DeFi最新2.9亿美元攻击、Kalshi对决Polymarket，以及《Clarity Act》能否通过？

There’s this amazing world where all capital markets can move on-chain, right? Where every asset in the world can have a liquid market, a transparent price, and a 24/7 order book. Where a retail investor in Kansas has the same access to deals, data, and disclosures that a hedge fund in Greenwich has.

That won’t happen if we don’t fix some of this stuff. So, I actually do believe the theme of this year is: do we get more pragmatic, or do we not?

We have the whole crew. King’s back. Santi, what’s up? Rob, what’s up?

Good to be back. You excited that Bitcoin’s almost $80,000, right? Feels like a bull market again.

I’ll admit I didn’t even know that. What’s the price at?

Hey, it goes to $80,000, and now $150,000 is back on the menu, boys.

$78,000. We’re ripping. We’re ripping. We love to see it. Man, I check prices less than I have ever checked them in my life right now. That’s probably a symptom of a bear market.

Rob, how’s the new office?

You’re building. You’re building. You’ve got the Blockworks rebrand. You’re building for the future. We’re all long-term investors checking prices every 15 minutes.

Rob, what do you think of our rebrand? Tell me your thoughts. I know you’re a great designer.

There is nothing at Dragonfly, from a design perspective or a branding perspective, that they allow me to have any opinion on. I am the last person who is allowed to have an opinion on design. Tom does literally all of it for us because he loves this stuff, and he was a PM in a former life.

That’s a nice design touch. Yeah, but it looks— I’m into it. You guys did a good job.

Dark mode. I’m all for the dark mode, man. Dark mode, baby.

1. The Fallout From KelpDAOs Exploit

All right, here’s what we’ve got. We have a packed agenda here. I want to talk about your favorite topic, Rob, which is Polymarket. We’ve got a little Kalshi-Polymarket drama from this week. Both of them are launching perps. There’s probably some oppo research that was done by Kalshi on Polymarket. A little—Shayne’s not wearing shoes to meetings anymore, it seems like. A little Adam Neumann going on.

We’ve got Kalshi and Polymarket. We’ve got Kelp DAO, Aave, LayerZero, the Arbitrum freeze, and all that kind of drama. We should definitely talk about that. We should maybe talk about these North Korean hacks in general. If you guys want to, we could get into USDai and CHIP and that launch that just happened. Then we should probably update folks on Clarity and what’s happening there.

If it were up to me, we’d spend the majority of the podcast talking about our rebrand, which is beautiful, but we won’t. So, let’s talk about Kelp DAO, yeah?

Let’s do it. Good.

Okay, I’m going to give a super layman’s version of this. Basically, the biggest DeFi exploit of 2026 just happened. There was around $290 million of this thing called rsETH. It was about a week ago, on April 18th.

Unlike some of the other hacks that have happened, this was not an isolated incident. There were all these secondary, second-order impacts that happened. An attacker, via a Tornado Cash-funded wallet, basically exploited—and I’m sorry in advance, I’m probably going to get 1 or 2 things wrong here—from my understanding, they exploited Kelp DAO’s LayerZero-powered cross-chain rsETH bridge.

There was a compromised RPC node, which fed false data to a single verifier in this 1-of-1 DVN setup. They forged a message, which unlocked or minted unbacked rsETH on Ethereum mainnet. The attacker then deposited a couple hundred thousand rsETH—maybe it was around 100,000 rsETH—as collateral on Aave.

This is where Aave gets into it. They deposited the collateral on Aave. They borrowed $200 million of real ETH across chains and then laundered that through mixers. The TL;DR is that they got the rsETH, which was this Kelp DAO staked ETH, put it onto Aave, and were then able to borrow real Ethereum. They got that out, and they sold the real Ethereum.

Where Arbitrum comes into this is, I think, there was a second attempt at around $95 million, maybe. Or there—I’m not actually sure which bucket of money this was, but the Arbitrum—

It was linked to the original bucket.

The original. Okay, so there was a second attempt of $95 million. Same hacker, it seems like, but that got stopped.

Then there’s Arbitrum. The Arbitrum Security Council met and did this emergency freeze of around $70 million, about 30,000 ETH on-chain. The funds moved to an intermediary wallet, and I’m not actually sure what’s happened after that.

There were very mixed reactions, right? You had Tae on 1 side going against Gabriel Shapiro. People were saying, “Is it really decentralized if you can do this?” A lot of other people were saying, “You’re an idiot. Of course this has to get frozen.” It’s very bad for DeFi.

Anyway, that’s my version of what happened here. I’d love to get your guys’ thoughts. It’s interesting timing, coming after we’ve been talking on this podcast so much about how you’re not getting paid enough for DeFi yields. Then this happened. It’s sadly very timely. Rob, Santi, I’d love to get your takes on the situation.

Number 1, I just want to say, you said you were going to do a bad job. You did a good job. That was good.

It was in my face every tweet for 6 days, so it’s tough to miss.

Yeah, yeah. It’s tough to miss. Listen, it feels like we’re a little late talking about this, obviously, because it’s been a week. So, it feels like a lot has been said already.

I think 1 of the interesting things here is that this is probably the first DeFi hack of this size that has multiple affected parties who are all trying to figure out how to move forward. It’s not just multiple affected parties that are big brand-name protocols and businesses.

The resolution here has become a lot more complicated than it would have been in most other situations. When Wormhole got hacked, it was pretty clear how to solve that. When Drift got hacked, it was pretty clear how to solve that, right?

What you’ve seen here is the same thing we talked about 2 or 3 weeks ago. This is not a smart contract hack. This is a hack where probably a person inside LayerZero was hacked, phished, or in some way compromised, which then allowed the attacker to compromise these LayerZero-run RPC nodes inside the security setup.

Following that, a lot of decisions—which, in retrospect, look very poorly thought out—resulted in what happened. There was the 1-of-1 DVN setup, and there’s finger-pointing on both the LayerZero and Kelp sides around who decided that and why it was set up that way.

On Aave, there was the collateral setup, so you could be at 93% LTV on rsETH. There’s also the question of what it means to allow rsETH to be collateral. There’s a lot of conversation around who ultimately bears the brunt of the full loss.

But we do know what happened: probably somebody inside LayerZero’s security setup got hacked. Then all of the resulting decisions made outside of that led to what happened here.

Now, I think at the end of the day, what’s interesting is that they all very clearly lawyered up right away. They’re all sort of pointing fingers at each other. I know there’s a lot of work happening behind the scenes to find a resolution.

It’s complicated because the incentives here are very misaligned. The Aave token has been getting absolutely crushed. The LayerZero token has been getting absolutely crushed.

You can probably make the argument that Aave is an innocent bystander here, right? You can argue about whether or not they should have allowed rsETH as collateral, what the right LTV should have been, and how they should have managed that risk. But at the end of the day, they had nothing to do with the hack. They were just the mechanism through which this attacker tried to exit.

Now, they might be the ones with the hairiest situation because all of these pools have continued to be frozen, and people haven’t been able to withdraw their loans or their deposits.

And they’re the ones with all the bad debt now, right? The bad debt is ultimately going to be borne by some people in and around the ecosystem—either lenders, Aave itself, or people around the ecosystem—but they’re the ones who ultimately have to drive that conversation. So, it’s a very complicated situation.

It tells you a lot about the interconnectedness of DeFi more broadly and the risk of these autonomous systems. Even if you aren’t the ones being hacked, there’s still the risk of these systems. I’m not quite sure yet what the resolution will be or when we’ll get it, but people are working on it, and it’s honestly a very sad situation.

I think markets always speak truth. Aave’s TVL is down 33% in 4 days. It’s down 60% from its peak, so it’s sitting at just shy of 15 billion. At the peak, it was 43 billion. Of course, Aave was not as affected here, but that just tells you, to Rob’s last point, that composability is a tricky thing.

I think we should talk about what the implications are for vaults and for DeFi writ large. The second point—so, that’s Aave. Aave has lost a third of its TVL in a matter of days and is down 60% since its peak in April of last year. DeFi as a whole peaked, I think, at 99 billion. It’s now at 85 billion, its lowest level in over 12 months, and roughly 50% below its October 2025 peak. That’s something we should keep in the back of our minds.

To put things in context, Tae is obviously a very vocal security expert. I know Hasib has tweeted about this as well—your counterpart, Rob. It’s a very adversarial environment. I was trying to piece together how much Lazarus pulls in as a percentage of North Korea’s GDP, and it’s staggering. It’s a professional, full-time unit.

This month alone, Lazarus has pulled in roughly 600 million. That’s what they can get off-chain, but every year they’re pulling anywhere from 600 million to 1 billion in hacks. That’s anywhere between 3% and 4% of North Korea’s GDP. That is massive. You are working up against a very serious threat. For all intents and purposes, Lazarus is the most important unit of North Korea as a country. That’s wild.

Yeah, Santi, I could not agree more. I actually tweeted this out. I said, “Everyone needs an enemy.” In crypto, it feels like we’ve been a little lost for the past year—an industry without an enemy.

To contextualize that, for years the banks were our enemy. We were like, “Down with the banks. Let’s debank the banks.” Now the banks are our customers. Then the SEC and Gary Gensler became our enemy, but now the SEC is our friend and is supportive of us.

I think this made it abundantly clear, if it wasn’t clear already. The North Korean hackers, aka Lazarus, are now the enemy of the industry. If we’re not able to fight against them and protect capital in the system from Lazarus, the industry will not grow.

Even though there was fighting between Tae and Gabriel Shapiro, I actually believe we’ll move past some of that fighting. Having a common enemy tends to bond people together. Even though we’ll have weeks of turmoil here, I think it will ultimately be good for the industry to come together, if we’re able to come together and fight against Lazarus.

One point there, Yano, I totally agree. We’ve always known about Lazarus’s existence, and I unfortunately don’t think they’re going to slow down by any stretch of the imagination. That’s at odds with the willingness of founders to experiment and create interesting protocols that fit in with DeFi’s interesting properties.

I worry—and we should talk about it now or keep it in the back of our minds—about how much permissionlessness is going to be constrained. What is the healthy middle ground? These are questions I don’t know the answer to, but I do think there comes a point where, as an industry interacting with regulators, we have to err on the side of being more practical.

That doesn’t mean you can’t interact with permissionless networks. It means timelocks, limits in terms of oracles, and slowing down a bit. When it comes to security, I think it’s a good trade if we go a bit slower to protect the industry and put in some training wheels, because right now I don’t see a path to recovering to all-time highs in TVL.

We talk about stablecoins a lot. Thankfully, I think stablecoins can be frozen, and I think both Tether and Circle have the ability to freeze them.

Everyone does, and I think that’s okay. Let’s not forget that DeFi and stablecoins are filling a huge unmet demand in most of the world that doesn’t have access to very basic financial products.

In many ways, I’m glad that we don’t have over 100 billion of TVL. It’s a huge honeypot.

Yeah, listen, I totally agree with you. Gabe’s obviously done a lot for the industry over time, but this idea that we’re going to have a fully permissionless ecosystem, with all of these protocols that essentially operate as businesses, and that they’re not going to be reactive to crime or regulators—it’s a panacea that a lot of people in the space are very ideological about, but it’s just not reality.

It’s not going to be reality if we want to get regulated, if we want this space to get as big as we think it will be, or if we want it to be the future of finance, as people talk about. Ethereum will always exist as a completely permissionless L1, and Bitcoin will exist as it is today. But there is a duty to stop crime if you can.

Basically, every protocol that has been successful outside of Ethereum and Bitcoin has the ability to do these things because they operate like businesses, and they are going to respond to pressure from regulators and law enforcement. At the end of the day, I care a lot more about building the new rails of a new future that is better for retail, better for customers, and allows global access to new types of assets than I care about whether it’s fully permissionless.

2. Fidelity Crypto Ad

Whether a 12-person Security Council can stop North Korea from taking our money is a discussion that, in my mind, does not matter. I know I’m going to get tweeted out for that because some people obviously care a lot.

3. Crypto Is At An Inflection Point

I don’t know. I agree, and this industry is clearly at an inflection point.

Santi and I recorded this great podcast that comes out on Monday talking about quantum and Bitcoin. We ended up talking a lot about Satoshi’s coins, and there’s clearly, to me, a right answer for what to do with the Satoshi coins.

It goes against what Gabe Shapiro would want to do. You can listen to it—maybe I’ll force you to listen to that podcast—but it’s the same thing that I think should happen here: We need to put protective measures in place and be a little more pragmatic with the industry.

This industry is at a tipping point. You take the blue pill, and we’re going to get three things: stablecoins, some institutional adoption, and perpetuals. That’s a nice world, but it’s not the world we spent the last decade of our lives building this niche technology for. We didn’t build a nice little niche technology that applies to a few small corners of the internet.

There’s an amazing world where all capital markets can move on-chain, where every asset in the world can have a liquid market, a transparent price, and a 24/7 order book, and where a retail investor in Kansas has the same access to deals, data, and disclosures that a hedge fund in Greenwich has. That won’t happen if we don’t fix some of this stuff.

I do believe the theme of this year is: Do we get more pragmatic or do we not?

I agree with all those takes. The most important thing in anything regulatory or in building—crypto is really good at coordination because you have certain guarantees and transparency—but we shouldn’t let ideology get in the way of implementing these systems, especially when most of the growth is going to come from things that inherently have much more connectivity with the real world, like RWAs and stablecoins.

You’re effectively tokenizing things that exist. There’s an SPV, there’s a legal structure, and it touches atoms in the meatspace. That’s just messier.

And that's okay because you still gain a marginal improvement. It's still a marginal improvement, to your point, Yano. It's more transparent, but you're still required to coordinate. For instance, if you have an issue with the Arbitrum Security Council, you have the option to leave the system.

I could tell you that 80% to 90% of the people interacting with Arbitrum see that and say, “Yeah, I actually like that,” because there's definitely a nonzero chance that something else continues to happen and there's a hack. I feel much better interacting with Arbitrum than with another system that's not going to act to protect my funds. You can look at who those 12 people are. If you trust them, great. If you don't, leave. You have many, many options. That's the beauty of crypto, right?

Can we go back to this KelpDAO situation? There's a really interesting thing that can happen. They can either socialize the losses, or they don't have to socialize the losses, right? The premise is that KelpDAO's rsETH were stolen, not minted. So it's kind of up to KelpDAO to decide whether to socialize the losses across all the holders, or have a subset essentially take the short end of the stick and get a little screwed here.

Wait, just to be clear, what happened with KelpDAO was that they minted unbacked rsETH, right? That unbacked rsETH was put into Aave as collateral to borrow ETH.

Yeah, which means there are now more rsETH, so the value of rsETH is lower than the value of ETH in the system.

Yeah. I think the value of rsETH is about 18% lower than ETH, right?

Correct. But I don't think that decision is going to be up to KelpDAO, right? Do you think it's a legal decision?

Well, I think people are trying to make it a legal decision. I think the person who's going to end up making that decision is Aave. I think that's what's going to happen. Not to get too deep into this topic specifically, but there are incentive differences between Aave, LayerZero, and KelpDAO.

KelpDAO is a protocol that had done a good job on BD and had gotten its tokens put into a number of different protocols as collateral and as something that could be used within DeFi. Following this, it's not clear to me that there's a future for KelpDAO.

Definitely. I agree with that.

Yeah, okay. Just to say that in a nice way. There's very clearly Aave, where people are the most angry right now, because those are the pools that are locked and people cannot get their capital out of. Aave has a large business, and even with all the drawdowns, I still think it's bigger than Morpho. It's still the largest money market, or lending protocol, on-chain.

At the end of the day, Aave has the most to lose. It also has the most to gain from trying to get to a resolution that makes people happy. My perspective here, and some of what I'm hearing in the background, is that more of the discussion is being driven by Aave. Obviously, it could just be a legal point, but I don't think there's any incentive for KelpDAO to make a decision here.

That's a good call. So if you're Aave, Aave's in this weird kind of prisoner's dilemma, right?

It's a very interesting situation.

But they're the one who has the most to lose.

Yeah.

On Aave, double-clicking there: whatever ends up happening—TBD—I do think it probably is KelpDAO and Aave trying to both socialize part of the losses. The supply caps and the risk-management parameters at Aave definitely need to change. I don't know if they've come out formally with more guidance, but that's something we should definitely change going forward.

Timing, right? We talked a couple of weeks ago about Chaos leaving Aave and the risk stuff there.

Yeah. I mean, if you're a traditional bank, you have a committee, and then you have capital rules.

What would you do?

All right, let's—sorry. Go ahead, go ahead.

No, no. Ask the question. I'll answer.

Well, just to be clear, I want to really double-click on one thing Santi just said there. We're talking about Aave—he's saying, “Would it be KelpDAO or Aave?”—but we're not really talking about LayerZero. LayerZero is the one that was hacked. It's a weird thing that has happened.

By the way, I think the most capitalized entity here that could socialize the loss is LayerZero. But again, there are really weird incentives in this situation, because they're infrastructure. They don't have retail customers. The retail customers are Aave's customers.

If I were LayerZero, I would give a loan to Kelp and get them in good standing, then get it paid back with fees over time. That way, KelpDAO survives, LayerZero maintains institutional trust and its ability to win over more customers—even though they're launching ZRO—and their brand doesn't take a huge hit if they do something here. Then Aave commits to increasing caps and maybe increasing its security spend, or somehow that also needs to change as part of this brokered deal.

I think 50% of LayerZero's app DVNs run a one-of-one DVN, right? I heard privately that they're telling everybody they have to change that now.

Yeah, okay, that's good. I wonder. I haven't talked to Brian. I think Brian Pellegrino, one of the founders of LayerZero, has been quiet during this period. Maybe next time we're recording, we should bring him on. I'm sure he's being advised to be quiet. He's not being quiet by choice.

Correct. Or maybe they're just trying to fix things under the hood before they become more public about it. You tweeted about Guy from Ethena. I was actually talking to him over the weekend. I also pinged Stani and a couple of other people to see if there was anything they needed help with.

Guy has been public about this stuff because they're also working with LayerZero, right? It would be interesting to have a couple of builders on down the road to see what changes from their perspective.

Right. But all this to say, Guy had a great tweet. What he basically said was, “I'd encourage all asset issuers to consider rate limits at the mint and redemption level, as well as customer rate-limit configuration on top of LayerZero OFTs. In a disaster scenario where the LayerZero DVN is compromised, you can at least contain the damage to $10 million per chain per hour before stepping into a shutdown of transfers.”

This is the key part, I think, Santi, which is really what you're getting at. It's a slightly annoying inconvenience for users 99% of the time, but a worthwhile trade-off to avoid going to zero. It's very similar to when you send a bank wire: you have to double-check that you're sending it to the right person. That's pretty annoying, but one out of 100 times, it saves you.

Yeah. Two observations, really quick. The Column CEO, a former co-founder of Plaid, talked about this really well on Invest Like the Best. He said, “A lot of times, what crypto's trying to solve is not a technology problem. The reason fintechs and banks haven't really implemented this faster is because we know there are certain clients who are just going to get hacked, and it's important that we're slow and make it harder for people.”

The other observation is, let me put this in an analogy that anyone can understand. Sometimes you have to make it a bit absurd to understand why rate limits are needed. Imagine that all of a sudden you could go to the bank and tell them, “Here's a piece of paper claiming that I have some fictitious real estate somewhere they haven't heard of.” Then you could borrow against that, and no one would be there to double-check that the piece of paper, or that parcel of land that magically appeared out of nowhere, was real.

Of course, that would never happen, because a bank would review and approve it. In a similar manner, I think that's what Guy is talking about. If you're Aave, rate limits are important because if you see an abnormal spike, there should immediately be a review process—a time lock there, almost.

Of course, there are going to be people who say that breaks the whole feature of DeFi and that we might as well go back to the real world. I would disagree with that, but that's essentially what happened here with the Kelp token. I would just make the point that, for a lot of the really serious founders in the space—the ones who have really understood that this is a trust business—security is part of that.

This is not just software; you are running a finance protocol or a finance business. In a finance business, security is at the core of what you do. Trust is at the core of what you do. People like Guy—and there are others; I think the Agora guys tweeted about this, and I know there are others who did as well—put these measures in place from day 1, with nobody telling them they should do this.

They put the rate limits in place. Despite the fact that the LayerZero documentation may have suggested something else, they used a 3-of-3 DVN versus a 2-of-3 DVN or a 3-of-5 DVN. These people were very thoughtful about making sure they had extra layers of security that nobody required them to have, but that they understood were important. Guy was one of those. To your point, Santi, there’s a whole world of DeFi right now where we have technologists who think they’re building software, and they don’t realize they’re building finance. When you’re building finance, the requirements and obligations are different.

I think we’re going into a future where there has to be consolidation around those founders who understand that and understand the pressure and responsibility they’re taking on.

I think, yeah, going forward, I would like to see this. This is something that I spent quite a bit of time on when Lend transitioned to Aave V1. I still remember it was nerve-racking. There was $1 billion migrating, and I couldn’t sleep that night. I mean, Stani’s balls of steel. I’m sorry, man, but managing that—there’s a lot of money at risk. I was thinking about that today: How does that guy sleep? I respect it.

The SWIFT messaging protocol, which is the equivalent of LayerZero here, does not have 1 verifier. It’s a group that messages and coordinates. If there’s a bank in Switzerland sending money to JPMorgan in New York, it goes through multiple approval checkpoints. That’s the same thing that should have happened here.

Practically, someone at Aave should have called Kelp—literally called someone and said, “Hey, guys, are you doing all right over there?”—because I just saw $292 million come in here trying to borrow immediately. What you could have done is slow down the borrow market. You could have probably frozen it and said, “Until we are absolutely certain that these 292 million receipts coming into Aave are sound collateral, we will not open the gates to borrow against the new collateral that came in.”

As simple as that, right? I’m not saying you immediately freeze it, but if you see an abnormal spike, it’s most likely going to be a hack or Justin Sun moving money around to yield-farm. You should still lock it. All of this could have been avoided with that. We can inconvenience Justin Sun, I think, to stop hacks. I’ll be okay with that. But that’s as simple as that, right?

Can we maybe round this out by talking about Aave’s token? It’s down, what is it, 14%, 15%? Morpho’s up 10%. Spark is up 140%. How much do you think this is a reshuffling of the crypto lenders? And is this a buying opportunity for any of those?

Yeah, I mean, certainly. Again, I think that if Aave doesn’t implement more safeguards, then it’s an opportunity for someone else to come in. Jason—or Yano—I’d actually be curious to get your takes on isolated markets.

Well, yeah, I mean, that’s where I was going to go next. I think the whole world’s going isolated. Obviously, there’ll be part of the crypto ecosystem that will continue to do these shared markets, but in the world of institutional adoption, the institutions want isolated markets. Every single one of them wants an isolated market.

There’s been this period of time where the non-isolated, or aggregated, markets have grown really, really large because they made complete sense for that period of DeFi. All of the growth that Aave and Morpho have been focused on recently has been winning fintech deals and institutional deals. We’ve seen things like the Coinbase Bitcoin Morpho market, which, through the CeFi app, grew to $1 billion very quickly.

We’re going to continue to see a lot of that. Basically every fintech in the world—if you’re talking to these people and having these conversations, all fintechs are looking at doing stuff like this—and not a single one of them is going to do anything but an isolated market. I think the reality was that this was already true, and this only accelerates that future.

To lend exact clarity to your question, Yano, siloed markets include Morpho, Euler, and Aave V3. V4 is this hub-and-spoke model. So they understand this, and maybe that is the writing on the wall.

To your point, Jason, I think if you’re a traditional firm coming on-chain, you will not deposit into a pooled market with shitcoins. Well, they will see it as a shitcoin. Everything crypto-native will just be looked at and haircut a ton, to a point where you have to go to isolated markets.

Of course, the trade-off there, Jason, has always been capital efficiency. But if you think about where we’re sitting, at least for Aave, at $14 billion, if you really want to break through the $100 billion mark, you go to isolated markets. The more interesting exercise is which isolated markets get the most liquidity. It’s RWAs. Figure is essentially an isolated market, right? It’s just HELOCs, which is a topic for a separate day, but obviously different quality is attached to all these different mortgages.

Let’s shift gears. Anything else on the topic? Cool. Maybe a question, Rob: Are you guys investing in DeFi these days? Bro is investing in DeFi. But what’s the biggest—what’s the latest? Ethena, maybe? What about DeFi infrastructure?

We’ve done a bunch of DeFi since Ethena. Ethena was summer 2023, so we’ve done a bunch of stuff—some of it public, some of it not. We’re big into Lighter, which is the one that people know, but we’re in a bunch of stuff, too, that hasn’t launched yet or is still in stealth.

DeFi continues to be something we believe strongly in. I think the world is changing, and the way DeFi exists is changing, but I’ve never lost faith in the fact that there will be finance on-chain and DeFi will be a large part of that.

I haven’t lost faith as much. But, yeah, I’m more of a true believer than you, maybe.

I believe a lot in DeFi.

4. Kalshi & Polymarket Push To Launch Perps

Okay, Polymarket and Kalshi. It seemed like the Kalshi team was doing their oppo research and decided to dump it all in 1 fell swoop this week, kind of taking some shots at Polymarket. The context here is that Polymarket and Kalshi both announced that they’re doing perps. I forget who beat the other to the punch by a day or something like that. I’m sure Rob will correct me. Rob, would you like to correct me?

1 of them did beat the other, yes, to the announcement. It doesn’t matter. They keep doing this thing where they can announce things and front-run each other. So they both do it.

Yeah, yeah. So, anyway, there was a Bloomberg article that said, “Polymarket Loses Prediction-Market Lead After Delays, Blowback.” The article said, “Gaffes, delays in product rollouts, and Shayne Coplan’s unconventional management style are contributing to the strains on various fronts,” said people.

Then there was another article, or maybe it was the same article. Let’s see. Yeah, I think it was the same article. It said, “While Coplan’s quirks are in many respects standard fare for charismatic tech CEOs, some business associates who declined to be named, discussing private meetings, expressed frustration with Coplan’s focus and timing. He’s regularly late to private meetings, has attended at least 1 of them barefoot”—best line—“and is easily distracted, texting and taking phone calls in the middle of the conversation.”

Very much like what Santi does on our podcast. I’d love a little context. As someone who at 1 point owned a news business, anytime you see stuff like this, it is typically from the main competitor feeding reporters. I’m not saying Kalshi’s the only one who does this.

I'm sure Polymarket also does this with Kalshi, but I would just take all of this stuff with a big grain of salt. I also saw another article that said something along the lines of, “I think Kalshi suspended the accounts of 3 politicians from trading on their platform.” It's a very good week. There's clearly a strategy internally to do a big PR push in 1 week.

So, I'd love to get your guys' thoughts. Isn't it all kind of petty? Why?

I mean, this is just business. You remember the Uber-Lyft days, right? The same stuff was happening. I will say, I do think there's a distinction. I would separate the perps from the stuff that's happening on the news side, because I think the perps are interesting and worth talking about as a business decision for both of them. They're both doing it differently, right?

Kalshi is obviously trying to launch its perps on a US exchange in a regulated way, through its DCO and DCM. Right now, perps in the US are illegal. Coinbase has its sort of synthetic perps, which are like 5-year futures. They throttle risk overnight and on the weekends, and that product has basically gotten zero uptake at the moment.

There's been a conversation for a long time about what the US will do from a regulatory perspective on allowing perps. Will there be a situation where the CFTC comes out and says, “Hey, listen, we want to bring perps as they exist internationally in crypto into the US. How do we want to think about appropriate leverage for retail versus institutions? Should that sit under current futures frameworks?”

At the end of the day, perps are really good retail products, but they're not really institutional products. Institutions mostly prefer a dated future or an option. You can see that with the fact that IBIT options just blow all other on-chain options, and also CeFi options, out of the water in terms of volume, and are doing probably almost all of the institutional derivatives volume in the space at this point.

I think it's super interesting, and I know Mike Selig, Chair Selig from the CFTC, has teased that there's going to be perps guidance coming out sometime soon. We don't know exactly what that will look like, though. I think there's probably some back-and-forth with some of the current US exchanges about what they want to see. I'm guessing some of them don't want to see perps happen because they think they'll lose market share, especially on the retail side.

I think that is very interesting. Obviously, there's a very big opportunity there, especially at the retail brokerages like Robinhood and others. The Polymarket perps are an on-chain product, so they're much more competitive directly with the other on-chain perps products that they're looking at, at least for the moment. I'm sure they'll be looking at their centralized-exchange US business as well eventually.

It's a little bit of a different business decision for both of them, especially in the current regulatory climate. I think it's smart for both of them to do it. Anywhere you have a captive audience and captive traders, trying to offer them more ways to trade is a good thing, not a bad thing. I would never say that adding a new product is bad.

There is a question around whether or not perps traders and prediction-market traders are the same group of people. I think there was some analysis done by some Hyperliquid folks that said maybe there's 12–15% overlap between the two. That isn't that much, but it's enough to make it worthwhile to add the product.

Santi, do you think Kalshi and Polymarket have any chance of having perps take off on the platform?

Yeah, I think so. I think they'll work. There was a good post by Ryan Watkins talking about the conversation around, “Will these folks eat Hyperliquid's lunch?”

Polymarket is probably better positioned here. I'm biased, of course, but I think they'll take off. Maybe we should reference that Ryan Watkins post, or we should have him on the pod. There was a pretty good back-and-forth on that thread. He obviously is more in the camp that Hyperliquid will win. That doesn't mean other folks can't build their own successful perps.

We know users like this type of product, right?

Well, go ahead. Go ahead, Santi.

I was going to ask about a comment he made that I haven't really explained myself. I'm not sure this is actually true, but he said, “Look, Coinbase perps haven't been very good. You can't just list perps as if they were any other derivative. You have to reproduce the architecture. Coinbase has already demonstrated this empirically with their lackluster CFTC-regulated, quote-unquote, perps products, despite plenty of talent and dollars thrown at it.

“As currently designed, they're long-dated futures with 5-year expiries, 3–10x leverage depending on the contract, and funding that only settles twice daily. Compare that to unregulated offshore venues like Binance and Hyperliquid.”

He's making the comparison that the product has been very underwhelming. I don't know if that directly translates or is applicable to Polymarket or Kalshi. Maybe it's a Coinbase skill issue, but I'm not equipped enough to have a perspective here, candidly.

Well, no, he's right. This is what I mentioned at the beginning of what I was saying: Coinbase launched this long-dated futures product and called it a perp, and it hasn't done well. That is obviously directly relatable to what Kalshi is doing, because Kalshi is also going to be CFTC-regulated. The product that they're launching is going to be through their US DCO and DCM.

The difference, or I guess the variable, is going to be what the CFTC does with the future of perps in the US. We know that Chair Selig and his team are thinking about this fulsomely. We know that they're trying to figure out a way to bring actual perps onshore in regulated markets. We know that the IBIT options market has done incredibly well.

I do think this idea that Kalshi and Hyperliquid are directly competitive is probably wrong. They're probably just completely different markets and completely different user bases, because Kalshi is trying to distribute perps to Robinhood users, right?

My guess—and I've always thought this—is that Robinhood does so much volume on the options side. Around 70% of it is zero-day options. These are retail traders looking for leverage. These are retail traders looking to get long, not because they have a general, well-thought-out fundamental expression of something, but because they just want to do something that a perp allows them to do even more cleanly.

For that user base, depending on what the CFTC allows, I expect that to do well. I don't think that's Hyperliquid's user base, right?

The Polymarket and Hyperliquid products are probably more directly comparable because they're both DeFi protocols, and they're both products that are going after a large international user base. They're also much more direct-to-consumer products. Kalshi isn't much of a direct-to-consumer product.

It will be interesting to see how this happens. The reality is that there are a lot of traders who are perp traders, who are Hyperliquid-first, who will sometimes look at a HIP-4 market and say, “Oh, well, I can potentially cross-margin.”

Right now, there's a lot of talk about HIP-4 markets, which are the Hyperliquid binary-option markets, having cross-margin with the perps side. But there's actually nothing in the documentation about that, and I think that will be quite hard to do. I don't actually expect that to happen anytime soon.

I'm sure we'll get some Hyperliquid traders who will trade those markets, especially mostly, if not entirely, on these crypto up-or-down markets, specifically—not the other stuff that people talk about with prediction markets.

On the Polymarket side, they have all of these traders who want to come and trade all of their prediction markets, who will occasionally trade the perps side. But I think this is not zero-sum, despite the fact that Twitter wants you to think it is zero-sum. Everyone's saying, “These people will eat each other's lunch.”

Prediction markets are going to get 10x bigger from here, and perps are going to continue to get way bigger as well. I don't know if they'll grow 10x, but I think both of them grow and serve different users.

What do you guys think of this from the investor seat? I know it's a little petty, and Santi, I think you're right. Rob, I do think you are right that this is just business.

If you guys remember Uber-Lyft—I don't know if you guys remember Operation SLOG, where Uber was having all these brand ambassadors—

They were giving them burner phones and signing up for Lyft, and then they would cancel the Lyft right at the last minute to take Lyft drivers off the market, essentially. They were giving them literal cash just to convince the Lyft drivers to come over. So I do think this stuff happens a lot.

But maybe from the investor seat—Santi, I think you guys see this—did you guys seed Polymarket, or are you invested in the A round or something like that? And then you guys have a big check. So when you see things like these things against Shayne, from the investor seat, do you guys care about this at all?

Yeah, I did the seed when I was at ParaFi. I think ParaFi is among, if not the second- or third-largest holder of Polymarket. And I know the team; since I left, they have been doubling down and participating.

Shayne's the kind of guy who, I don't think, is paying much attention. I'm skeptical when a founder pays too much attention to the competition and focuses too much of their time on that, publicly or privately. It's like, dude, just own your product and you'll crush, right?

I don't think it's zero-sum. So, again, I think it's about—look, to throw dirt, it is a bit unfair because, I mean, you could argue that Kalshi has thrown a lot of dirt Polymarket's way, probably more than Polymarket seems to. I'm biased, but they have a whole apparatus around it.

Yeah, yeah. So to me, it's always weak. It's like—I don't know, you probably know this more—but who was more focused on the other? If you're truly winning, you're in your lane. You're unbothered, moisturized. You're just crushing, right?

It obviously hurts when someone's trying to go after you in a very malicious, ill-intended way. You should respond to that aggression, but I don't know, man. Both of them have built nice businesses. Stay in your lane. Win on fair grounds. Don't do—be a good sportsman is all I'm trying to say.

Yeah, I think what really bothered me, to clarify my point earlier, is that I think generally opposition research is fine, right? It's business, and generally continuing to highlight why you think you are better than your competitors is totally fine.

I think where I've really struggled has been what is factually a lack of, or an extreme willingness to lie. It's not just, “Hey, we talk about how we're different.” There's been an extreme willingness to fabricate stories and lie about the things that are happening between the 2 different parties from the Kalshi side.

There is an ethical standard in my mind that continues to be breached. That is not okay. I've been very vocal about this on Twitter and X and other things, because absolutely go and tell people why you think you are better than your competitor. That's great. But don't send out your head of comms to go and lie about something that happened between the different platforms that is very clearly and easily verifiably different.

There was an article yesterday about some of the volume between the 2 parties that was factually untrue, and this Barron's reporter didn't even check it. He clearly got it from Kalshi and didn't even check the work, because it was obviously untrue. We saw this yesterday, too, in the Bloomberg article. I chatted with a few people about it, because there was a paragraph that said, “Polymarket has basically zero volume. There was de minimis volume in their U.S. app, and they haven't been able to really grow that at a time when smaller competitors have been able to.”

If you go look at the data, they've done over $1 billion in the last 30 days on the U.S. app, far more than any other competitor. There are little things like that where it was clear that somebody picked up something that was basically written for them and didn't do the work to put it into a story.

I didn't care so much about the story itself. I cared about what seemed like a lack of—nobody did their job at Bloomberg on that article.

I will say, reporters are fiending to get content here and throw dirt. This is every other week. I'm sure you guys get hit up all the time, and it's probably more of an—unethical stuff is bad. Obviously, don't do it. If you're going to criticize someone, back up your stuff.

The problem is, media doesn't care to check. But you know, I don't need to tell you this, man. This is why Blockworks exists. This goes back to the famous Napoleon line: “Never interrupt your enemy when he's making a mistake.”

I think Kalshi, time and time again, just—it’s a bad look, right? All these things are easily verifiable, and if they're not true, then it's just a bad look.

Yeah. Yeah, I mean, I think I had a point. Go ahead. Go.

Well, just to give you one more point on this: I got hit by a reporter yesterday afternoon on another negative Polymarket story. It was, “Oh, hey, I heard from a couple of people that this is true.” It was so obviously untrue that I knew it was factually untrue in the most obvious ways.

I asked the guy, “Am I the first person you're talking to who is not aligned with Kalshi?” He said, “Yeah.” I said, “And you're telling me you're going to publish this tomorrow?” I didn't understand, because if he had asked 3 people who were not aligned with Kalshi, he would have known.

Some of this is happening right now because it is a story and a rivalry that has so much interest in the public. You're getting people who are just willing to pick up things that their people are sending their way and put them into print.

I was talking to some of the Polymarket leadership about this yesterday. They didn't staff up, and they haven't really staffed up, the comms department. In fact, they don't have an internal head of comms. They're trying to hire one right now. They put an offer out for somebody, but they've been using outsourced agencies for this stuff.

This was never what they expected to be spending their time thinking about, and it's not what Shayne thinks about. It's not what he cares about. On the Kalshi side, this was part of the strategy from the beginning. They've been doing stuff like this for a while.

One of the things I said to some of the Polymarket folks yesterday is that we probably just need to take this more seriously, because I can now tell there's a constant barrage of opposition research and negative stories coming that are very clearly coming from the other side.

Mhm. We should have someone like Matt, or the Paradigm guys, or someone who's in the Kalshi camp come on here, because I know you guys are biased on this.

Corporate rivalries often spill into the public eye through marketing and comms, and I think there's a right way to do it and a wrong way to do it. I don't know if you guys have listened to Acquired, but these stories often come up on Acquired. I was just listening to the Coca-Cola episode, and one of the most famous rivalries is obviously Pepsi and Coke.

There was the Pepsi Challenge. In 1975, Pepsi started airing commercials showing blind taste tests where consumers preferred Pepsi. That actually pressured Coke to launch New Coke, which was one of the most disastrous launches in the history of consumer goods. It was only because of the Pepsi Challenge that they felt pressured to do this.

There's a way to do it that can be very savvy and strategic, but you need to do it in a way that leads with integrity, I would say. I don't know if Kalshi is doing the wrong things. I know some folks over there, and they seem like great people, but I'd love to hear from them about their comms and marketing strategy.

So maybe moving Kalshi—

But that's the thing. Go ahead. Over time, stuff comes out, so you can rarely get away with it.

5. Will The Clarity Act Pass?

Yeah. Also, a quick update: one of the listeners wanted to know about Clarity.

I think Galaxy’s Alex Thorn came out and said 50/50. I’m not sure what Polymarket is showing, but it was a sub-50% chance that it passes the last time I looked. I would say that we are above a 50% chance of passing. And, Rob, I know you’re pretty deep—nope, you’re shaking your head. You’re saying we are now below 50%.

I’m going with the over. I’m going with over a 50% chance that this passes. I think I said it on the pod before, but I always thought we were about a third, so a 33% chance. When I talk to people in D.C.—lobbyists, people who are actually there every day or spending a lot of time on this, who are not what I would call people who are paid to be hopeful about crypto—it’s the number that I have been getting all the time.

I think Rebecca Rettig said this when Santi and I did a podcast with her, and maybe that was in February, but all of those people have seemed to be very aligned that it’s maybe a 33% chance. The story that I hear continues to be that that’s where we are today. The longer this goes on, the lower the chance is.

I think there was an article that came out, maybe in CoinDesk, that was like, “Oh, well, there’s still time if we get this through July.” I’ll tell you, when I was in D.C. last month, every single person said to me that if this is not done before Memorial Day, it is dead. Again, this is an option, right? There’s theta decay happening right now.

I think Polymarket hit a 37% chance at one point yesterday and then popped back up and is now down to 43%.

I expect this to probably settle somewhere in the 30s for the next month, and we’ll see what happens.

You expect it to settle at what?

In the high 30s.

It peaked at 65% a little over a month ago. I can tell you I’m getting the same read from people: every day that goes by, it is more unlikely that it gets passed.

Listen, I want it to get passed. I’m obviously not paid to be bearish here, but I try to be a realist, and I just get less confident every day. It’s down to a 42% chance now. It probably came down while we were recording. People probably heard us. It was Will on the other side trading it, but I think—

A whale. Will was like, “I see an arb. I see an arb.”

Yeah, well, we have a history of degen producers on the show.

Joseph’s out here, just retired in Alabama or Arkansas or something—

Buying multifamily properties. No doubt.

So, okay. Well, CLARITY—we’ll keep you guys posted as we hear more.

6. What Is USDai?

Last but not least is USD.AI. I don’t know if you guys followed this, but USD.AI launched, and I knew very little about it. I only knew about it because I remember the Framework guys—Michael and Vance, who we’re good friends with at Blockworks. I think they invested in them after they had pivoted from being an NFT protocol. I was like, “All right, let’s see how this thing does.”

We’re in USD.AI as well.

You guys are in it? Okay, nice.

Every now and again, there’s a new coin that performs incredibly well after launch. Oftentimes, things are down-only after TGE, and USD.AI launched the CHIP token and skyrocketed. I think it traded at over a $1 billion valuation.

It’s hit every narrative, right? The AI narrative, and they actually have real customers and revenue and a real business. It seems pretty cool what they’re doing.

For people who don’t know about it, my understanding is that there’s the USDai stablecoin and the sUSDai yield product, and that’s tied to AI compute and GPU lending. Rob, you’re an investor, so correct me.

It’s collateralized loans against compute, or against the GPUs themselves.

Got it.

Essentially, it is a proxy for compute demand. That is what it is a proxy for. Clearly, right now, that is a very hot topic. It’s topical for people trying to get exposure to that, and it is a problem. We are compute-constrained, and people are also CAPEX-constrained, especially if you’re not one of these big foundational labs.

All the credit to these guys. They’re really good credit people. Their last protocol was NFT-related, which is true, but it was specifically about credit and how to do different types of credit around different types of NFTs. They’re really smart and really good at this part of the market.

They saw early that there was going to be demand for lending against GPUs and hardware, and that wrapping a token around it gives people access to that and to yield related to it makes complete sense.

And is this David Joyce’s thing?

Yes, exactly. It’s David. I’ve met him a couple of times.

Quick question: Who are they lending out to that does not have access to credit in the real world? If this is a very hot, investable category for most lenders in the real world, what gap in the market are they filling that is not being met by traditional allocators?

I would push back a little bit on the idea that this is an easy market for startups. There’s obviously lending against compute for one of the foundational models and a lot of the big companies, and you’re seeing a ton of demand there. But for earlier-stage AI-linked startups or compute-linked startups, it’s still something that a lot of people are trying to get their hands around.

These guys have a very attractive cost of capital. They’ve done a very good job underwriting this part of the market. I don’t know if it’s public who all of their customers are, but it is earlier-stage companies that are adopting this market. They’ve processed about $100 million in loans, I think.

They’ve got a lot more TVL today, but they have capacity. They’re being thoughtful about who they roll out to and how they underwrite, so they have more capacity than they are extending in loans.

Is that a quality of the business? They have, I think, over $1 billion or so that I saw in the pipeline, and they’ve extended 10% of that. But they have more deposits of idle stablecoins than they have people willing to lend against GPUs and H100s?

They’ve got $61 million in active loans out. They’ve got another $59 million that has basically been signed, and they just haven’t sent the loans. They’ve got a pipeline of over $300 million that they’re talking about right now, and they’ve got about $350 million of capacity at the moment.

Capacity, yeah.

Content of the week: Project Hail Mary. Did you guys see that movie?

You said that last time.

Did I say that last time? I think so, or maybe I said I wanted to watch it.

You said you wanted to watch it.

Oh, no, you told me when we played paddle. That’s what it was. I told you—I was like, “You’re taking our offline conversations to our podcast.” I was like, “I don’t think I’d do that.” That was a great movie.

Mike said it was awesome, too. He said they nailed it.

Santi, what do you got? What’s yours? No, you go first.

I’ll give folks a non-finance, non-crypto podcast. There’s a really good episode of Invest Like the Best with this guy Alex Karnal, talking about GLP-1s and everything that’s going on in health from an investor’s lens. It’s the newest episode of Invest Like the Best.

He runs a fund called Braidwell, a life sciences investment firm. The guy’s been investing in biotech for 25 years. It’s a really interesting podcast.

You’re on the peptides train?

Not yet, sir. Not yet. But half the people I know are on it. It’s crazy. I remember those group chats of friends who are really early to things. Five years ago, everyone started doing GLP-1s, and I was like, “Oh, this is crazy.” Now everyone’s doing them, and that same group chat is literally 80% peptides.

Are you guys on peptides?

No.

No. No peptides. Maybe I should, though. I haven’t been working out as much recently.

Dude, after my paddle game, Rob—oh my God—I need every peptide I can get. Santi, I had the worst game of paddle I have ever played.

Yeah, I did beat Yianni in paddle in 2 straight sets over the weekend. I’ll put that out there.

Right. God.

My content: I’m continuing on this media-cable terrain. Obviously, I read Going to Be Wired by Malone, then Ted Turner, then Barry Diller. Now I’m moving on to this book called The Gambler, a story about Kirk Kerkorian.

He was probably one of the wealthiest guys that no one has ever heard of. At one point, he was the largest player in Vegas—not just because he owned a bunch of real estate, media companies, and studios. He came from nothing, essentially.

The book is called The Gambler by William Rempel. If you think you’re a degen and have a high risk tolerance, go read the book.

MGM, that’s right.

Yeah.

Yeah. Epic. Great recommendation. Great book.