#312: Cerebras: 2026's biggest IPO?

0:00Study and play. Come together on a Windows 11 PC. And for a limited time, college students get the best of both worlds. Get the Unreal College deal. Everything you need to study and play with select Windows 11 PCs. Eligible students get a year of Microsoft 365 Premium and a year of Xbox Game Pass Ultimate with a custom color Xbox wireless controller. Learn more at windows.com slash student offer. While supplies last, ends June 30th. Terms at aka.ms slash college PC. I'm going to talk about a company that turned down Elon Musk. They spent$200 million solving a problem most engineers said could not be solved.

0:39And then just a matter of a few days ago, walked into the NASDAQ, pulled guns from the metaphorical holster and raised$5.5 billion in the largest US tech IPO since Uber in 2019. team.

1:00Welcome to Stock Club, the podcast where we find and discuss stocks that anyone can buy, many of which we believe will go on to grow your wealth. Before I welcome my friend and colleague Mike, I just want to remind you listeners that we, off the back of last year's great success, have decided to run Investicon this year, 2026, actually. Interestingly, we call it Investicon Year Zero because we wanted to just gauge if there was interest, if there was excitement, and if there was results, and we had all of those things in abundance. So this year, on the 27th of August, in collaboration with the Erlingas College Classic American Football game held in Crow Park, we are running Investicon separate to the gig in Crow Park.

1:47It's being held in Dublin's Market Bar, which is probably the coolest bar in what's effectively the world capital of bars. And we have something far cooler than a location and far cooler than the lovely tapas food you will be served if you attend. But we have some of the most revered and successful stock pickers in the world joining us for the day where we are going to have chats and laughs on stage. where Mike, my colleague, and I interview the greatest minds in the investing world. So what do you think, Mike? How are you this week? You all ready for that? I am. You're a great salesman now. I don't think the market bar is the coolest bar in Dublin.

2:31I'm sure plenty of bars will be listening to this with some objections, but no, I'm good. It's pretty cool, but I guess I'm such a good salesman, I forgot to say, go to investicon.ie and secure an early bird ticket because they are going fast and we are going to have a laugh and if you attend this year you get first access to buy next year should we do investicon year two so get on to investicon.ie which is the tld top level domain for an irish website.ie do you know and not to distract this whole podcast but do you know i think i'm pretty sure it was a antigua had the tld.ai and that's how right become their entire uh entire economy essentially is selling tlds to websites and companies and all the rest so gdp with the dot ai domain and then gdp without it is basically how they report the economic figures pretty much yeah so they looked into that one like ireland with the aviation industry do we include the plain leasing or not uh should we give them both numbers well there's more work being done in ireland now than just being given a given a domain name but uh yeah that's uh yeah that's my fun fact for the day and it's a top but that was a lot of fun i think we've just doubled our listenership while you're telling us that mike it was topical for today people are ringing their cousins and friends and father you should tune in wait to hear this for a fact so thank you mike for bringing us that little sparkle of joy mike can i can i talk to you about mipo yeah yeah and as i was saying it's topical with the antiga maybe they actually have to go and pay antiga for a top level domain these guys we're talking about today so we're talking about cerebras yeah so why would they why would they well i suppose you're right of course i mean the nature of their business sorry okay keep up emma well so let me rewind and tell our listeners though i'm going to talk about a company that turned down elon musk they spent 200 million dollars solving a problem most engineers said could not be solved and then just a matter of a few days ago walked into the nasdaq pulled guns from the metaphorical holster and raised 5.5 billion in the largest US tech IPO since Uber in 2019.

4:59And if you haven't heard of it, the company, well, you have heard of it. I said a few seconds ago, the company is Cerebrass, ticker C, B, or S. And if you have not been paying attention to this company, I think now is the time to put it on its radar. What do you think, Mike? Yeah, I was a bit, I'm not shocked. I'm not shocked by anything that happens in the market. But just a company of this size going public that I hadn't really heard of, like the only rumblings were like a potential Nvidia competitor. For it to go and be the biggest IPO in seven years, it's pretty impressive. But I suppose it's indicative of the current market we're in, where there is clearly demand for AI names and new shiny AI names with the kind of customer, the customer book that service has and the hype around it the fact that it is being seen as an video competitor and a genuine one because because nvidia has become so dominant obviously people are looking at their their margin profiles you know if a company like nvidia which deals in hardware can have you know 70 75 gross margins that's someone else's opportunity yeah i mean i was in france about two three weeks ago where i was chatting with a new friend who i will just refer to by his given name which is bob and bob is a very very senior law person in the united states one of the most senior and he was a i think an angel investor in cerebus and he told me the amount he invested and when i did it back at the envelope in my mind i suppose of what it was worth now I ran out of zeros yeah I was like well done Bob that's a great investment and to my own embarrassment I wasn't really aware of the business which kind of prompted me and he was quite enthused by the forthcoming IPO which has now happened and I dived in and as I usually do Mike I have a little bit of an origins story about this business so can I indulge you and our listeners for a couple of minutes yeah please do because it's a relatively new company I think it's only it absolutely is It was only founded in 2015 by five engineers, which is hardly surprising.

7:1211 years to a$100 billion valuation is kind of nuts, isn't it? It's unreal. It's unreal. So these five engineers had all previously worked together at a company called C-Micro, which one of the five founders, Andrew Feldman, and another guy called, I think it was Loterback, started in 2007 and they sold that business to amd in 2012 for 334 million dollars so not only do these engineers have their chops and know how to do things right but they also made a few books meaning that they are probably even more fearless for round two because when you've taken a sizable amount of money off the proverbial table, you can move forward with new knowledge and the strength that you're sorted.

8:00So these are proven operators who'd already built and sold a chip company and they came back for a second act with something that's far more ambitious. And it is, was, is that the entire architecture underpinning AI computing was all wrong. So that's probably a big bodacious claim and for i suppose people external to the industry like you and me mike it's kind of breaking news you're like oh interesting well the pitch that they made when they brought it to early investors um was really simple and you could argue pretty devastating and it was that gpus actually suck for deep learning and they just happen to be a hundred times better than cpus so we're all we talk about gp graph graphics processing units gpus and they were there the one upmanship on the cpu central processing units but that insight which is now kind of obvious in retrospect was pretty radical at the time and that's what cerebrus was built on if gpus were a workaround as opposed to a solution then there was a massive opportunity to build something purpose designed for AI from the ground up.

9:15And the answer they arrived at was a wafer scale engine, WSE. And it is in every sense, a genuinely strange and audacious piece of engineering. And that term, wafer scale engine, is something that I think we're going to hear a lot more about as we move forward through the AI age. so um that's the that was the pitch and that's the background but maybe i should just talk a bit about the technology mike but i don't know what way we should bring it you're gonna have to talk me through it because this is all getting a very technical for me already i'll let the physicist well it is no it is technical and when you start to read through their paperwork you have to push through the eye glazing syndrome because um it's very technical but here's how i would describe it I suppose, to a child, most chips are very small and they're cut from a silicon wafer, just the way you'd like cut tiles from a floor or for a floor.

10:17And Cerebrus threw that model out. The wafer scale engine that I mentioned uses the entire wafer. So rather than printing all these chips on a wafer and then getting out your microscopic scalpel, you basically use the entire wafer it measures 46 000 square millimeters and it can contains four trillion transistors and 900 000 cores and by putting everything on one surface it eliminates the need for complex interconnections between chips it kind of makes perfect sense and it delivers memory bandwidth 7000 times that of a standard GPU. So the engineering challenges involved in building this thing were really extraordinary.

11:08And anyone who might have studied stereolithography or the printing of chips or assessed chip makers in the past will know that this is something more than just printed on a big wafer there, Mick. The team had to invent this new cooling method to prevent a chip of that size from burning when it's drawing power. One thing we all know as listeners and as pedestrians in this investing world, AI world, is that these chips generate huge heat. And in previous podcasts, we've spoken about the cooling challenge. Well, that's exactly the same challenge that exists by inventing or building this mega chip, that it needs to be cooled down.

11:52and the team had to as i say invent a new method um so they invented a machine that could drill 40 screws into the wafer simultaneously without cracking it but i'm which i'm pretty sure in its own own right is a feat though i just write it it's like oh without cracking it there you go i'll write that down but i bet you the engineering team that without cracking it bit was probably a lot of intelligent people chewing on the end of pens. Well, we drill it from the other side, make and see if it cracks. Sure, you may as well throw it over there on the other side. We'll see if the drill cracks it.

12:28Oh, geez, it cracked. Oh, no, sure. I tell you what, we drill it from on top. No, you're not making any sense. So anyway, they burned through $8 million a month for years. So it's like we need another 8 mil, folks. We need it. So they whipped through 8 million a month for years, which is probably somewhere along that trajectory my friend bob from france said here i'll give you some mills and isn't he glad he did so most people in the industry figured it just couldn't be done then in july 2019 after this exhaustive trial and error they finally produced a product that worked and it was the first generation wse1 and that was announced in august 2019 and the payoff was that Cerebrus claims its inference platform achieves speeds of 15 times faster than other GPU solutions and on certain workloads far more than that and just in May of last year so 12 months ago independent benchmarkers confirmed Cerebrus beat NVIDIA's Blackwell on Lama 4 inference processing over 2 ,500 tokens per second versus Blackwell's 1 ,000.

13:44Wow. So suddenly. Well, I think that's the headline there, isn't it? That's the headline. That's the headline. There you go. Hold on a minute. We were all looking over at this multi-trillion dollar behemoth called NVIDIA, while these laddies over here with their fancy drill have suddenly lapsed NVIDIA's performance on that. And speed is, to say speed matters enormously in AI is an obvious thing, but it especially matters in AI inference. And when a model is generating a response in real time, latency is everything. And that's Cerebrus' killer advantage. It's now becoming more valuable. So for anyone who's used AI, like, so my own user experience when I've used AI and I've used, I've subscribed to them all.

14:35If I put in a prompt into GPT and I put the same prompt into Claude, which by the way, has a new expression. The first round of AI working, people are like, oh, AI slop. Then along came this thing where people are using multiple AI engines and using them to fence each other. um there was another expression but it was like kind of um well it was like ai contradiction but when i put something into both engines i find claude while its output might be superior to gpt and i'm not so sure it takes way longer and we've we're now in a world where it's like give me what i need right now so anyway this is the the end user advantage so the journey to going public for Zerber.

15:21It's quite interesting. It was anything but a straight line. Yes. It filed for an IP. You know this, don't you? Well, yeah, it sounded kind of mad. They backed off the diving board pretty much. Oh, yeah, you're right. I mean, they filed for IPO in September 2024. And that was the only time I saw their name up until a few weeks ago when Bob said, hey, look at my bank balance. And I didn't even look at them at the time. I just saw them on the NASDAQ. on the nasdaq website there's this conveyor belt of ipos built out month by month and i remember seeing it there but it withdrew its submission after its prospectus was really heavily scrutinized largely due to this massive reliance on a single customer in the uae which is a microsoft a microsoft-backed business uh called g42 uh don't know anything about them but that was that kind of like hold on a minute you've effectively got one customer are you sure you want to do this So the IPO was further delayed due to review by the Committee on Foreign Investment in the United States, CFIUS, if you will.

16:28And it was triggered by a$335 million investment from these guys, G42. And that review dragged on through late 2024 into 2025. And the company refiled in April this year, having diversified its revenue base. Though it is worth noting that the UAE still looms large. Two UAE companies, the Mohammed bin Zayed University of Artificial Intelligence and G42 account for 86 % of the 2025 revenues. And that concentration is clearly a real risk. It's the first thing any serious investor generally scrutinizes. So you and I talk about companies. We do it every week, year in, year out. And very often bargaining power of a buyer is one of those porter seven forces that can make or break a business.

17:19And if you've got two big buyers, it's a little bit discombobulating when you decide I'm going to throw dollars down and invest in it. Especially like as in, you know, the problem, the initial problem was that there was 84 % or whatever from G42. And now it's 86 % from G42 and a university. in the UAE. I imagine that's the same revenue. Yeah. But again, this isn't a company that you're buying for its past results. You're looking at that headline figure, two and a half times the performance of Blackwell chips, and that's what's sold this to the levels it has. Because I think it only did do less than a billion in revenue last year.

18:01Is that correct? Yeah, it did. It did 510 million revenue in 2025, which was up it was up 76 percent on the year before um and it did swing from a net loss of around 841 million uh in 2024 to about 87 million in net net income uh in the last financial year and that path to profitability uh with this breakneck growth rate is what has institutional investors excited so it's a profitable business growing like crazy i would suggest though that they wouldn't care less about profitability in the future they would care much much more about sheer production and scale so if we have this chip this magic little piece of engineering that is two and a half times the industry best the industry dominant nvidia chip we need to make as much as possible as fast as possible and outstrip demand that that is selling all the building and sell it yeah i couldn't agree more i mean like what we're talking about here i think to the lay person like myself on matters to do with chips is could this eclipse nvidia or at least create a competitive threat that's substantial enough to take serious market share from nvidia or even get to a place where nvidia goes we got to buy that thing over there um albeit the complexity that goes with that now that it's a floated entity because on may 14th but we're recording on tuesday 19th of may so just five days ago cerebrus opened for uh trading at 350 bucks a share which was well above its ipo price of 185 uh raising 5.5 billion dollars as i said at the top of the piece and as i also said is the largest tech ipo in years and the stock closed about 311 which was up 60 percent 68 percent on the ipo price but i just noticed as of this morning which is slightly dated when you listen to this podcast uh it kind of has fallen since then um but as you already said market cap is floating around 100 billion dollars and the offering was reportedly oversubscribed by 20 fold yeah and you know that's just not the polite institutional appetite thing um it was real everybody wanted a piece of this and uh and as these things go the ipo minted two new billionaires the uh the guy i mentioned already founding ceo andrew feldman uh whose stake is now worth well over three billion and the cto sean lie at 1.7 billion him so um i think so that's the history until last uh a few days ago five days ago so um but i think where the story gets more moves from interesting to hopefully relevant is by talking about its prospects so i might uh i might give a few minutes on that mike 100 yeah because that that's where the story is that because then you're not buying it is for you're not buying last year's 90 million 90 million dollar profit for 90 billion dollars do you know what i mean like it's it's where it's going for sure that's true so what is the bull case or the the um the story for saying that this is an investable asset now despite now being worth 100 bill straight out of the dust um and the answer i guess it isn't just a hardware company it has diversified through a couple of landmark deals with both Amazon and OpenAI.

21:36And the OpenAI deal is a cloud compute agreement that's worth more than$20 billion running through to 2028. And AWS, Amazon Web Services, is deploying the Cerebrass chips in its data centers for developers running AI models. Meta is using Cerebrass to power the Llama API, as well as Mistral and Perplexity and Mayo Clinic. and the customer list is just growing really fast. They're throwing on a giant respectable mega brand to the list at a rate that's just unbelievable. And the company's cloud revenue, which is customers renting, I guess, compute power by the month or indeed by the year, accounted for 30 % of 2025's revenue, which was about$150 million.

22:26And that was way up from the year before, but everything was way up from the year before but that recurring model if it scales and if it's sticky which i expect it is really totally changes the financial profile of the business now the bear case the argument for saying ah yeah but hold on hold the bus this thing isn't as bright as it sounds it really centers on what we already said the customer concentration competition from nvidia who at this stage and scale and size are the dominant force and the sheer cost of building out a data center infrastructure at scale is absolutely astronomical. And this is a business that's capital intensive in a way that makes software businesses look easy and cheap.

23:10And the bigger picture is that the AI inference market is enormous and only expanding. Speed is the differentiator that matters i think the most um certainly at the model output layer and cerebrass for all the complexity of getting to this point has really built something that nvidia doesn't really have the direct answer to and as i said at the top of the show they turned down nilon musk they spent a decade solving a problem that the industry broadly or modern engineering mind said was impossible and now they're public they're profitable they're partnered with open ai and i would love another thing that begins with p so um potential are they worth 90 ish billion that's the debate um but that is the that is the story of starburst and i think it's it's it's going to be one we all follow over the years yeah i'm very surprised you mentioned they turned down Elon Musk.

24:09I'm very surprised it hasn't turned into a bidding war because this seems exactly what the companies, Nvidia, yes, for sure. But I think there would be a lot of antitrust and buying up a competitor or whatever else. But really, the people who would look at this kind of company would be the company spending hundreds of billions on Nvidia every year and want to cut that out and turn that into a profit center instead of a cost, which would be the hyperscaler. So Google, Amazon, Facebook, Tesla as well. I'm surprised they haven't really come in for a business like this because it sounds like exactly what they need.

24:42I know Amazon and Google are specifically, maybe Facebook too, specifically going out to build their own chips to reduce their reliance on NVIDIA. So it is interesting. Obviously, there is a bit of a drive from the founders to go out in their own way. He's saying they've already turned down Elon Musk. But it's a fascinating business. And I think it's very indicative of just the current market, the strength of the AI trade business that's, what, 10, 11 years old, reaching that kind of valuation so quickly. And I thought it's very interesting as well to just see an IPO of this scale because you haven't for over half a decade.

25:22Do you know what I mean? And it makes me look at the wider IPO market right now because it's just kind of fomenting with potential. This is our second episode about it in a month. We talked about SpaceX, which is lining up next month to be the largest IPO of all time. And then Cerberus comes out completely, smashes all its own expectations. Like it just shows that there's real demand there. There was only 31 tech IPOs in 2025 and there were 121 in 2021. And obviously 2021 was a very frothy time, but like the demand for new companies wasn't there for a long time. Cerberus kind of could be the first domino we see, especially because it's the pure play AI IPO as well.

26:10It gives it a lot of credence to the demand there. I keep saying the word demand. I always do this on the podcast, but it could open up the floodgates. You know what I mean? I'm actually surprised we haven't seen more come to market yet because the popularity of the AI trade right now, the amount of money being thrown at it is just nuts. but then obviously we go down the line and I kind of, you know, you're looking at private AI names, the big two are OpenAI and Anthropic and all the talk is really around their potential IPOs and, you know, Cerberus is very minnow compared to those lads so it got me thinking about it and I think very much the roadmap is on, maybe not as much for uh anthropic but 100 for open ai like the thinking is now so they've just come off the back of this lawsuit with eli musk which which they won which was thrown out you know um um so once they've done with that distraction the focus has to go straight to go in public it's long term it's had a target of q4 this year for the ipo and uh and yeah i i believe that it's time to capitalize on the momentum here because when it comes to IPO timing is everything you know Cerberus absolutely nailed it will open AI um so not and like obviously the timing here as well was it came with roadblocks I suppose not to dwell on the trial but you know it was thrown out due to a statute of limitations and not the actual subject of Musk's claim so basically Musk filed uh the lawsuit too late after open ai became a for-profit business and you know also after he set up his imagine how much money was spent to figure that 100 percent my wife turned to me last night and said how much money did they have to squander to figure that they've broken or whatever the correct term is the statute of limitations exceeded it yeah i mean come on you would think that the an intern in the law office would go read page eight of the 555 000 page manual and go oh it's too late but now marched into the court heaven knows it just it just didn't make any sense no and then even like you know what's funny about this and i always think about this is just the value of time so we've got elon musk sati nadella and sam out and how much time this is taken up out of their lives for a set for a simple like as you said statute of limitations oh yeah sure you could have done this up until 2022 too but you're too late now good luck throw throw the case out we'll see you later you know it was mad stuff um but with that out of the way now we're very close to like you know less than a half a year away for what could be trillion dollar valuation according to wall street i think the last valuation uh opening i got on private markets was 850 mil 850 billion and then you imagine the hype around going public would exceed that.

29:14Now, in fairness, though, there is this kind of cloud around Sam Altman and the trial did not help things at all. He was described as a liar by multiple witnesses and just the trustworthiness of him is really in question. OpenAI is already set for a House Judiciary Committee over his potential conflicts of interest with his personal investments as well. So that's kind of another opportunity to literally and figuratively put him on trial in front of the public, all coming right before this potential blockbuster IPO at the end of the year. But it's weird to kind of speculate over what will happen.

29:59Like that's probably the last obstacle before an IPO can really gain traction. this committee, one that, you know, it's not going to stop OpenAI in its tracks by any means, but it just would further muddy up Altman's reputation, especially in the eyes of the public. I don't think it's going to stop anything. I think there'll most likely be an IPO in late 2026, maybe early 2027. But it will come along with this kind of growing traction against Altman and against OpenAI as a company because he's so representative of it as a whole. He is one of these totemic founders like Musk, like Zuckerberg, that becomes very much his personal reputation is tied up within OpenAI's reputation.

30:46So it's interesting. It's not the clearest path to IPO compared to other companies. And we're talking about SpaceX. That doesn't have it either with this floating around with XAI. but then much cleaner is Anthropic who's kind of avoided all the mudslinging and becomes more and more attractive within the context of OpenAI and Altman's deterioration, I suppose just reputation-wise at least and then obviously it's just performing like absolute gangbusters as well both its model is outperforming kind of OpenAI and then the financials, it's been on absolute fire its 2026 revenues are expected to pass 45 billion which would be up fivefold from last year's nine billion and they're raising money like crazy just three months ago they took in 30 billion dollars at a 350 billion dollar valuation and then the week just gone they raised another 30 billion at a 900 billion dollar valuation so casual you know what's that 550 billion dollars in the space of three months you do the maths there it's about what like uh two billion twenty billion dollars a day give or take a day yeah not bad um not bad at all and it gives an gives us an idea of scale as well when these massive companies do come to market so that's 10 times the size of services ipo in the space of three months and two fundraisers and that was the largest ipo since 2019 so it is fairly nuts when like we're talking about the size of servers and then we're like but wait till these lads come wait till open ai comes wait till anthropic comes wait till spacex comes now with anthropic there is less talk of potential ai there's no real comments from management they haven't come out with anything official of a long-term target of when to go public and then they've just raised 60 billion dollars worth of cash so the need isn't as present even though they're obviously spending money like crazy one can only imagine uh but when it does come is sure to be immense because because they're coming in the shadow and this is going to be kind of this this this troika of just scale uh that we've never seen before in the public markets in terms of ipos because they're coming in the shadow of the impending spacex ipo which we talked about we did a full episode on last month uh reports are now coming out that they're targeting a june 12th date it's looking to raise 80 billion dollars at a 1.7 trillion dollar valuation so what's that it's geez 16 times what um what servers just did it's just the numbers are no way comparable to anything we've seen before and they're going to come out one after the other after the other i'd say i'd say it'll be about six months maybe between three maybe that's early uh early targeting for Anthropic.

33:39But yeah, with that day targeted, we should see a prospectus out very soon. Most likely knowing our luck with this podcast between now and the time we record. But yeah, we've never seen an IPO of that kind of scale. And then to just kind of have three of them, not in a row, obviously, and I'm doing a lot of speculating, which I said I wouldn't do about the Anthropic IPO, but to see the three of them coming one after another to the public markets, it's pretty nuts like um but yeah it's interesting with anthropic and open ai open especially coming in spacex's wake because there's obviously that underlying tension between musk and allman who clearly hate each other um yeah so to kind of have the two rivaling ipos happen in the same year as well is uh it's all a bit mad like um but yeah we'll finally see if the demand is there for this kind of gargantuan IPO because you'd have to think yes you know if Cerberus was 20 times oversubscribed the institutional demand is there we know the retail demand is going to be there anything with Elon Musk is going to be there in terms of retail demand and then just secondary market demand as well so yeah it's it's nuts we won't really see what's happening until later in the year SpaceX will be the the kickoff I suppose but it could be the largest year for IPOs we've ever seen on public markets and just the amount of money being raised the amount of money thrown at this ai trade the space trade as well uh it really is uh maybe like nothing we've ever seen before so yeah those those three companies the public markets and it's also it's a shame as well i'm repeating points i've already made on this but the fact that these companies are doing while they're growing on the private markets and their returns are exclusively ring-fenced to institutional investors who are worth you know at the very least over 200k and have experience and blah blah blah it is taking money out of retail investors so yeah that's another point and just a mini rant but just the fact that these companies are growing so big on public market on private markets and then almost the the elite institutional investors are using retail investors exit liquidity kind of doesn't sit well with me really but yeah it is it is going to be an absolutely nuts IPO market going forward.

36:01Well said and at the very least for all these trillions of dollars flowing from one hand into another in the imminent IPOs it's going to give us a lot of talk about on Stock Club because every week you and I slack each other and go right what we talk about this week yeah i don't know pinky what do you reckon brain so where is so it's so good there there's a few podcast uh podcasts coming folks stay tuned yeah um that was very interesting mike uh come here trivia so if you had to work as the right hand man of sam altman or elon musk where you shadow them non-stop for three years who would you choose jesus christ you can't choose him sorry oh i don't know neither of them sound very um i know but if you had to come on mike maybe i'd go with altman i guess um just out of morbid curiosity i must I know less easy you almost could have you setting up your camp bed in the office like good luck no problem no problem slept in worse places than his office floor right are we anything else uh are we finishing up there asher look we'll finish up there Mike oh well I do a following pro I know I have nothing to say about profit I tell you I do I have something I can say about profit the rebalance is happening tomorrow, Friday 22nd of May.

37:34The results of profit are absolutely astronomically good, and it is going to possibly sell one, two, three, four of the positions and replace it with a corresponding number of buys. It might not sell anything, but the one thing profit does is it outperforms the market most of the time. So if you want a deal, well, I don't even know, before we increase the price to four thousand dollars a year which you would probably most people have to chip in with family to buy because the cost of our processors are going up if you want a deal email frank at my wall street.com um yeah there you go oh yeah and and also buy tickets for investicon and also post us some money so it costs a lot of money our ipo our trillion dollar IPO is not coming up um not yet there you go yes not yet but there's still time I saw a comment on one of the episodes saying is Frank a real person and unfortunately he very much is uh we know we know to uh but uh just to us every name we use is real Frank yeah Frank is a real person Frank is a marketing genius who built and sold a couple of businesses um one to Google and another to giant telco and he's our marketing cmo and he has a team but we always send it to frank because none of us try and interpret what in fact makes for um a acceptable discount so yeah frank's the man frank and also he has an email frank the tank at my wall street.com but i think he only uses frank at my wall street.com there you go on that on that note i think we'll finish up frank does exist uh i couldn't have made that up on the spot i probably could but i didn't that's true Emmett, thank you for joining me and thank you everyone for listening.

39:27We'll talk to you next week.