
The Exchange, along with a little help from two surprise guests, takes a deep dive into the blockbuster Cerebras IPO as shares start trading at the top of the hour. President Trump was “fawning over the Chinese Communist Party”, according to AEI’s Derek Scissors, and that could give Democrats an edge in the midterms. Plus, a win for the crypto industry.
Loading summary
A
Hi, I'm Kelly Cavagnaro, Managing director, head of North America Institutional Distribution at Janice Henderson Investors we believe working together is
B
the way to work better.
A
Like combining your portfolio plans and our in depth strategy, your valued assets and our valuable insights. Your mission and our vision working in harmony to seek the right investment opportunities. JANICE henderson INVESTORS Investing in a brighter future together, is your business achieving its current strategic goals?
B
Are operations as tight as they should be? Are finances in a realistic place for expansion? These are questions investors ask. That's why ey Parthenon brings an investor
A
mindset to help executives reinvent business models
B
for long term growth, reset strategic goals, optimize operations and get finances in order. Let us help you reimagine your enterprise
A
EY solutions that work in practice, not just on paper.
B
You're listening to the Exchange. Here's today's show. I'm Kelly Evans. Welcome to the Exchange and at this very moment, one of the biggest IPOs certainly of the year, if not of the last couple years, as you heard Brad Gerstner there say a moment ago, has just opened for trading. The company is Cerebras. It's a chip maker. It priced last night at 185, which was already above the expected range. $350 appears to be the open price. That's that top left Colum there, that top left stat on your screen opens at 350. Pretty good for those who got in on the IPO. Its current price is 385. Even better now for those who could buy in right at the public opening. We love to see this companies that are able, of course, in the hottest part of the market as well to upsize the offering, take it public and do even better. And then for the people who bought it, now for the the public, it's in your hands. You can do with it what you will. 385 is the current price for Cerebras. Christina Parts Nevelis is at the NASDAQ with all of the very latest on this one. Hi, Christina Kelly, Demand was so intense
A
for this IPO that there were 45 buyers for every one seller. Like you said, the company sold about roughly 30 million shares last night at 185 bucks. So that's a raising of about 5.5 billion. Normally an IPO opens with about 10% of those shares trading hands, but you saw this opened with 4.3 million shares changing hands. Now we're seeing it open at 350, trading at 385. The lead book runner, Morgan Stanley, had to hold off a little bit because they thought the price was a little too high, but the demand was just so high that they just had to go to market. To your point, this is the biggest US tech IPO since Snowflake in 2020. In this case, Cerebrus raised $5.5 billion last night, now valued at roughly 56 billion. Fully diluted at that IPO price at 300. And at $350, that valuation approaches $100 billion. So why all the demand? Kelly? Cerebras is an inference chip company. Inference is a part of the AI model where responders you respond to users in real time. The company makes a chip which I happen to have, it's quite large with memory built directly onto the silicon. So it essentially eliminates the bottleneck that slows down the GPU system it sells not only the chip, the full computing system that runs with it and also runs its own cloud services, which that's they're hoping is going to be a major segment to grow. OpenAI signed a $20 billion compete deal back in January. AWS has a binding term. She, you know, no financial details just yet. There are some big winners. Tiger Global, for example, made roughly a billion dollars on this position in just seven months, according to a source. CEO is behind me. He's definitely made over 2 billion. But investors, if you're looking to get in, should watch out. The lockup period is August. End of August, 84 million insider shares could hit the market, dwarfing the public float. Another 80 million become eligible in September. But you can see the stock price holding steady at 385.
B
Incredible. A little too steady. And I'm glad that we've now put up the headline that it's halted for trading. Christina, I don't know what to see
A
that all the noise.
B
We don't often see things flatline unless. So who knows what's happening in terms of the mechanics there. But I think all of this indicates that there is huge demand. It's so nice to see. I mean our audience knows very well a lot of the IPOs of the past year or two have been private equity backed. They've come to market. It feels a little bit like a move of desperation. They haven't traded well on the open. They've kind of moved sideways in general. This is very, very different. This feels like the first major AI related IPO that is now in public hands. Yes.
A
And two points the reason why you're seeing this halted for trading. This is the reason why they were stalling in the first place because of that just mismatch between buyers and sellers right now. So that is why the stalling happened. 385 but to your point about, you know, so much demand for Cerebras, there is so much demand for memory right now that companies all across the globe are trying to look for alternatives for cheaper memory opt. The fact that a company like Cerebrus is putting the memory directly onto this massive silicon like sheet dinner plate, whatever you want to call it, you know, would cut costs. This does make it a little bit less flexible though, especially when you start using smaller systems, when you start using physical AI. And of course we know that Nvidia is going to be launching their Vera Rubin chip that will have the Groq technology, also memory directly on the chip, eliminating some of those bottlenecks too. So there's no doubt the competition is fierce. But the market is so big and everybody keeps talking about how agentic AI is just blowing up demand, especially within the inference space. And that's where Cerebrus, Nvidia and maybe
B
others can play putting memory on the chip. Christina, stay right there. We were able to bring back Brad Gerstner and Scott Wapner. Brad, appreciate it very much. Love your reaction here. A remarkable moment, truly.
C
I mean, you know, listen, everybody's getting caught up in the heat of the moment. We have a supply demand, you know, imbalance here. The company's going public at over $100 billion. You've seen the IPO price, you know, that's gotten walked up. I think the reason for that is people, the long term excitement. We're in the age of inference. You have to produce tokens in order to get intelligence. This is one of the best companies at producing tokens at scale and fast tokens. By removing that memory wall, if you will. You know the challenges. Nvidia, this is going to be all tides rising right now. I think some of this exuberance I get worried about. Scott, like I said on the show with retail investors, I would not be yoloing into this thing at 400 bucks. I see. It's okay. It's trading down a little bit now at 325. That feels better. A retail perspective. I imagine you'll see institutional investors begin to move in, you know, at or around these levels.
D
Would you say that, that you couldn't have timed this IPO any better for the moment that we appear to be in? Right. It's all about AI. Some have suggested it's literally the only trade in the market. Kel, you have to think about that kind of thing. Right. The bankers, they timed it, the management team, they timed it perfectly to capture a moment in time and they end up being the first, as Kelly was saying, Brad, the first of the real big AI IPOs of many more to come for sure.
C
I mean listen, 18 months ago we pulled this IPO, if you remember. Right. The Biden administration would not allow this to move forward because of cfius. Right. Remember, the UAE is an incredible and important partner here to Cerebras. They're also building out stargate in the UAE which is super important for OpenAI, Anthropic and others. And so we couldn't even get it public 18 months ago. Today you see what, how much has changed when it, when Jensen said inference is going to 1 billion x as we showed in that clip, that was the tell. The whole world shifted from pre training and flops to inference and overcoming that memory wall. That's what Cerebras does so well. This is going to be a long term story with these guys. I'm telling you, they're business builders. Their product pipeline is chock full. I'm not trying to talk the stock up from the prices it's trading at today. I've already warned retail investors. But this is going to be something that plays out over the decade, not out over the next decade.
D
I get the feeling you're almost trying to talk it down a little bit because it's in uncomfortable air. It gets a little harder to breathe. Kel, you know, if Kelly, if this is a test case for what's to come, in some degree it's look out above, not below because there's just so much excitement about all of this. Anything AI related as you know and cover every day on your own program.
B
It's a great line.
D
Just has been straight up to the right.
B
I would quickly ask Brad, but I know we have to let you go but Brad, who is next? Everyone's talking about space X, but I'm talking about now that this has happened, does it actually do, does it hasten an open air and anthropic? You know, I'm just trying to fill in the blanks here. Those names are already so mature. But is there anyone else early stage that this IPO frenzy could now kind of draw into the public space?
C
Well listen, you know, kudos to Andrew and Bob and team for having the courage to go first. Right. They didn't know in the middle of an Iran war whether or not, you know, the markets who are going to be sitting where they're sitting. They had the courage to step out now just like they did, you know, 18 months ago. That will open the way for not just the mega caps like SpaceX, Anthropic and OpenAI which are all coming right, but it's also going to open the door for us to encourage some of the smaller and medium sized companies. As Scott knows, I want to see retail investors in this country get more access to these things and I think there are going to be more coming as a result of Cerebras having the success.
D
You don't think some of those smaller companies are afraid to come out in the like swim in a pool with whales? I just wonder what they if they're being discouraged actually to come to market and being crowded out rather than encouraged to go public. In the wake of all of this,
C
my hunch is today in Silicon Valley my portfolio companies are all saying hey, we have a shot to do this. You don't have to be the biggest, you don't have to be a trillion dollar company in order to go out. We can get access to cheap and durable capital. They all need capital to continue to scale and grow their businesses. This is a great example. It is a textbook play. Kudos also to Mor and Stanley for their leadership and getting this thing priced. This is not easy stuff. The, the people wanted allocations in this IPO greatly exceeded 25 times over subscribed. So I think this is a good indicator of what's to come. The next one's going to be SpaceX and, and, and you know, everybody wants a piece of Elon.
D
Yeah Kelly, if you'd allow me one more. I just wanted Brad to just quickly go over the lockup structure on this because it's unique. Yeah, it's unique. It's not normal for what we're traditionally used to seeing. Correct. And that's for a reason.
E
Correct.
C
So you know, in a traditional IPO we had what we called a six month cliff lockup where six months after the IPO investors like ourselves, insiders would be able to sell. The problem with that is as these companies get bigger and bigger, it has this massive overhang and that overhang gets arbitraged by quant firms and other folks. You don't want to have all that supply coming to market all at once. This one is very different. In fact off I'll post a chart on Twitter as to what the lockup looks like because it's a little complex but basically it's a rolling lock over a course of six months. It's called a Dribble lock. So they'll dribble, be able to dribble out shares early. Employees or employees in the business I think can start selling shares within the next couple days. Investors and others over the course after their first earnings report, which I think will be five, six, seven weeks from now. And then it will be dribbled out over six months. So it's like 15% release, 15% release and then 6% each few weeks after that. So you're not going to have a big overhang on this where all that supply splashes at once. And the other thing it does, Scott, this is clearly supply constrained. It's going to get supply into the market faster.
D
Kel, thanks for letting us hang with
F
you for a little bit.
B
Oh my gosh, this is an honor, truly. Thank you both. Really, really appreciate it. Perfect timing here right at the turn of the hour. Love getting your reaction. Scott. Thanks Brad. Thanks very much for those details on Cerebras. A dribble lockup. Everybody should adopt that. Let's get back over to Christina Parts and evil at the nasdaq. Christina.
A
Yeah, to that point, the dribble lockup. So it's 84 million shares at the end of August. That would be the first tranche of class B shareholders that can put their shares into the market. And then in September, October, it's another roughly a little bit over 80 million. So that's a substantial amount that we need to consider whenever we are buying into we meaning, you know, our viewers right now buying into a publicly traded company. The other issue is that Brad did mention back in September a year ago when the company, or over a year ago, 2024, when they tried to go public first time, the issue was the customer concentration, concern with the ties with the UAE. Just last year, 86% of total revenue still only came from two customers from the UAE. Of course that is changing. The company is growing their cloud services. They signed on OpenAI AWS. We're hoping that's going to translate soon into actual numbers so we can see how that's going to affect the revenue line. But that is still somewhat of a concern. The other major concern too is the moment Nvidia launches the Vera Rubin offerings with the Groq technology in there. What does that mean for Cerebras and their, you know, stance that they're, you know, x amount faster when it comes to high bandwidth HBM3 the memory on their chips. So how, how can they leverage that when you have Bear Rubin, a much smaller ship coming to to market as
B
well as the MO too. And that's what you just said. I mean 86% of their revenue last year came from two customers in the UAE. You know, there's, there's, there's a little hair on, on the story there, but still an incredible moment here at 368, 370 or so is the print right now. Christina, thanks. Let's bring them back after that halt. Yep, exactly. Went all the way back down around 352amoment ago. Dan Primack is the business editor at Axios. And Dan, a year ago, a little earlier maybe it was around January. I asked you if we were going to start to see a bigger wave of IPOs. You said it wasn't the year. I posit the same question now. Is this a bigger, more significant moment or is this just going to be a cerebras, a SpaceX, kind of a one or two off?
G
No, I do think it's a bigger moment right now. You know, you've seen this week for example, I think Some, there's been about 9 billion worth of IPO value just this week and we're what on Thursday there had only been 18 billion going into this week. Things are definitely picking up and that hasn't just been this. It's. There was an energy company, there was another infrastructure play, lots of stuff around AI. There was also a health care one. Now we do seem to be seeing it. But that said, when the year ends the numbers are going to be all skewed because of the big three. Right. Space X and then probably OpenAI and Anthropic.
B
SpaceX, OpenAI, Anthropic. Do you think OpenAI and Anthropic are coming before year end?
G
I think they want to and I also think each one wants to go before the other one. Maybe just in part to be able to tell the story first and a little bit of concern that some demand will get sucked out or some people will, you know, make their bets and pick one and they'd probably pick the first one.
B
That's fascinating. So OpenAI and Anthropic, we love this rivalry that they have between them. They're also in a rivalry to go public and be the one to tell that AI story. And that could be coming along with SpaceX and Cerebras. And I guess Cerebras is telling us there's plenty of investor appetite for all of this.
G
This Cerebrus is definitely telling us that. Like right. And, and Brad talked about the timing or Scott talked about the timing being so good. It really is. This is a chip company. You know, when this thing originally got funded as a startup a decade ago, venture capitalists wanted almost nothing to do with chips, let alone memory chips. There had been a bunch of memory chip failure companies. So this is a huge success in this. But if you look this year, kind of the rally since March has really been driven by chip makers. This totally plays into that. It obviously plays into the AI strategy. It's worth noting though, it's tried to go public twice. Brad's right. About 18 months ago it tried to go public. There was National Security Review that kind of slowed things down. But it also pulled an IPO filing this past year, late in the year, well into the Trump administration. That was maybe because of the government shutdown, maybe not. Either way, it worked out very well for them, timing wise.
B
So. And that's. Do you know who Dan directly benefits in terms of the VC ecosystem from this?
G
Yeah, I mean there's a lot.
E
Right.
G
This is a company that's raised billions and billions of dollars. Benchmark is one of the biggest ones, kind of one of the landmark Silicon Valley firms that invested in it at the very begin. But like Tiger Global invested in it just about eight or nine months ago. It did that round when the company pulled its ipo. Tiger put money in, then they put in more money this past January. It's gotten a huge value on that. Fidelity, which you don't think of as a venture capital investor, they led around. They made a lot on this too.
B
So then now I have to imagine, because it's kind of unique, like you said, that in order to have a Cerebras today, you had to be doing this a decade ago before anyone had ever heard of ChatGPT. Aside from OpenAI and Anthropic, are there other companies waiting in the wings to go public here or are these just a couple of unicorns?
G
No, I mean there's a bunch. And whether they'll go public this year or not, Databricks is an obvious one. Right? Which is a, which is an AI infrastructure play on the data side that's already valued in the private markets hugely. That's going to come out soon. There's a bunch coming. The big question, and it got raised a little bit earlier, is will the huge ones, these trillion dollar companies, right. SpaceX, OpenAI and Anthropic might they suck a lot of the money out? People do have allocations for new issues. People have equity allocations. Could those suck the money out? Cerebras doing this suggests maybe not. People Aren't waiting. But that's a question that I think we'll all have to wait to see
B
who wins and who. I mean, if anyone loses from today's
G
ipo, I mean, nobody loses today because the thing's still going up. I mean, there are some questions for this company, right? For example, there is the power they have to deal with. There's the customer concentration issue that they have to deal with. There are real things. But nonetheless, right now this thing's going up, right? It was supposed to price originally. I think it went 1:25 to 1:40. And then they raised the thing and then they raised it again and now it's, it's through the roof.
B
Right. So Cerebras, by the way, the IPO price was 185 last night. It opened at 350. And so we're showing the 89% we're showing on the screen there is from the IPO price. If you're in the IPO, fine. You don't have to worry about it if you bought on the open right now we're right about the, the flat line. Dan, I just want to point this out. We've traded all the way up to 385. That's where it was halted when it opened around the turn of the hour. We're coming back down now to the 350s range again, as long as it's well above 185. I'm just saying for those who bought in at the open after it was public right now it's going to be a question mark about whether you're in the green or in the red.
G
That's right. But I mean for anybody, obviously for any of the venture capitalists, this is extraordinary, including and for employees at the company. Yeah, but we're what, we're 10 minutes into this thing, 15 minutes in. This becomes a long term play. If you're buying this to day trade on it, then, you know, good luck to you. But this is a long term play and it's a long term bet on the future of AI in the country. I will say one other risk factor in here are data center moratoriums. Not a huge deal for this company yet, but it is something you have to watch.
B
Aside, would you echo what I can't remember if it was Brad or someone else who made the comment. Is it only AI that's working in terms of the IPO market and the whole market? I mean, we've talked about kind of the, the soggier IPOs. We've seen private equity Backed names in a number of different industries the past year or two. Do you think this lights a fire under all of them, or is it AI specific?
G
I think it's still AI specific, but. But that's really broad. Right? Again, think of all the energy companies that are, you know, providing power to data centers. So that's an AI play, but it doesn't have to be necessarily. I think it's primarily an AI story, but I think it's a very, very broad AI story.
B
All right. And then finally, 3:37, I just want to point out, is the price now. So it's, it's coming back down to earth a little. Would expect a lot more volatility today. I mean, does the dribble lockup that Brad talked about help you think the trading pattern? Because they're going to try to only unleash shares a little bit at a time.
G
It should. And I'm guessing that lots of banks and lots of other potential issuers are watching to see, particularly in the months ahead when that dribble starts to actually come out, what the impact is and see if it should become kind of a new industry standard, at least for large issues.
B
Yeah, it sounds like a good idea to me. So I don't know if I'm missing anything thing.
G
I, I don't think you are. No. I, it, it strikes me as a good idea, too. Not quite sure why it's taken until now to, to do this.
B
All right, Dan, we'll let you go. I know you have to go write this up and do a billion more things, so thanks very much. Dan Primac joining us there from Axios. Again, Cerebras just below the open price. I just mentioned that because we love the public market investors, too. It's well above the IPO last night, popping, as we say. Chris Crisanti is here on set with me. He's the chief market strategist at NAI Capital Management. Michael Kanch is a chief investment strategist at Piper Sandler. Welcome to. Chris, a thought on the significance of the IPO that we're seeing here right in front of us, a big one today.
H
Well, IPOs are like life, Kelly. Timing is everything. And I couldn't think of a better moment for an AI semiconductor company to come public after six weeks of semiconductor companies just going through the roof. And as one of your guests said, it is the age of inference that we're now entering. And this is a chip made for inference. Okay, so all of those things put together, but yet we have a company that's now trading at 200 times sales. So I would say buyer beware. There's more things coming. There's competition in this space which is great for the, for productivity but you know, it's priced for super perfection right here.
B
Cantra, what are your thoughts?
F
Well, I think this is good news. I mean clearly for the market to see such a large IPO in a space that there's a lot of excitement for it certainly gives investors another name to diversify this AI theme. Obviously they're all correlated but I think again it gives another way to diversify out the current positioning. But it's the first of potentially very many large IP or larger IPOs we'll see throughout the year. So it shows the capital markets are open for companies and that's good, that's good news.
B
Michael, does it become the kind of bell ringing at the top signal like oh remember when Cerebras run out at that valuation and now it's trading at 150 we can all write these headlines lines.
F
Yeah, you know one of the, one of the things that took place in the late 90s, roughly in late 98 and certainly through it peaked in the March of 22,000. The S&P had 12% of its weight in non profitable companies. We're sitting around 3% today and that number has been going down over the last several years since 2023 as obviously the bulk of the capitalization in US large caps are very profitable. Cerebras is profitable. So that doesn't make that go up. Maybe later this year we'll see some less profitable or unprofitable names but that's a number that I'd continue to watch. I don't think in many ways we're at the peak of this cycle both from a macro economy or from an AI capex spend.
B
Chris?
H
Yeah, I agree with my friend Cantro who's terrific to see. Again, what I would say is that this is act one, we're going to have act two of the anthropic. We're going to have act three with OpenAI and act four the big closer with SpaceX. So there is lots of things waiting in the wings. It's hard to believe that this won't be seen 10 years from now as part of a climax. How far we have to go up from here remains to be seen. But I also believe with Cantro that we're not at the peak right now. There's more excitement coming. We're building, right?
B
Exactly. We're in that building phase and so, so you know, this is a kind of specific niche of the chip space. They're putting memory on the chip and Nvidia is working on a competitor to that. What are your thoughts about the semiconductors broadly after the run we've had for the past, is it almost what's mid May, six weeks now?
H
Sure. Well, you see, intel has tripled in six weeks. So a value investor like me would tend to say, oh my gosh, this is like a meme stock, but it's really not. Kelly, there's real reasons for this happening. This move from training to inference is a big deal. And it's probably the biggest evolution certainly we've had since the Internet and I think since the railroads. So these semiconductor companies that are taking off now, intel would be one of them, but Serbia's today would be another because they're focusing on inference, that we're in inning one of inference. And so I have to do the valuation. But it's not a meme stock. This is the real thing. And you got to put your pencil to paper.
B
Even though it's trading at 200 times, was it revenue or are there earnings?
H
Yeah, there is earnings. It's trading at a thousand times earnings. But the question, of course is not what it is today is will they be in market leader like Nvidia in five years? And that's the big question. Hard to answer, but. But unlike Nvidia, there's a bunch of competitors nipping at their heels. So stay tuned.
B
Michael Kanchowitz. It's also, this is a sector technology where normally in an environment of rising interest rates, rising bond yields, and I used to drive Mike Santoli crazy because I'd always ask, are rising bond yields, you know, going to hurt the tech stocks? Because they used to have a little bit of this yin and yang feel to them. And here we have, it's just full steam ahead. It's rising tech stocks, rising bond yields, oil prices. I don't know if that matters one way or the other. Is this different because the earnings power is just so strong? The story is so blatantly obvious. It's not like the 2010s when it was the era of ZIRP supporting the fang trade?
I
Yeah, absolutely.
F
I mean, earnings strength and magnitude is the protection against higher rates in oil. And we're seeing that, you know, in, in 2023 through 2025, we saw small caps get whipped around day in, day out because of what interest rate. And now their earnings are going up and they're a little less correlated to interest rates because There are actually some earnings growth in small caps. The first time in a few years, tech has the strongest and broadest earnings growth which offers a huge moat against any macro volatility and uncertainty. And in the last six weeks, besides all of the good news, the deals and the strong earnings in the tech space, we have seen earnings, I'm sorry interest rates in oil remain high in areas like, like retail, home building where they're much more rate sensitive. And there are many more questions have obviously struggled. So I'm not surprised the market's been narrow in the last six weeks given what rates in oil and tech news has been and going from here to get the market to start to broaden out here and maybe not just reward semiconductors and tech. I think you need to see interest
B
rates come down, Chris, on a day when the Dow was back above 50,000 for the first time since February. Nasdaq, I'm sorry, Nasdaq and S and P again at record highs. Kind of same question to you. What is your feeling broadly about the rally and what this signifies?
H
I think he's exactly right. I think we need to see a broadening of this rally. But you know the war has, I think the war has ended about four times now so far. But if the war really ends, I think you will see rates come down obviously you'll see oil come down not quickly but slowly and that would be a terrific recipe for seeing a broadening the rally. We need housing, we need financials and but, but I think there's a good case to be made that we will see that.
B
All right, come back, we'll talk some Microsoft next time. I know that was your under your
H
a lot going on.
B
Yes, exactly. Cerebra to bring our viewers up to speed or those who are listening as well. 326. It was pinned at 385 after opening at 3:50. So we've come down a ways since the turn of the hour. Michael and Chris, thanks very much. Coming up, a grave warning from Chinese President Xi Jinping setting the tone for the first day of the US China summit. What that signals about the relationship between these two nations.
J
Next,
K
What would you like the power to do?
B
Don't worry, you got this.
H
Whoa.
K
Hear that?
B
I did it.
K
That's the sound of you helping your child find confidence that lasts a lifetime. Bank of America invites kids 6 to 18 to join golf with us for a limited time. Sign them up for a free one year membership, giving them access to discounted tee times at thousands of courses. As we champion the next generation. Who dares to ask, what would you like the power to do? Restrictions apply. Activation required. CPFA.com Golf with us for a complete
E
details on May 29th. From focused features and the producers of Darkest Hour comes the untold true story of D Day. Facing the wrath of nature and with the largest seaborne invasion in history at stake, the fate of the war rests on the shoulders of two extraordinary men.
C
One impossible decision.
E
Featuring powerful performances from Andrew Scott, Brendan Fraser, Kerry Condon and Damian Lewis, the Untold True Story of D Day only in theaters May 29th. Rated PG13 may be inappropriate for children under 13. Experience it in Dolby Cinema.
H
Dish has been connecting communities like yours for the last 45 years, providing the TV you love at a price you can trust. Watch live sports news and the latest movies, plus your favorite streaming apps all in one place. Switch to DISH today and lock in the lowest price in satellite TV starting
B
at $89.99 a month with our two year price guarantee.
H
Call 888-D dish or visit dish.com today.
B
Welcome back. A news alert in Washington now with Emily Wilkins. Emily, what's happening, happening?
L
Hey, Kelly. Well, the crypto industry just notched a major win with the Senate Banking Committee going ahead and advancing a rules of the road crypto bill that has really been the industry's number one priority. It's been what they've pushed for. They did it with all Republican support. Two Democrats did join and everyone on the committee, both Republicans and Democrats, committed to working on this bill to further build bipartisan support in hopes that it can pass the Senate. But really, this is just a, I'd say a stunning win in some ways for the crypto industry, given that you had the banking industry, you had unions, you had law enforcement groups all come out against this bill and you really saw the crypto groups be able to flex their muscle, flex their power. Part of that was because of a grassroots campaign. I talked about that a little bit yesterday with Coinbase CEO Brian Armstrong, but I also asked him if the outcome of these votes should be impacting how the crypto industry gives in the upcoming midterms. Remember, they have that $200 million war chest and are ready to play a role in some of these primaries. Here's what Armstrong said when I asked him about how that crypto pack should be spending their funding.
M
You're going to do. And you know this committee vote is important. I think even more important is the full Senate floor vote.
L
We'll be, of course, keeping a close eye on if the Senate is able to get to the 60 vote threshold needed and when that vote might be scheduled. Even if it does, they will still need to figure out a path forward with the House. They passed a completely different version of the bill last year and all that would need to happen within the next couple of months before attention really turns to the elections and legislative activity grinds to a halt around here.
B
Kelly, it looks like the final compromise allows for rewards or incentives for kind of bona fide crypto activities, but bans passive interest that mimics bank deposits. So they were able to bring, you know, I know Coinbase was originally against it because they weren't going to allow what do they called yield and it sounds like they're allowing some forms of it.
J
Yes.
L
Basically what Coinbase and others basically agreed to was that if you are able to spend stablecoin in that case you'll get rewards. But if you're just holding stablecoin the way that say you'd hold money in a bank deposit, that's not going to allow stablecoin holders to get those rewards. That's what Banks said was going to be too close to interest. And they said it could potentially lead to millions or billions of dollars being taken out of bank accounts and then banks not having that capital to make lo. Now, Banks said that there is still a loophole in the law that they are concerned about. There were some efforts today to offer amendments, but those are rebuked. That might be one of the things that's going to have to get tweaked a little bit if this bill wants to get enough Democratic support to actually continue to advance and pass the Senate.
B
Great point, Emily. For now, Banks with Coinbase shares up 7%. Meantime, if you're looking for any indication on how strong the AI trade has been, look no further than 6. Cisco CSCO on pace for its best day in 24 years on strong earnings and guidance with AI at the heart of it. It's up 12% right now. Evercore's Amit Daryanani joins us now. Amit, you join. You covered the stock, you raised your price target today and you think this is a breakout moment. Why?
E
Yeah, this is. I feel like it's kicked back to the 90s with the networking boom. Back again right now. Listen, I think there's two things happening at Cisco and probably at a broader ecosystems level, right. Right now one is you're seeing the AI infrastructure tailwinds, right. So as the hyperscalers are adding more capacity, Cisco is benefiting by selling them Silicon Solutions, their optical solutions and so on. That's a nice business is going from 5 billion to 9 billion. The bigger excitement I would argue telling with them is going to be as enterprises start to future proof their campus, their networking solutions to become more AI ready. You could see a very long tail of demand tailwinds of upgrade cycle at Cisco which could really drive multiple years of double digit low teens kind of top line growth for this company. It's really more on the enterprises getting AI ready I think is a big story for them as they go forward versus anything else.
B
And I'm just going to restate this again because network and infrastructure. So what you're saying is they are going to benefit from companies looking to future proof their campuses to be AI ready?
E
Yeah. This means two things. One is you want to deploy AI, you want to deploy agentic workflows in your corporate environment. You need to upgrade your network, your network time supported. The other element of this is if you think about things like Anthropic's Mythos and all the vulnerabilities that it can take advantage of from a security basis, you also want to make sure your network is future proofed and can protect itself from all these breaches and vulnerabilities as you go forward. Right. So it's almost a two pronged thing on why enterprises have to upgrade their network. One is being AI ready, but the B is actually protecting yourself from all these security threats that could arise.
B
Right. So your price target is 150 now that stocks at 114. And again, you know we say best day in a couple of decades, it's up 12%. I mean this isn't exactly the roaring strength of intel, but could it be?
E
You know, it didn't fall as severely as maybe other names did. So perhaps the recovery won't be that spectacular. But I do think that this could have durable legs as you go forward. So I don't think it's a one quarter perspective. It's going to be a multi year tailwind. As you start to see not just the AI tailwinds, broaden out the hyperscaler to the new clouds, but also you start to see not just big enterprise but also medium and small enterprises upgrade the architecture quickly.
B
What's the deal with the layoffs? I mean this looks significant. They're cutting almost 4,000 jobs.
E
Yeah. One of the challenges, and I think Cisco and others that will do this will have, is AI inherently is a bit more of a lower gross margin business than everything else that you do. And in an attempt and in the Right way to protect the operating margins and free cash flow. They have to optimize opex. One way to do it is you deploy AI internally to automate tasks that you can. I think that's where a lot of the researching efforts are coming from. The other part is obviously reorienting your workforce a bit more towards Silicon One optics. And you're trying to optimize your headcount as a result of that.
B
It's interesting that you say AI is a lower gross margin business potentially. And I mean that's significant even as they're pivoting to capture some of that growth. Okay. On Apple there are reports what's going on, some bad blood between them and OpenAI.
E
You know, the relationship was never exclusive to begin with. Right. And I think that that always set the, you know, at least a fear that hey, how long will this really continue? And Apple and Gemini are, I think or Apple and Google, I should say have been partners for much longer. I would imagine that's the way they want to go as they go forward. Forward. I do think there's an element and if you read that article, there's a lot of this which is, you know, OpenAI did buy IO, which is Johnny, a company that posts a fair amount of talent from Apple. So I'm sure there's a bit of a, you know, friction from that basis as well. But I do think Apple wants to have a more broader open ecosystem. So you and I as consumers can hopefully over time go and choose which LLM you want to use, be that Gemini or OpenAI or Anthropic.
B
You did a 105 page deep dive on all things Apple and your takeaway is outperform.
E
365 outperformed 365 days of the year and the stock will get to 365.
B
I think investors would love if they outperformed 365 days of the year. Amit, thanks very much for joining us today.
J
Appreciate it.
E
Thank you.
B
Amit Daryanani from Evercore. Coming up, Presidents Trump and Xi seemingly friendly in public, but the Chinese President issued issuing a stern warning. We have details from the first day of this key summit. Next, Take a look at the biggest IPO at least of this year and possibly of the last couple and a very indicative one as well. We're talking about Cerebra Systems, a chip maker. They do AI chips that have memory right on the chip itself. They opened today after strong demand in the ipo. Demand was so strong in fact that even after opening at 350. It was briefly halted for trade up at 385, just past 1pm Eastern. It's come down from that level. We're around 330 right now. So we're below the open price for the public markets, but again, well above the IPO price. You've heard our guests saying they do expect this to drive a host more companies who could be part of this AI narrative to go public, including the biggies this year. A race between anthropic and OpenAI and then SpaceX, of course, perhaps the largest. Meantime, President Trump and Chinese President Xi Jinping wrapping up their first round of talks in Beijing today in a summit covering trade, Iran, tariffs and Taiwan. Both leaders smiling for the cameras. But there seems to have been some tension behind closed doors. For more, we're joined by Derek Scissors, Asia economist at the American Enterprise Institute, and our own Eunice Yoon, who is live in Beijing. And Eunice, what do we know?
J
Well, Kelly, Presidents Trump and Xi are going to be sitting down for their second bilateral over tea and lunch on Friday after a full day of activities. Today, early at the state dinner, President Trump had delivered remarks inviting President Xi to the White House on September 24th. It's unclear at this point whether or not President Xi will accept that invitation. During the day, the two had met at the Great hall of the People where they sat down for a bilateral and the preliminaries were quite positive. The two were talking about improving stability, you know, injecting more stability into the relationship and avoiding conflict. The Chinese readout had described this as a new vision of strategic stability. But the big news out of the meeting was President Xi's surprisingly strong statement on the issue that he finds most important, and that is Taiwan. He said if handled badly, the two countries risk collision or conflict, pushing the whole relationship into a very dangerous situation. Taiwan independence is incompatible with peace in the Taiwan Strait, like fire is to water. Now, that stern message underscores the fragility of this truce. We've also learned more about what was discussed on Iran. President Trump told Fox News that President Xi has committed to withholding military equipment from Iran and would like to see the Hormuz Strait open. Both of those comments, though, we have seen before. Also, a President Trump told Fox that President Xi agreed to order 200 Boeing jets. The market, though Kelly, was looking for 300 hundred or more.
B
I love when we get to this level of precision. And Boeing shares are lower, but Cisco is helping the Dow anyway. Today, Eunice, the Chinese make a big effort to choreograph and project a certain very select image. But how is everything playing out in reality?
J
Well, you know, there's always a big difference as to what the Chinese like to present and what's actually happening on the ground. So we've seen the authorities staging several events, including, say, the children who were waving flowers when President Trump had arrived. But meantime, even here personally, there were Chinese police who were banging on our doors during certain live shots. People were, some of the authorities were actually below me from the balcony trying to distract us with green laser lights. The Senator, Rick Scott, had an interview earlier today that was blacked out. Journalists have told us who have been coming in say that their apps won't work, they haven't been able to call taxis. And a lot of that is because this country is still very much segmented off because it is a security state and a lot of the international service here are blocked.
B
Wait, wait. So you're telling me that while you were giving a report for cnbc, which is carried both here and I presume in China, that Chinese, Chinese security forces or police were beneath your balcony flashing a green laser at you?
J
Yes, they were public security, security. And so they were trying to get us not to be doing our live report on the balcony. So we saw laser lights, and I was talking to Joe and I started seeing all these green laser lights, you know, in front of the camera. But we were just trying to ignore it for the moment so that we could finish the report.
G
Wow.
B
I don't even know how to interpret that other than that. I think that level of pressure indicates how strongly they feel about controlling the narrative. And Eunice, we appreciate you joining us, making the time throughout the day today. Thanks. Eunice Yun in Beijing. Let's move to my next guest who says the most telling development from this trip so far has nothing to do with trade, Iran, or Nvidia's chips. Derek Scissors is an Asia economist at the American Enterprise Institute. And what is that telling Development. Derek, good to see you again.
I
Good to see you, too. I was very surprised that the president invited General Secretary Xi back to the US so quickly, September 24th. Of course, that may not happen. I even say it probably wouldn't happen. But he put this on the record that he wants a return trip six weeks before the midterms. And you've seen a lot of the fawning that goes on in these trips by both sides. It's a surprise to me that the president, who's the leader of the Republican Party, wants to fawn over the general Secretary of the Chinese Communist Party. Six weeks before US Election. So there's a lot of stuff that we expected. Stern words on Taiwan. Order some Boeings. Harassment, harassment of journalists. This is all standard play. A quick return trip to the United States in an election year is a very big signal from President Trump that he values the China relationship apparently more than he values some of his Republican colleagues chances in November.
B
Why? Because what, what is the big risk if this invitation is accepted? And again, it may just be one of those people who says, yeah, yeah, yeah, we'll, we'll get it on the calendar. We'll, we'll, we'll find, find a date.
I
Yeah, it absolutely may not be accepted, but it's, if President Trump is obviously taking the risk that it's accepted, and then you have a situation where the head of the Republican Party is, as what happens in these meetings, fawning over the general secretary of a Communist Party, telling him how wonderful he is, how strong he is, how smart he is, how much he likes China, which is what he did on this trip. A lot of Republican candidates are incredibly uncomfortable with that position. They may or may not say it publicly. And more to the point, for them, it's a wide open door for Democrats to say, is this, this is what you want. More Republicans fawning over the Chinese Communist Party. And, you know, it just hands them an unnecessary election issue. If you did this November 24, it wouldn't happen this way. So I just thought the timing of this, it's not necessary. It's only a few months from now. They hardly need to meet so quickly. I thought the timing of this was, I think, unintentional, but nonetheless sort of a snub of Republican Party politics for the midterm election.
B
Why might there some urgency to the matter from the President's point of view?
I
You know, I think there's a. He wants to lock up. What he's gonna say are the successes of this meeting, which is the Chinese again, promise to buy soybeans. They promise to buy Boeings over a long period of time. They say they're interested in us lng all of this. I think if you're following, this is old news. I think the President wants to make sure it happens as soon as possible. Those are the gains he's looking for here. I hate to put it this way for everyone invol, but if the Iran situation drags on, maybe you want to have an earlier meeting Again, this looks like he values what he can get from what he can get, possibly from China, more than he values the political optics back in the U.S. two more
B
quick questions, one on Iran. What comes out of this, if anything?
I
So far, nothing. I mean, what the Chinese have said is they want this Strait of Hormuz to be open. Well, that's what they wanted three months ago too. Obviously we have a Chinese pledge not to sell military equipment. Equipment. I don't know that I entirely believe that. They're obviously not going to ship large scale military equipment to Iran, but they can ship it to other countries who then ship it to Iran. So, you know, we didn't get a deterioration, we didn't get a Chinese attack on the US Position in Iran. I don't think that will happen in the next few months with all the meetings that are, that are at least scheduled. So I think the US can be satisfied that the Chinese aren't undermining our view of Iran. I don't think anything pushed the ball forward.
B
Finally, on the question of Taiwan, this is where this sharp language and warning came. President Xi said it was the most important issue in U S China relations and he said quotedly, reportedly, quote, handle it well. The relationship holds, handle it badly. The two countries risk collision or conflict. This is after yesterday Eunice Yun reported there was some hope in Chinese media that the U S would agree to oppose Taiwan independence, which I thought seemed unlikely. I think most people would agree. What's your take on the status of the Taiwan Taiwan question?
I
Well, I am reassured that the President hasn't said anything about Taiwan and the White House has made a point of this. I hope that continues through the meeting tomorrow. Given that he hasn't. It's very obvious from the Chinese side that they're going to make a big deal out of it. If you won't, then we will. And what the signal there is is you want some things from us, you want those purchases speeded up, you want us not to make trouble in Iran, you're not going to say anything about Taiwan for your own reasons. We're going to drive the point home that we can be coop, we can listen to your requests, but if you mess with us on Taiwan, all of it goes out the window. So I think in the absence of a US Statement about Taiwan, you are going to get a strong Chinese statement. This is all language, but nonetheless you're going to get stronger Chinese words about Taiwan to drive home the point that the rest of our discussions are conditional on the US not encouraging the Taiwanese to do things the Chinese don't like.
B
All right, Derek, on that note, we'll leave it there. Really appreciate it. Appreciate Your time, Derek.
J
Scissors.
I
Thank you.
B
With AEI coming up, the biggest IPO of the year and it did not disappoint. Cerebra shares soaring. After pricing way above the range on huge demand. They're currently trading around 325. What message is this IPO sending Silicon Valley? Are the pipelines about to burst with new offerings? We'll dive into that next. Soaring on its IPO, up more than 70%. But people who have watched this show for the whole hour, they saw it at 385 and we are down at 310. So I'm just saying what will it mean for Silicon Valley? And are we about to see a flood of IPOs? Let's bring in Slovenures general partner Sam Lesson. Sam, it's great to have you here. What's your take on how this IPO is going?
M
Well, number go up, I mean I think is the upshot. You know, I wonder how many people retail buying service even really understands what they do. But I think that's kind of besides the point, which is we're in this moment where people are dreaming of infinity and they're desperate to not miss. Right. And so what you're seeing, you know, the retail market has been boxed out of all the interesting stories. They're not an open AI, they're not an anthropic, they're not in world of money's being made. So my God, if there's an opportunity to be on the storyline, you know, you can only lose all your money.
B
I, I'm also just thinking about this being a chip company. This is hardware, this is factories, this is fact labs. That's not like hey, a guy creating a piece of software. You know, I just, are there a lot of other chip companies we should be thinking about in, in the IPOs, is this a one off and is everything else going to go back to, you know, a few people with a great idea about a way to deploy AI?
M
No, I don't think it's a one off at all. I think the real question is just look, why go public? You know, these companies don't need to be public to raise money. So then the question is, are you trying to get, get public to establish a narrative in the public market? Is it a read like what is the point?
E
Right?
M
And I think that's, that's a really interesting question. This will obviously excite a lot of people. You know, an Elon tweet about your company or about space data centers sets your valuation at a billion dollars. And this going out there Also, just like, is great. I mean, there's a lot of great private capital to be raised off of this. Well, and the excitement around it. The question is like, you know, again, are we racing for public companies? And why, I think will be the real question.
B
Brad Gerstner's take on this was, you know, it's great to have what, you know, we call permanent capital. And I'm sure that must be true, that on some level it's worth it to tap that important source, you know, and not just be relying on private markets.
M
Maybe like, again, private. I agree that permanent capital is wonderful and Brad's an awesome guy and very, very smart. But look, I mean, the history of do you want to be public or not? I feel like every few years switches right now, if number goes up, there's incredible retail demand. You know, retailers are willing to believe and buy stories and narratives because they might be true in terms of value equals infinity. You know, great. The question in my mind is, what will evolve that? What changes that? It's been a very bad time to be a value investor for a long time at this point. And the reality is you have this feedback loop going where the number goes up because the number goes up. And so to the extent this is true, I mean, like, look, I'm, As a small SpaceX shareholder, this is a great sign. I'm very happy.
A
Absolutely.
M
Like, number will be big based on what story?
B
All the value investors are in Microsoft. And I say that tongue in cheek because even as a value play, I mean, that's where people are looking. And so what else do you see coming down the pike? We mentioned the biggie's anthropic OpenAI SpaceX. Are there smaller companies that we should be looking to as well?
M
There are, but to be honest, do we care? I mean, the irony is those three IPOs, when they happen, will be so much bigger than all other IPOs, period, that it's kind of hard to get that excited about anything, to be honest with you. That's coming down the pipe. So I'm sure, look, some people will take this as a sign to go out. I promise you, the investment bank anchors are thrilled and they're going to run around the valley and say, now's your moment to be a public company and raise money. You know, should. Cerberus could have had twice as much money, it sounds like, from their ipo if they had felt like it. Do they care? Probably not. But again, I just. I think the value mentality on this is. I personally think it's very hard to care about anything other than the $3 trillion potential IPOs that in theory are going to happen in the next year.
B
And then those carry their own risks because by the time you get to that size and valuation, you know, it reminds me a little bit of the Uber story from years ago. But we'll have plenty of time to get into that. Sam. We'll let you go for now.
J
Thanks.
C
Nice, Sam.
B
Lesson from Slow Ventures 309 for Cerberus. That's it for us. And Power Lunch will pick up the coverage with Brian Sullivan right after the break. You've been listening to the Exchange. Make sure you're subscribed to get each episode every day, same time, same same place.
K
What would you like the power to do?
B
Don't worry, you got this.
L
Whoa.
K
Hear that?
B
I did it.
K
That's the sound of you helping your child find confidence that lasts a lifetime. Bank of America invites kids 6 to 18 to join golf with Us for limited time. Sign them up for a free one year membership, giving them access to discounted tea times at thousands of courses as we champion the next generation. Who dares to ask, what would you like the power to do? Restrictions apply. Activation required. CBFA.com Golf with us for complete details.
Episode: Cerebras Opens for Trading, A Midterm Misstep, and Crypto Clarity
Date: May 14, 2026
This episode centers on three major stories:
Throughout, host Kelly Evans and a range of expert guests break down the significance, risks, and market reactions to these fast-moving events. The tone balances enthusiasm and skepticism, with plenty of real-time market color and notable moments.
Segments: 01:00–24:26 | 34:34–36:34 | 47:10–51:37
Opening Day Frenzy: Cerebras, an AI chipmaker, debuted with one of the most anticipated IPOs of recent years.
Product & Market Context:
Competitive Landscape:
Lockup Structure:
“This one is very different… it’s a rolling lock over a course of six months. It’s called a Dribble lock.”
— Brad Gerstner, 10:35
Risks & Warnings:
Heavy initial customer concentration (86% of revenue from just two UAE clients last year).
Future competition from Nvidia and others.
Lofty valuation:
“It is the age of inference… but yet we have a company that’s now trading at 200 times sales. So I would say buyer beware. There’s more things coming.”
— Chris Crisanti, 20:34
Upcoming insider share releases that may add volatility.
Market Reaction and Implications:
Seen as a catalyst for more tech/AI IPOs.
Could set a template for lockup structures and how late-stage VC-backed companies enter public markets.
“Everyone’s talking about SpaceX... this will open the way for not just the mega caps like SpaceX, Anthropic and OpenAI... but it’s also going to open the door for us to encourage some of the smaller and medium sized companies.”
— Brad Gerstner, 08:50
Debate over whether public markets just want to join “the narrative”—“number go up,” as Sam Lesson jokes (47:56).
Stampede Into the Stock:
“There were 45 buyers for every one seller... the company sold about roughly 30 million shares last night at 185 bucks. So that’s a raising of about 5.5 billion...”
— Christina Parts Nevelis, 02:02
Warning for Retail Investors:
“I would not be yoloing into this thing at 400 bucks.”
— Brad Gerstner, 05:43
On the IPO’s Symbolism:
“This is the first major AI related IPO that is now in public hands.”
— Kelly Evans, 04:04
Segments: 38:02–47:10
Background: President Trump and President Xi Jinping met in Beijing, discussing trade, Iran, tariffs, and Taiwan.
Tension and Theatre:
“While you were giving a report for CNBC... Chinese security forces or police were beneath your balcony flashing a green laser at you?”
— Kelly Evans, 39:59
Main Takeaways:
Taiwan Tension: Xi delivered a stern warning—“Taiwan independence is incompatible with peace...like fire is to water”—emphasizing its centrality to US-China relations.
Election-Year Optics: Trump’s invitation to Xi for late September drew criticism as politically risky for Republicans, given how public warmth toward China could be leveraged by domestic opponents.
“It’s a surprise to me that the president, who’s the leader of the Republican Party, wants to fawn over...the Chinese Communist Party six weeks before a US election.”
— Derek Scissors, 42:00
Iran and Trade: No significant breakthroughs—China’s pledges on Iran deemed only symbolic; aircraft orders smaller than hoped.
Segments: 29:02–31:59
Legislative Breakthrough:
Political Ramifications:
“This is just a, I’d say a stunning win... you really saw the crypto groups be able to flex their muscle, flex their power.”
— Emily Wilkins, 29:08
“For now, banks with Coinbase shares up 7%.”
— Kelly Evans, 31:59
Segments woven throughout, especially 20:34–26:47, 32:27–36:34
Chip Sector Mania:
“Earnings strength and magnitude is the protection against higher rates and oil. Tech has the strongest and broadest earnings growth which offers a huge moat against any macro volatility.”
— Michael Kanch, 25:04
IPO Market Outlook:
“I would say this is act one, we’re going to have act two with Anthropic, act three with OpenAI, and act four, the big closer, with SpaceX... it’s hard to believe this won’t be seen ten years from now as part of a climax.”
— Chris Crisanti, 22:39
“There were 45 buyers for every one seller... raising $5.5 billion.”
— Christina Parts Nevelis, 02:02
“The age of inference... that’s what Cerebras does so well. This is going to be a long term story with these guys.”
— Brad Gerstner, 07:06
“This one is very different… a rolling lock… called a Dribble lock.”
— Brad Gerstner, 10:35
“Timing is everything... couldn’t think of a better moment for an AI semiconductor company to come public after six weeks of semis just going through the roof.”
— Chris Crisanti, 20:34
“Taiwan independence is incompatible with peace in the Taiwan Strait, like fire is to water.”
— President Xi (via Eunice Yoon), 38:02
“It’s a surprise to me... the president wants to fawn over the Chinese Communist Party six weeks before a U.S. election.”
— Derek Scissors, 42:00
“I wonder how many people retail buying Cerebras even really understand what they do. But I think that’s kind of besides the point... we’re in this moment where people are dreaming of infinity and they’re desperate to not miss.”
— Sam Lesson, 47:56
End of Summary