal is it's so interesting to see kind of some of the immuno-oncology drugs that have done so well with the Western drug makers, the companies that are based in Europe or based in America.

China is really upping its game in biopharma. It's really, really putting a lot of details into place for reform to help its own drug makers, to help them with R&D, to help them get new drugs through clinical trials, sometimes to fund the work that they're doing with these academic centers. And I think that's going to raise the bar, not just for new medicines and pharmaceuticals that are introduced in China, but also for those new medicines that are introduced in China but are then applicable to the rest of the world.

I think this is truly an international market. You've got regulators on every continent that want to have control over their own regions, but there's nothing wrong with saying a drug that's developed in Europe or the United States or in China can't go out there for review and for marketing and for sale everywhere in the entire world. I think it's generally good for patients everywhere that you raise the bar on all of this.

I'm really impressed with what China and the biopharmaceutical industry has been doing, especially in these last couple of years. Keep an eye on that one.