Is NVIDIA's $500 Billion Wall Street Deal a Genius Move or the Next Housing Bubble?

Welcome back, fools, to another episode. We just had Nvidia team up with Wall Street to unlock $500 billion for AI investment. Now, obviously you guys know me. I'm extremely extremely bullish in the AI market, and I can spend this whole episode talking why I'm excited, but I think that's the wrong thing to do. I think it's great to see both sides of the story and for that reason we have Emily joining us to kind of throw a lot of cold water in the excitement of this AI story that I have. Now before we continue with today's episode, if you want market beating stock picks from our analyst, make sure to check out the pin comment and the description. Using that link gets you a promotional offer as our thanks for being a viewer. Thank you and let's get back to today's episode. Hey Emily, [clears throat] welcome back to the channel. >> Yeah, thanks for having me. I I love being a wet blanket. It's one of my more toxic qualities. So, I'm I'm really excited to discuss this partnership that Nvidia has launched because like I I see why people are excited by it, but like the pessimist that lives inside my head constantly is also just yeah, always throwing cold water on these sorts of news. They're like, "Well, is it really as good as the market thinks it is?" >> So, before we we kind of take a look at some of the bullish and bearish takes, Emily, can you kind of explain what what is actually happening here? >> Yeah. So big picture, this probably doesn't come as a surprise for anybody who's following the AI space, but there are a lot of uh people trying to build data centers, trying to get involved in the AI game, just to develop their large language models and access to capital because this industry has been moving so quickly has been one of those barriers to entry even for like the largest, more established players. And Nvidia obviously sells, you know, GPUs and chips that are used as part of these data centers and they want to be able to meet that demand. But if somebody can't afford to come out and buy those directly, then they need to find outside capital. So what Nvidia is doing is it seems like they've signed someus and now just to be very clear, these are memorandiums of understanding, right? They're like handshake deals. They're not promises, but it's upwards of $500 billion with third parties. um these are Wall Street banks, asset managers that are agreeing to basically work with Nvidia to potentially underwrite loans to some of these big players in the AI space. And the premise that this is kind of running off of is that these AI GPUs or the GPUs used in AI processes are within themselves valuable assets that can be used as part of an underwriting process for loans. and Nvidia is trying to put some capital behind that themselves up to 25% of some of these deals to kind of provide um substance to the banks that would otherwise be financing to say, "Hey, we see long-term value here." >> Yeah. So pretty much one of the big the one of the ways that I was trying to understand it is before the market believed that GPUs would be kind of these extremely depreciating assets and banks would be like there's added risk because I don't know where the value of this GPU would be 3 4 5 years from now and for that reason we add that risk towards your financing side. Now, it seems like not necessarily to this extent, but kind of like in real estate, right? And again, I don't want to compare a GPU to to a house, but the way that it seems like they want to make it show uh showcase is look, this GPU is not necessarily as crazy depreciation asset as we thought it was. It's not a house, but it is a product that holds a little bit more value than originally expected, and you have this revenue collection out of it. So if something happens to that data center, I as a bank probably have a little bit less risk because I can take those GPUs and be able to sell them out or or or do whatever with them at a better value than I thought was the case before. Is that kind of like what what we should be thinking about, Emily? >> Yeah, exactly. And I you know, Jensen Wong kind of mentioned this in the letter that he sent out to investors, readers, uh which by the way, AI gener I'm just so tired of reading AI generated content. I'm so sorry. I understand when you're when you're the CEO of Nvidia, you're using it. Um, but come on, talk to us directly. But yes, so obviously this is something that Hang and his AI provider of choice talked to us a little bit about when it comes to u the value that's being derived from older GPUs and specifically pointed to the GPUs that, you know, Nvidia sold back in 2020, which are still being in use today, 6 years later. And it was kind of the mindset back in 2020, even just a few years ago, even just a few quarters ago, that these would be assets that would rapidly depreciate as new GPUs hit the market. but they're still being used for things like inferencing. And you see that with a lot of these hyperscalers that are extending the useful lives of their GPUs, reducing the depreciation that they're experiencing on their income statements because they're still getting use out of like they're still being used. So like there is a very very real, very legitimate argument to be made that we should not be arguably depreciating the GPUs that are being sold today, whether they be Nvidia or otherwise, at the rate that we were when they were initially sold. Like the world environment has changed. But I think your comparison to the housing market and mortgages are is is a really good one. We actually saw Larry Frink who was kind of, you know, the the pusher of mortgage back securities back in 1970 saying that this is a revolutionary way of financing the same way mortgage back securities were. And he meant that as a compliment. Like I understand hearing mortgage back securities. I think everybody listening probably like okay great. Are we just heading towards some great, you know, some version of the housing >> mortgage crash? >> Exactly. Exactly. But but like I will say there was a big substantial difference and I think the perception in the 1970s of the safety of the value of the asset class that is real estate which is historically one of the least volatile asset classes and it was perceived as one of the least volatile asset classes back in the 1970s versus the chip industry today which has a strong history of being highly cyclical and you can like make the argument that it's no longer cyclical which is maybe what this is trying to do but the differences is in the perception like do I don't know do Do you Jose believe that the chip industry is no longer cyclical? >> Definitely not. The only thing I believe is that this cycle might be a little bit longer than prior cycles. Uh but I I I I do believe there will be a there will come a time that you will have kind of >> an overbuild. It happens in anything in anything in life. Um uh so so um yeah. So that sometimes worries me, right? is like if you want to shift the mentality, I hope we don't completely go too much into like GPUs are now Elsa's kind of mentality in this portfolio. >> Yeah, it's it's you start to build um a house of cards so to speak, right? If you start to do that and now I will say I think it says a lot about the third parties that Nvidia is doing this with that they seem to want some Nvidia backing behind it. Now Nvidia said, "Look, this isn't circular financing. We're just helping incentivize like we're building third parties. That's why we're going through these third parties." But they're still committing up to 25% of these case by case, they haven't said how many, but case by case projects that could be built out as part of this this partnership. And um that is important because it says to me that the banks and the asset managers that are part of this deal were maybe not willing to obviously on their own go out and underwrite and finance a lot of these GPU and data center purchases that are happening by you know you can use coreweave or other you know AI anthropic openai whatever it may be these these LLM developers um without that partnership behind them. So, it is interesting. Out of all of the partners that Nvidia has moved forward with, the one that I know the best, the one that I follow the closest is probably Brookfield Asset Management, there's a lot of fools here, the ticker is BAM, who like this company. They're an asset manager, really wellrun. If there's one thing that I know, and I'm quoting one of my co-workers here, but if there's one thing we know about this company, it's that they're pretty strict about their underwriting standards, and they tend to really favor themselves with their deal terms. Like, they have that leverage historically. So when you think about the way that Brookfield has historically financed their deals, their experience, their management team, to me it says if you are somebody who is, you know, one of these AI developers who's going to be coming out taking advantage of thisou with Nvidia and its partners, the terms of your deal are probably pretty heavily stacked in the favor of those third parties rather than you. But your hands are tied. You have no other options right? >> So Emily, I want to jump into maybe some of the bullish and bearish takes out of this. And obviously we covered a lot. One of the biggest things is now these NeoCloud players do tend to have the opportunity, it's not a guarantee, but an opportunity to be able to get these financing at better rates than before. I mean, if if anybody follows Core Weef, I think this company mentioned that for the upcoming quarter, they're expected to pay 900 around $900 million in interest alone next quarter. Um, on the bright side, they also did provide a lot of data points behind this. This was one of the companies that mentioned that A100 that you were just talking about, uh, Emily, that came out in 2020 that they just got a contract on this to extend till 2029. Uh, so it it's a lot of where we are in the market, but you are seeing that a GPU can have a lifetime of potentially 10 years um or or or close to it. So I think that's the the biggest thing is is is the ability for these finance companies to get financial uh opportunity. If you are a semiconductor investor, the bullish take would be not just necessarily for Nvidia because you're no longer just buying one chip anymore. You're buying a whole AI data center, a whole AI rack that comes with the networking. If you increase the amount of data centers that get built out, yes, it seems really good for Nvidia, but it's also going to be very strong for all the networking companies that are providing the networking solutions to Nvidia. all the optical plays, all the memory plays because you're increasing this infrastructure buildout, increasing the the amount of supply. Now, obviously, that can lead maybe to that over supply issue in the in the future. Maybe this was the turning point that that provides that. Um, but I'll share some of the bearish states to you. To me, those are the two main bullish points is better financing for the NeoCloud players. And while it does seem like Nvidia is the only winner because Nvidia solutions provide meet so many different components, it's a breath of the whole semiconductor industry benefits from this at least on the topline number. >> Yeah. And I will say like I I have a lot of sympathy for the position that to your point like Neocloud providers at Nvidia are in here which is they have firstline visibility into very real demand and it is quite literally their job to figure out how to meet it. So for anybody who is like, you know, myself included, who's like being skeptical of these things, what would you rather have them do, right? Simply not build it out when everybody and their dog is coming to them and begging for capacity today. So you know, they their job is to sell GPUs and that's exactly what they're doing. So they're doing exactly the right thing in that sense. But it does beg the question of as we as retail investors, h what do we do, right? Because what's best for Nvidia may not be best for us as individual investors. And I think that's kind of where maybe some more of my bearish takes come in. Um, the first one being Nvidia has an incentive to extend the useful life of these assets. Now, Nvidia is typically the one coming to the people purchasing telling us how long their assets useful life is. Now, do I think Nvidia is making that up? No, of course not. And I think they have some data to back it up. But I do think that's obviously a clear conflict of interest. We're hearing about the expanded lifespan largely from Nvidia itself and using examples that they're pulling out um to justify the underwriting of these projects. Now I I will say I I am also concerned. I mean this is like I said a cyclical industry and so when you're doing that you're extending the useful life. The difference between a chip which maybe we're extending from 7 years to 10 years to be generous. It's still not the same thing as land which historically depreciates on what a 50ear schedule. like the difference there is still pretty substantive and we don't really have a great idea for how long this tail is going to extend. But, you know, the bigger concern to me is kind of on the side of these deals are ultimately only as good as the people taking them out. And I do think it's a little bit of a a sign here, if I can say that, that the neo clouds need access to financing so desperately because Nvidia and Jensen Huang in that letter himself says that like the real value that's being derived from AI is the usefulness of AI and that will hopefully generate revenue. And if we're not seeing sustainable revenue be, you know, generated by these companies even this many years into their lives, then what does it say that they're having to get backstopped by Nvidia to raise capital for their debt plans to continue to expand? It's I still don't think we have seen the ROI that is necessary to make me build a lot of confidence in the sustainability of a lot of especially the smaller ones that don't have the same access to capital, but a lot of the smaller players in the space. >> Definitely. Hey Emily, I want to kind of focus on we touched a little bit about it of what it means for the whole semiconductor industry. I think um we can make some direct names and even a top level overview of how this can impact certain companies. I'll start off first uh Emily and then I'll pass you away. Maybe you have some names or maybe just a top level overview of what you think this means for just the rest of the semiconductor and AI industry. Um one of the ones that kind of has a little bit worry on my end Emily is AMD. So AMD is a huge competitor in the semiconductor industry. The bearish side of me goes. So what ends up happening if this is not a whole AI chip, the whole chip market has increased in life and usefulness. It's only Nvidia's chips. Does this provide or create um a hump or a delay for big companies to go, do I really need to go buy that AMD chip because I'm not going to get that same type of financing as I'm going to get for Nvidia? Or the other take is Nvidia's paving the road and then AMD can eventually go around this road that Nvidia has already paved and kind of create the same type of financing with its own type of partnerships. Um so so I I I wouldn't say with AMD it's it's it's very flexible of where it where it can go. Um the the core ways of of this world uh bullish and bearish state bullish is now you get financing at a better rate hopefully the bearish state it opens up the wall for competition to come in that might have not been there uh before. So you're going to have the added uh competitive player. Um, I I I think one of the only true true winners in this market right now with this true is a manufacturer player like TSMC because they don't care if it's a Core Reef, if it's a Nebus, if it's a whatever Neo player, they don't care if it's Nvidia, they don't care if it's AMD. As long as the man is there, they will be able to make chips and they will be able to make money. Uh, so emulator to close it out. Any thoughts on maybe just the rest of the industry outside of like your Nvidia and your banking uh place here? Yeah, I will say I actually think a lot of the concerns with AMD while totally justified are maybe a bit overblown. So AMD obviously their core business is providing CPUs. They, you know, recently got into the GPU game. And this is not new though for AMD. Like nobody is going out there and being like, man, I wonder which one's better, Nvidia or AMD. AMD for years now has been sweetening the deals that it makes with stuff like performance-based warrants. That's a big risk for investors. That's all I'm saying. um they have been literally sweetening their deals to get their GPU sold from day one. So I don't think this deal alone in Nvidia's case is going to be the reason why one of these NeoClouds like oh I guess I'll use Nvidia's GPUs because I have better access to financing versus AMD's because AMD has been trying to fill those holes to convince people to use their chips in lie of a Nvidia. In fact, if anything, you could almost read this as like a yellow flag for Nvidia because it feels a little bit like they're taking the AMD playbook now of incentivizing of sweetening their deals to try to get people to I don't think that's actually what happening, but like if I was a real contrarian, that would be my take. [clears throat] Um, but I yeah, AMD is clearly still losing the game here. >> Like that say, mindful mindful Nvidia is copying AMD. And that's how we're going to end this episode, fools. I hope you guys enjoyed. Let us know in the comments below if you are what are your thoughts about this and bullish or bearish stakes. We love to read all the comments and I feel other fools would love to hear other commentary as well. So Emily, thank you again and fools take care. Have a good day and see you all next