为什么基本面终于开始胜出，以及这位 DeFi 投资人正在把钱投向哪里｜The Thesis

People get tired of memes, right? They're like, “Ah, there's no fundamental in this.” Because what happens at some point is, when there's a lot of memes, you don't really know what the differentiating factor is between this meme and that meme. Both could be good memes, both have followers, whatever, right? And so what you trend toward is fundamentals, because that's the only anchor you can really look toward to tell whether this project is good.

Does this project make money? And if it does, you know, it's a good project. Only the ones who have real usage and real revenue grow to the top. All of this points to a K-shaped outcome where 90% of investments die, then 10% of investments float to the top. Whether it's memes or fundamentals, it's the same. It's just that 90% of them don't have any real backing, and then 10% have good metrics, you know. That's how stuff like Ethena and Hyperliquid, et cetera, rise to the top.

I'm DeFi Dad here with Nomadic. Today's show features investor and DeFiLlama senior research analyst 0xKyle. Kyle, thank you for joining us. How are you doing?

Hey, guys. Thank you for having me. Happy to be here.

Yeah, dude, awesome to have you. I've been following you on Twitter for a while now, and we're doing this new series we're calling The Thesis. We're just trying to bring people on who are showcasing real edge in the market and have skin in the game.

Today we're going to talk about a ton of stuff I've seen you talking about. I want to get into the Ethena debt sort of stablecoin X trade. You've been talking about inference capital markets, and we want to know what that means. We want to hear about maybe some theses you're building around new tokens that you haven't been talking about quite yet.

Again, of course, none of this is financial advice, but the goal here is just to understand how good investors think and hopefully find a little more edge for not only ourselves but our audience. I want to get into a bit of your background, because in your bio it says you previously worked at DeFiance Capital. DeFiance, if you've been in crypto for a while, is a pretty legendary crypto firm headed up by Arthur Cheong. I don't know if I pronounced the name correctly, but I want to ask you what it was like working there and any insights into working closely with a guy like Arthur for some time.

Oh, yeah. It was great. It was one of the best times of my life. I really enjoyed working with the DeFiance team. Actually, I'm going back to their office when I go back to Singapore. It's just a cool place to work together. I'm still pretty close with the team there.

Working with Arthur is actually pretty fantastic. You get to watch an investor up close and really watch what he does and learn from him. He's very well known for being a high-conviction-bets kind of guy, taking them to the maximum. I feel like it's something I know, but it's very rare to see in an institutional setting where you go ham on a single position, and that's how you make the majority of your returns. That is how I believe you should always invest.

Many times when you work at all these funds, people are all about diversifying risk. You only hold a maximum of 5% of a position, and you hold plenty of different positions. Whereas Arthur is very fine with holding one position, but a double-digit position—like, I don't know, 10% to 20% to 30%. Okay, I've never seen 30%, but 10% to 20% on a single position. I think that's very rare to see.

I was curious: During your time there at DeFiance, were you there in 2022? Obviously, there was the 3AC blowup, and I know there was some relationship there between DeFiance and 3AC and Arthur, of course.

No, I was not there. I only joined in 2025. But, yeah, there's the whole relationship. It was pretty insane, actually. I don't have the full story—I'm not the most knowledgeable on it—but basically, Arthur climbed back up from zero.

The entire fund got clawed back, and then he started Fund 2. He basically made multiples on Fund 2, so, you know, kind of a beast.

Kyle, you've been talking about something on X that's near and dear to me. I feel like I've been a fundamentals investor since I came into crypto. I don't even know how many years ago it's been now. Jeez, I hope it's not 10—maybe 9.

But fundamentals just never really got their full bid, I would say, and the full attention they deserved, maybe until now. So maybe walk us through what you're seeing. I think you had a tweet saying we're moving from memes, thin liquidity, and grift into this second phase where people are starting to hunt for real signal—things that matter, things that make money. Talk us through this movement that you're seeing.

Firstly, I would like to caveat that this is not a new view. Many people have been saying this for quite a while. It's the holy grail of liquid investors, because all liquid investors want is to bet on good projects, and no one really wants to put any sizable amount into a meme coin.

Meme coins will still exist, obviously. They're still the vehicles for things to go up the most. It's just that, firstly, people have no edge playing them. Secondly, they do not offer any sort of fundamental long-term growth, which is what you kind of need in capital markets.

The fundamental thesis is basically that crypto has good businesses, and what you want in good capital markets is steady, long-term growth over time. You can only get that when the underlying business grows a certain percentage year on year.

So why is the S&P so strong? Because they're basically betting on the U.S. economy growing year on year, and this has worked for the past 20 years. Why are the Mag 7 so strong? Because they're monster businesses that just keep growing.

Nvidia, right? Jensen Huang recently said, “It's interesting because Nvidia's the largest company in the world, but they're also one of the fastest-growing companies in the world.” The idea of growth behind the company is what you need for sustained price action over time.

We finally have that with Hyperliquid. Hyperliquid is really the one coin that actually exhibits that kind of up-only chart, where you DCA and you'll be fine. I believe this is going to trickle down broadly into the crypto ecosystem, because at the end of the day, capital markets are very—what's the word?—reflexive.

What's going to happen is people see Hyperliquid, investors demand more, and businesses implement these outcomes. Then their price action sort of reflects this. That's basically the broad thesis.

Now you're seeing more and more people do buybacks. Three years ago, buybacks weren't a thing, right? Then Hyperliquid came, and it was like, “Oh, guys, we're doing 100% buybacks,” or 99%, if you want to be pedantic. Now everyone's like, “Oh, we're buying back this amount a month.” It's not 100%, but they're still doing something.

Another thing is investor relations. Now you have more and more different ancillary tools like Blockworks, and DeFiLlama has its own investor relations, where projects come to us and they're like, “Okay, can you do an IR report?” Then you're forced to audit them, right?

When you're doing this audit, you know, it will look bad for you if there's some siphoning of money or something like that. This forces companies in crypto to have some sort of alignment. All of this is building into the idea that we're slowly getting better and better fundamentals.

I think we have this back-and-forth of everything working to nothing working. We go from believing that every idea is a good idea to nothing absolutely working. The 2017 cycle was an everything-can-work-on-blockchain cycle. All you need to do is create a token.

2018 through 2019 was total despair, realizing it's all fugazi and none of it works. Then we have DeFi Summer in 2020 into 2021. Now everything's working, but they're actually really poorly designed protocols, barring just a few. Then we go into 2023 and 2024, and nothing seems to be working again—or 2022 especially, I guess.

I felt like the last cycle, though, we were betting at the start on things that were working. The problem is that a lot of folks missed that so much of it had been propagandized in their face with VC money. There were all these new DeFi protocols that were new iterations or next-generation-type protocols from the last cycle.

The last cycle, 2020 and 2021, was all organic and being launched at light speed. Every new protocol was launching, there were new tokens, and it was all a cash grab.

But then you have folks who are raising money from reputable VCs and trying to put forth a better product. This time, I feel like last cycle was really a cycle where there weren't as many net-new investors using that on-chain. So, I follow what you're saying. I think there's a setup here now that we've seen a lot of the VC-pumped narratives start to die off and real grassroots successes emerge. I think Hyperliquid is that ultimate grassroots success since the days of Bitcoin and Ethereum.

That, to me, is Hyperliquid's legacy here. Hyperliquid got people to believe again that a protocol that's a thriving business on-chain, that generates revenue and fees, and can return value to token holders can actually work. So now we're in a place where we're seeing other teams that hopefully we're going to talk about that are thriving and seem to be attracting attention.

How, again, do you summarize that thesis of yours—that we needed to suffer through the nihilism around memecoins in order to get to the stage of believing in and valuing positive-sum DeFi applications and protocols? The Hyperliquids of the world, the Ethenas of the world. We've been covering Derive extensively on the podcast, Lighter, and so forth.

I actually have a good image to send you guys.

We would love that. You're welcome to screen-share it.

Yeah.

And for folks who are listening, we'll do our best to describe what's on screen here. Kyle, whatever you can do to describe it, but otherwise, take your time.

Oh, screen-share. Okay.

And for folks who are listening, we'll just do our best to describe what's on screen here.

Yeah. So your question was, how do memes lead to this?

Yeah. How does the mania around memecoins lead to this resurgence in investing in positive-sum applications that are thriving businesses on-chain? These are, again, the Hyperliquids of the world, Lighter—there are lots of examples here. I'm trying to stick to the stuff that's near and dear to you. I know Ethena is another one of those.

In DeFi, yield shouldn't come from chasing dashboards or timing markets. KPK runs non-custodial vaults and on-chain funds built for stable risk-adjusted returns. Capital is allocated across curated strategies and managed by automated agents that act within seconds under predefined controls. No custody, no manual intervention, just disciplined execution on chain. If you're allocating capital and want sustainable yield with stronger controls, visit kpk.io. The next generation of transparent dollar yield is here.

Axis brings you eight years of institutional experience providing global liquidity, generating durable returns. By synchronizing prices and liquidity worldwide, Axis captures delta neutral opportunities without directional risk. Deposit into Axis to improve global liquidity and earn 10% to 20% net APY. Over 1,800 unique wallets committed $67 million to the Origin vault. Stay tuned for the next phase and learn more at axis.to.

Apex is the first dividend-backed stablecoin protocol, bringing enhanced public market yield on chain, backed by strategy's perpetual preferred stock stretch. With APYUSD, anyone can earn dividend-driven real yield live on Ethereum. To learn more, go to apyx.fi.

Okay. I think this goes into how market participants view markets and how markets work. Essentially, memecoins are the left-hand side of this, where you have one memecoin that goes up a lot, like Dog With Hat or something. Then the next one is Shark Cat or something, and it just spirals all the way down.

This is what memecoins do because there's no infinite bid. What you have is one king that goes up a lot, and you see this every single time. It was Anthem, then it was Pawns, and then it was—okay, I guess Pawns isn't a memecoin. But it was AI, Artificial Inu.

Then it spawns a bunch of other stuff because people don't want to buy the main coin. The next coin goes to, I don't know, $100 million if the first one goes to $500 million, and then $30 million, and then $10 million. It just spirals all the way down. The memecoin goes up and up and up until it reaches the top, where it physically can't go up anymore, like Doge at its peak.

These things have to peak because they don't have an underlying business. At the same time, there's no ceiling to them, but there's also no floor to them. I think that's very important to understand. That's how these things go down 99%. There's no ceiling, so they can really go up very fast if everyone piles in. Then there's no floor, so they go down a lot.

Trump coin is a very good example, right? It was an insane event like nothing before, when Trump launched the coin. It shot straight up in the span of hours, and it really peaked. The next day it was down 15%, then it was down again, and now it's, I don't know, around $2 or something like that.

Memes are one vehicle. I wouldn't say that they directly lead to each other, but they kind of do. People get tired of memes. They're like, “There's no fundamental in this.” At some point, when there are a lot of memes, you don't really know what the differentiating factor is between this meme and that meme. Both could be good memes. Both have followers, whatever.

What you trend toward is fundamentals, because that's the only anchor you can really look toward to tell whether a project is good. Does this project make money? If it does, it's a good project. So people get tired of memes, look toward fundamentals, and that's how you get what we have today.

When you look toward fundamentals, your next question is, “Okay, between fundamental projects A and B, what's the difference?” The next metric will be revenue, how much money they're making, what's their TAM, and all that shit. Crypto is also really good at making a lot of meme fundamental projects.

Just now, when you mentioned all the VC narratives, they're kind of fundamental-ish, but they're not fundamental in the sense that they probably say something sexy, some insane narrative—AI agents, for example. AI agents trading and that kind of stuff. Then it launches a coin, but when you look at the metrics, no one actually uses them.

The crux of the idea has some merit to it, which is what attracts people. In the fundamental section, there's also this dispersion where a lot of people copy each other, but only the ones that have real usage and real revenue grow to the top. All of this points to a K-shaped outcome where 90% of investments die and 10% of investments float to the top.

Whether it's memes or fundamentals, it's the same. Ninety percent of them don't have any real backing, and 10% have good metrics. That's how things like Ethena and Hyperliquid rise to the top.

I would say that this cycle, I went from a place of not seeing any real fundamentals behind memecoins to at least appreciating that there are a few colleagues we know who we believe are actually searching for some signal in the sea of chaos that is memecoins.

One of them is Michael Nadeau from The DeFi Report. He did some research over the last few years and came to some conclusions. I don't know when he ultimately made these calls, but I think his entries are up something like 10× on a few of these coins. I really came to admire the signal he was looking for and, again, some of the metrics on-chain. Everything he's looking at is on-chain metrics.

My takeaway from that was not that I should necessarily be investing in memecoins. My takeaway was that there are folks finding theses in a very chaotic market. Eventually, when opportunity has dried up in some of those memecoin markets, I've seen them shift gears to looking for positive-sum, value-accruing types of DeFi applications and tokens.

I think that, to me, has been the evolution of why we're now in a place where it's actually cool again to have a thesis and have some conviction to invest in something meaningful. Again, for us, I think we've been talking a lot about some of the obvious plays here, like Ethena, and Derive has been a huge focus for us lately.

Yeah, what you guys are saying totally resonates, and I've had this problem with memes. I've never been able to trade them, in part because there's no fundamental indicator to tell me when I should buy and when I should sell. I know DeFi Dad was referencing some people trying to use metrics like Michael Nadeau, and then—what's the other guy's name? Oh, forget the guy's name.

He was like the meme coin Jesus guy. He's got long black hair. Do you remember?

Oh, Murad.

Yeah, Murad. He was looking at things like average holder size, how dispersed the supply is, holder retention, and trying to apply all this stuff to memes. It's a noble pursuit, but it's still not really passing my personal sniff test. Maybe that stuff can work, but ultimately I'm navigating in the dark with memes and with stuff that has growth fundamentals, like revenue. Looking at TAM and all that stuff is more tactile—something you can latch onto.

I want to get to some of these trades that I've literally been seeing you make in the open here, Kyle. One of them was Ethena. Ethena has this DAT called StablecoinX. The ticker is kind of confusing because when you talk about it, the ticker on Nasdaq is USDE, but that's also the actual stablecoin, USDe. Maybe walk us through what you were seeing there.

To me, there was a ton of asymmetry. I also joined this trade. There's another guy on Twitter, Evan S., or whatever his name is, who was really early to promoting this. But I kept looking at it and thinking, "What the heck don't I know? Is there a massive overhang here that I don't know about? Why is this thing trading way off mNAV?" Maybe walk us through what you were seeing.

Shout-out to Evan, who put me on this trade when I read his tweet. When I read it, I was like, "Oh, this is a..." I was so disappointed in myself because I had the same idea; I just didn't believe in crypto enough. We can get to that later.

This was really at the bottom of the bear market, when this thing was like a dollar or something. The core idea is that there will be a DAT that rises to the top this cycle. We see the same thing every time: something new and novel comes out, and the first wave rises to the top in terms of breadth. Then there's a second wave, when it crashes 90%, but there are a few good gems you can pick out, and those go up only.

This is another famous chart that I've constantly tweeted about. I've tweeted that this chart has made me more money than anything else, and it's true. This is the first wave of attention and hype, and then it goes down like 90%. If the project is good, there'll be a second wave where it eventually trends to all-time highs and breaks all-time highs. Everyone writes it off here, but fundamentals are still growing, and then price catches up.

Specifically in the case of Ethena and StablecoinX, there's a lot to talk about, including how it works and their warrant structure, which I'm probably not the best person to talk about because I don't have my notes on hand. But the TLDR is that a DAT is a vehicle that accrues value by buying the underlying token.

The traditional debt model doesn't work for most altcoins because this model came out with Michael Saylor and MicroStrategy, where he just keeps accumulating Bitcoin. He sells equity, buys more Bitcoin, and increases Bitcoin per share. Recently, he set up a few different instruments, including STRC and all these new dividend-like instruments that pay some sort of yield, along with all kinds of mechanics.

But I always found the idea to be very one-dimensional. It really only works for Bitcoin and assets that are Bitcoin-like, where you just have to buy and hold them, and that's all. With altcoin debts, though, there's an opportunity to do something different. You can figure out ways to bring the altcoin to the masses besides buying more of the underlying token. Obviously, you still buy the underlying token, but there's so much more you can do.

For example, PER, which is Hyperliquid's debt, has a CEO who is an ex-Barclays executive or something—some pretty high-up financial official. I don't know what he's doing, but a good example is that Stanley Druckenmiller's fund was reported to have bought PER because you want underlying exposure to Hyperliquid. On the Ethena side, there's basically room for the DAT to do something similar.

Firstly, it can attract traditional investors and people who otherwise wouldn't touch the underlying token. Secondly, there are a lot of business lines you can tap into when you have a traditional vehicle. You can build access and utility for Ethena's digital dollars through the DAT, which you otherwise wouldn't have. You have this second vehicle that does all of this.

That's how I see it, and that's how I think you can accrue value to an altcoin beyond just buying more of it.

I think two of the benefits there to reemphasize are that some funds and investors just can't touch the underlying token. Let's assume that Stanley Druckenmiller's family office isn't going to buy HYPE itself, but they're willing to buy the PER debt so they can get exposure to the upside.

However, they're not just getting exposure to the underlying HYPE; they're also getting exposure to DeFi yield. I've always thought of those debts as wrapping the DeFi yield in a package that allows these bigger investors to get exposure to it. DeFi yield for the public markets—I think that was a term we saw being trumpeted by the SharpLink team.

This is part of the reason we got so excited about ETH debts, because there's native yield there. You can earn yield, and that can help pay for all the administrative fees. With someone like MicroStrategy, as much success as Saylor has clearly had, he's essentially creating instruments and selling products that are asking investors to lend money so they can buy more Bitcoin. That Bitcoin will ultimately go up, and investors will have a greater concentration of Bitcoin per share through that.

They're very different approaches. There isn't a yield there. I've heard talks that there is a yield because sometimes they're selling options as well, but it's a much leaner, cleaner trade to me. It's a much cleaner trade for HYPD with Hyperliquid, a much cleaner trade for BitMine and SharpLink with ETH, and I think it's a much cleaner trade for what Ethena is doing with StablecoinX, where you're getting exposure to ENA.

Just to piggyback on that and go back to your chart too, Kyle, the one chart where exuberance gets way ahead of fundamentals and then comes crashing down: during the debt phenomenon, the hype cycle that was debts, a lot of people were like, "Oh, yeah, this DAT experiment is over." You heard a lot of cynicism and stuff, and I was pretty adamant the whole time: "Hey, I think there are some very useful DATs."

HYPE D, like Hyperion, was one of them, and I loved the Ethena one as soon as I heard about it. I was like, "Okay, this kind of makes sense to me." But there's always that sort of reflexive nature with opinions, too. It's either, "This is good," or, "This is bad." As DeFi Dad was saying, it's such binary thinking, and it's hilarious to me.

There's a token out there that doesn't have a debt. I don't think I'm allowed to say which token it is.

I can't say which token it is either, but I have a very reliable source. To go back to our other point, there are people who want to buy these tokens but can't because they can't hold tokens; they don't have the mandate to hold them. Apparently, there's about $500 million of pent-up demand for this certain token.

Maybe I'll tell you guys that offline afterward. It's a very reliable source, but they can't buy it, and they want to buy it. They want there to be a DAP. I really want to see whether that materializes.

Getting into your Ethena thesis and what you're thinking, I wrote something recently—I think it was a tweet. Between the fund I used to work at and my own angel investments, maybe I've seen 70 to 80 investments materialize from start to finish. You come into investing with these ideas: "The founder's really important," whatever. Of course, that's obvious.

But I go through this cycle of investing and going through the ups and downs of 70 or 80 projects, and I come back to this: the founder is all that matters. The founder and the team—that's literally all that matters. You can have a great idea, a flavor-of-the-month idea, but chances are you're going to have to pivot and iterate 5 or 10 times along the way. You're going to get smacked in the face. All this shit—all this adversity—is going to come your way.

So, maybe talk to us about what you think about Guy Young from Ethena, just in how you look at investing in people ultimately.

Investing in people, now that you bring it up, I've never really given much thought to it, but it's actually incredibly difficult, because they have to get it, and you have to get it that they get it. When I say “get it,” I've always wondered what “get it” means, because this term was introduced to me by a guy who's ex-Wall Street. He's always like, “They just don't get it.” When he said that, I kind of understood what he meant.

But I think what it translates to is business sense—that very specific skill: the ability to convert your determination and your intellect and your skills into a tangible business, and have the business sense to run it well. Guy obviously has a lot of this. Jeff obviously has a lot of this—Jeff Yan from Hyperliquid.

And I think people always shit on Coinbase. I don't know if—sorry, can I say companies on the podcast? I want to say they shit on Coinbase.

Oh, hell yeah.

Okay. So people always shit on Coinbase, and for good reason, right? They're always like, “Their customer support sucks. Their exchange always doesn't work and freezes the funds.” Brian Armstrong is kind of seen on crypto Twitter as this guy who's not very receptive to user feedback. He's not a very move-fast, break-things kind of guy.

And you can kind of see the stock, right? Now, Could is running laps around Coinbase. Could Chain, for example, is doing what Base Chain should have been. But Base Chain didn't become it because Base doesn't get it. Vlad gets it. That's why Vlad is such an incredible founder compared to Brian Armstrong. That's one.

Another is Jeremy Allaire, the founder of Circle. I used to be a very big Circle bull simply because it's the only way you can bet on a stablecoin, and it's pretty cheap in the sense that, compared to their whole TVL—I mean, if you're into TVL-market-cap stuff. But it was a very simple idea: an N-of-1, pure play of stablecoins, the second-biggest stablecoin in the space.

But the more I ran into it, the more I realized that Jeremy also doesn't get it, get it. He's not a very move-fast, break-things kind of guy. That's why I think there's room for a second stablecoin to beat Circle, and that could be Guy. Guy really gets it. He's a monster. He's definitely going to do everything in his power to really grow Ethena's TVL and USD's market cap.

Yeah, one more thing to add to that has become more apparent to us lately. This is again thanks to one of our more recent podcast guests, CoolCrypto, combined with our own research and our own experience using the product.

I really didn't realize that while Derive is eating into Deribit's market share, I just didn't realize how asleep at the wheel Deribit was. Funny enough, they were bought by Coinbase. I'll say I have a much less negative view on Base and Coinbase, but I still see the same sorts of issues that you do. I see a team that, for whatever reason, is less receptive.

I look at Robinhood; it looks like they're moving lightning-fast to listen to that user feedback and implement those changes. Again, I know it's difficult too, because there's a lot of stupid feedback being fed to us on the timeline, especially with folks getting loud and just screaming about things on Twitter.

I'm absolutely seeing a similar dynamic play out, and maybe it's less of a knock on Deribit and more of a compliment to an extremely strong founder in Nick Forster—someone who's been building on-chain options since before they were in vogue and has been basically in the mud just trying to make this primitive work. Now it's finally paying off. Just in the past year, they're finally seeing—

You know, volumes. I think notional trading volume was up something like 5× year over year the other day. Fees and revenue have skyrocketed as well.

It takes much more than just a great idea, and even great execution, because these markets just keep evolving. Honestly, I feel kind of like a fool to throw stones at Coinbase because I also marvel at what they've built. They have enabled so much in terms of crypto products that are available to centralized-exchange customers, but they've also built so much on Base, and they've tried to lean into DeFi in some sort of hybridized approach before I think a lot of the other exchanges believed in it. So, yeah, I'm seeing a lot of the same themes there.

They have been executing the DeFi mullet really well, tapping into letting people borrow against their Bitcoin and borrow USDC, and stuff like that. I'm sure they've fought a lot of the regulatory battles that I don't even know about behind the scenes.

But I was thinking too, just looking at Circle, going back to Circle and Ethena's business models, man, they're inverse business models. Being the salesman for USDC is so much harder than being the salesman for USDe because they just give the yield back to the people. If I was a sales guy, I'd be like, “I can sell that USDe product pretty easily,” because everybody kind of wins.

Selling USDC is like they sit on it and take the rewards, right? I know they're doing a lot more unique pass-through stuff now where they realized where the puck was going. They had to make that incredible deal for Hyperliquid. Basically, “Hey, we're actually going to pass through 90% of our yield to you guys now because we see you're such an absolute winner.”

But anyways, I want to move us to something a bit different. Another thing I've seen you writing about is something called inference capital markets. There was this BlackRock piece that came out recently on AI and crypto, and I saw you quote-tweet it. What are you seeing with this term that's kind of new to me—inference capital markets—that's exciting you?

Okay. Galaxy actually wrote a very, very good piece called “Inference Capital Markets.” I'd highly encourage everyone to read it. Essentially, what I personally mean by inference capital markets is more the fact that I want to find ways to long inference. I'm very, very bullish on inference demand.

The term “inference capital markets” specifically refers to an on-chain project that you've probably seen me shill. But the general, top-level thematic that I'm very, very bullish on is simply inference demand. If you like, I can explain more.

Yeah. I would love for you to flesh that out, and maybe I'll inject another thing into this. There's this whole craze for L1 proof-of-work tokens again all of a sudden, and it's stuff like Pearl, Knock Chain, and Tao. People are trying to sell this thesis of AI money because, of course, the TAM can be huge if you just call it money.

There seems to be a there there. There's something interesting happening, and this feeds into your inference capital markets, because some of these players are providing inference as a service as part of the mechanism design. There are teams on TAO, like subnets, that are doing this as well.

I don't know if you've looked into these tokens that I've just mentioned much, but do any of these play into it? Flesh out how you're taking action on inference capital markets. What are you buying? What are you looking at? What are you excited about?

Yeah. So, for projects, every time I look at a thesis, I always try to go top-down. Firstly, inference, right? Why is inference demand? In the entire AI stack, what is the lever that you can pull that has the most torque? That has shown itself to be inference, because inference is really Jevons paradox right now.

Demand is just skyrocketing. If you see the data on OpenRouter and stuff like that, it's just nonstop. Then you saw Muse, right? Muse came out, topped the charts, and was basically downloaded more than ChatGPT in the same amount of time. ChatGPT was a killer app when it came out. Everyone downloaded it. Then Manus came out, and I was like, “See you.” Then it was just like, boom, right?

That was when I knew my thesis was playing out, because what I said in my “Artificial Fugazi” article that I wrote a while back was the idea that the value chain of AI is going to accrue to something else. I specifically said Grok Bot was a very, very good example of where I think it could go, because I think Grok Bot was massively successful, and I think people were sleeping on it.

And then Muse came out, and Muse proved that the Grok Bot model worked. This idea is that people don’t want to set up the agents on their own. They just want to talk to something, and it does it for them. So I think you should go long inference.

The next step is, okay, how do I go long inference? Obviously, the whole semis value chain has gone up because no one knows what to bet on on the software side. They just long the underlying compute that does this, like Nvidia. Then you can go up the chain all the way to memory, PCBs, and the underlying commodities, et cetera.

But in crypto, there are actually a few coins, like you mentioned, that do very different things and offer very different upside to longing this inference demand. Like you said, there’s a proof-of-work one. I’m quoting a Galaxy article here; they categorized it pretty well. Firstly, you either own inference access, mine with AI work, or lend for GPUs. That’s what crypto has right now.

My biggest favorite is really owning inference access because it’s the most direct lever you can pull. If tomorrow I tell you, “I think inference demand is going to 2X from here,” what’s the most direct way to express that view? You want to long a frontier model, right? Because if inference demand goes up, these guys are using something, and they’re probably using a frontier model or some open-source model.

The thing is that you can’t really do it. Anthropic is private, and OpenAI is private. You have Zhipu, which is listed in Hong Kong, and that’s the only vehicle. You can long the Chinese models through Zhipu, Alibaba, or Tencent, but they’re not really my favorite vehicles because they don’t have the U.S. premium. People also have very weird perceptions of the Chinese models, so it’s not clean.

But then you have Venice, and Venice is a very simple token that’s literally directly correlated to how much people use it. It’s literally a model. I think this is very unique because nowhere else can you get access to a token that you can directly bet on inference demand growing. Venice does that.

Eric Voorhees, the founder, posted a hockey-stick chart of its growth. It was something like this, and then it went up, and that’s where we are now. Then, obviously, you have the other side of it, which is the more degen side. My pick for that is Orbio, but I can go into that separately.

Take a step back for a moment. In case anyone hasn’t spent more time on Venice, you mentioned that Venice’s token is correlated or tied to demand for inference. Can you tell us more about what that means? How are people using the token, and what are they getting? When you say “inference,” in case someone isn’t familiar with it, what does that mean?

When I say inference, it means that when you type something into Claude, what happens is inference. You run a trained model to get answers.

When I say Venice is tied to inference, I mean that VVV is Venice’s main capital token. You stake it to earn more VVV, or you stake 100 to get a pro subscription. The most important use is creating DM, which is their compute token. Each DM gives you $1 of Venice usage per day forever on any model, and you create it by locking up your staked VVV.

There is some usage here, on the technical side, in terms of what VVV gets you. But the way I see it is that it’s great that the token has all of this, and what this translates to is that the market perceives this token to be the de facto vehicle to express its view of AI.

For example, in the past, if you thought privacy was going to go up, you would lock Zcash because it’s the purest expression of the trade. In the same vein, when you longed AI, people used to long TAO, like you mentioned, with subnets and stuff like that. But there’s a much cleaner and simpler long where you can long the model—the app that people use directly—and that’s Venice.

My belief is that if you want to express AI, the best and cleanest expression of the trade is simply longing Venice. Going back to the technical side of things, revenue burns VVV, so VVV actually has some buyback. It’s not a lot, but I think it’s really good because there’s some token-equity alignment.

Demand for DIEM also locks up VVV. The more people want DIEM, the more they’re going to lock their VVV up to get $1 of compute. It’s not a very complicated model. I think the most important thing is getting more and more people to use Venice, which is what I was very skeptical of when it first came out.

I was like, “Okay, the privacy model. Who cares?” But now you’re seeing privacy become very important because all the frontier models are trying to restrict your chats. If you say something wrong, they’re like, “Oh, we can’t help you with that.” You can really see that in the demand and usage of Venice that Eric posts.

With the Venice token, people are actually staking or locking up VVV to use Venice the way that you might use Claude or ChatGPT, right? That’s how simple and pure the trade is for people who view VVV as an expression of demand for AI.

It’s the most retail-friendly application that all of us associate with Claude and ChatGPT—with AI—and people are using a token for that here. That’s why I think the token has had so much success this year, and I think it’s been a pretty popular trade.

Going back to what you were saying, people search for these narratives and say, “Okay, this is the AI trade. I’m going to buy this token because it’s associated with AI.” We saw that, like you said, with TAO. Render got a big bid as well for a while.

Going back to Venice and VVV, there’s almost a spectrum of being a fundamentals investor. For me, Venice wasn’t clean enough because it still had that token-equity issue. I bring this up because there was a point where the price dipped a bit, right when they did an equity raise, and everybody was like, “Wait a minute. Why did they do an equity raise? We thought the token was the main value-accrual method.”

You had this conflict: you’ve got equity, you’ve got VVV, and you’ve got DIEM. I wasn’t sure if it was going to work out, and I guess I mid-curved it because I don’t hold Venice right now. But it’s almost funny, too, because you also mentioned that you can bet on people betting that this is the AI trade.

Even if I’m not totally convinced in it myself, I can say, “Well, I think the market is convinced enough.” That’s almost enough sometimes to say, “This is good enough. I don’t need to get the perfect checkbox of token-holder rights,” or whatever else I’m waiting on the sidelines for.

I want to go back to Orbio, too. I wrote up Orbio, I think, 2 weeks ago, and you called it the degen side of the barbell of Venice. I want you to get into it a little bit more to see if I can maybe make a catch-up trade on not owning Venice and get into Orbio.

It’s degen because it’s on-chain, first of all. It’s sub-$100 million market cap, which, depending on your appetite, can be considered degen.

The TL;DR is that Orbio is one of those tokens that I got into at around an $8 million market cap. When you’re buying $8 million market-cap tokens, it’s like we said: it all falls on the founder. The idea itself doesn’t matter as much as the founder you’re buying.

This is a guy who’s from the NFT Perp team, which produced Elon, the founder of Pump Fund, and a few others that I can’t remember off the top of my head. But I was reading tweets about him, and it was like, “Okay, this guy’s a goated builder.” People were saying, “This guy’s good.”

The idea behind Orbio, at the beginning, was simply that you could sell the inference you don’t use to people who need it. It was an inference marketplace—a real, literal meaning of inference capital markets. If I want to buy $10 worth of AI credits, I just go there and buy it.

I’ve been looking for something like this for such a long time because if inference demand is going at 20X, or whatever, you can expect there to be people who really want to buy this—big users who really want to buy it. Some sort of marketplace like that is a really novel idea.

As we know in crypto, novel ideas get a lot of premium because they’re things that haven’t been done well at scale before.

That was what Orbio started with. And now he has created something way beyond what I thought he would: a launchpad where I think his grand vision is some sort of a thought—I don't know. I haven't spoken to him. But it's all connected, right?

First, you sell inference. Then what he did was create a launchpad where you can launch an agent that buys the inference on Orbio. That makes this a reflexive, self-sustaining loop, where every time your AI or whatever runs off inference credits, you can just buy some from Orbio. It keeps building and keeps launching some sort of really, really reflexive AI ecosystem.

I did not expect this when I bought it, but it's kind of crazy. Very simply, the thesis is: good, great founder; good vertical; novel product; sub-$80 million on-chain token. We know that on-chain tokens can run up to very high valuations. My belief is that this could really be a billion-dollar project. Okay, maybe not a billion, because Pond ran to around $500 million. So at minimum, maybe $500 million. I do believe this could be some sort of on-chain hit that we've seen before.

It's funny because I truthfully haven't known anything about Orbio other than the name, and Nomatic recently wrote it up in DeFi Frontier. We used to just cover DeFi yields, but given the state of the markets over the last year or so, we've been pushing to do more research on new projects and new tokens—more due diligence on what might be good investments for this next cycle.

What I'm struck by is that there are lots of different sectors here that I do want to bet on. One of them that I feel slightly underexposed to is what you're covering here: the AI sector within digital assets. How are you thinking about which sectors to place bets on? How do you think about where you prioritize your time and, ultimately, your investments right now?

Do you have any framework you've been thinking through in terms of where you want to place your bets? A few cycles ago, it was like, “Oh, you want to bet on an AMM or DEX. You want to bet on perps. You want to bet on a money market.” Now we're thinking about some of these new and exciting applications, something like Venice or Orbio. How are you thinking about which sectors you want to bet on this cycle?

1. The Winning Sectors Emerge

Great question. Fantastic question. Top-down, it's basically 2 things. One is product-market fit, which basically just means underlying growth. Two is that the growth has to come from the intersection of TradFi or something outside of crypto—something evergreen and consistently growing—because we all know crypto revenues are very, very fickle. OpenSea was a unicorn one moment and zero the next.

This is going to be covered in the article that I'm releasing, but basically, S tier, in my opinion, is inference, tokenized stocks, and DATs. I've also included privacy, even though privacy doesn't really have that, just because it's Zcash. I admit that I have zero Zcash, so I'm not the best person to ask. A tier is perp DEXs and stablecoins.

Those are the 2 tiers that I've written. I want to emphasize that I really, really, really believe liquid investors are going to have a great time because we're spoiled for choice. There are so many real businesses being built. Again, we're reaching this point where you have demand coming from outside of crypto, and you have these tokens that are doing really, really well.

It's a great time to be alpha-picking and doing some alpha generation right now, just going through the weeds to find these tokens. Tokenized stocks and RWAs are obviously very, very big. I think one token that I've seen no one on the timeline talk about, except for the big funds, is Backpack. Backpack is up, I don't know, 4 or 5 times—some crazy amount—from the bottom.

Backpack is one of those businesses where, when I read about it, I was like, “Whoa. I have to buy this.” The founder is really, really long-term-oriented, and it's a timing thing. All of markets is eventually a timing thing, right? You could have the thesis right. All the liquid funds had the thesis right that fundamentals would eventually get better, but they were just too early.

Last cycle, you were too early. Even in 2024, I think the reason the bear market was so brutal was because you had Trump getting elected, you had Gary Gensler out of office, and you really started to have TradFi adoption. Everyone was so hopeful about this.

Then what happened was that the Trump coin launched and DATs launched—and DATs, the worst version of DATs, where it was basically a bunch of crypto people dressed up in suits selling to other crypto people, thinking that they were the suits. A lot of money got lost.

You had all of these factors that eventually led to a complete loss of hope across the liquid funds. People were like, “Oh my God, regulation is finally here. We kind of have a pump.” I think that was really a timing thing. We were just too early again.

This time, are we too early? Maybe. We could be. But I would argue that the data points show we are finally reaching the inflection point. A broken clock is right twice a day, so I'm just going to keep saying this until we finally get it. I think we're finally getting it now.

Hyperliquid has numbers up and to the right. Perp volumes are growing steadily, month on month and year on year. Perp DEXs are finally having their moment. Perp DEXs are new—dYdX came out in 2021, GMX, et cetera. It's not a new concept, but they were all too early, and obviously they didn't have as good of a product.

Hyperliquid came out, and perp DEXs are steadily growing. Everyone was like, “Perps are only a crypto instrument, and TradFi isn't going to adopt them.” TradeXYZ came out, and they did equity perps. It's all timing, and now the time feels right to me.

Another is tokenized stocks. Again, I believe we already had them back in 2021. I don't know if you guys remember Luna—the UST chain had this protocol called Mirror, where you could buy Tesla-mirrored stocks. So it's really not new.

But now all the banks are pushing it, right? BlackRock coming out and saying tokenization is the next big thing. Every single bank, in its sell-side reports, is saying, “Tokenization this, tokenization that.” Obviously, they kind of have to because the people at the top are pushing for it. But at the same time, it's real and legitimate. There's real interest from traditional finance in this.

If you could condense everything that I want people to take away, it's essentially this: bet on stuff that has underlying growth in verticals that are seeing net-new, steady-state demand grow. Then get in the weeds, find those projects, and bet on them. I think it's going to be a really, really exciting time for crypto investors.

Dude, I totally agree, and I love what you said about this being a great time for liquid investors. I personally think we're now seeing this flip. It's been a VC game for a very long time, but I think sometime in late 2024 or early 2025, it started becoming obvious that there was a lot more liquid opportunity coming down the pipe.

I think it's almost the inverse now. It's a liquid time, and VCs are just going to keep doing more of these Series A and late-stage deals in a big way. I kind of think some of this early-stage stuff—the whole industry is just consolidating. We're starting to see the winners emerge, so it's going to be harder for new, fresh ideas to really become unicorns again.

I'm not saying that's not going to happen, but there's just less blue ocean than there was before. I couldn't agree more. Even DeFi Dad and I have had so many conversations about, “Hey, I think we need to tone down angel investing,” and we've basically cut it off this past year.

Maybe the odd exception there, but it's been mostly realizing that there are liquid opportunities out there. You mentioned Backpack. I don't know if you want to get into that a bit more, but what I want to go to next here is: What are the next winners that you're looking at? New tickers that maybe you haven't spoken about in depth.

I know you have a subscription service on X, so we don't want to blow all the alpha that maybe you're sitting on or thinking about, but we want you to blow it all, to be honest. So, yeah, what else are you thinking about? What are you eyeing up to add to?

As I said, we are entering a very K-shaped world, both in investing and economically. In investing, what's going to happen is people will just flock to the few good projects that really have fundamentals. Hyperliquid is a great example, right? Everyone buys HYPE because it's growing, and it's the only project in that vertical for now that's growing. You have Lider in all of this, but Hyperliquid is undeniably the top right now, and so it's accruing all of these flows to it steadily.

It's like a reflexive flywheel where it gets more flows, the price goes up, fundamentals grow better with it, and then it just keeps growing. Literally like how MAX7 works, right? It just goes up over time. So I think my new investments are actually not that crazy. What I'm doing right now is just trying to find the projects that have a similar sort of growth and bet on them for the long term—just buy and hold. That's really my advice.

I think people are going to try to time the market. People are going to chase all of these things, and it's just not going to work out well because maybe you catch it, and you chase something else, and you catch it, and you chase something else, and then it dies. That's how these things end up. Whereas you can just buy something that goes up steadily.

Venice is one, obviously. Venice is actually my largest one. I really, really like it a lot, so I want to buy a lot more. Backpack is another. Collector Crypto, you know, that one has been heavily, heavily discussed by me. I posted an article way long ago. The token hasn't really done that well, but I continue to believe that it's one of the best businesses in the space. I really, really like the team, and I really, really like the vertical of tokenized RWAs on the gacha and collectible side. I think that makes a lot of sense. So, those 3 off the top of my head.

One thing I'd be sad not to ask you before we let you go is that we didn't ask you about any of the majors. There was a post you had around, I think, September 12. So this is just a few weeks ago. It was about, “Has the market bottomed?” What does Bitcoin look like? What's ETH look like? Just a few comments that you had here: The BTC monthly chart and weekly chart still look good.

This was, again, just several weeks ago, and I believe Bitcoin was lower than where it is right now. As of this podcast, it's trading around $84,000. You also mentioned that ETH is the cleanest bullish chart. Anyway, not to recap all of this, but any final thoughts to give us on what you're seeing in these majors? Bitcoin, ETH, SOL, Zcash. Those were the 4 that you were covering here.

I think it's actually really simple. I posted this a few days ago, but basically October is kind of a pivotal month because that's either when you're really early cycle or when you're late cycle. So, very simply, if we are at the end of the cycle in October, the next few months will probably be bad. But if we are at the beginning of the cycle in October, the next few months are probably going to be good. That's my view.

I think you can really, really safely say the bear market bottom is behind us, both sentiment-wise and price-wise. I think what people are most scared of right now is that we're having this 1-month bull market, right? Because we've seen it happen before, where it's this very, very short period of exuberance and everything comes back down.

It's at this point where I kind of just have to say, you have to believe in something, right? I think a year ago, I would also be very on the fence about whether this is the start of a new cycle. I'd be more fearful, actually, because at that time, I really didn't have faith in crypto. Everything had just crashed. But now, you just kind of have to believe that we are in a new cycle. All the data is pointing to real adoption, and so I continue to choose to hold.

For majors, I don't really have a strong view of where they go. My view is more that, as long as Bitcoin doesn't go back to, I don't know, $58K—that was where it bottomed—as long as it doesn't go back there, majors can do whatever they want. Obviously, there's still some correlation, but there has been a very, very good showing of how Bitcoin can be in a range, can go do its own thing, and then the strong alts will continue grinding up.

I think that is a trend that will continue to grow, where we are not so tethered by Bitcoin's performance. We will be, to some extent, if Bitcoin does a 5% or 10% day. I think it's natural to see alts crash. But overall, I think it's going to be less and less likely that you need Bitcoin to go to all-time highs for a new bull market.

Bitcoin could remain at $90K or $100K, and alts could still do very, very well. That's my take. If Bitcoin just does its thing, totally fine. Alts can still do well. And I actually think that's what's going to happen this cycle, where Bitcoin might not make a new all-time high. I think that's very, very possible. But I still think alts are doing well. Plus, we have so many good alts now.

Yeah, I see the untethering that you do—the untethering this bear market of really strong alts from Bitcoin's performance. Although we saw pretty similar drawdowns in most of the major tokens, some less severe than others, Bitcoin and ETH had notably less severe drawdowns versus what we saw with SOL, which I think went all the way down to around $80 or $75 there.

The silver lining to all of this was seeing some of these newer projects thrive. So, HYPE making higher highs this year was really not expected for me. That was not something I was ever going to bet on. I really felt like we were just going to see the same thing play out: Everything was going to suffer because Bitcoin was suffering. And I think it's a sign of the digital asset markets actually maturing.

So, credit to those teams and most of those tokens. I wish I had bet on those. I feel like I waited for the markets to more clearly bottom, and now I've been starting to more aggressively enter some of those trades. So, yeah, just excited and happy to see that sort of evolution in the markets.

But, guys, otherwise, I think this is a great place for us to start to wrap up. So I want to remind our listeners, they should follow Kyle on Twitter. His handle is zero, Z-E-R-O, xKyle, K-Y-L-E. Also, you should check out DeFiLlama. Again, sounds like you're contributing to all of the great work that goes on there at DeFiLlama. I know Nomatic and I are super reliant on DeFiLlama AI. Such a great tool for just getting to better understand the on-chain metrics behind all these DeFi protocols. Plus, they make it super easy now, because now you don't have to go looking through charts and all. You can just literally ask the DeFiLlama AI for whatever information you want.

Kyle, thanks so much for joining us. I really appreciated the insights you shared, and we would love to have you back in the future. I want to give you the final word before we go.

Yeah. Thank you. I really, really enjoyed this. I don't come on podcasts much, but I really, really enjoy yapping. I do think I have to keep my answers more concise, but otherwise, I always love sharing my market thoughts.

I'm just so excited, guys. I think there's really so much opportunity out in markets today, in crypto markets in particular. I can talk about TradFi markets too, but admittedly, I'm less fundamental regarding that. But, yeah, I'm very, very happy to talk markets anytime.

Thanks, everyone for tuning in. To stay up to date with future episodes, plus get expert tips, strategies, and exclusive content, subscribe to our free newsletter at the-edge.xyz.