Can DeFi Bounce Back? | Sam MacPherson & monetsupply

Transcript

Hey everyone, quick disclaimer before we get into today's episode. Nothing said on Bell Curve is a recommendation to buy or sell securities or tokens. This podcast is for informational purposes only and the views expressed by anyone on the show are solely our opinions, not financial advice. Our guest and I may hold positions in the companies, funds, or projects discussed. All right, everyone. Welcome back to another episode of Bell Curve. Um, today I'm joined by Monet Supply and Sam McPherson of Spark and Excellence Dao. Welcome. Welcome to the program.

Yeah, thanks for having us. It's been a week. It's uh it's been a week. Um but uh obviously we're going to be talking about the honestly probably talking a lot about the Kelp Dao / Aave situation that's unfolding in real time, but also wanted to just take this opportunity to kind of set the stage and talk about DeFi as a whole here um cuz it's been a tough month, uh tough at least last 30 days or so. There's been about $500 million in hacks uh mostly between uh the Kelp Dao situation and Drift. So Let's focus on the Kelp what actually happened here cuz it seems like the sophistication of these attacks are getting much, much better. I don't know if it feels that way to you guys, but

can we just start with like an overview of what actually happened here? Who were the actors? There was there's kind of a Kelp Dao part of this. There's also a Layer Zero. Uh there's a DVN, uh potentially an operator error, potentially Also, guys, I don't know anything. No one from any of these places yell at me. Uh but what actually happened here? Yeah um so you're you're absolutely right, Mike, that uh North Korea is really upping its game. Uh the like sophistication of these attacks is getting quite alarming. Um starting with the the Drift hack and now this RS Eth one as the postmortems have come out, we've learned uh like how how deep these penetrations are into these organizations. So, yeah, just a a quick overview of of what happened with the

rETH exploit. Um, so what happened is that there is um, a bridging mechanism for rETH to go to other chains. This is called a lock and mint bridge. So, basically the rETH is is natively issued on Ethereum. And so, if you want to go to another chain like Arbitrum, you would then lock the rETH in the this is LayerZero bridge on Ethereum. And then it would send a message which would mint the rETH. It's like an IOU for the backing that's sitting on Ethereum. So, what happened is that uh, there was a forged message um, from the security mechanism within the LayerZero ecosystem.

Um, again, this North Korea was able to penetrate deeply into the infrastructure layer here. And um, you know, there's a number of issues that like that could have been improved here, but uh, yeah, so they were able to basically uh, forge a message, take all of the backing of the rETH on Ethereum, and then they uh, they moved it out to various places, but the majority of it went into the Aave core instance, um, which was then used to mint uh, ETH, or sorry, uh, yeah, to borrow ETH in the market, and then they were able to exit into ETH. And so, the the name of the game here is for North Korea is they really want these like non-rate limited large pools of money they can get. And then once they have it, they need to exit. And so, exiting

at size This is why Aave was a big target here because of the ability to move over hundreds of millions and able to exit into ETH, which is viewed ETH on Ethereum is is a very like like decentralized collateral. It's there's no way to like freeze it or anything. And so, this this is why they've chosen to go through this mechanism. Okay, so I'm going to follow up here with the obvious and maybe naive question that I feel like everyone's asking, which is why is there so much money being secured by, you know, very small multi-sigs, essentially, and also, how difficult would it be to implement something like rate limits?

This has been something that's been discussed for a while. I also know, in the theme of, say on the show a lot that Cosmos was kind of got everything right here, and I know that protocols like Osmosis actually did this. They rate They uh you know, limited the amount that you could bridge out of that protocol many years ago, but maybe it's more complicated because, um, there's composability, um, within at least Ethereum-based DeFi, like why why have some of these fixes not been implemented here, I guess? So, I've I've heard, uh, at least from like, uh, you know, public posts from different projects who are using various bridge infrastructure, maybe LayerZero OFTs and such, um, that they've, you know, people have implemented, uh,

manual kind of like rate limits. Um, so, I think that's, you know, in hindsight, that's a security feature that clearly should have been in place. Uh, you know, it's like a it's an inconvenience um for users in some cases if you want to bridge large amounts, but the the risk-reward of of rate limiting is definitely very good. Uh, I think like one example, uh, guy from Athena was mentioning that they have a, uh, like a rate limit of how much you, uh, USDE that you can transfer between chains. That obviously makes sense. Um, the other the other kind of weak point here was that the, uh, Kelp DAO team used, uh, like a one-of-one DVN, which means that you're just relying on the LayerZero Labs,

uh, sort of core infrastructure. Um, if they had used, you know, a DVN set that required two of three or something like this with different independent operators, uh, it would have been this much harder for North Korea to like compromise each of these independent organizations. Um, so, I think that was that was clearly a failure as well as as using a one-of-one implementation was, you know, not the right choice. And is the reason to I mean, is the reason to use a one-of-one in this instance just as simple as convenience? I mean, is that kind of the answer?

I think we're we're sort of uh this is like up for the data we're like learning what the reasoning might have been, but what people what I've seen someone suggest is that it might have been like kind of the default configuration when you're setting up uh like layer zero bridging infrastructure. So, like if you just click through or like, you know, follow the dots or whatever with the default uh you're going to end up with one-of-one. Um so, you know, it they should have been familiar with the infrastructure they were using and set it up properly, but like uh you know, guiding people to making the right security choices decisions also is important. Yeah. Um this is also probably a good time to remind folks uh

that you guys are both at Spark, which is competitive to some of the um you know, protocols and things like that that we'll be discussing today, but you know, we'll keep everything balanced and fair. Both of you are really good at that. Um okay, I have a I have a question. I I was listening to a um you know, a talk with James Prestwich who's talking about this who whenever he speaks, I always listen. I don't always totally understand, but you know, again, this is kind of on this theme of this has been solved in TradFi, but it feels more technically complex, especially when it comes to bridging. And it's been interesting to see uh some former advocates like big advocates of kind of like the L2, you know, road map uh like the roll-up-centric road map um kind of

talk about moving things back to um the L1. And actually a lot of these discussions that we're having today remind me part of why I wanted to get the both of you on the program here is like some of this was risks some of these were risks that have materialized that the two of you flagged to me years ago. Um and there were many kind of like semantic security debates about this kind of stuff a couple of years ago and it all I feel like nothing really happened at that time. So we all concluded maybe none of this stuff matters. But obviously it really does. So I'm trying to zoom out of maybe the specific details here and you know, what can we infer for you know, the future of DFI here? Is the problem just operator error and we have a lot of young teams? Is it

the fact that we're doing bridging and bridging in itself is kind of risky and a little bit insecure? Do we need more like even when I look at how like operating systems and computers solve their virus problem? There was a lot of vertical integration and full ecosystem building. This is why Apple solved it before Microsoft. Because Microsoft's actually more open. So I'm just curious, you know, when you look at if we had to start to kind of zoom out of some of the specifics of this individual situation and start to talk about the state of DFI. What what needs to change? What needs to change here?

Yeah, so I I think it's definitely operator error. Like you can do DFI custody in the right way and it's been actually some of the oldest protocols have been doing this the right way, which is governance under time locks and where there are multi-sigs that have like custody of funds, they need to be under time locks. And then for immediate reaction time, you have these like freezing multi-sigs that can react right away, but you need this like the custodian of the assets to always like any sort of like unexpected movement that can be done with the assets, you need this under a time lock. This is a this is a massive deterrence to North Korea. Now the reason I mean most people are doing this good. Like they're doing it very well. But the

unfortunate thing is you know, there's a lot of people in the space and it all takes is one person to not be following best practices and it's just a big story and everybody thinks that you know, DFI is is is you you it's over. But like I I think the if you if you follow the recipes for doing custody and DeFi correctly, I think it can be quite robust. On the bridging side, I think again, like this is an operator error of using a one-of-one DVN. This is not a good security practice. So, Sky, for example, uses a four-of-seven for bridging through Layer Zero, and I believe this is an incredibly robust setup.

I have high confidence in it. Um So, like you know, it's not so much about the the brand that's associated with these or like you know, specific It's like the security setups need to be robust. And so, this not only applies to your own organization, but for any collateral that you onboard as a lending protocol, you need to be aware of the security practices that these collaterals are are taking as well. Um so, yeah, I guess big picture, DeFi is not dead. I think we're going through the ringer right now, but I think we'll all be stronger on the other side.

Yeah, I would I would maybe add that um I think a lot of people were caught off guard by the fact that like um you know, Aave, which is kind of like a core pillar of the the industry, um through this smaller relatively collateral RSEs, uh it had exposure to you know, what turned out to be a pretty unsafe like bridging setup. Um and it's pretty challenging to like map out this whole dependency graph of you know, you're just a depositor in Aave ETH. It's you you know, on the face of it it seems like it should be a very safe asset, um but you're actually exposed you know, through a couple layers of intermediation to uh you know, a one-of-one DVN bridge. Um that's

something where I think we can do a lot better. I know there's you know, a bunch of teams are are actually they've already been working on this, and you know, uh mapping out all the different operational risk factors. And that's something that actually, I mean, I think it can be kind of a team sport as well where once we are mapping out all these dependencies, like this can become kind of an open-source resource for everyone, and it also will help drive best practices if, you know, each collateral asset, um you know, it's just kind of out there what their, you know, sort of admin setup and dependencies are, um it'll be a lot easier to kind of just everyone gets guided into having safer setups, having, you know, better uh time locks and and managing their operational

risk better. Yeah, I mean, you know what this kind of reminds me of from a sentiment perspective, ironically, is kind of reminds me of, this is maybe DeFi's equivalent of what was happening, I mean, this happened multiple times, but like the the 2022 FTX blowup for CeFi. Um and if you remember the sentiment at that time, you know, the feeling was, why would I ever have any money on a centralized exchange which wasn't like Binance or Coinbase. And, you know, I remember all of these prognostications that these CeX's were going to die and blah blah blah, and now when you look at it, they're super healthy part of the ecosystem and they've all survived and or many of them survived and or thrived and or, you know, are thriving, doing fine. It feels

like there's a little bit of that um happening with DeFi at the moment, which um you know, even from some of our analysts who are super crypto native, uh you know, pretty hardcore users, it's like, just doesn't feel safe to be in DeFi right now, but I feel like I I completely agree with what both of you are saying. I This is where I'll get into some of my own thoughts here, but I feel like this could lead to a sort of consolidation in the industry um as well around uh protocols which are battle-hardened and have better security practices than others. Like, there's a little bit of kind of weeding out that needs to happen here, I think. But also, people have really short memories, and I think we love short memories, too. So, I'm just curious, you know, how should listeners be thinking about um the

general landscape and outline of DeFi here? Do you see consolidation? Are operators going to vertically integrate to reduce dependencies, you know, in external like the infrastructure layer and what North Korea's been able to do is pretty concerning, I would imagine. But yeah, how do you expect the landscape of DeFi to sort of change from here? Yeah [snorts] this is definitely a wake-up call. I've been getting very alarmed at the sophistication of the attacks with North Korea. Like with the Drift attack, there was 6 months in advance where they were like proxies of the DPRK are like meeting their team in person and building trusted relationships. Um like it's quite scary. And so like

I think, you know, I'm not I don't know for sure, but I think these attacks are kind of accelerating because of AI. It's It's probably giving them a lot of leverage on their team to be able to, you know, more easily infiltrate into organizations. So uh we're in general and I don't even think this is just a crypto. I think crypto's maybe on the forefront because of all the the money at stake, but more widely, you know, you see the the the Mythos announcement with from Anthropic and they're like there's thousands of critical security vulnerabilities in every major piece of software. I completely believe that. So now I think we're entering the phase where it's like uh yeah, there's we're going to have to be like very careful about like structures we're using and how what our threat models are. Um a lot

of things that are are vulnerabilities that were sitting there that were unexploited are now going to be much more likely to be exploited. Um but I think once we come out of this and especially we can use AI on the defensive as well, like that everything is just going to be way more hardened, which is actually a much better end state. Because it's probably that a lot of state actors actually are exploiting these things, but we just don't know about it at this stage. So like I I I personally view this like it's going to be a lot of short-term pain, but I think in the long run this is going to be this is going to be very beneficial. Now, specifically in in crypto, I know we're taking a hard look at like what a a lot of the dependencies are.

Um there are still like threat models can still be reasoned about. It's not like everything is is like on the table right now. It's like we we know how to do certain security practices well. And so, it's just it's about like like taking another look at and rethinking about where are all the dependencies on infrastructure, what is securing the funds, and just being very very critical and not assuming that Lindy is is some as safe a protection as maybe previously assumed. Yeah, I'm curious like I actually maybe just one more comment about I guess like current state here.

To This is trying an analogy and kill time on you guys, but you know what? I kind of feel like the the moment in crypto is crypto in many ways is an industry, but it's also kind of an kind of a fringe almost like political ideological driven movement. Those are sort of the roots. There are these very deeply held ideologies to the point where you know, if you had kind of an unorthodox idea, you couldn't even say it on Twitter a couple years ago without getting absolutely roasted. You still will today. It just doesn't matter as much. But now what's happened is actually crypto has won. So, it's moved from fringe into the mainstream. And everyone's kind of like, "Okay, great.

This like weird outside DeFi thing, we're going to do this now." And now just like any other ideological movement that started fringe and moves into the mainstream, you're being confronted with reality. Like as a thought experiment, let's say the three of us led the October Revolution in the Soviet Union. We convinced everyone, "Hey guys, give up your personal property, you know, come on. Let's be communist. It's going to be great." And then we're leading this revolution. It's like, "Oh [ __ ] people aren't showing up to work. It's like, okay, well, maybe we'll make some compromises here and there. You know, I feel like that's what's happening at the moment. And I mean, it's just interesting. We're talking about this before the program, but Arbitrum took what I or Offchain Labs, I think, took what I thought was a great move, which was they froze about, you know, between around

$65 million worth of these stolen assets. But I see some people saying that, well, this is multi-sig fi instead of defi, and we kind of can't I to me, I interpret this as this is the ideological backbone of crypto colliding with the real world. And I think that's what's causing some of the weird sentiment drift, and people are like, wait, was this stuff good, or do I need to Like, how do How would you guys respond to that general, you know, kind of analogy to this moment in time here? Well, I think um you know, North Korea not having $65 million, and that money going back to users, like the outcome is unambiguously good. Like, that's

that's very good. Um anyone in that position, like, I think if you were on the Arbitrum Security Council, like, you may feel some type of way philosophically or something, but you have this like fix everything button. Are you really not going to push it because of philosophy? Um like, of Of course you're going to you're going to solve the problem if you can. So, I think that's um it makes perfect sense they did that. Um but I think it's, you know, I see why people are uncomfortable, too, especially with this, you know, we've been talking about the L2 roadmap for the past year or two, and like, are L2s Ethereum? Like, um in some important respects, clearly they're not. Like, you know, Ethereum can't They don't have a security council

to like freeze hacked funds. Um L2s, in a sense, like they have this power so now there's kind of like a social expectation that if they can use it, they they they will. Same thing for I I thought this this went totally under the radar, but I thought it was fascinating. I think everyone would agree with this, but it just shows how much the Overton window has shifted, which was for the Drift exploit, there was a lot of flak that was directed at Circle for not freezing those funds. It was like a 3-hour period where they could have frozen the funds cuz they were in USDC.

But I remember a time where there were all these debates, and then Paolo from Tether kind of retweeted and said, "Look at how much we've done to support, you know, US authorities in freezing funds." And there was a time where that was really swept under the rug, and that was super embarrassing for, you know, centralized, you know, stablecoin issuers, and it's just I personally am in that camp. I think we should not be funding North Korea, and I think you can make pragmatic decisions. I'm just remarking on how big the Overton window shift has been in a relatively short period of time. I mean, this isn't 20 years, you know, we're talking about maybe 18 months or or 2 years. And I I think it's positive, basically, would be my commentary on it.

Like, I'd love to not fund a bunch of hackers but Okay, can we talk about risk within DeFi here? Um it's kind of difficult to even even if you were to look at something like uh Drift versus the Kelp DAO exploit. Feel like people zoom in uh really deeply on the actual mechanics of the hack, which I understand, but I think it's pretty difficult to reason about Risk is very multi-dimensional here. And so, if you look at something like Drift, it was a very sophisticated setup, but it was a multi-sig hack, right? And it impacted Drift primarily.

Most of the drama um and kind of intrigue about how this is all going to play out on the Kelp DAO side of things has actually been focused on Aave. And so, Aave didn't have an improper, you know, configuration here or anything, but the risk from Aave's perspective is that, you know, they had onboarded I don't even know like RS ETH as a um collateral type, right, within the pool and there's kind of a um uh a challenge to the business model of uh you know, having multiple different uh types of assets in one pool here. This is where I'm going to remind everyone again that you guys have a competitive product here to Ave, but I want to start to unpack uh you know, the implications of what this is cuz Ave is the flagship

lending product for the industry and you know, like others like Camino and Morpho um you know, have been kind of going towards this more modular, you know, isolated markets. Although those have their own challenges, but I'm just curious here we start to unpack some of the implications for you know, Ave, what potential other spillover there might be here and you know, does this point to a change in the lending model, like the dominant lending model within D5 at least? There's kind of um you know, different protocols or lending markets or risk curators whatever you want to call them. Um they take different sort of like risk return trade-offs. Um they have like a different level of risk tolerance that they're uh enforcing on behalf of their

users. Um and it's not always super well documented or like uh the brand's uh association of a certain market might not actually be aligned with the actual sort of risk approach that the market is taking. So, just to give like an example on Spark Lend, we we actually off-boarded RS ETH in February just because we decided it wasn't within our you know, our risk preference. Um but even beforehand we had low rate limits for depositing it and then we never listed it in the uh like efficiency mode which would let you loop it with with high leverage. So, you know, even in a case where we hadn't delisted it, the the last radius would

have been like much smaller. I think in all likelihood we we wouldn't have lost any money. Um so I think that's, you know, that's an approach that we've taken consistently is just to like try and have like a really slim market, few collateral assets. Um certain other protocols, you know, they might have dozens of different collateral assets that are listed. Each one of those presents a different tail risk sort of factor and you have to like um you're exposed to all the dependencies of all these different assets. So, um yeah, I think that that's, you know, there's been a gap I think where people are just looking at the brand value or what's bigger as um how they decide what they think is, you know, a safe or a low-risk market, you know, low-risk

DeFi. Um when you actually need to like look at what are the assets in the market, like what's the approach of the risk curator that's uh affiliated with the market. Um I I think from this whole situation that people are going to be taking a much, you know, more sort of objective look at how how the risk uh preference of these these operators is working. Okay, I've got another question for you guys around um and again, money I'm kind of going off of uh some of your tweets here, but I know this is not something I have super detailed understanding of, but it seems like, you know, in the case of this hack, there is a loss and there's going to be need to be a decision made about how that loss is socialized, right? Does this um you know, you've seen Aave come

out and say, "Hey, all of the rSETH on ETH main chain is fully backed." Um and which leaves the L2s saying, "Well, hey, like am I going to just be fully holding this bag?" And regardless of the decision that gets made here, I just find myself it's interesting that it's a decision as opposed to if this were some sort of corporate bankruptcy, there would be no one would be deciding anything. There'd be very clear rules, right? Which would be based on, you know, the capital stack. You know, you'd have your banks get paid back first and then it'd go to mezzanine and then, you know, preferred equity and that type of thing.

And is that the type of thing that one could fit into some sort of legal framework? And so and it's just a weird weirdism of crypto that it's kind of global and there's no jurisdiction here. So does this fit into the existing, you know, framework or is this something that we're going to have to build in crypto? Like I'm just Or do you think it shouldn't be rules-based and we can make these sorts of decisions but what do you think about how this is all playing out regardless of the actual political decision that needs to be made here? Uh I mean, I would say it definitely it should be rules-based and like as a user when you're depositing like it's it's better for everyone if

you have clarity on exactly where you sit in the sort of the preference stack. Um it's something where I think nobody nobody wants to like consider the possibility of a major loss within their protocol or like even kind of like flag that to end users. So it's something that people, you know, there's a natural aversion to actually addressing this issue head-on. But it's something that's super important. I think um you know, even looking at the Sky ecosystem, I think there's like a reasonably well documented sort of like waterfall of like what happens if there's losses like first the sub Dow like Spark would, you know, cover the loss up to their capital and then it goes to the Sky balance sheet um and then there's like you mint Sky to pay stuff back and you know,

there's there's a waterfall, but it's even with Sky it's, you know, it could be better documented, I think, and like more explicit. And I think most other protocols like it's uh it's just not even like listed anywhere. It is it a money is this a like so Ave hasn't called it this and also this isn't what this legally is, right? So don't quote me lawyers, but you know, with the umbrella, you know, architecture, they kind of have something that looks a little bit like junior capital, right? You have people that are staking to individual pools that are essentially getting paid more yield to take on more risk. So, that looks a little bit junior capital-ish to me, which actually credit to Ave, they

rolled this out think about a year and a half ago and it's actually been super helpful in terms of mitigating the potential losses. So, is this the type of thing that's going to get built in a crypto-native context? Do we need to go to lawyers and legal jurisdictions to figure this kind of stuff out? I I think we all agree it should be rules-based and not like, I like you or based on your political clout, but what's the actual mechanism that you think then the industry is going to coalesce around here to enforce this stuff? Yeah, I'd say it's like two parts. So, like you need actual capital available to cover losses, which you know, it could be your treasury, it could be you have like a first-loss capital, like a junior tranche that where you're sourcing, you know, external deposits and you're

they're getting a higher return for taking higher risk. I think both of those are good. But then you also need to like just clearly specify, you know, what is umbrella covering as an example? It that actually is pretty well specified, I think. You know, umbrella, the ETH umbrella thing on on mainnet just covers Ethereum-related losses on mainnet. It doesn't cover, you know, if there was stablecoin losses, it doesn't cover Arbitrum. Um so, I think Ave is, you know, I think they've done a pretty good job in in that respect. But then like, you know, thinking of all the dependencies, you know, Kelp DAO, it's sort of still unclear whether the OFT rseETH is like the same as mainnet rseETH or whether

they're they're different. Um so, you kind of almost need to like force these tough questions onto everyone that you're you're dealing with as well. So, what are like, this is always a dangerous question to ask, is what are the knock-on effects of this? Is this over? Um obviously there could always be another crazy North Korean DeFi hack here, but one thing that's been a little bit unique about this particular situation, you guys are more versed in banking history than me, but I can't really think of an example where there's been this kind of weird soft run where you have a bunch of people that might want to get out of their positions, but they've been stuck because the utilization rate of the

pools that they're in is 100%. Like are we are we just in this kind of suspended state of animation here? I mean, is that actually going to work? Because people are kind of like, "Okay, it's been a couple days. I've cooled down." I I don't know. I've I've I'm struggling [clears throat] to come up with a precedent here, and then I'm also trying to understand what other, you know, knock-on effects there might be. Yeah, so the mechanics of what had happened, as I mentioned, like so the rseeth was deposited and then a bunch of ETH was uh borrowed. So this made um and again with the ambiguity of who takes the loss, is it like 100% to the users on L2? Is it a haircut across all the This is ambiguity am uh ambiguity kind

of uh like makes it so that people are uncertain. What if What do people do when they're uncertain about if there's if there's losses in the market? They'll pull all their funds. And so this uh led to ETH uh reaching 100% uh utilization. Um but maybe not everybody's aware of this. This actually had a a compounding effect in that uh there's a lot of positions on Aave because they rehypothecate the ETH collateral quite heavily. Um this liquidity issue now turns into a solvency issue on the USD markets potentially. So that the there's a lot of positions there that are ETH collateral borrowing stablecoins. And so normally when the market is is there's there's enough liquidity, the liquidations on a ETH-USD position can

function just fine. The problem is is that with an impairment in the ETH market, uh if the price of ETH drops, there's no more liquid ETH to sell and then recap like recapitalize or basically pay off the loans that are reaching margin calls. So, this can uh potentially makes the uh you know, primarily the USDC and USDT markets on the market uh at risk as well. And so, this is why everybody then started pulling uh the funds from the USDC and USDT markets. Um and so, it led to this uh basically walk up of of these and it's also cascaded out to other stablecoin markets. All of them are at 100% utilization since, you know, early Sunday, maybe even late Saturday.

So, how is this going to like this is where I uh get tripped up. And this has been the this has been the long criticism of DeFi um around that that I actually have a pretty difficult uh time answering, which is the way that this has been fixed in TradFi. Basically, having a god mode button on the currency. So, to you the exact language that you used earlier in this podcast Monet it like if you have a button and you can press it to fix something, then uh obviously, you're going to press that button. That's the choice that people in TradFi made a while ago and it kind of led to the conditions that led to the birth of DeFi. I I think this is the

thing that's a little bit difficult to reconcile and, you know, even you know, to be fair to Aave here, right? Like some of the risks that are present for Aave could also in theory be present for like a Sky as well, right? Which is the risk to me is there's no FDIC. There's no, you know, agency like the Fed which brokers agreements if there's ever, you know, the DeFi equivalent of, you know, a bank run. It's it's kind of unclear what ends up happening in that situation. There isn't, you know, there isn't um there isn't precedent, basically. And so, obviously, I'm asking you guys to hypothesize about things that are hopefully never going to happen and and don't have president or president D5 at least, but how can we It feels like we're not going to get away from the

model of lending overall, you know, where you have depositors and then one, you know, one balance sheet, right? That you're making loans out to. And we can always harden our practices um both on the kind of smart contract security side and on the onboarding collateral side to make sure that, you know, we're pricing risk properly. But these things happen. And so, I I guess the the question that I have from a long-term perspective is how do we solve that god mode currency problem that we solved in the banking system, but probably not a trade-off we'd want to accept here. Yeah, I I think, you know, D5 and just the whole crypto industry it is kind of playing on hard mode because um we can't rely on just like moral hazard and someone

bailing us out. Uh if anything, like maybe the closest uh we have in crypto to like a lender of last resort would be someone like a Binance or a Tether where they're extremely profitable, they have a huge balance sheet, they have a lot of sort of excess reserves where um they can backstop uh losses. You know, 5bit got hacked, Binance uh had a loan out to them to make sure that they could still cover withdrawals. Drift is like in some sort of recovery program with with Tether.

Um so, it is possible, but like there needs to still be whoever Whoever experienced a loss, there still needs to be like uh you know, some sort of going concern so that it's like a proper investment by these these companies. It's not just like a pure handout like it might be with a government. Um and, you know, if you have a situation like currently where there's, you know, three different potentially at fault entities who are all kind of blaming each other and it's not clear like where the liability falls, it becomes a lot more challenging as, you know, someone like a Tether to like say, okay, we're going to backstop this, but you don't know who is actually giving you like an equity stake or something like this. So, it's pretty challenging situation.

Um I guess maybe like one other um thing I'll say is like, you know, if if you do have losses, like trying to like apportion them fairly as much as possible is is quite important. Yeah, um what about that though? Like is is sort of the traditional finance system like any better? And I I think like what happens with these big like government bailouts? It's basically just moving the risk of the banks. It's further concentrating it into a fiat currency. So, like I would not really view like the US and it's like massive debt problems, the deficit problems as like the shining beacon of like the solution to everything. So, I I think it's it's different trade-offs, right? I forget who to attribute this quote to,

but there's the saying about risk that you can't destroy risk, you can only transform it. And I think the compromise that we have implicitly made in tradfi is to transform a bunch of short tail risk into long tail risk. And I get that this is a crypto opinion. I think this will survive our entrance into the mainstream, but I would connect a lot of the populist politics and weird [ __ ] going on to some of these long tail risks because the net effect of god mode on the currency has created the perception and reality of unfairness, unequal treatment, wealth concentration, all that stuff. So, political opinion, but I do believe that. Um so, maybe in terms of defi, to get a little bullish here cuz everyone's

been so down. Um like just pattern matching to having been in crypto for a while, everyone has to take their turn going through the ringer. Um I I can't think of a single sector or even company that's been successful that hasn't. Congrat like at the last crisis of in the Maybe just to zoom out for folks, I know everyone I I also was very skeptical of the four-year cycle continuing. This looks exactly like the four-year cycle. It certainly feels exactly like the four-year cycle. And this could just be the bottom of a bear with a different part of the ecosystem that's moved into the mainstream is thus feeling the stress, but they'll emerge.

And the opportunity for DeFi here is everyone still believes this. This is obviously going to work. Um but maybe we don't need as many uh DeFi providers and maybe some startups that haven't been around for a long time and don't have their security practices or aren't able to, you know, risk manage properly, maybe they shouldn't exist and there'll be some natural consolidation and they'll, you know, we'll come out of this the other side. Um You know, I'm also reminded of someone who gave me some advice at Blockworks, which we haven't always followed, but uh solve a hard problem. Make sure that you're solving a hard problem. This is a hard freaking problem because I mean you guys know this, but if you're building in this space, not only do you need to be at the forefront of risk management and have a lot of financial expertise,

but you also got to be at the forefront of security and opsec. And that's just there aren't that many people, you know, that organizations that can do both of those things. So, I'm going to end my rant here in a second, but DeFi is solving a hard problem. So, the folks that emerge from this, I would imagine, are much stronger, have more competitive market share. I don't know. I think it's I mean, don't listen to me and then put a bunch of money in DeFi and lose it, but come on. It's like a I think it's a good time to be bullish DeFi here. Um For sure. I like I think fundamentally what we're building is an incredibly resilient system. And so, how to build a resilient system from the ground up, you basically have to do it in hard mode, like as you say, which is um

I think [clears throat] like smart contract bugs were like massively prevalent in the 2021 cycle. But those are largely going away. A lot of what the hacks that are happening now is more operational error. And with each hack, it's it's very unfortunate for the people involved, but like everybody's looking at this and they're improving their practices. So, what I think the outcome is of all of this is that, you know, TradFi kind of gets around this by having all these humans in the loop and it but it's it leads to an efficiency and just like um you know, it's just a different set of tradeoffs. If we really want to build like a very robust financial system that's resilient in the age of AI, we have to do it in hard mode and we're going to we're going to it's going to be a learning painful

experience in the short and medium term, but the end result is going to be something that's so battle-hardened and ideally, you know, there you know, there needs to be degrees of humans in the loop with all types of lending. DEXs are kind of the exception where they can operate a bit autonomously, but lending you'll kind of always need humans. So, but there that doesn't necessarily mean you can't improve upon the status quo with with the tools that we have available. As I mentioned like security practices of having like you know, time locks and you know, there's multi-sigs or like non-custody of assets the one that is allocating, freezes. All of these practices are well-known at this point and we just need to do the hard work of like implementing and make sure all these

systems are robust. And so, yeah, I'm I'm you know, I'm biased of course, but I'm incredibly bullish on the future of DeFi and I I I I really hope we can get these hacks under control cuz I don't think it's a good look for anybody and if we really want to be the future of you know, TradFi coming on chain, we need to sort this out and I think we will. Yeah, I would I would say also that um >> [snorts] >> you know, there's the there's an aspect where like TradFi it's so people-centric that like it's kind of like a ship of Theseus where like if you look 30 years down the line like sure there's still you know, Chase Bank or whatever the case is, but like the actual operations of it are completely different. So, it can't

really become Lindy in like a meaningful sense when I think with DFI like we will probably arrive at some smart contracts as an example like these contracts are free from bugs like they basically like they're not going to get hacked. That's something that's I don't know if it's possible in you know tradfi that stands today but like um it's like an amazing feature of DFI where like um we can actually you know progressively build like a safer and safer system. We love ship of Theseus reference here on Bellcurve. Um that's very welcome. That's classic Stern to me. Um okay, I would love to understand uh so

I mean maybe to lead with my own perspective here as well. I think that the news cycle is very short. Uh I think people have short memories. I would be in the camp that like first of all we've been talking about ideological like the ideological drive bumping up into reality. The mother of all of that is really happening with Q day and Satoshi's wallet. That's a pure that's a completely non-technical purely social decision. So maybe we can end on that cuz that's if the theme of the podcast is you know kind of ideological ideas bumping up into reality is like the mother of all that all those examples.

Maybe we can end on that. But help listeners understand um from your guys best perspective best guess here what is the timeline um around fixing some of this stuff? Because you know the naive layman or even you know someone who's relatively in the industry but maybe isn't as in the weeds you know the perspective would be hey I mean I thought up until last week at least you know Drift was definitely a blue chip. You know they have great backers. I never had any reason to believe that their security practices weren't awesome. I I just I don't know uh you know who's got a one of one you know DVN set up somewhere and I I don't understand even if that's not me I could be in some other pool and get screwed here. So, first of all, how do we fix this stuff?

Um is there some sort of risk standard that needs to get created and adopted by the industry? Um and then what what's your what's your best guess on the the timeline here? Are we going to be dealing with this and talking about this for the couple years, couple months? Like how how do we put the quantum around here? Yeah, for sure. So, this is something I I tweeted about uh pretty recently and I've been tweeting about for months now is uh we we need sort of like uh independent risk ratings within within DeFi. Right [clears throat] now, there's like a like an incentive problem with uh with like the managers of these risk curators, I guess is the term that's used. Um so, like basically the ones who are optimizing for growth, revenue are the same ones that are underwriting the

risks of the collateral that are being onboarded. And so, this is not what happens in in TradFi. There's like the risk side and the growth side are two separate and like, you know, walled off entities such that the growth side submits a proposition to the risk side and then the risk side from purely risk like sort of perspective says if it's if it's a go or no-go and what the concerns are. So, we could we could definitely achieve this. This has happened in the case uh with a lot of places in the past. And so, at Sky, we're actually uh uh spearheading an initiative here where um the underwriting process uh for new collateral goes to Sky. So, us as a sub-DAO in Spark, we're on the growth side. Like, we're sort of although we do

do the job of like, you know, risk management now um and we will in the future, but it's it's also important that there's an independent team that is not like uh like focused on growing Spark's uh revenue and TVL and stuff like that, that's an independent check on us. And so, Sky is doing exactly that. And um already there's a risk framework in place where uh it's a very mechanical way you can run through all the technical and market risks. And basically, it's sort of a traditional credit underwriting framework. And then this will be sent out to independent online arms length entities, some crypto native, some actually in tradfi. We're looking at some big names in tradfi as well to do this type of underwriting. And so then this will produce like a credit score that you get

from like S&P Moody's that they have in tradfi. So we should adopt the same exact same sort of methodology. I think it's quite useful. And then the awesome thing about DeFi is that because the space is transparent and like everything is pretty open, we could just open source this and like you know, here's Sky's view on like these various cloud rules. You know, maybe people disagree with us, but there could be some other major player who has their own credit underwriting framework. And then we can combine all of these resources into a single index much like L2BEAT does for L2s that's independent and kind of like a check on everybody doing best security practices.

And I think this is going to make the industry much safer because then, you know, if you're not doing best practices, and you know, we're not perfect either, it's important to know where your flaws are and so that you can focus and improve on them. And so like I think this is a very achievable problem. I think the current situation is going to accelerate this initiative. We've been talking to the Ethereum Foundation about this about making a you know, a joint effort on everybody's part because nobody wins when these types of situations happen for DeFi. It's just a big L for DeFi as as a whole. So like nobody wants to see this happen. So I I think I'm very hopeful that we can really kick this initiative into high gear you know, quarter three, quarter four, have

something out in the open by then of this year. I completely agree. I think, you know, open source and like risk mitigation and best practices just kind of trying to make the advancement of the industry a lot more open and kind of a team sport where like everyone's individual efforts are like basically compounding against each other. Uh it's like how, you know, open source software has been just a game changer for the pace of technological advancement. I think we can do similar for for risk management and visibility into all the dependencies.

Um I mean, I just to like turn back to also what you were saying about uh like Q-Day and like stuff like this. Uh I think it's it's um you know, it's timely and in a sense, like the pace of technological advancement is accelerating at such a crazy rate. Like, you know, people think that Q-Day could be coming within 5 years. So, the whole crypto ecosystem needs to find quantum resistant signing schemes and implement them. Um and then, you know, looking at like smart contract security or like operational security, um you know, AI is advancing so fast that I I do think, you know, um all of the like sort of loose hatches need to be locked down, probably in the

next year or two. Um or else, you know, anything that's anything that's vulnerable is going to get found pretty soon and hacked if we haven't locked it down. So, uh it's a it's a very urgent uh issue. I was actually going to follow up and ask you guys about this cuz Sam, I think you were talking about um AI previously and a lot of people have connected this to Mythos. Um now, my own opinion on this is I don't see why Anthropic would release that to North Korea first. It wouldn't be on my list of uh beta customers here. But, it does it does feel like it's hard to escape, first of all, the very salient message that Anthropic has been saying.

And then, also, yeah, I mean, correlation doesn't equal causation, but as the LLMs have gotten better, these hacks have certainly picked up. So, I'm just curious how much of a connection like how should listeners weight that connection there? Yeah, I agree with that. I don't think it's actually Mythos, but AI in general, um you know, North Korea, I don't know you know, I've no idea what their operation is like, but like assume there's some number of humans that are incredibly intelligent that are like working through and doing these hacks. But, you know, humans are you know, I did security and and you know, university and stuff. It's incredibly like it taxing on human minds to kind of go through binary and stuff.

It's it's it's super hard. Computers AI are are very good at this. And like with I think there's been kind of a in the last 6 months we've really seen a change in AI where it's now doing a lot of the coding. And this agentic AI is now to basically autonomously like go out and do these tasks and it's it's getting it's getting like quite good now. So like for me like you know, I just think it's there's certain number of human beings that are doing this are now being leveraged like probably 10x. And this is why we're seeing this increase in in speed here. You know what this reminds me of a little bit is there's a I don't know you guys familiar with Renaissance the Jim Simons hedge fund that has unbelievable over returns like

66% after they kicked all their LPs out. Well, there's a book that came out a couple years ago called like the man who mastered the markets or something like that. And they talked about in the early days he experimented a lot with different algorithms and stuff like that. And he made it work on commodities and currencies I think, but just tried over and over and couldn't get it to work for equities. And they were you know, like a month away from shelving it where when Robert Mercer is also pretty infamous guy these days discovered there was like one tiny error in the code and then he fixed that and the whole thing ended up working. And anyway, the rest is history. But I I wonder how many stories there are like that. And like this whole world is going to change in a post LLM society. But

Okay, final let's just end on something fun that's not related to DeFi or hacks or Q-Day. So here's here's my cursory understanding of Q-Day and the challenge here. Whatever it needs to get passed on the Bitcoin upgrade side of things will get passed. Right? Like I'm not I don't think the whole Bitcoin blockchain is going to suddenly break down and it's not a technical challenge for my understanding, it's a social challenge of there are some old wallets in particular, the Satoshi wallet. And the question is, okay, if all of the other wallets are opt in, then everyone will be fine who opts in to upgrade their Bitcoin wallet, but Satoshi's in a weird spot because that person or group of people can't update that cuz they might be dead or if

they did if they were still around and they upgraded it, that would be a risk cuz everyone would say, "Holy crap, you could sell all these coins." And so, it's most likely that this decision is going to have to get made at the level of Bitcoin core devs, which is going to shine a light on Bitcoin governance. And so, the hard part about this I think is that on the one hand, it feels totally intractable to just let Q day happen and let some hacker come in and get a hundred billion dollars worth of Bitcoin and dump it on the Actually, in their S-1, Coinbase actually listed the Satoshi wallet as a risk to their business. To just give you a sense of how known this risk is. On the other hand, to mitigate that risk, they would have to do something like burn or freeze that

wallet, which they can do, but it would tell the whole world and you hear all these guys on Wall Street talking about how Bitcoin is mathematically protected and then it would kind of force them to admit it has nothing to do with math, it's all social consensus. And there's a So, this is kind of the ultimate like an asteroid is heading for the Earth, do you change Bitcoin Bitcoin's code and the 21 million and they're actually going to have to make this decision. So, I'm curious, where do you guys see that shaking out? Yeah, I mean, I think this is another example of like outcomes versus philosophies, right? So, like if you actually look at the end outcomes, like in my view, the you know, where you let people upgrade, but then

if you don't upgrade, you know, let it rip, don't burn the coins, is the same outcome as if you burn the coins. Because the end result is you as a user who has lost your key have your coins disappear, right? So, like I I like really yeah, that's kind of how I view it as a practical outcome, but I do understand the argument. There's like, you know, 21 million coins and everything is sacred, but I'm I'm I'm usually in the more practical camp on these things. I agree with your assessment, Mike. I think Bitcoin will My base case is Bitcoin will deal with this. I think now that there's a clock, again, this is another like wake-up time. This Q-day, I think people are like, "Okay, we need to go now." And I'm I'm I don't I'm fairly confident this will be worked out one way or another.

There's too much money at stake otherwise. Um but yeah, for me I'm I'm burn the coins seems like the way to go. I mean, it'll um it'll be a fork, like, you know, uh Bitcoin and Bitcoin Cash. Like, there'll be two sides. One of them will freeze the wallets, one won't. And then, whichever one is worth more keeps the ticker. And my bet is that the the one where they burn the coins is probably going to be worth more cuz there's not like maybe 4 million Bitcoin getting massively dumped on everyone's head. So, yeah. I mean, it's I think yeah, there's it's kind of a no-win situation, unfortunately, because it does, you know, challenge the founding myth of Bitcoin a little bit,

but ultimately, like, it's still, you know, it's it's a strong social consensus that's protecting you. Like, people are not going to, you know, willingly do this. It's only because like there's such a huge stake. So, I think it's, you know, Bitcoin's still solid. I think Bitcoin's totally fine. Um I do though think that they will burn those coins or freeze them or do something. They're not going to let a hundred billion dollars of Bitcoin get burned. But what I think the Sorry to use another analogy here, but the other analogy of an ideological movement doing this is like the church, which everyone loves to compare Bitcoin to. I think it'll be fine. I mean, the church kind of went through this when they had to reconcile with the Roman Emperor and like you and I wouldn't care about this, but

in Christianity, you can only worship one God. Roman Emperors were version like worshipped as gods. How do you reconcile? They actually killed themselves for like 100 years to decide that, but then it turned out the Emperor was an instrument of God. No problem. Story makes sense. Bitcoin's a strong story. They'll come up with some story that we'll all buy into on the other side of this. What I do think might be interesting is there'll be a lens. I think a lot of people have bought into this high-level tagline that math determines the Bitcoin supply, which is not true. So, I think people are going to be like, "Wait a second, you changed this? How'd you do that? I thought math was securing this." And I think there will be more emphasis on first Bitcoin governance, but then more broadly L1 governance that ends up succeeding in general. And the same

rules that we're talking about for uh lending and, you know, senior and junior risk and all that stuff, I think that'll come to L1s. Um anyway, that'll be my long-term prediction, but we'll see if that plays out. All right. Not what we're here to talk about, but guys, this has been a ton of fun. Uh thanks for hopping on the pod. If folks want to follow more, you know, uh or find out more about Spark or Sky or anything like that, what's the best way to get in touch? Uh yeah, so you can follow uh the I mean, Twitter's use or X is usually the best place. Uh spark.fi for the main account. Uh I'm Hexanaut and uh Monisaputra is we're all on X. So, follow us there.

Awesome. All right, guys. Excellent pod. Appreciated to you coming on and uh hope to do it again sometime soon, hopefully in happier circumstances, but uh we'll do it again uh sometime soon. Yeah, [snorts] thanks for having Yeah, thanks for having us. Take care, guys.