E226: Revolut $75b valuation after 72% yoy revenue growth, 35 million customers; AI App investment cases - Sierra AI Agents and OpenEvidence

0:00Clint Sorenson:All right, gentlemen, three super interesting topics today. Actually, well, maybe we'll talk about one of them in kind of by itself, but then the other two we can perhaps join together. So first, we're going to hit on Revolut. So they just did like a tender round, okay,$75 billion valuation. These guys are growing fast, man, up 35 % year over year, active customers at$35 million now. They did$4 billion in revenue in 2024, which was a 72 % year-over-year revenue increase. And then$5.9 billion forecasted for 25 and$9.3 billion forecasted for 26. So this is a 12.7x revenue multiple if you apply that$75 billion on the$5.9 billion for the 25 revenue forecast.

0:52Clint Sorenson:So the Revoluta FinTech company, right? They do online bank accounts in Europe primarily, but they are here in the United States. So maybe, Evan, I'll start with you, man, because I know you love fintech. So what do you think about Revolut? Is this something you like this valuation? Is it too big? What do you think? We've talked about Revolut before. Oh, yeah, lots of times. As I say, I think right now, this valuation and this round makes quasi-sense because like all fintech valuations, they are cyclical based on the market dynamics outside the macro dynamics primarily. And so look, when people are investing, when people feel wealthy, when there's dollars flowing through the ecosystem, fintechs look great.

1:38They see lots of volume. They see lots of investing. They see lots of consumer sentiment. That's all good for fintech. We had that sentiment for about 18 months now. Will we have that for the next 18 months? Remains to be seen. When we start seeing a downturn in the macro, I expect to see a downturn in fintech valuations. I do think 12.7x value is totally not outrageous by any means. But these fintech companies, these are really hard businesses to scale. We're talking pennies on the dollar that you need on every transaction that you earn. You really have to have a massive amount of scale to generate that kind of revenue.

2:16They're clearly doing it. I look at, I think a very good call for them from the global perspective is the neobank we see down in Brazil that is crushing it right now. They continue to just make tons of money, grow their customer base, and they're just redefining the way FinTech works abroad. And so I think there's a nice playbook for Revolut to follow that. So if you're a buyer in that thesis, I think it's pretty interesting. But here in the United States where they're trying to penetrate, they're struggling. It's hard. There's a lot of competition here.

2:44Clint Sorenson:Right, right. So Clint, some of the new news that came out, because Evan's right, we've talked about Revolut in this round specifically a couple weeks ago, right? But this round, the new news that came out this week is that they were doing some internal workings with this tender offer so that they could stay private for longer, right? So here's another. It's a$75 billion company. Now they're signaling that they want to stay private for longer. I mean, what do you think of this trend? And it seems like these tenders, these kind of structured tenders or like periodic tenders are happening more frequently now with these larger companies.

3:21Clint Sorenson:And also the CEO teams are kind of like signaling to the market there's no desire to go public, right? Like as an investor, like what do you think of that? Is it okay that these companies have no real plans to IPO and they have such a big valuation?

3:35Nick Fusco:Yeah, I mean, on the valuation point, I don't understand any of the math anymore. I mean, I got pitched a deal, a quantum computing deal trading at 300 times sales last week. So I just, I don't even understand the math anymore. That's very high. But no, I mean, because if you do the math on 12 times sales, that means they got to pay me 100 % of their profits. At least they have revenue, right? Yeah, for 12 years. Exactly. 100 % of their profits for 12 years without any kind of R &D or expense. And I understand the market's rewarding growth. They had enormous re-user growth. You cannot knock the business model.

4:10Nick Fusco:And so taking the valuations out of the equation, the company seems to be performing very, very well. And I think the market's rewarding that and liquidity is there for companies that can perform. And so if the liquidity is there for the companies that can perform, why go public? Right. Unless you need to do a massive capital raise or it's just the company is better suited for the public markets. There's a lot of cost. There's a lot of red tape. There's a lot of bureaucracy. There's a lot of extra expense and regulation that goes into going public. And if you can give liquidity to your employees, I think I saw on this one, it was like up to 20 % employees could sell in the tender offer.

4:48Nick Fusco:I think that's a great deal for employees, giving them some liquidity. And hey, I'm all for it. I'm all for companies paying private as long as possible, private over public.

4:58Clint Sorenson:Okay. That's interesting. So, Nick, let me ask you this. I mean, structured tenders are becoming a real thing, right? I mean, now it looks like Revolut's going to start doing tenders for their employees. SpaceX does a tender every six months, right? Open AI's done a lot of tenders. Who else? Stripe is doing tenders, right? I mean, like, this is like Canva, I think, just did a tender, right? I mean, this is like becoming a thing where companies are doing tenders every year, every six months, or at least on some kind of like periodic basis. or natural basis so people can get liquidity employee and early investors.

5:38Clint Sorenson:Is this healthy? Do you think this tender is healthy? In my opinion, yeah. Do you think it should be more structured? What do you think about all this stuff? Go ahead.

5:48Nick Fusco:I think when you're able to serve up something so consistently as, let's say, SpaceX is the best example out there, every six months, the employees can count on it. Then they're not out there fishing around in the secondary market on a bespoke or like an a la carte basis for individual lots. They kind of know what they're getting into. And for the company, it's assurance to their shareholders, which are also employees in a lot of cases for these tenders, that they're going to stand by that and keep doing it. And it doesn't get you nervous about holding those shares for an extended period of time.

6:25Nick Fusco:So it sets your duration a whole lot longer for being with the company, I think. And the odd part out is, and I'm happy you tossed me this one, was when I was at market that became IHS market, a big part of them not going public was that they wanted to be big enough to not be pushed around within the public market as well. So yeah, they could have access to better debt facilities and things like that, but they wouldn't be gobbled up immediately. And the fun fact there is I was there for about 13 years, and I think 11 or 12 of them, We had annual tender offers and it was great. It's not like everybody ran for the exit.

7:05Nick Fusco:It was, hey, I needed to put a down payment on a house or something like that. Let's take some off the table. But everyone believed in the company, believed in where it was going. And when the management sees that and identifies with all the staff, I think it's a really symbiotic relationship. and it keeps ticking a nice trend line of where the valuation should be and you get to feel it out. So I won't wax on too much, but I think having tender offers periodically with some regularity is a really strong signal from a private company.

7:40Clint Sorenson:Yes. I mean, that's the piece I think that's important for both employees and investors. Because if you know that the tender's coming, right, then you can plan for it. And it also too, I think as an investor, like if you're an employee, that's just great. It's like now it becomes part of your compensation model to like get free option to pull the trigger. You can decide if you like the valuation or not. Do you need the money or not? Like whatever, right? If you're an investor, it's also kind of nice because you get like a proper mark on your position, like an institutional mark on the position, right?

8:12Clint Sorenson:You kind of know what it's worth. I can see from like the CEO's perspective and the executive team's perspective, you lose a little bit of control, I suppose, right? There's two ways to look at it. One is you get the transparency because you know now what your company's worth. That's, I think, interesting, right? You can think about what it's worth or maybe have conversations about it. But when money starts trading hands, that's what it's worth, right? That's right. That's interesting. You also lose a little bit of control, too, especially if it's structured. But I mean, look, I got to tell you, if you're a$50 billion plus company, even$10 billion plus company, I mean, come on.

8:50Clint Sorenson:That's like a huge bit. The cutoff from small cap to large cap is$4 billion, right? If these companies were public, they'd be in the S &P 500. Like, these are huge companies. Like, you got to have liquidity. You got to have some kind of liquidity. I think you just have to. I don't know if you have a choice if you're like a CEO of a company and you want to stay public or you got an IPO, right? My only argument against that is that I do think that the companies do need a healthy revenue stream and distribution. Yeah, fair. Or else you're running the name of the name. situation where you have companies that are employees that are joining they've been there for less than 18 months and you've made them multi-millionaires and now they have little incentive to sit around and now you're trying to figure out how to retain them and you're listening evan when you start when you so then is it fair to say when you start your like structured tender program you need to have like confidence as a ceo or executive team that you're going to be able to like you know the revenue train the revenue growth like for this foreseeable future, it's going to be good.

9:52Clint Sorenson:It's only entertaining talent.

9:55Nick Fusco:Absolutely. The brain drain at some of the fallout companies, the big AI companies, is going to happen just because Mark Zuckerberg wants to step in and offer people$50 million. I don't know that a tender is going to defeat the purpose there. It's just a factor in the whole idea. But also knowing that I could take some value and go, it is a little, I don't know. I think it works equally well as it does as a detriment. Because if I was at OpenAI, I was like, what are they ever going to do anything with this weird structure of profit, non-profit? And then Mark Zuckerberg steps up with a paycheck, I'm just going to walk anyways, potentially.

10:38Nick Fusco:Right. Yeah, fair. It's a tough call.

10:39Clint Sorenson:No, but I could see just like you wouldn't want to IPO into like a shit market, right? Or if your company's not doing well, you wouldn't want to IPO. I mean, companies like prepare themselves for the IPO.

10:49Nick Fusco:Yeah, there's like a dollar-crossed average component you could do now.

10:52Clint Sorenson:It makes sense that you would start a structured tender when it makes sense for your business, right? So like, you know, I think things are going to be good for the next couple of years. So we'll start the structured tender process. That way, you know, it's not going to be like a net negative, you know, for us. So, okay, that's good. Guys, let's talk about AI apps. All right? So two AI. So you guys know that I've been on the fence on AI apps because I just feel like it's too early, right? And I think everyone kind of feels that way too based on our prior conversations. But there's a couple of them.

11:25Nick Fusco:You and Evan have been the most positive on them for sure.

11:27Clint Sorenson:Well, I mean, I think in the long run they're going to be huge, but I just don't know where things start and stop. But some of the stuff we're maybe getting, yeah, we're getting some glimpses of clarity, right, I suppose. So there's two companies that hit the radar this week that I think are interesting. One is Sierra, which is an AI agent. And Brett Taylor is the CEO of that company. He's the chairman of OpenAI. And this dude was like, he was everywhere. He's like, he's the best. The Silicon Valley, you know, golden child. And then OpenEvidence is a medical, like a medical large language model solution.

12:04Clint Sorenson:And like 40 % of US doctors use OpenEvidence. That's a huge amount of market share. Okay. So here's the difference though. So Sierra is one of these companies. So they did a$10 billion valuation. Okay. And they got like, you know, Green Oaks, Iconic, Sequoia benchmark. That's like, you know, the who's who. And then Open Evidence did around at 6 billion and they have Kleiner Perkins and Google came in. Right. So, but here's the thing. Sierra is using OpenAI, Anthropic and Meta. Okay. So it's sitting on top of those companies. Now, you could argue that maybe they have like kind of a consultative approach with their clients and they have some like incredible clients at Sierra.

12:47Clint Sorenson:Some like really great brands are using their AI. It's AI customer service reps, effectively, AI agents. So that's interesting. Okay. And we can like, but I want to talk about that with you guys. Open Evidence has their own large language models. So that's like an app, right? I view Open Evidence as an app. I view Sierra AI as an app, right? But Open Evidence is building their own models. And I got into that because everyone's spending billions of dollars on models, right? They have a different approach. It's kind of like a deep seek approach. They have different models for different specific use cases.

13:24Clint Sorenson:They're training things. So it's almost like they knew going in they were going to solve for one thing, right? And then they were able to box it in and spend a lot less money training that model, right? because it was so laser focused on doctors, right? And medical diagnoses. So what are your thoughts? Just maybe an update, Clint. Maybe we start with you, man. Like, what do you think just added on its face about AI apps? Is it still too early? Is there like this? The reason I brought these two companies up is Sierra is doing kind of what we've seen in the past. They're building on top of the large language models.

14:00Clint Sorenson:Open Evidence is doing it on their own model. like is that what do you think of this are you leaning towards one way or the other yeah i mean

14:10Nick Fusco:i'm i'm more of a fan it's hard on a blanket statement to just say open you know uh apps on ai are not good or or they are good yeah um i've always thought there's some disruptive use cases for ai and it looks like open evidence is more of the direction that i would lean just because there's a moat there. There's an established moat. They have a proprietary LLM or model, whatever, um, that they're using and training on a specific use case. They're super targeted and focused. And that's kind of the way I see the evolution because again, all of this is going to come down to data at some point. And that's where the war is going to be fought.

14:47Nick Fusco:When you start to get close, when you start to have all these open source, now open source businesses start to close the gates because they're worried about competitive pressures because everything's race to zero. I think that that's when you're going to realize who's swimming with a bathing suit on. And I think these, borrow Warren Buffett's term, but as I believe it's going to be the ones with proprietary models, right? That are super targeted, like legal, accounting. I can see a financial advisory one that's just trained on that data source only. They have their own data. They They've built the data infrastructure necessary.

15:27Nick Fusco:They've built the model on that, and they're running full speed on that one thing.

15:32Clint Sorenson:Okay, so let me ask a more, like, pointed question. So you could build on a proprietary – is your thing more about the proprietary data set, or is it more about, like, the actual model itself? So, like, you could take an open source model, like a Metalama, right, and train that on a proprietary data set. You can even do that with like OpenAI, XAI or Anthropic, but then they kind of like own your data. Then they own it, right? Yeah. Or is it like, hey, I need to own my own data and own my own model.

16:03Nick Fusco:I think you have to own your own data and own your own model on that data because ultimately, at the end of the day, what's going to stop these larger platforms with more capital from just disrupting you in their next release?

16:16Clint Sorenson:Got it. Got it. Okay. That's very interesting. That's very interesting. Okay, Nick, let me ask you this, man. Do you think that - Now that does not - Hold on one second. That might not apply to Sierra

16:26Nick Fusco:because the CEO is the chairman of OpenAI, so maybe they have a thing going on. Yeah, I understand. But in general case, yeah.

16:35Clint Sorenson:Yeah, okay. So Nick, let me ask you this thing because I'm seeing this in a couple of these different AI app companies. So Sierra, my understanding with Sierra is they kind of like parachute into a company and then they build out like a solution, right, with their customer. It's not like they just, it's a SaaS business where you like go to SierraAIagent.com and you like sign up and now you're ready to go and you kind of do some tweaks or whatever and you're ready to go. Like they got to like do work with you to make this thing work. You know what I mean? It's a consultative approach. It's very much like an enterprise sales type effort.

17:11Clint Sorenson:Is that a moat? Do you view that as a moat, right? That like kind of deeper connectivity or do you think, go ahead, what do you think?

17:22Nick Fusco:Yeah, yeah, yeah, absolutely. If it's something that you're internal to the company and you have their proprietary workflow or data as Clint sitting on the data piece, I wholeheartedly think that's a key benefit because if you were not doing that, and let's just use an example of, I had lunch with someone at the big four this week and they were saying, yeah, the challenges of going public, Like we just charge so much more for very similar work. And I was like, oh my God, okay. But let's say you parachuted one of these agents into a big four and you wanted to shift him to the next one and the next one and the next one.

17:59Nick Fusco:There are absolutely nuances. Some are better at media. Some are better at fintech. Some are better at different areas of the market, consulting, the accounting, auditing, what have you. But your moat just kind of dissipates. if you're able to translate the same work across a different firm. So I think, yeah, if you can keep everybody's eyes out of looking into what these agents are able to learn, it's like a really tight IP protection on your staff. When you have somebody join your company, they can have two weeks and go to the direct competitor. Or if you work for a quantitative hedge fund, sometimes it's like a year, whatever it is, right?