On The Brink with Castle Island - Felipe Montealegre and Noah Goldberg (Theia Capital) on Investing in Internet Capital Markets (EP.668)

Episode Date: September 22, 2025

Wyatt sits down with Felipe Montealegre and Noah Goldberg of Theia Capital, a fund investing actively in liquid crypto tokens with strong fundamentals. In this episode: What drives crypto market valuations? Fundamentals or supply of capital? The imbalance between venture vs hedge fund activity in crypto markets Navigating a market where assets become liquid early in their lifecycle Talent in the crypto industry What makes exceptional founders and team members? Using LLMs in investing and general workflow Future themes Theia is excited about

Today, I sat down with Felipe and Noah of Thea Capital. Thea is a crypto investment fund focused on investing in liquid tokens over long-time horizons. I've had the pleasure of knowing the Thea team since they got started in early 2022, and I think of them as without a doubt some of the best in the business. Our discussion touches on identifying opportunities among a wide frontier, crypto's institutional capital base, and the extent of venture activity over hedge funds, finding the best founders, and exciting themes going forward. I hope you enjoy our conversation.

Matt Walsh and Nick Carter are partners at Castle Island Ventures. All of these expressed by them or the guests on this podcast are solely their opinions and do not reflect the opinions of Castle Island Ventures. Guests and host may maintain positions in the assets discussed in this podcast. You should not treat any opinion expressed by anyone on this podcast as a specific inducement to make a particular investment or follow a particular strategy, but only as an expression of their personal opinion. This podcast is for informational purposes only.

Brought down by bad mortgage investments, Lehman, which has 25,000 employees will be liquidated. The federal government loans of America. American International Group, AIG, $85 billion. This is a different kind of market, and the Fed is asleep. The federal government is stepping it to stabilize Fannie Mae and Freddie Mac, the two mortgage giants that have been threatened by the housing crisis. The Bank of England has pumped 75 billion pounds more into Britain's ailing economy with a new round of quantitative easing. You print a couple trillion dollars, and all of a sudden people start to worry.

So out of this worry, we have something called the Bitcoin. Bitcoin. Felipe and Noah from Dea Capital. Thanks for coming on the podcast today. I'll start off by saying, I've an immense respect for you guys and what you built. I'm looking forward to having this conversation. I think you guys have built an exceptional reputation, the crypto ecosystem, and excited to hear your latest insights with that in mind. I'll start off with, I think there's a lot of noise in the crypto and crypto asset space right now. Digital asset treasuries, people are talking a lot about that. I think people are trying to assess what has value in the frontier of tokens, especially when we're going to see more on-chain assets than your future. So I wanted to know what do you make of that current landscape? And in that wide range, how do you guys find qualified opportunities and what makes a qualified opportunity for you?

Why, thanks for having this on. Again, we echo the sentiment. We have immense respect for you and the Castellan team. So thank you very much. Yeah, on that question, I'll take a cut. I don't know, has a lot of thoughts on this. But we've been investing in the market for about four years now. And one of the nice things about the crypto market is you have to really speed run a career

in investing in terms of your emotional control and like ability to interact with volatile markets. So when I was doing private equity investing, I would read all these buffeted letters where he talked about the one or two years where prices were above his, really far above his estimates, what that did to him emotionally and I'd read that from other investors too. But it's never really happened in my time in the industry. Things are more or less sustained year in year out. In crypto, you have massive volatility. So you get to experience the full range of deep fear when Moon is collapsing and GPTC is trading 30% under and also see people talking about AI agents and wonder to yourself, is this a big thing that I'm missing? And you get to do that multiple times in a year.

I think right now we're in a frothier period. That seems to be the case across all financial markets. It didn't start with us, which is a little different from the other ones. Work more of an offshoot of what's happening in broader financial markets and equities, gold. Today, the Fed might be lowering rates into an all-time high market. We don't know the backdrop. But it does seem like most tokens we look at today are overvalued based on our estimate of going forward fundamentals and growth. And that's not the majority.

That's just most things that we're looking at are passes today. We don't take a view on markets. We don't time markets. We don't try to predict where they'll go. But we do try to keep our underwriting box constant overtime. So we just naturally find ourselves buying a lot more when markets are going down or depressed because a lot more things just fall into our box of good value. And we find ourselves passing a lot more in times like this

when there's just fewer tokens that fit into our box. No, I want to hear your view as well. But Felipe, just on that, I wanted to say quickly, I completely agree that you see more of these, what I'd call market regimes and crypto. But I'd also even just say in this period called the last five or six years with the crypto market's actually been mature, it's been a tumultuous market even from a traditional landscape.

you're a bit older than I am, but I started following markets closely in probably the late 2010s. And you look at like 2016, 17, 18, 19 is a lot of the same. But then you look at like 20, 21, 22, 24. You've seen four or five different markets in a period of this many years, which is very abnormal in traditional markets. But Noah, please, I want to hear your views as well. I agree with Felipe sentiment on the market and that it's generally the opportunities are abreastly priced. And I think part of that, if I were to steal man, why the market has coalesced on this aside from broader financial market dynamics is that KPIs on chain have generally

inflected quite positively. And also there's a regulatory tailwind for our market. So, for example, Athena, USDA outstanding, is grown from maybe $5 billion to $13 billion over the past three to six months in Abe has all-time high barrow activity. And so you see KPI is moving upwards, but simultaneously, I think that given that the base asset prices in our markets play such a large role in the on-chain metrics for financial applications, you have to attempt to have an opinion on what's driving those prices. And as you alluded to earlier, I think that the digital asset the treasuries are quite material as to how markets are moving. So for most of 2024 and the start of 2025,

Microstrategy was dominant digital asset treasury strategy, perhaps not fully correlated, but Bitcoin outperformed E-thensel, and a lot of the flows into Bitcoin were coming from Micro Strategy. And then when Microstrategy started issuing many different instruments and its premiums and have started to compress, we saw weak in BTC and simultaneously we saw Dats emerge for Ethan's soul. And so there were net positive flows coming from Dats into Ethan Sol and perhaps a bit of a headwin in the short term to BTC.

So I think that probably is a large driver why Ethan Sol have outperformed in the short term. And given that Ethan Sol are the largest collateral assets in lending markets and the basis rate, for example, plays a large component and yields on chain. There are these reflexive dynamics. I think what I'm personally wary of is that while it's hard to properly ascribe fair value to ETH and Seoul, given that they're not necessarily cash flow-oriented assets, it's hard to model out. You have to wonder whether the current prices are sustainable if the debt buying pressure subsides

in the short term. So I think that's all their dynamic play as far as you have the KPIs inflecting upwards. Part of it is tailwinds in our industry. Part of it is flows into the base assets. And those drivers of flows probably are in the midterm. sustainable. So I think there's some risk on Ethan's sole prices and as a byproduct, a lot of the defy projects that use these assets as their primary generators of revenue. Why, if you don't mind, I double click on this question a little more. There's a lot here. I think that the other component

that makes this a tough question right now. And since the election is, I think our thesis at the idea is that we are in the early stages of S-Crib adoption for the broader technology, the internet finance technology. So I think in our view, in 15 years, years, a lot of the world will run on chain, whether it's in lending markets, trading, remittances, stable coins, what you will. And that's obviously, those are really, really big numbers. So when you're in an early part of an S curve, you shouldn't be too spooked about something double in or tripling. It can actually grow 100x from the base. That is the case, again, for some of these base markets like stable coins, real world assets, lending on chain.

And we now have the green light from the regulatory regime in the United States to embark on that. It is kind of a clean cut with a situation a year ago where you just were now allowed to build these products. The tricky part is that there has been a lot of money lost in liquid markets trying to get the beginning of an S-curve that wasn't there. I think we've all seen this, narratives where the world is going to change. I mentioned the AI agents as an example. Honestly, defy in 2022 was that super cycle S-curve that didn't pan out.

Every new L-1 is the beginning of a new S-curve of adoption. People have really lost a lot of money in liquid markets trying to play that game. But again, you need to be right about it because you lose a lot of money if you're wrong. Things re-rate fast and the liquidity is gone on the way down and you can make a lot of money if you're right. So I think we try to really balance out the two forces of understanding. Look, there is this regulatory barrier that's been lifted and we have this huge opportunity ahead of us.

And there are some big winners that can take advantage of that. I will discuss some of that in this podcast. I mean, the names that you met, we've discussed in the past like Maple Oiler, Morpho Camino. These can be big winners over time. but you don't want to get it wrong. You don't want to mistake reflexive activity. Soaking going up, incentive rewards, driving KPIs for real growth. Because again, that can go away in an instant and liquidity is gone when the paradigm shifts.

I think that's a great point. Things move fast in this industry naturally and it can go very well and it can go very poorly. Felipe, to your point, I completely agree with the perspective that something like a 2 to 3x outcome can happen quickly. and that's not even a sign of a rationality necessarily. It's just that we're inflecting quickly. And also because of the early liquidity behind a lot of these assets, you have market discovery faster. And that's no different than the sorts of valuation step-ups you saw in early software, you saw on early internet. So I think that makes a lot of sense. I wanted to pull on a few different threads that you guys threw out already because I think there were several points

worth double-clicking on. One, as we're talking about the way that this market behaves and token values widely, and Noah, you touched on this, so maybe I'll ask you to circle back. My viewpoint, you can look at equity markets, you can also look at token markets, is that there are two very broad ways to view what these things should collectively be valued. One is a little bit more bottom-up where you're assessing what a business should be worth. You're potentially applying a multiple. Whether that's value or growth, you're trying to determine its own value or relative value. And the other philosophy is a bit of more the top-down, like market liquidity, which is how much capital is falling into this market. Stan Drucken Miller takes a bit of this kind of

approach. And relatedly, you'll have these massive asset managers. Crypto has their own. They're just smaller scale that are assessing whether they want to be in government debt or whether they want to be in equities or whether they want to be in other sorts of loan financing. And I'd say this market-wide liquidity analysis is more of that ilk. It's looking at where is that capital going to gravitate and how will that impact the valuations of companies today. So how do you guys weigh these two philosophies of saying whether crypto protocols should be valued according to their own fundamentals versus as something of a product of the amount of liquidity or capital in the crypto market? I think that the two ways that you're framing things are correlated

in that I think that liquidity in an asset class is partially a function of the attractiveness of the opportunity, and that is based on bottoms up analysis. So if a market has a lot of growth prospects and trades relatively cheaply, I'd imagine that more capital will flow into allocator's hands and flow into these assets. I think that as far as framing of pure valuation, personally I scribe to the view that the only way to value an asset is based on future cash flow is discounted to the present. And I think there are ways to assess what an asset can trade at any given moment. To your point, I think liquidity plays a large role in how assets us trade.

And I think that there's a fair amount of alpha in having a general view on where flows are going to move. But that's generally a game that they have chosen not to focus too much more time on. Yeah, to pick, I think you'd pick one if you're building a firm. Are you trading flows or are you buy a whole firm looking for fundamentals? and we're firmly in the latter camp. We're always underwriting three-year models. We hope that we can hold companies for much longer than three years to continue to execute. I think whenever we get into a new investment,

our hope is that we're looking at something that can, we can grow for 10 years with the management team as they take on a really big market. And that's been the way that a lot of funds have made great returns over time. I would argue that the base rate of doing well as a fund is much higher if you're doing long-term fundamentals of investing than if you're doing flows. it's very easy to blow up trading flows. People have done it well, drug another to your point.

It's a small minority of superjudices that have done that well, whereas the reality is if you go to any good market P firm, any decent VC firm, you see a lot of people who have done well playing the game that we're playing, which is picking good companies,

good management teams, good markets, holding for the duration of the growth period. And you guys, founders will speak very highly of you for this, rightfully. So you're very involved, typically, with companies and management I don't know if I want to say management companies, management of protocols or of projects.

Why have you taken that approach besides it being a creative to the investments that you guys make? And do you see others in the market doing that? Did you see a gap in doing so? Or how do you look at that sort of investor behavior, a little bit akin to some of those strategies you mentioned, Felipe, of being middle market or working with folks in a traditional finance sense? Yeah, we, I think one, that was both of our backgrounds. So by background with mid-market private equity, you worked very close to with companies.

In that case, you'd own 100% of the company. So basically, you're getting every dollar recreation. So it's a strong incentive for the sponsor to work with a portfolio company. Noah worked in venture capital. I'll let him expand on it, but also work closely with companies. But we also just really saw a big gap. We're a liquid fund. Usually we're buying tokens where unlocks have starter or well underway.

It's not a hard rule, but that just tends to be where we play. And what we notice is that the VC ecosystem was very well served. in the market. When you're private still, you have to launch a token. There are a lot of people in crypto who are willing to talk to you and give you advice and a lot of very high quality VC funds. While once the token launches, we've noticed that most of those VC funds move on to the next batch of companies and these management teams that are now quite experienced, know the markets well, don't have anyone to really speak to, or maybe they'll have one or two people to speak to on a routine basis and just kind of showing up as they are saying, we're happy to just be helpful,

help with blocking, tapping on, recruiting, structuring, thinking through market dynamics, sharing analysis, and in exchange, let us help you think through strategy questions, was an appealing value proposition for a lot of these liquid-turbing founders who just didn't have them to speak to anymore. And for us, it's really valuable for two reasons. One is we run a pretty concentrated portfolio. So any one time, there are, try to run between 10 and 15 companies, but at any one time, they're like seven to 10 companies that are really going to drive our return for the next 12 months just because of the way sizing shapes up or growth your entries shape up. So helping those companies at all really makes it material impact

in our portfolio if it can be helpful. And the second is just we get really valuable market color from these relationships. Because usually where we're playing the 100 to billion dollar token range, like the market leader in a segment, you're talking to the best founder in that segment, usually, or a top three founder in that segment. And if you speak to them week after week, see what they're concerned about, see what they're thinking about. You really get a lot of valuable contexts that you just can't get any other way. So I think for us, I'm very happy that our portfolio companies are happy that we're involved, but we're really getting a ton of value from it. It's something that they're giving us more than we're giving them. So I appreciate them

taking a time with us, really. Yeah, I guess I'll add a bit. I think that there's a lot of, to Filippe's point, brand and network value that come from being supportive to portfolio companies and that I think that although we play in liquid markets, a lot of the time we need the opt-in of management teams to want our support and be willing to hear our feedback. And additionally, a lot of this industry is network-driven. there's very few major players in the space.

Having relationships to these players and them feeling confident that when you make a recommendation to work with a specific service provider to take that introduction, I think is quite helpful. And additionally, when we're trying to source potential investments, having founders want to speak with us and be on their cap table is something that's accrued to the fund, generally speaking. I agree with a lot of what you guys have said, frankly. One point I wanted to come back to was, I agree, Felipe, with your characterization, I'd almost call it the over-venturization of crypto. Everything's a venture fund. And I think this stems, frankly, to the appetite of capital from the outside looking in, where allocators are comfortable with a venture structure because you're shielded from the direct volatility a little bit. You only have to believe that crypto will be sizable in the future versus now.

But I'm curious, my point there would be, from your guys' view, I agree that valuations are still pretty expensive across crypto liquid markets. You can only infer what that means for the private markets when we're overvalued in the liquid markets and we have a disproportionate of venture participation. That means we have to be massively overvalued in some forms on the venture side. And with that in mind, most of the capital being on the venture side, who do you guys look at as the downstream investors of what you're doing, especially when we have this very disproportionate capital stack where most of the money is in venture? Important question. And I think that's some level. That's the heart of the fundamental

strategy. We consider ourselves to be the holder of these companies and our exit liquidity, so to speak, is happy customers paying for a service, providing revenue to the business. And that model is not innovative. It works well. Look at any. company on the purchase balance sheet. Buffett is not selling to an ex liquidity, larger investor. When you buy Coca-Cola, you are getting your ex-liquity from the customers who pay Coca-Cola and you get that money back as dividends. Same goes to Apple. Same goes with any major company that's become too large to have your buyer on the back end. And that's why we're so focused on revenue, cost, structure, cash flows. Because there is no right now, there is no downstream investor

to buy our tokens on a story. We need the customers to buy the products of our companies. to give us cash, to give us cash flows, to support our valuations, to give us an exit. So that's really the heart of it. I think this idea of exit liquidity has been a huge problem for the industry, in my view. It's created this haze of just greater fool theory everywhere. How can I spit up a narrative to sell this to some outside investor? When is retail coming? You hear that all the time. When is retail here? When is retail coming? That's basically asking, when will greater fools come and buy my tokens that have no value for me. And we just refuse to that game.

We know that they're not. We know at our size, at our size, and where we play in the timeline, it's not going to happen for us. So we just really need to think through the fundamental question on these businesses. One really interesting point is why is there so much venture in crypto versus liquid? And there's been a lot of debate about this. There's one real view that there should be a lot more liquid. That's a fair view.

But I think that there's structural reasoning. That's very easy to blow up as a liquid fund. Internally, we call this AAS patients theorem, which is basically, if you don't blow up and you just keep building your skill set, at some point, you'll be the guaranteed top desal player in the Ligot space. And I think that's true. And basically, as since we started, there's been 30% of trituary in Ligot fund's per year. It's an insane number compared to any other kind of asset class. And the reason why is the markets are so volatile.

It's so easy to be positioned around. You can blow up by taking leverage. Of course, you know that. I think in Ligurit Torkutorkan market's taking leverage is basically suicide. I don't know how people do that. You can blow up by going short. Look, it's very appealing to going short. Sometimes you see something that's vaporware value to $5 billion, but then you get a ripple outcome. Your fund is over. You can blow up through high watermark. That's an interesting one people don't think about. But if you have a team and you have a 10x year on some very speculative trade and it's down. This literally happened with virtuals, by the way. It peaked on December 31st. It peaked at carry cut off. And then it went down 80% after that. And look, if that was a big part of your fund and you have a team of 10, people are now. looking at your fund saying, hey, you're not going to clip carry for five years. So it's a hard

position to be in as a fund. And then people also just temperamentally, people who invest in run liquid funds often, smaller ones. Many people don't want to retire. Honestly, we don't want to retire. We want to work till we die. Or this I do. I know it is too, but I don't want to speak for him. But some people just, you have a good year. You make $5 million, $10 million and you retire. And then you're out too. So there's not a lot of liquid funds at track of goods to go raise that capital. And I do think a lot of it is preference for VC, the VC valuation policy. I don't even think it's for VC. I think it's for the valuation policy of getting to the site with companies worth a recorder and kind of suppressing it all that way. But I do think a lot of it is

just there's not a lot of liquid funds that are around. And a lot of people, people have come into the liquid fund industry, but it's just the failure rate is so high. It completely makes sense and I'm aligned. I wanted to touch on your point about on-chain adoption, and Philippe Hades mentioned this and internet financial systems. The main question I have there is, I think we see a lot of companies, existing companies, traditional companies at scale, that are discussing putting assets on chain. And this has been talked about for a while, but in the wake of recent regulation, genius, and hopefully clarity coming up, I think it's much more reality.

Do you have strong conviction that Web 3, I hate the word term Web 3, crypto-native protocols and companies will be the beneficiaries? Of this, or do you have any concern that the next wave of on-chain finance might not benefit proportionately these companies and protocols that we're so familiar with? I personally lean towards the latter for 90% plus of companies, but I also think that this is a power law-driven market. So I imagine the 10% of companies that can cross-a-casm to offer products to a broader audience will be worth a conservative. amount more than where they trade today. But I think that a lot of the way that

crypto-nativism has played out has been very insular. So a lot of the products that have been built for crypto-natives don't necessarily resonate with the broader market. I also think that's okay. I think that if you look at Hyperliquid, for example, the top 2,000 traders account for very large amount of revenue. So I think that when you look at a trading platform, it tends to be power law driven as well, where the most active traders are driving most of the activity. So I don't necessarily think that for some of these projects, you need a substantial increase in the amount of users, just because I think that the users that can use these products and generate a lot of revenue already using them. I think that depending on which market

segments we're looking at within on-chain finance, some will work as back-end technologies for existing fintechs that perhaps have built-in compliance and regulatory features, have consumer applications that users already have like a defined behavior with. So I think that's one way across the chasm. But in general, I think that companies in the crypto space today are going to need to find ways to have their products more easily resonate with a broader audience in a way that I don't think they currently do. And if I added one thing to that, I would say, is crypto teams need to understand that the amount of the work ethic and level of seriousness is required to cross into fintech is just substantially higher than what they have been used to over the past five years.

This kind of remote first, work a few hours, live in Bali culture that Heptu has been in, is purely a product of regulatory hurdles keeping Fintech away from market. And now that those hurdles are gone, and these 1,000-person work 12 hours a day teams in San Francisco and New York are coming into the industry, you'll have to match that to compete. And there are a lot of founders who do max that, and they work hard and they're thoughtful, and they don't try to coast off being Ethereum OG or Twitter account. I think they'll do well because they have a head start. But people who think they can sustain their prior lifestyles and compete with Vlad from Robin Hood

or are going to be sadly mistaken? Yeah, there's no free lunch. No. How do you go about identifying these sorts of people, the sorts of people that you guys want to work with and the leaders that you want to work with? I'd say it's not that hard to know who matches the criteria. The Greenoaks founder had a nice line in a invest like the best podcast. Yeah.

Yeah. He said every bad founder is different, but all good founders look the same. And I think that's approximately true. Like, people who are hardworking, present well, obviously are doing what they love and building in a market that they want to build in for decades. And all they think about is their business and their competitors. When you see them, you know who they are, right? This is Marius.

This is Marius from Camino. this is SIF from Maple, this is Paul Pomerpho and those teams, Michael from Euler. Like, there are people who are born to build these businesses and I would take them over anybody on the C-Suite at Robin Hood any day and a one-on-one matchup. And you know who they're not. People who are trying to retire or chasing narratives. People who are angel investing, spending half the time angel investing while being the founder of a pre-revenue startup.

People who are just kind of throwing out jargon and hoping something sticks without having a real view on the market. I think I'd say this is one of the easier things to separate down our middle and say these are the founders that can really build a big business and you have founders that just are trying to maybe extract, maybe sell a dream and get out, or maybe they're honest builders, but they just don't have the work ethic and the commitment that it takes. I agree with everything Felipe said. The only thing that I'll add is that given the stage of companies that we're typically looking at, I think it's a bit easier for us to identify founder quality relative to pre-seed or seed, and that we also have a track record for how the founder

has reacted to a lot of different events within their business. So one example that I think has been a pretty good pattern matching exercise, has been founders that have survived a bear market in crypto and are continuing to work and have not made any unethical decisions or short-term decisions while being put in a tough position. In general, that has been indicative of the ethics, strive and founder, et cetera. I think that's the other angle here in that. We have the market do a fair amount of screening for us

before we tend to see opportunities just given that we're investing in projects that are, I would say, around the equivalent of like a series A in the venture space. I agree, no one. I think you guys probably look at these opportunities as much as anyone else in the market, but I think there's an underappreciation of people.

and Michael and Sid, who you mentioned, I think are the perfect example of this, but have a lot of stick-to-itiveness and now market reps and a proven track record. I'm curious for your guys' view. I unfortunately think we're at a bit of a lull in this regard, but I think we'll come back in the number of those people that we have in the industry and that we're attracting. I think there are a couple reasons for that. I think the meme coin mania had an adverse impact on the talent and profiles of people that we've drawn into our industry, which I think is temporary, but I do think it was a headwind. And I think that there are a lot of really interesting things happening outside of the crypto industry. When you think about AI, we think about robotics,

that also draw talent relative to maybe what you would have seen three, four years ago. So I'm curious for your guys' view on how many people are you seeing like that, are you meeting where you feel excited to potentially work with them? And how is that a headwind the tailwind in your investment approach and strategy? We are not seeing enough of those people. I know you're not seeing enough of those people either. It's a huge market problem. I think the first point you mentioned is actually more important.

The stigma, meme coins, NFTs, the TikTok, crypto influencers. It's just become an initiative that's associated with a lot of people who really good builders don't want to associate with, unfortunately. And I think the way to measure that, and no one's done this, I'd like to see it, is create the same deep profile on Hayinger dating app and the same person. and one, but just do A-B testing. In one version, he's an AI founder. In other words, he's a crypto founder,

and then just see what the responses are. I know that's kind of funny, but that drives human behavior. It speaks to the issue, yeah. It speaks to the issue. And that's also reflected in how people say that they're fintech founders, not crypto founders at weddings

when they're meeting their girlfriend's best friends. That stigma is real and people know it. What I'd say is, I'd like to say that people need to be more serious, so it goes away, but that kind of happen. Permissionless, be the way it is and they'll always be the some cringe edge of it by nature for a while. What I'd say to a founder is just think about opportunity to competition ratios.

It's my favorite argument. Yeah, I completely agree. Crypto is one of the biggest opportunities, I think, in the world today. It's on par with AI founders. Probably a lot of AI is going to open AI. I think that's a different of it. But I think for founders, it's on par with AI. For capitalized hitters, it's on par with tech or more interesting in terms is optimity size, but there's just not a lot of competition here. So I think if you are a smart, hardworking, serious person, you can find your way in the top decal of the industry pretty fast if you come into crypto. And the stigma issue will be gone in five, ten years, and you'll be in a pretty good spot. So I think it's a great place for founders to come. That being said, people,

obviously new founders don't fully agree with me because they're not enough here. I agree. I think a lot of people, particularly younger people, indexed towards what they think the largest opportunity is today. And there's not an appreciation of, to your point, relative white space within a category, relative competition, relative to the opportunity set. Yeah. It's a very flawed way of thinking.

Yeah. You need to think about how many people are going to be competing with you. That's part of why I left private equity. It was still a great opportunity, right? It's every private company in the U.S. But the amount of competition was just vicious. Yeah. I worked at a $10 million hedge fund.

I thought it was a very hard job. Yeah. Like, everyone's trying to do what you're trying to do effectively. And in crypto, you have these massive scale market inefficiency still, like the fact that some 80% of the money goes into venture investing. Yeah. And their private equity doesn't exist as an asset class. What kind of asset class has that set up?

Yeah. Do you know this Michael Maldeson luck versus skill continuum argument? Yeah. It's interesting. I'll say it for anyone listening. If nobody knows what they're doing, it's all luck. Imagine you're playing Catan or poker. If no one knows what they're doing, it's all luck.

If a few people are doing, but most of the people don't know what they're doing, all of the money will go to the few people who know what they're doing. If you're playing katan or poker, any game in the world, and you know what you're doing and you're playing with your 10-year-old cousin, you're going to win a disproportionate amount of time. But then once everybody knows what they're doing and they know the game theoretical optimal move at every step, it becomes luck again.

Because if you're playing with people who know how to play the game perfectly, again, I don't know how public of katana is. I'd always think of that. If you're playing people who would not have to play perfectly, it's your role. That's it. That's what determines the outcome because it's being played at the most efficient level. And I think a lot of industries are at that all luck because everybody is so good level right now. And I'd say, you know, hedge fund, private equity are getting close to that.

There's people who are better, but for the most part, everybody knows how to build a model. Everybody knows how to run the channel checks and use whatever data sort set is necessary. So you just don't have that ability to create really large skill gaps with the rest of the market. And I think in crypto, you can still do that if you come in at the stage market. Yeah, I love that framing. You don't want to be at that table. That's achieved perfect efficiency. This is a nice segue because another topic I wanted to get into with you guys is a literal

processes and creating edge. And, I mean, it can be said for very efficient markets, I think probably even more effective in less efficient or less mature markets. But how do you guys look at in your context, creating edge and putting into place? processes, whether that's via LLMs, whether that's via other internal tools or processes to give yourselves an advantage. Yeah, this has been my favorite topic over the summer. I think for a while, we basically won fully on LLMs right now as a shop. I'd say we've done a really good job of becoming an LLN first and that's manager. And I think that's the beginning of a big trend as well

in itself. So basically when we started, they actually before TTPT, believe it or not, we decided to become a Python-first modeling shop. That was what we called at the time. We were like, okay, data streams here. You can just reuse code a lot. So let's bite the bullet. Let's not use Excel and PowerPoint like we used to. And let's just build all our models in Python.

And that was hard, but it was worthwhile, because you could build a model for AVE one month. And then instead of going through this process, we do with Excel, where you have to get new data, put in the model, and like have a discussion, you could have this much faster process where you just hit refresh on your Python model. And then automatically you have the new numbers going through your model with a new price. And we've been doing this since we start.

So basically all of our underwrites, you have a live model. You can hit refresh on it and all new data will flow in. And we can have a get-on-the-icey discussions afternoon and say, hey, we haven't looked at this company in a while. Let's talk about it. And on the spot, we're having a discussion with the latest numbers, updated models. that's already a big improvement over what happens in traditional shops like kind of use Excel and PowerPoint. But then we got lucky because the end of our second year, TatsypT came out, and all of a sudden, being Python first really started to make a big difference

because we could just build a lot more financial models in much less time, because obviously LLMs can write a lot go-free. So when you ask what does it mean to be an LLM shop, I think first it means that LLM's are writing our models. We say, hey, build me a model for this company, this is how I think about it, here are the four or five-mead drivers. I want five cases. Here are the five cases. That's an LLM building that for us.

We're talking to the LLM as if it was a super analyst and give me a comments. When it comes to building analyses, bar charts, line charts, historical data, data-packed work, will say, hey, I want to see these 10 numbers. You're automatically in the position, even though Noah and I are both analysts, we're doing the analysis ourselves. We're in the position of a senior partner of a private equity firm with 10 analysts working for him. because we can just spend all day asking for analysis and getting them back immediately.

And that's a big, big change. If you think about the economics of an investment manager, there used to be these really nice economies of scale where if you were a billion-dollar fund, you can afford 10 analysts, you could just do a lot more analysis than the three-person shot like Thaya, and that's no longer true. You can have 10-Aless. They can't do that much more analysis than the two of us can do anymore because the LLMs are just so good at getting it done.

I'd say right now in an individual day is doing, I used to work at a shop of 10 people doing analysis, but we're probably doing in a day what that shop would do in two weeks and working people when they were 10, just because of how efficient LLMs are for financial analysis, especially when you add in streaming data. I think it also just changes not only the economics, economies of scale for firms, but it also changed the way you get to think about doing analysis in the first place. Because before the world was analysis constrained, you were doing it. doing a deal, it took you four weeks to build out a really big model with Power a

deck and a data pack. You couldn't do 10x that might work. It took a month to do that. So you were going to do the same kind of high ROI analysis for every deal. You're going to do your LBO model. You're going to do your three-year revenue model, whatever. Now you can really have five instances of cursor running five analysis at the same time. So you can start to throw in more and interesting models. And that's going to be, I think, a hard transition for one that we're trying to really get ahead of. You can build a traditional TAM model, you can do S-curb penetration, you can do an S-IR model, which actually is a model where you have recovery, people leave the product. You can run, you know, it's a lot, but you can run multiple different lenses and start to get into the idea of

like stacking these models together and averaging them out with certain requirements. To get better answers for your predictions. There's a lot to dissect there, but we've gone from a world that's analysis constraint to one that is not analysis constrained if you start L.E. them first and use those to build models. And I think we're really just trying to lean into that. A couple questions and follow up. The first one is, how many financial firms do you think are operating this way? And I'll mention a second to just so you can address both. But the second is, do you feel like it's equally enhanced your capabilities on the qualitative research side, whether that's following the social presence of a company or assessing individuals' online

presence? It's certainly enhanced our capabilities on the qualitative side. The other side of modeling is Every question is just deep research, chatypt, or just conversations with chatypt. That's just a lot faster, too. Used to me that you had to go find an expert in the industry, ask them questions, set up a call, ask some questions. So on Monday, you want to research the industry, and by Friday you had your first cut after three or four expert calls, you spend a lot of money doing that. Now you just get on with taxisputy and say like, hey, explain the home equity investment

mark to me, walking through growth over the last five years. What are the regulatory issues? and you can get two hours of doing that or three hours of doing that is a lot of work. There's a lot of knowledge. And I use this has to be for everything, for everything qualitative, either deep research or not. We use it a lot on the qualitative side. I'd say on the qualitative side, the big constraint for people is just the amount of hours they dedicate to deep work a day. And that's an interesting question that we're wrestling with.

Most investment firms are set up around this meeting type day. When you show up, you have a firm-wide meeting, you meet with some porkos, you meet with some, founders, you meet your friends who are also analysts, you do your model at some point during that day. But the reality is that that schedule was like an evolutionary offspring of a different time before LLMs, where that was the highest ROI would spend your day. But now I think the highest ROI day needs to include four or five hours completely alone with an LLM, just asking questions. It's hard to carve that out. So we've been trying to force ourselves to carve those four, five hours a day out, do you spend with LLMs asking questions?

Because it's by far the most signal and knowledge transmission per minute. It's hard to do that. Think about getting four or five hours out of your day to spend with an LLM and doing high effort research. And then your first question was how many firms are doing this? I think for small firms, we'll see a lot of firms doing it, like us. I think for big firms with hierarchies, it's really hard because the senior team is used to getting a certain type of output. If you give them Python, they'll be, I don't know what this is. Don't give me this.

just build the Excel model. We've been doing it this way for 20 years or 30 years or whatever. So I think that big firms are going to be really actually hard causing the cross because the senior team will just not want to go out of the way to learn a new modeling technique where a new way of working where they could make a big mistake in the meantime if they miss an assumption because they're just kind of not used to the way that the code looks on the page. Most of these people are pretty comfortable.

They have a lot of money. They've been doing it for a while. They have a way of doing things. My guess is small firms. You'll see a lot of them doing it. Big firms will take a lot longer than people expect right now. Yeah, I think you make important points. One on the path dependence.

This was actually a point for the same podcast you mentioned from Invest Like the Best with with Neil Meadow, but he made an important point where he said a quality of the best founders he's seen is their ability to block out everything that's not the one or two most important things at the moment. And to your earlier point, I think the difficult part of taking the time to set up Python workflows to take five to six hours a day and deep research. research is not actually doing it, which I think people might think so. It's saying no to everything else, which could be your alternative marginal opportunity cost of time. Yeah, I think that's exactly

right. I want to touch on a few points as we start to wrap up here. What do you guys think of the regulatory environment and how has that impacted your job? How do you look at it on a go forward basis? It's definitely been a large tailwind for our industry. I think that we've started to spend a lot of our time looking at traditional financial markets and understanding where the inefficiencies are and which ones will be first to tokenized assets or tokenized funds, whatever it may be, to kind of blend that into how we see the growth of the companies that we're investing in today. Just because I think that the proliferation of crypto-native assets making their way into defy is continue to play out.

It's probably somewhere in the middle of the process, a large percentage of the native assets are on these various defy applications. So at this point, I think the marginal growth has to come from exogenous assets that are currently on blockchains. So now that regulatory barriers are starting to be reduced and there's more experimentation by traditional firms, it's definitely an area that we're trying to get ahead of. Additionally, around that vein, I think speaking to early movers that are traditional financial firms coming to the space and looking to get involved, so whether that's Janice Henderson that's working with centrifuge or Apollo that's tokenizing ACRED.

I think understanding how they see the space and where their interest lie is also important. So I'd say it's really just understanding what the major drivers are going to be for the companies we're looking at as a byproduct of the regulatory direction moving in a way that we see as being favorable. Outside of that, it's keeping track of where the regulatory direction is moving. And so you have these bills that are constantly moving through Senate. So reading through them, understanding the implications. for current market dynamics in case that these things are disruptive.

So on the stable point side, Circle and Tether are the two major players. Circle historically has had a pretty big regulatory note. Depending on how the Genius Act played out and other stable coin bills played out, it had pretty large impacts as to whether Circle would continue to have regulatory moat or whether that regulatory mode would be declining. So you naturally need to get ahead of how those policies may play out as well. So speaking to regulators, I think has pretty high. ROI as well. Yeah, it'll make sense. It's going to be an exciting, I would say 24 months,

and I think we're only seeing the tip of the iceberg on the stable coin and related front. I want to leave off by tossing you guys one more point here and an opportunity to share some of your experience and beliefs, but I want to know what's one investment thesis or idea that historically you feel like you've gotten right, one that you held that didn't or hasn't panned out, and then one that you feel most strongly about looking down the line. And feel free to tackle those in whatever order. Makes sense. I can start with one that we have strong conviction going forward.

It's one of the more excited ones that link. Some good podcast topic. Metadal is the Futurearkey platform on Solana. We've publicly made two large OTC investments into Metadow. I think it is one of those really big ideas in Hito. Can you speak to what that means? It basically, Futarki, is a governance system where you have a decision. So let's say the decision is, should Apple bring back Steve Jobs as CEO?

And then instead of having a vote, you separate out two markets, one where you trade Apple stock in a situation where Steve Jobs does come back as CEO, and one where you trade Apple stock in a situation where he is not. The Pepsi CEO remains in place. And just trust me for now that the markets work, the true markets, based on that condition. And then you look at the two markets and see where Apple stock trades higher. And that determines the outcome of the decision. So if in this case, presumably the market where Steve Jobs comes back will trade at a 40%,

let's just say, premium to the other market. And the market has spoken on what is the better decision for the stock price. And that ends up feeding into the actual decision. Just one really quick example. Say Un-Swap wants to turn on their fee switch. Or better yet, say Un-Swap wants to send revenue to U-N-WBelabs. Instead of having that be a decision made by the team or by a governance vote, you just put up the unit token and see whether it trades higher in a case where revenue sent to Microsoft Labs or higher in a case where that motion fails. And it's not sent back.

I think two new shop labs. Again, two examples. But basically, the reason we think Futarky is there's two views on Futurkey. One is strong Futurkey, which is like the market will make a better decision on everything. Just always defer what you have for lunch or how you design a product. If whether you hire or fire this VP to the market, we don't believe in kind of strong future key. What we believe in is future key as a board of directors. So when it comes time to changing CEO, making an acquisition, selling the company,

releasing a second token, restructuring the token, we think that future key markets should decide in that case. And the reason we think it's so important is because the essentially, the market essentially acts as the replacement as an enforcer of token holder rights. So this goes back to something that we talked about all the time, which is the token holder problem. Right now, token holders don't have regulatory protection so they get fleeced all the time. Cashfuls are sent to the equity entity instead of the tokens. Teams launch a second token where they have more an outside share just to kind of make a profit. Teams pay themselves way, way, way, way too much as a way to just withdraw funds from the treasury.

All these problems are really, really serious problems in token markets. We think they're existential. I think we've talked about the ad nauseum in the past. They make it very unattractive for capital flowing to token markets. And a futarchy kind of board of director's system solves all those problems. Because all of a sudden, it's like, oh, should we launch a second token where the CEO has 50% of supply, the market automatically will vote no on that because the token, the main token will trade lower in the case where that happens.

Or should we do M&A with some centralized exchange where they pay out the team? They pay the team a huge big bonus, but the token gets nothing. Token would obviously vote that down. it becomes kind of trivial to vote down attacks against token holders through this mechanism. So we see it as an enshrining of the token as the objective function for a token business, which is very important. That happened in the U.S. capital markets when in a court case called Doge v. Ford, where Henry Ford tried to donate a lot of profits to charities of his choice.

And the Doge brothers, who were investors in Ford, sued him and said, your job is to maximize shareholder valued. In the Supreme Court of the U.S., agreed with him. that's where the fiduciary duty to maximize shareholder value comes from, and it's worked great in the U.S. capital markets over the past 200 years. We think that future key is that moment, is the doge versus forward moment in internet capital markets. And without it, I think you live in this mean coin trader land where nobody wants to buy and

hold because they're just not protected. And with it, I think you have actually have a better mechanism than you have in U.S. capital markets for protecting shareholders. That's why it's important. Now, why is it going to take off? just frankly, like if you raise money on future Arctic markets, investors are more willing to give you the money because they know they protected. We had an experiment called Mountain Capital.

I don't know if you followed that at all where this team raised a few million dollars to invest in Solana Accelerated Projects. And they were able to raise through Medadal and it went quite well. And then there was some strategy drift along the way where the team was not going to be involved as some investors like us that they would be. Maybe their deals were going to be different than they initially composed with Solana Accelerated projects. And that's the type of thing that could lead to a big fight in regular internet capital markets unprotected,

or it could lead to very unhappy investors. But in this case, there was just a vote to return capital. And every single person who held kept their money kept 100% in principle. So that shows you what the protection means. And then as an investor, looking at an experiment, I'm way more willing to invest in it. If I know that my downside is I can pull back funds if the team, if it's not what I thought it was, then if the way that we really normally worked is if you, You invest in an experiment in internet capital markets, invest in a token, and it doesn't work out well, you're getting a zero.

So the math on that is just a huge difference. But anyway, that's our go-forward idea. No, anything else on past for present? Yeah, just on pass ideas that I think we've gotten right. I think that we noticed that lending and credit was a more attractive market post-election compared to volume-based businesses on chain. I think there's a few things. So there's the regulatory change and what that means for both sets of KPIs. So when you look at volume-based businesses, meme coins and other on-chain assets that have

price discovery that are on-chain, a very high turnover relative to their market cap and are generally somewhat ephemerally priced. So they make for relatively low-quality assets, but very high-volume-based assets, whereas the marginal assets that occur that exist publicly outside of crypto tend to have much lower volatility, while you might see a very large amount of assets come on chain and dollar value, it won't necessarily amount to a significant amount of trading activity just because they're going to be lower turnover. We see the marginal assets coming on chain as being much better collateral assets than

the assets that exist on chain today, but don't necessarily see it leading to a substantial spike in volumes just given that the current amount of volumes that occur on trading trillions of dollars a year is quite large relative to like the aggregate amount of volumes that are going through traditional financial markets, whereas the total amount of assets that sit on chain is a fraction of 1% of the total financial assets that exist broadly. And I think that's generally starting to play out in markets and that volume-based businesses, even with a lot of activity that's occurred this year, have not gone much past their all-time highs in Q4, 2024, whereas TVL and Barrow outstanding has gone to new all-time.

time highs for many of the primary lending businesses in the space. Yeah, I think it's a great insight. And for what it's worth, I still love lending as a category on a go forward basis with this in mind and the number of assets that will come on chain that to me are conducive to following to those sorts of businesses. This has been incredible. I want to thank you guys for coming on again and I hope to chat with you again soon. Thanks for having us on. Thank you for happiness. Thanks for listening to another episode of On the Brink with Castle Island. To find out more about Castle Island, visit castle island.Vc. To listen to all of our podcast episodes, please go to On the Brink dashpodcast.com or just click on the tab in our website. Thanks for listening.