Institutional DeFi & Yield in Modern Markets | EthCC Panels

Transcript

What's up, guys? Welcome back to another episode of Zerx Research. Today, we have a very, very special episode. Um, actually two episodes combined in one. Today, we're going to be sharing with you guys the um podcasts that were recorded at ECC with um partnership with Carpot Key and uh the Blockworks advisory team. They um they did they recorded two podcasts on lending, vaults, everything defi um for uh for ECC and um I hope you guys enjoy. Hey all, Block Works co-founder Michael Apalledo here. Quick break to talk about something we've just launched, Block Works investor relations. As the market shifts toward institutional capital, investors want more transparency, more standardization, and a higher level of professionalism. But the traditional IR

model is slow, manual, and not built for how crypto works. If you're building onchain, your data is already live. Your business is already transparent. The challenge is turning that into a clear, credible story for investors. That's exactly what we're solving with Block Works IR. It's a single platform that brings together real-time analytics, branded investor portals, and hands-on advisory support so you can communicate what matters. If you're an onchain business looking to level up your investor strategy, check out Blockworks investor relations at blockworks.com/investor- relations. All right, back to the episode.

>> Nothing said on Zox Research is a recommendation to buy or sell securities or tokens. This podcast is forformational purposes only and any views expressed by anyone on the show are solely our opinions, not financial advice. Pocachio, Ryan, and our guests may hold positions in the company's funds or projects discussed. >> Welcome. Really appreciate that you could make it here. I'm Marc of KPK. We were born five years ago building the infrastructure we needed back in the day to to manage Dow treasuries and also actively manage some of the largest ones like ENS, Agnosis, A

and we've been deploying billions without losing funds and The this is really the second event we done um scaling defy. The first one was uh last year in dev connect in Argentina. I think it was a very different crowd in that moment. I would say last year I would say was mostly engineers, protocol developers and I see I see different faces.

We also have different co-hosts today. We have um Gearbox, Oiler, Anchorage and Redstone. So really really strong teams building and and yep. So the main thing I would say is right now the the crowd is different. I see we have more asset managers, infrastructure providers, institutional providers.

So that's that's a a pretty big pretty big change. Today we're going to have this um uh another thing that is going to be different is a format for the event. We're going to have a a live streaming of of a podcast by ZeroX Research by Block Works and it's going to be moderated by Anastasia and and Sylvio. So So after a while we're going to have two sessions. It's going to be a conversation. It's going to be recorded.

you can hear it afterwards. So, so that's going to be the main the main thing. Afterwards, you can go outside. We have a very nice very nice, you know, today is a yeah, nice weather. So, hopefully you're going to, you know, enjoy this. Um, so before starting, I want to name one thing and this is that really crypto got boring. I would say and it's something that is painful for for many of us especially for an industry that's been traditionally pure hype based on vibes

based on speculation right especially today we have we are very deep in a in a crypto prices bare market even though some people want to to deny it. So what is different this time is really that I think is the the first crypto prices bare market where where we have uh where we're basically boring right it's not it's not interesting. So I mean people think know are we going to are we going to survive it right because it's not cool anymore and still you know we see we have $300

billion in stable coins almost $30 billion in real world assets and growing. We have Black Rockck building on top of unis swap. We have Visa settling on USDC. We have the genius art. We have Mika and I would say even though some of us you know not like many people today are not so much interested yeah in Dows in the original the original things in in let's say not so interested in in governance

token and and these things. Still, I think that the crowd today is much more interested in let's say settlement asurances, how we're going to have uh insurance for deposits, how we're going to how we're going to manage risk and make sure we're building this infrastructure we we want to build. And having said this, I want to mention what's what's scaling DeFi means to us. For us, it's really open source smart contracts can be verifiable.

So basically, if anything you're building is just a black box, it's just basically just stratfi with extra steps, that's not how we're going to build on top of each other. Another important thing is risk management. We know you know doing proper risk management and operational security is not just having somebody making good decisions fast is having the infrastructure required to to act in zero blocks delays. You put predefined rules verifiable rules. So that's the that's the industry. That's something that can can work uh really at scale and

with all of the real world assets, you know, this trillions of dollars that Larink keeps talking about. I know for some people really this is just another another uh distribution channel. I think you know what really we're missing is all of the economy of derivatives we can build on top of these of these Legos. I think this is really what is going to make a difference of assets on chain and and all of this you know that the the way we're going to scale this thing is by making it uh secure

by making it open transparent you know bringing the Ethereum values to the financial industry. So many of these things are you know are are not uh exciting are you know everything when it tends to be long term is sounds a little bit boring boring work is good boring work scales and yeah people who are here they're building that so yeah now we're going to move to to to Sylvio and Anastasia thank you so it's great to be here. And uh if

I got boring, I hope that this panel won't be as boring. Um but yeah, I'm going to try to do my best. Uh these are all very familiar faces, but nonetheless, I want each of you guys to have a 30 second introduction. Uh and then we can get started. >> So yeah, um just introduce myself, so I'm gonna skip it. >> Uh hi everyone. Marcia Blacken with Anchorage Digital. I I lead growth and strategy for Porto, our institutional self-custody wallet. Many of you know Anchorage as the first federally chartered bank for crypto in the United States, but we've grown the offering in terms of infrastructure going beyond

just staking and governance to self-custody trading. Um self- custody is a means through which our clients on the bleeding end can access DeFi. Hi everyone, my name is Oya. I look after EMIA for Ono Finance. Ono focuses on tokenization of liquid financial markets. We're building what we call Wall Street 2.0 by bringing traditional financial markets on chain. >> Hey everyone, this is Jonathan uh CEO at Oiler Labs. Uh we're building a D5 boron lending protocol uh and vault infrastructure uh on D5. Very happy to be here.

>> All right. So it's very clear that stable coin have PMF and someone said stable coin are like the WhatsApp moment for crypto. So I wonder if you guys were to pitch tokenization and to financial institutions but also as well to defy protocols what will be your pitch like for stable coin is instant settlement global money. It's pretty obvious what the benefits are. uh for the WAS I feel like it's still a bit fragmented and not very obvious and each of you will have different perspective. So maybe we can start from Jonathan and uh yeah >> yeah I think the pitch is a little

different if you're talking to institutions or retail. Um, for institutions, I would say it's utility, efficiency, and speed. Um, cuz crypto and DeFi really allows a 24/7 market. And Trafy moves at a I would say on average slower speed and on weekends is very uh a lot of the financial instrument doesn't really trade on weekends. Um, and for example, a lot of the institutions can uh hold a lot of treasury bonds or low duration private credit instruments. and uh you leveraging defi to boost up their yields uh or borrow against them use collateral. So there are a lot of lot of more tools uh that these institutions

can do through IWA. Um for retail I would say the pitch would be open access. Um for example uh we are a partner to right so a lot of the retail currently don't have access uh to US stock market if you live outside the US and that's just one example. uh a lot of the financial instruments on the fixed income on the bond side uh if you are not a high net worth or accredit investors in the US you are not able to invest or get access to these uh financial instruments. So I would say um the biggest opportunity unlock uh to retail is um the access to a ve a wide array of financial institution um and also their products. Um yeah

>> yeah I I'd agree with that and I think the trajectory that RWA is following is very similar to stable coins. So stable coins initially were attractive for remittances more retail use cases and now all the payments companies anyone who's doing payment processing or fintech and payments is looking at stable coins. Similar with RWAS I think the first initial client group was regions who couldn't easily or cheaply access US markets. So that we we saw the adoption really quickly. But then on the institutional side, I think the utility point is really important. And I think RWAS has the potential to really bring multiasset

to asset managers where it's not just restricted to listed markets, but you can do crypto, commodities, private credits, um, stocks, ETFs, all via using this one tool. >> Yeah. And I I think the pitch to institutions is that you know DeFi proved the model like you know in the last seven years we have been using Ethereum, Bitcoin, Salana, XRP as the collateral asset and it it's really stress test the system in sure in terms of ensuring that it's able to work and function and deliver the the exact parameters that you would expect in terms of getting liquidity. And so real

world assets coming on chain that just you know stabilizes in terms of you know the the the addressable market that's trillions of dollars that are are going to come on chain and match the liquidity being provided by the stable coins. >> Yeah, I think all of these are really good points. Collateral is super important. Accessibility to more markets that's something of course is super super attractive. my mind. We have the the first time in history the opportunity to to really have a global financial system that is not uh that is really integrated and not necessarily um

compartmentamentalized in let's say countries. So, so that's really the first time where if we do it right, we can have access to to all of these collaterals, but in a more transparent way that could potentially uh if everything is connected, ideally we will be able to to predict and and avoid uh larger systemics systemic risks. Talking about systemic risk, one very very obvious concern that most people have is transparency and risk management. So this is like a

multifaceted issue. So I wanted to get your uh guys opinion on this and like if you have a magic wand and will solve like problems in this area, how would you go about it? Maybe I start with Marsha for this one. Yeah, in terms of transparency, you know, I think we, you know, we have a first party, third party issue where oftent times like tokenizations happening where, you know, underlying collateral is being in the legal structures through which they're being, you know, developed does not really necessarily confer legal rights to the end user of the token. And so when you're you know dealing with you know acquiring the token and then you

know something happens to you know the data or something it becomes illquid because you're not able to you don't have the legal right to redeem at the first party level. Instead you're this third party holding a token that doesn't have much liquidity except in secondary markets and then you become a taker in terms of your access to the capital. So it is a I I feel like there's concerns around you know being able to tap into the global markets if it still remains permissionless and and permissioned and your legal rights don't pass through um to your to the token holder.

>> Yeah. And uh how about the on like legal right? How you guys are thinking about these? Are you actually owner of a stock if you own an older stock and how you guys are approaching this? >> So our model is what we call a tracker certificate. So token holders get the performance um and the corporate actions of the underlying stock or ETF but not the beneficial ownership. There's a lot of criticism around or debate around native issuance versus the wrapper model that we're doing and we can do both but we chose this model because it's actually we found it's preferred by the clients and I think if it's done correctly in a bankruptcy

remote structure where the rights of the token holder are clearly defined and you know what to do when things go bad then it's a pretty secure structure. I mean in traditional finance it's very similar to total return swaps where the bank or the issuer of the swap holds the underlying and they just write a swap to represent the performance. So it's it's a model that works. Um and I think the first example of this is stable coins. They use a rapper model for dollars and everyone knows they worked so far unless you do algorithmic stable coins. It's that's not >> and yeah this is interesting and I think like the way you structure these assets is very important that has like

downstream consequences to the whole D5 markets and so to that point I wanted to ask Jonathan a question so like how do you see these tradeoffs at the issuer level impacting the D5 markets and also following up also the your opinion Marcel on this. Yeah. So, uh on this we do work with a lot of um risk curator partners. Obviously, KPK is one of the close partners that we're working with on a lot of these RWA markets. Uh understanding how risk are expressed on DeFi is a top priority because um we're introducing a wide range of different financial assets on chain. uh that includes tokenized equity that includes private credit uh treasury bond that

some we're also discussing kind of introducing commodities on chain as well and each of these assets trade on different speed and they have interest they have different kind of risk curve and characteristics. So for us to open markets and allow user access uh to these financial tools, we have to partner with highquality risk curators and managers in the space who understand the risk. Uh some of the some of the bottleneck that we're currently seeing are um in terms of the price oracles but also in terms of how uh these assets how frequent these assets are priced to the market. Um with longer for example longer duration bonds um there is a

bottleneck with the inreal time um price curve and also marktomarket um and uh we call it nav um so it does create a little bit of issue of handling that but I think with tokenized equity we do have a pretty graceful solution with crossorracle partners like uh chen link and restroom uh yeah I think I think this is a this is a space that we're diligently like working and uh evolving on with a lot of close partners and uh I think it's uh it's a top priority for us to really understand the risk very well uh to protect the the users. Yeah.

>> Yeah. I would say on on top of that something that is critical for this is understanding where are the the liquidity um restrictions and how we can enable redemptions because that's at the end it's uh when we have sometimes misallocations or or liquidity we can have wrong prices that can trigger liquidations and you know those liquidations can trigger some other liquidation. So, so basically the risk uh properly done uh understanding what are the the variables that are going to trigger certain effects especially when there is

uh a hack or a high volatility event that's something that is u critical in order to to exactly protect remove liquidity early. So you know even even when there is one of these um one of these situations there is a a possibility to to recover you know that's something that many people don't know because there are fractions of liquidity you can recover when anything happens so there is always a lot of value on on doing proper risk management. Yeah.

>> All right. Yeah I agree. And um my next quick followup question is do you think risk is more a matter of like the fact that the there is no like a common transparency is more uh a matter of being the market being fragmented or just the issuer like not being transparent enough. Um where do you think the problem lies? Um if you have any takes any of you like obviously there are different asset classes and different standards. Um but where do you think the the problem lies?

Is more on the issuer side is more on the just market structure that we have right now. >> I'd say it lies in the market structure. Um because you know like Ana gave a good example of how like structuring things and in in terms of a market structure um it's a liquidity supply match and when you go into real world assets there isn't yet a price discovery mechanism for the assets when you know they're offchain and onchain and so you have to rely on you know the data that you're being fed and you know there's if the mo the the market's moving fast. So, I I kind of think about what's going on in

the latest like credit crunch where there's so many there's been some pri private credit firms that are, you know, gating access to their funds where redemptions are only about 11% a quarter, but they're seeing higher than average redemptions. And I wonder what this what this looks like when you know the RWAS, the private credit, they're they're issued on chain and what that redemption flow would look like in terms of you know a vault or you know an asset a risk curator like how do they manage that?

>> I see. >> Yeah. Um I think to build on your point Maria um I I do agree it's it's a lot with the market structure um some of the high risk point um usually share a few characteristics so if the market is not super uh efficient meaning that the trading volume or liquidity is pretty thin uh that's a risk factor um two is if the price oracle or price fees are not always up to date or has a delay that's also another risk factor. Uh three is with if the asset has long duration long redemption period um that's also a risk factor is so if if a certain type of RWA that has multiple of these characteristics I will consider higher risk um yeah so when we're

evaluating uh which market to open on oiler platform uh we do that in coordination with our risk managers uh and taking all that into consideration uh if we feel like the market is for example the liquidity is too thin. We don't have the capability and capacity to offload or liquidate in a smooth way in volumes in real time. Uh we sometimes would delay certain markets and waiting it to grow more organically and see more organic demand so we can make sure when the market's being stressed we have ways to handle all the ramification that comes with it. So um yeah that's what we're currently seeing. Yeah.

Okay, so obviously we are talking about the 30 billion market of which 1.5 is allocated into DeFi. I was wondering if you guys have any take of what's going to be the asset class which benefit the most from tokconization. Um obviously we have now credit, we have stocks. Um so I'm really curious to know what's benefit more from tokenization but also DeFi as well. So maybe we can start with the OA for this. >> I think the way RWA has evolved so it was stable coins and then tokenized treasuries or yield bank stable coins.

Private credit was done previously but now it's coming back and being done probably in a more sensible way. uh stocks and ETFs and I think the next iteration is going to be derivatives and structured products especially because more institutional players are coming on chain they want to get creative they don't want to hold just S&P 500 ETF like you can trade um per against that or hedge your positions with per against that or have more uh creative structured products when you have all the different components of it on chain.

>> Yeah, makes sense. Spots come first like like with D5. Spots come first and then you build all the building blocks and um yeah, same question to you Marcelo. What what do you think is the assess class with the most potential to to scale next within RWAS? So really the um the ones we are seeing that are having traction of course stable coins are the the easy ones but I would say yeah commodities we seen traction with gold especially and and there is potential for real estate even though it's it's proved in the past the tricky very tricky one but of course has huge potential as a as a

collateral. Okay. >> Yeah. I feel like commodities isn't spoken much about um especially in this environment where um you know we are seeing a lot of participation in like hyperlquid over over the last month when you know prices were spiking people wanted to hedge their positions. Um I I'm interested to see where commodities go. >> Yeah. the the problem with commodity I guess it's very hard to find the a good issuer structure because will be like someone that gives you like barrels of oil uh so I'm going to need to move with the derative just you have these so uh that's hard and um to your point

Jonathan you were making before you kind of mentioned it but secondary liquidity and primary liquidity is very important and depending on the assets that there is no a single way to go about it I was wondering if you see any best practices um or like this is even very connected with aexis and the propam's uh work right now right so um just wanted to know your thoughts there >> yeah so what we're currently solve is to is mostly focused on the private credit and a fixed income asset class because um the most of the bottleneck exact uh exist in that asset class right now cuz they all have different kinds of delay in terms of settlement time frame and

also redemption time frame. Uh that's just uh we're we're not used to that type of delay on DeFi. We're used to autonomous settlements. Things can just be done in seconds or even quicker. Um so I think some of the best practice I'm actually been talking to protocols like 3F and a few others that working on RW looping and working on some uh free market um infrastructure to allow a to allow more participants uh especially from the liquid fund and also market maker to come in that naturally trades the underlying assets these bonds these uh private credits. So they can all chip in and they all have economic incentive as well in this whole diagram uh to chip

in and provide liquidities and offload um and and also putting on positions as we need um so so that we as defy users and arbit token uh other token holders won't like have to wait for months or weeks uh to to get out of their positions. Right? So uh those are some of the the things that we're seeing. We just need more fluid free market structure uh better liquidities from institutions that currently trade these assets naturally on a large volume and to find them and coordinate with them uh in a network in a decentralized coordination way. Yeah, I think that's uh that's something I'm thinking about.

Yeah. >> Okay. And um do you have any other points on this on the like secondary liquidity? How you guys are evaluating the secondary liquidity for different assets? what's working and what's not. >> Yeah. One thing that it's um connected to what Jonathan was mentioning, they definitely need market makers, somebody who can be an intermediary, somebody can who can um eventually inject liquidity if it's if it's required in in moments of stress and also u insurers. That's something we've seen of everybody, especially our clients. They're looking for some sort

of reassurance, especially a first uh line of insurance. And I think all of these things are going to allow to to to mature this Yeah. our sector. Yeah. >> Yeah. Um this brings me to another point. So many institutions actually enter the space through custodians uh like Anchorage and it's also a place where a lot of DeFi activity financial activities originated today. Um obviously FWAS have different compliance requirements and this is also another value proposition. Um I wonder if the future looks more like a huge permissioned environment or what would

be the consistent between the permissioned and permissionless aspect. Um so yeah I'll start from you once and get get your thoughts there. >> Yeah on our side it's it's more about the like operational compu uh and compliance controls that are in place. um you know clients are weary of using self- custody to interact with DeFi and so for like lending or borrowing activities they want to hold their assets in qualified custody um so securing their loan with park or Camino using BTC in custody or stake soul in custody um but of course you know custody is a walled garden um so there there's only so much that we can do and

in terms of supporting our clients that are on the bleeding edge, we've offered our self- custody solution which allows them to directly connect with any protocol that they want to interact with. So, you know, so if they want to go further, they have a browser extension to do so. >> I I think I think we're likely going to have both. I think both are going to are going to grow. I think we're going to have this permission economy like they exist today and they're going to replicate but we're also going to have this permissionless that is going to act as a as a sort of a bridge with with within this this uh permissioned um markets just as a sort of an arbit arbitrage like just like

today we have um you know um shadow economics that tend to bridge different permission markets. So I think we're going to have uh both and and both are we need both to to have some sort of equilibrium. >> Yeah, I think it depends who who you're talking to really. Like if you're a prop fund, you're just managing your own money, then maybe permissionless works well for you. But if you're managing external capital, your investors are going to ask who your counterparties are, the venues you use, and what sort of protections they have around them. So the fully permissionless open system might not work for one but

it's pretty much required for others. >> Yeah, I think I agree with everyone. Um so uh we are infrastructure provider. Uh we've been focusing on permissionless technology for years and that's always going to be and continue to be a focus for us. Um but there are growing interest on the permission vaults uh solutions from institutional side so they can stay compliant and feel safe uh from the regulatory environments. Uh so yeah we are one of the few lendium boring protocol that can build uh fully permissioned vault infrastructure and tech. Uh we re recently announced a partnership with securitize as a first use case of that. Uh there are way more a lot more partnerships with uh custodial providers and uh other tokenization platforms and issuers like

that. Uh I think that's the adoption of the permission vaults will continue to go uh continue to accelerate I think. >> Okay. Uh yeah this is very interesting and um I wonder what's how the counterparties will evolve like so my my question is uh what do you guys think is the like main user today overwas and will be the next marginal users um where is the next leg of growth for this market coming from maybe I start with the oily for this >> yeah I think as I said at the beginning it was primarily retail But I think the next significant user group will be asset

managers who are moving their strategies on chain. They obviously have a lot more capital. They trade in bigger sizes and they can get a lot more creative with what they do on chain. And once that side is demanding this then all you know the prime brokers, banks, market makers they're trading with, they would also have to look at options to facilitate this. >> Could you expand on this? What what what do you think is the like end customer for these asset managers? I'm thinking about like fintex that provide uh like instant yields. I'm thinking about so like an asset manager of a bank of this world that starts to use DeFi. What do you think is the the whole value chain?

How do you imagine it? If you look at ETF issuers for example, they can have fully onchain ETF issuance and then the clients or you know retail customers who buy that ETF then they can have exposure to not just you know what you would normally get in the stock market but whatever can be traded um onchain. >> Okay. Um, Marcel, do you have any other com comments on this? Like on the next kind of cardies that are coming um and why they're coming on chain to be honest, I don't know where the

demand is coming from. No, we're we're building the infrastructure. We make sure it works and Yeah. Okay. >> Yeah. I I I can add a little bit. So um I can add a little bit. So um you guys probably have heard uh a lot of the centralized exchanges crypto centralized exchanges allowing people to trade to buy to purchase and get access to uh tokenized equities as well. Uh so I think that will um on board a lot of emerging mark uh emerging country uh uh retail investors that get can get access to a lot of uh uh US stocks to start

with but hopefully more uh country stocks markets um through uh crypto exchanges and tokenization. Um I think on the retail side this would be a big adoption point. Um on the institutional side um I think they just wanted to further diversify their portfolios on chain uh because they're more sophisticated. They look at their onchain portfolio from a holistic perspective and they shift their positions from risk on moments to risk off. So if it's a riskon moment they will switch to more Bitcoin ease more cryptonative assets. If it's risk off they will probably hold more treasury bonds. you'll probably hold more tokenized gold on those type of asset that behaves um counteractively to risky assets right so I think for institutions

the more options you have arguably obviously the better uh for retails I think the centralized exchange tokenization stocks is a huge unlock I think will be a huge adoption um and we will see what the fintech partner uh can unlock as well I think it's probably just going to start with stable coin But they're probably going to allow tokenized a stock as well pretty soon. Yeah. So, >> okay, I see. All right. I think uh I'll do the last question and this will be about where do you see the space in three five years from now? Uh what other things that have happened and uh where where's where what are the next steps or like this journey? So yeah, maybe Mars

is if you want to take this. Yeah, I I see the space moving in a direction where, you know, like the fintexs and the crypto exchanges, they're, you know, tokenizing real world assets and delivering liquidity to their their client base. Um, no longer are you gated from accessing a line of credit against your equity or a line of credit against your um, you know, your private credit assets as well. Like I think a lot of that has been very blocked. Like retail users have been blocked from using equity. Unless you have like an account with Meil Lynch and significant assets for example, you're you're not able to tap into the liquidity. Whereas in a

permissionless system, you can go ahead and deposit your tokenized assets and take out a line of credit. And I I think where we've in the space has right now been on like very you know varied interest rates that move according to the market but we are developing systems where there's fixed rate um infrastructure where you can accommodate that on chain. So I'm excited to see where that goes. Do you have any concern like if the sector fails like what's could be the the cause >> the sector fails? um you know in terms of smart like smart contract risk always comes up but I'm I feel like where we're

moving in the direction with you know AI like smart contract risk starts to really decrease as we're able to pressure test all of the different area surface areas of attacks and so I think we're getting much better there but I do have concern over you know you know assets coming on chain that aren't you know as liquid as they should be. Um because that could cause contagion. >> I think two things will happen in um 3 to 5 years. One, we saw RWAS give access to onchain users give onchain users access to assets

offchain. So real estate, stocks, ETFs, um, private credit and now I think the opposite will happen. So institutions who are not on chain will move for the previous reasons we discussed and do more things uh on chain and I think secondly we will maybe not stop but definitely lessen referring to these as tokenized assets and others as just assets. I think it will mostly be the same thing but tokenized in the background. Like you'll have traditional brokerages or retail trading apps just having a stocks offering but it will just be tokenized stocks in the background for a more efficient and

seamless product. >> Yeah, I think um in this next three years I see two different directions. For retail it's going to the route of simple is better. um basically oneclick earn solutions or oneclick borrow against your stock of tokenized equity. So everything has to be one click. Everything has to be super easy uh risk managed of course. Um and people don't need to understand what actually works in the underlying. They only need to know this is available only through crypto and it works uh and you're earning yield on your financial assets and now you have a lot more tools. Um and as DeFi builder we just need to get integrated to all different avenues that

includes centralized exchanges, custodial providers, fintech, wallets everywhere basically allowing a very like good front end easy to understand solution. So that's on the retail side. On the institution I actually see opposite uh things will get more and more complicated. uh we need to combine the strands of trifi and defi meaning that we're allowing defi's act advantages but also we need to make sure we have all the trifi tools that the institutions are familiar with for example on the border lending side fixed rate fixed term very common place in trifi not as common on defi but we also have our own products that's not very common in trfy so what

we're building on the institutional side is combining the strengths of both efforts. So the demands for institutions will evolve. Uh they be right now they're doing some simple staking. They're doing some simple borrow land. Um but that that needs will continue to evolve and become more sophisticated in the next couple years and we will stay on top of it and build all the tools that they need to do very advanced financial portfolio management and trading strategies. And I would say one last thing is um uh let me think just lost my train of thoughts. Uh yeah I would say one last thing is um

for for financial institutions um I think the future is going to be uh risk man uh risk measured but also highly capital efficient solutions. Um and there are there are things that they are learning from crypto too. For example, the 24/7 stock trading markets. Um you guys are seeing like large uh stock exchanges saying they're adopting tokenization and crypto technology. U this is a brand new space for Wall Street for Trifi folks which is an industry of a century plus a century plus old and they're learning new things from us. They're paying attention and investing actively. So this is what we as DeFi builders are uh super excited to kind of show them what crypto can really

do with them together. So um yeah, the the way I see this, I think in a few years we're going to start having some assets that end up being more liquid onchain than offchain. And I think there's going to be eventually a tipping point where all of the infrastructure is going to migrate to to uh blockchain pipelines and and we are as a users we are not going to realize about that. It's just going to be abstracted and that's going to be the the moment where yeah you know defi is going to stop being a thing. It's just going to be finance. That's that's how I see it.

>> I love that three years is super ambitious but that would be incredibly bullish for like the whole audience here. All right. Um, time to wrap up for the first panel and was great having you. Great conversation and I think you can see the recording in a few days in a ZeroX research podcast and thank you. >> All right. Um, so the one who is not here, it's his own or her fault because u DeFi maybe or crypto is getting boring. But I think today's topic is not boring because we're talking about capital efficiency, yield, and risk. And I think APY is the word that is thrown away a lot. Um, and I have an amazing

company and I have so many questions and I'm super excited for this talk. I will let's introduce you guys and uh let's start with you Deian. >> Hi, I'm He. I'm the head of risk curation at KPK and I'm mainly involved in everything related to vaults and yield generation. >> I'm engineer. I'm at urine as a solutions architect and ern does vaults vaults with APY. You'll hear about that more but that's enough. >> Nice to see you. Thank you for being here. I am Mike Masari. I am head of atom and cuedora at Reston.

>> Hi everyone. My name is Yugis. I'm from Gbox. Gbox is doing lending. >> All right. Um so maybe a little warm up before we go into details. I basically want to know the decomposition of um capital efficiency, yield, risk and everything about it. And you kind of present maybe a different angle of the world or lending. And um I'm curious how do you define capital efficiency? as a as for example for an ordinary depositor.

>> Yeah. So yeah indeed it depends who you ask. So for example for like a depositor in my opinion is just like the amount of yields um they generate. So the interest they generate per unit of risk. Well, typically what um a retail or like a depositor looks at um mainly if they're not so deep into the space is just a headline yield to be honest for like for a protocol is different like capital efficiency is like how much of the capital is being borrowed out on the lending market for example. Um so yeah it really depends on who you ask and what they're looking at.

engineer. >> Yeah, I think the key word that you just mentioned, Gil, is per unit of risk. And I think these units of risk are very poorly understood in the space. Um maybe not by some of these other analysts here. Um >> maybe not soon. We'll we'll see what happens. >> Yes, we'll explain it soon. Um but yes, this idea of capital efficiency, basically the idea that your assets are doing something. Uh let's say the the opposite of capital efficiency is assets just sitting and doing nothing. Um so rather than define capital efficiency, I'll just define the opposite and let you uh infer from there.

>> Yeah, not much to add here. I think this is a great question because you're right, engineer. This is topic that is poorly understood in web three. We're going to talk more about it today, but at the end of the day, capital efficiency comes from the delta that there is between return and risk, right? At the end, this is what happens not only in DeFi but everywhere. And and now in DeFi, we focus a lot on the return part of things. And mistakenly, maybe some of us thought about capital efficiency as being the return of things without thinking too much about the risk. Now we're seeing a bit more of this balance. So yeah, definitely interesting time and interesting conversation to have

>> uh for lending markets. Uh I agree with Gay and engineer that idle liquidity uh keeping it as much as less as possible I guess that's one all the sides of capital efficiency another side I guess comes for borrowers like another side of protocol and here I guess uh this here angle of efficiency is how much it cost for you to rotating capital to do whatever you want uh to switching from one strategy to Another cost it also usually comes with some inefficiency in the market. You always do like uh swaps uh have some redemption fees etc. So how much you pay for it it's also uh affects

your efficiency uh of your capital. >> Okay. Uh completely agree. We will dive deep into risk return tradeoff. Uh absolutely. Before that I want to ask you uh something different and it is um with respect to incentives because um capital is incentivized and I'm curious what is the structural difference between incentivized APY and organic APY and maybe maybe I can start with ILGS.

Uh I guess this is very good questions to be honest and it mostly depends on the like where our markets uh standing right now. So like couple of years ago uh there a very big difference cuz a lot of people are very bullish on tokens and different uh different parties who uh farm tokens actually had like the internal models and they estimate value of tokens in different manner especially like when tokens are not liquid or have some constraints. Uh and people will be able to like to play with this uh and different sides like depending on case they could be right or wrong but

different people have different model. It's creates some uh some play and magic in the market that like different players creating actually like uh uh like order book. Yeah. like some some uh value of uh strategies in that manner. Right now I guess uh uh a lot of people just considering uh incentives like in very short-term manner and in that case I don't see very big difference right now uh at least from my perspective uh for incentivized strategies and like uh purely organic yields. Uh so I I I see that a lot of uh

existing players just uh see it like pretty much the same and they consider it in very short term and even they if there is some incentivized sometimes they like consider it in even in like pessimistic way because you always when you farm like you have some uh some time before you claim rewards. So they always like consider it like in pessimistic way. But yeah again like we saw that happens all the time in the market. Yeah. Like we are in very volatile spice and uh yeah so people like always playing these numbers and I guess like this change like hope soon but who

knows. >> Uh quick show of hands. Who likes free money? Oh that's a lot of hands. Uh, I think that's all you need to know about incentives. We see a lot of money moving around because of incentives. Um, when you don't have incentives, life is hard. I think that show of hands explains it all. >> Yeah, I think it's also very easy to discern between the both, right? Like uh organic demand and organic yield, they scale with TVL uh while like incentives, they scale with the treasury of whoever is providing them. So uh for instance now like PayPal is doing huge incentives on uh PYUSD.

Um but I consider them more to be let's say a marketing expense rather than actually organic yield is also driving borrow demand. So I think you can easily like see how they scale with TVL and draw your own conclusions. >> Yeah I completely agree. I think that uh you know you mentioned time and and time is really the the key differentiating factor here right. So if you if you're incentivizing yield which I do 100% consider it to be a marketing expense then at some point they will end right. So the longevity of the yield changes obviously by definition organic yield has got more longevity than incentivized yield on average. So I think that the

key differentiator is time. At the end of the day we all like free money. So, and when you have an incentive, there's no specific reason in a specific time interval to put money on incentivized yield instead of non-incentivized yield. But at the same time, you know, it's probably not money that you will keep there in the very very long term. >> Okay, I agree. Uh maybe a small follow-up question. If we have incentives everywhere, do you think there should be kind of um an approach to show how much of the yields coming from incentives that come from organic or is it kind of information nobody wants to disclose?

What's what's your take? >> I'll take this first. Short answer is yes. I think that the more information someone gives to the investor, the better. So this is as a blanket statement I think is always true when it comes to how this happens. Obviously there's multiple ways to make it happen. There are some protocols that do uh and already differentiate between what is you know APY and what is incentive on top of that APY. Uh there are some curators that do as well. U obviously it's just a matter of how accessible the information is. Like when you do as an investor research, eventually you're going to get to this answer. So it's not to me about whether or not this is

obscured because it is not obscured, but it's about how accessible this information actually is. >> Yeah, I agree. Like I really like the approach of like false FYI or Defi Lama where you have these dashboards with yield and you clearly have this breakdown between what's organic, what's incentives. Um and also between the incentivized there's a big difference because some of them might show like uh a future FDV value of their to be released token which is a lot of the times very much inflated. So you kind of have to discount it as well. So yeah like it's a tricky like uh comparison sometimes because yeah you don't really know what a value is going to be in the future.

>> It's very unstandardized right? So that's the real problem. It's the depending on which asset it is and where it is like the information comes in a different way. That's the real challenge. >> Yeah, I think uh the key word in this question is should uh even if we say something should be done, there will be many actors that do not do this. I can give a very simple example. If you are going to any DeFi app, you will see potentially a APY number. Probably you will trust this APY number. The question is is it accurate?

And >> is it like the six hour or the 24hour or whatever? >> And even even then sometimes it's uh like an estimated value or maybe there's a token incentive and this token is not yet liquid and so it's using an estimated value or perhaps the value is stale. Uh the front end has bugs. Whoops, the uh TVL the APY is inflated. That was a bug definitely. So a lot of the time what should happen uh is different what than what is happening unfortunately >> I remember all of this old talks about API or APR if you remember like uh so yeah uh we always act like from transparency side I I think we should

yeah in our industry uh to push it uh and to like show in transparent way uh because yeah again like code is a law That's I guess how defy like get its initial traction and that's on what we should like stand for. Uh again like in future I think like with all of this AI automation stuff it maybe less rely on user interfaces like what's built by uh developers itself because like all these numbers can be checked on smart contract level of course like it's very difficult to usual user to check it uh but I think uh in future we'll see all of this like

infrastructure that helps users whatever like engines maybe uh to to see like all of these numbers like and check is it a real number on UI or not >> because that's a good thing like like most of us are using C4626 so you can actually query the share price over time and you don't have to rely on reported numbers to see like the actual returns in the vault. >> Okay. Uh engineer already mentioned bugs. Um uh we always see that >> it's a feature not a bug.

>> Yeah. >> Bug. >> Okay. A feature. Um so what is then a source or a feature of yield destruction? Maybe Gil you go first. >> Okay. Um yeah I would say like so for let's say underperformance or just general getting wrecked on the yield. I think the the main factor is the underestimation of tail risk events like we saw a few of those over the past few weeks and months. We had like um the resolve incidents very recently and I think there in such a scenario um it's often something that wasn't uh properly monitored or that was missed from security reviews etc. Um and is this one

single event that can change the whole um yeah the net yield that the depositors get like some vaults got hit um even today like they they took a big cut on um well due to a market removal from the market um due to the affected resolve exploit and suddenly you can have like instead of getting like four or five% yield uh it's now at minus 99% APY over the past week. So there's definitely like these tail risk events. I think that's the place where you will lose the most money. Nobody's going to lose money from like underperforming for like a few hours just to pull funds to safety. You're going to underperform by getting like hit in these exploits etc.

If you don't have the proper monitoring in place um which I should give a shout out to for doing this publicly um because you guys like you you open source even like all the the monitoring scripts. um which I really like. Um we have our own in-house automation but like the I think also the more public we can share stuff like the better for the whole space uh in general. >> We'll be doing more of that. We're just getting started. >> Mike, >> I think it's three factors here. There's monitoring, there's time, and there's notification. And so monitoring is like understanding what it is that is happening, right? So the risk or or where this risk comes from or why it exists or why it doesn't exist. Then

there is time because of what Gil mentioned right. So obviously this needs to be something that gets calculated over time and then gets timely updated. And third is notification which is notifying the interested party when something is going arrive for whatever reason. Right? So this is something that for example we're working on on a on a threeprong attack because a single one or two other tree is just not good enough right otherwise you can find yourself in a situation like you know the resolve situation that happened last week now what to monitor it's the crucial part and it's probably what we're going to talk the most about today but I do

think that in order for us to have a complete risk profile in order for us to understand very well what risk means they needs to have three. >> I think there are like also like two sides to look at this problem. Yeah. So first one is from like user side uh and yeah like again like if you are small user of course like it's too difficult or like sometimes uh cost too much uh to maintain all the infrastructure. Of course you you can't do it and to just uh sometimes like put money and pray that everything goes well. uh for bigger players like uh I don't know like KPK for example they run a lot of their own infrastructure to check what happens

with the deployed capital yeah uh and for pro from protocol side it's another angle so in protocol I guess uh our like teams that build protocol itself they have very deep understanding like what happens inside and sometimes they have like their own angle and understanding like what potentially could go wrong or even like if doesn't uh lead directly to losses but potentially it could like in next stages could lead for something bad and I guess has some like this internal monitoring uh uh infrastructure for each protocol is crucial to have and unfortunately not

every defy protocol right now focusing a lot like on this and I guess this is something where we can as an industry and like what we can do better and improve here like uh doing some standards sharing some best practices between teams and focusing on it more cuz yeah like again like it's not something that hidden like uh by usual user and not directly uh affect on your metrics and TVL because yeah very few players like who chatting directly with team can understand like okay what's happened under the hood and what kind of infrastructure built under the hood to maintain all potential bad cases. Uh and yeah uh but anyway like I guess this is

crucial to uh to grow industry like from uh current stage to uh Marcela's vision like in in in three years like to replace all current threat infrastructure and like be settlement layer and uh financial layer for for the world. So yeah. >> Yeah. And it's also a lot of invisible work, right? That might be why not so many teams focus on it cuz ideally you never need it. Um but it can save like uh your TVL because on gearbox for example we we run this like insolveny monitoring developed uh by the gearbox team and what happens is if there would be a shortfall in the protocol um the

markets gets automatically passed without human intervention. So it's kind of a fail safe to stop the worst case scenario from happening even when everybody's asleep uh etc. So and it's something it's nobody knows because it's happening behind the scenes but it is there. I'll be quick uh I guess I'll alter the question which was about APY or yield distortion and earlier we were talking about uh per unit of risk. I think there's actually a lot of risk distortion. I'll give an example. If you go to Morpho, you will see a lot of vaults and a lot of APY numbers. You won't see anything about risk. This basically incentivizes the curators to pick the highest risk opportunities that

give higher APY numbers, and you don't see that. So, it's it's distorting the yield, but it's really distorting the risk underneath. >> Maybe that's going to change soon. No, I have a feeling. >> That would be great, actually. Okay. Um, so I think we already kicked off the topic of the risk and I'm super excited about um and I agree with you that uh risk monitoring, risk transparency, risk clarity is something that should be open for each actor in the whole lending whatsoever starting from depositors, risk creators, borrowers and I'm curious about uh maybe it will

be a longer question what is Actually um how do we price risk? Uh which metrics do you monitor? Um for example you mentioned morpho we have lending vaults that have um protocol bounded strategy. So you are only in the protocol and then you have vaults that are cross protocol. So actually you can um create a very crazy strategy. So you compose the yield across um vaults, AMMs and everything that you can you can find suits to your strategy but I think you also kind of compound on risk whether it is a smart

contract risk whether it is the asset risk whether it is um credit risk. So I'm curious about what is your what's your opinion on risk pricing and composibility here? Maybe Mike you can you can start there. >> Yeah. So this is a very interesting question and funnily enough I think that the risk pricing part of it it's more up to the curator than it is to the risk organ. risks exist right they exist in everything and it's something that uh again it's just about how visible and how accurate it is in terms of reporting right when you do anything in your life

whether it is financebased or not there is a level of risk and some people are very very good at doing independent research and figuring out exactly what the level of risk is some people are less good and some risk are much more difficult to estimate so the first part of your question which is how visible It should be to which person the answer is very simple. It should be as visible and as understandable as possible to everybody. That's I would say we can take all of this uh as a blanket statement that should be true in everything in an ideal perfect scenario perfect word. I will add that independence is a very important factor here because having selfassessment of risk it happens in

trfy a lot where people like you invest in a trfy asset or in an ETF and it's risk you know seven out of 10 safety like and you say why and they say well seven out of 10 it doesn't make a lot of sense right and and it's important that it's transparent and is verified independ dependently okay that by a a third party at the very least if not by anybody. So that is that is the additional factor. Now how to price it is a result of the risk profile information right and and there obviously you know higher risk doesn't necessarily mean worse right so the more secure investment they're not necessarily inherently

better than the least secure investment people should have the information and should be able to make their own mind on how to build their portfolio based on what their personal risk profile is information should be there and pricing should be also there based on what the curator for example wants to do with the risk you know sometimes and I I talk with a lot of curators not being one so I'm the least expert person to talk about this but sometimes you want to have a more risky vault for example well because he generates better yield in a specific amount of time and that's acceptable and it's an acceptable level of risk if everyone is informed the problem like engineer was saying is

if the information is not there. So if you know there is a front end bug or there is like a misassessment or whatever reason why the information is not transparent. But more risky doesn't mean worse if it's transparent and is independent. >> K. >> Yeah. So maybe to come back to like um like from a curator standpoint like how we do the risk assessments. Um I think for us it's easiest if everything happens onchain and there's as much as possible information and it can go very advanced. So for instance one of the latest markets we created on morpho and onboarded in our high yield vault. It's um based on a product by origin protocol

and it's basically combining um their own infrastructure. Then they um combine it with lido. They kind of buy deep or they buy um underpriced st eat and they go in the native unstaking queue to redeem it and that's how they make a profit which is very profitable in uh volatile times. Um under the hood they use the urine vault um that's not listed anywhere so people don't know but like we know um and yeah they cycle liquidity either in lending markets or in um like lost state E. So for us the good thing is it's a lot of complex interactions but we can all monitor them. So for this one specific uh we have 300 smart

contracts we are monitoring just for this one collateral market. So if anything goes wrong, it can go from um even like the the native unstaking queue of Ethereum, anything happening with Lido oracle with the chain link multisig with all of these uh components involved, we will automatically exit the market on Morpho within uh one block or like as soon as we get the alert in autonomously. So for these type of things, I agree like transparency is key and I think we should try to keep DeFi as open as possible. Um because we see now a lot of these yield bearing stable coin vaults. They might do some less on exchanges. They might move it into wallets that are um where the custody is

a bit more a gray zone like who is holding the signing keys etc. Um so from a curation perspective it helps us a lot to price this risk and to determine like the LTV for that market or maximum allocation if we have that information. So yeah, I would just like advise anyone that's building something, please follow like the standards that are there and try to do as much onchain as possible. >> Okay. Yeah, I think uh to add to what was just said, there is a bit of a difference between DeFi and multisigfi.

In one of these two, you basically give your money to a multisig and just hope they do the best with it. Um whereas if you're dealing with DeFi, you should be able to fully verify where everything is on chain. And I will say for uh year, the main two uh ideas related to how we look at risk, it's who controls the funds in terms of centralization, decentralization, and where is the money? If we can't answer these two questions in a way that indicates the value is safe, then really you shouldn't be putting very much trust into it. So just two simple ideas. Who controls it?

Where's the money? Um these are good guidelines to start with. And I see to you guys as well, it's okay to get exposure to risky stuff as long as you can transparently verify it and get out in case something bad happens. Also I think uh one more thing to add is uh also crucial to have some diversification. Yeah. Like uh because it's always some like probabilities theory that any asset actually can go to zero like something could happens some smart contract could exist uh and be executed uh not in a good way. So it's always some uh prob probabilistic uh game and uh good way is to have like some diversification between different

risk profiles like having exposure to risky asset is fine. Yeah. If it's not at 100% of your assets. So having some like uh instruments uh some tools to properly uh build this diversification I guess is a good way to both like uh asset issues like curators and protocols uh to mitigate some bad scenarios. Yeah. And I mean in favor of the curators it must be said that most of the vaults actually work like this right. So there is a diversification of asset is very uncommon to have like one single risk profile in the vaults and every single asset is like particularly risky, right?

It would not also be a very smart thing to do. Uh but yeah, I do agree that you know understanding it is the first step and then composing it is a second. >> All right. Um you mentioned diversification and I think now as um LWA are entering the space I think it gives a big add-on to portfolio that is managed for generating yield but also I think it it gives you a little bit a different kind of risk that is not on chain and maybe not very much trans transparent and I think also we have like also

more um let's say movement to offchain exposure. So I'm curious how do you see this off-chain risk is being priced? How can you match this off-chain risk with onchain risk? And how do you then build a strategy to kind of have a healthy um complex strategy yield sorry risk tradeoff? Yeah, good question. I can like take uh the first part of it. Uh because we're building currently a market for uh oiler institutional uh using securitized assets. So these are tokenized money market funds or private credits or

private uh sorry corporate credit funds. So there you indeed you have like a lot of offchain exposure. It's um credit lines. We don't have full insight into the book itself but um of course you can be quite conservative in the LTVs you set the different caps you set to how much you can allocate to it. So what we do is we allow um regulated or uh KYC participants to borrow against their existing assets uh in markets where we have like the full control over the interest rate model, the LTV, the LLTV um and as well as like the caps that are set on that. So that means that you can actually build more complex strategies

on top like a MV vault like a morph vault that allocates to different strategies. Um and for that we use like redstone oracles uh for example and yeah there again you you try to go into as much detail as possible. We try to find the data sources that can provide you with the right monitoring tools. So we will be for instance pulling like the oracle data from redstone compare it to other secondary liquidity uh market prices if those are available. Um and yeah again like it's it's about getting enough data input to make this possible.

and being not too too crazy on the LTV side. You don't want to yeah give somebody like even though these assets they only update typically once a day the nef. So you're kind of um safer as both a lender and a borrower than in a lot of um like defi platforms where you can have like continuous price deprecation. Um so you typically do have some time to react like we also act as liquidators in those markets. We can step in redeem the assets. Um but yeah, there's a lot of additional like components that in DeFi we were not really used to that you have to take into account but it can be done in a good way uh as long as people that like uh have to estimate the risk don't like

be overly optimistic because bad things will happen one day or another. engineer. >> So I think anywhere you put your money regardless if it's fully verifiable or not there is al always some element of trust. So going back to these two main risk items that year looks at who controls the money and where is the money. Even if you can verify it fully on chain, you have to trust that the blockchain works, that your computer is working, that the the screen isn't showing some bad data somehow. Uh so there's always trust and I think if there is a way to get this off-chain real world asset yield in a way that can

be trusted, uh then it gets very interesting. Uh the puzzle is how to do that. I think that's something that's much easier onchain because we can we have some trust assumptions with onchain data. When we're dealing with offchain, there's always additional entities or parties involved. Um, but that's what I would say the main item here is trust at the end of the day. Uh, it's hard to avoid it. I agree. Not not much to add here. like the only I think valuable added input that I have here is that as part of the risk assessment both should be considered right so off considering off-chain factors and onchain factors it's important and they should also be

weighted in a different way right so transparency and accessibility of information is a factor of risk so if information is more trusted right or more controlled by a single party or less decentralized is inherently more risky than if it isn't. So I think a proper risk assessment should consider both factors. I should also weight them against each other in a way that it makes sense just to spit out like as as we were saying before a standardized risk profile which is the main problem like now we don't have that it's very difficult to understand it because the information is fragmented it's very difficult to not only access but also

interpret now if we standardize it obviously it needs to be done in a proper way and needs to be done in a way that considers all of this. So I completely agree. >> I have two answers like uh shortterm and longer term. Uh in short term I guess uh like few parties already building some solutions to check what happens in thread file. So like it uh allows like uh to understand like where money is. Yeah. like uh guys from Redstone building it, Lamarisk also uh checking books uh of like where sitting uh exact assets for different uh depth

obligations. So this is I guess like first steps uh at the end in longer term I guess uh we are talking like and we will see uh fully digitalization of uh financial stack and right now like it's something hidden and can't understand that what you can do just trust okay this uh big financial entity uh it has good reputation so I can trust my money uh and put in put into it uh in longer term I guess we will see like uh all the data like from uh like high level to very like end level where money goes, how it how it acts, where it sits, what

kind of limitation it has and all of this like financial stake will be digitized and we can see like in a very understandable way in very like standardized way. Uh so that uh later like user maybe not any end user but at least like more sophisticated players will be able to understand uh where this asset sees what kind of risk and hidden risk uh it has and deploy capital in more properly and more like uh uh in better way.

>> Yeah I agree with you guys. Uh I have one follow-up question here as well. Um, Mike, you mentioned and I guess you as well standardized risk measure. You mentioned trust. So, can you give me one standardized risk measure you would agree on today that would be available for everyone in real time? >> Okay. So there's there's many obviously and there's a number of um I would not call it maybe risk measures I was call them risk parameters.

Um one is depth on liquidity like that is a very immediate thing to understand right so if something becomes in liquid there's a huge factor of risk that I think is probably the one that the most people in this room could agree on. I would I was pausing to try and figure out what would be the most agreeable and I landed on this. >> Uh I say that like all the stake uh all the financial stake it should be rebuilt a bit. Yeah. So uh right now uh what happens like when uh thread fivey money comes I mean like thread fivey products comes on chain it's actually some legal rappers of existing products and these

products works in the same manner as they worked before like in threadfy and uh in longer term I don't think this is like good way and good way I think that we should like rebuild all this product in a bit different manner like like when uh internet uh came into our lives. Yeah. Uh if you just uh put a website uh for some shop or whatever like service, it doesn't work well. Yeah. It works well when like it comes with all the supply chain that has like Amazon or any other uh big player now have. And in that case it increase efficiency on the whole of the system. And in I think something similar should happen like in

like like five years maybe 10 years. Uh uh yeah I agree there also like on the RWA side like liquidity is a big issue. Minting and redeeming is a big issue and it's currently all going through traditional finance rails. Okay. Some projects also allow minting with like USDC but like the underlying assets still require like US dollars on a bank account. So yeah, I agree like there we can make a lot of progress and without resorting to let's say Zimbabwe real estate as our golden standard but really build something like sustainable. What I love like in old uh defy yeah it's fully like liquid and velocity like very uh very high. So like for example when you

on board uh Ethereum BTC for example as collateral in lending markets you always can understand what happens inside it because you have a lot of and very deep dex liquidity for these assets and uh they like they are fully liquid and it's very understandable from risk side what happens with these assets uh and I think something similar should happen with threat foss like uh anyway like even if it's private credit. It still has some uh like you can understand like what's uh how how it cost right now. And one of the problems like why uh we have such big redemption periods that this market

inefficient in terms of like uh speed velocity in terms of like uh rotation capital and I think such kind of things could be improved uh by using blockchain technologies and practices that we had in in DeFi. uh maybe something new like some new products will like uh uh will appear here to improve that. But anyway, I think that's the direction where we should move. >> And as for risk metrics that everyone can agree on, I'll go with the very simple one I mentioned earlier. Do you know where your money is? If you don't know where your money is >> at 5:00 am in the morning.

>> At 5:00 am 900 p p.m. midnight any time of day. >> But you to be honest like you always know where money is. Yeah. The problem is that data in some like in some levels just disappeared. Even if it's private credit for example uh at the end this is some money that was given to some uh entity and you have a data what happens inside the entity and the only problem that this data doesn't move on the next level. So for example, bank doesn't have this data. So he knows that okay something happened like 6 months later and then it could problem like for portfolio and then uh it come like with some problems for RWS etc like in one

year for example but you have all of this data like the only problem to gather all of this information together and uh make it like transparent and available for different parties to be able to measure risk and understand in real time what happens in the whole like in the whole economy. Yeah. >> Yeah. Full full stack of of understanding and also you know can you move your money for where it is now? Like oh it's there but like can you actually move it somewhere else? >> And and like you were saying if uh there's only a monthly report then you really don't know where your money is at at 5:00 a.m. at midnight. Yeah. You only know on the first of the month where your money is.

>> Yes. The problem with monthly report you can automate it and do second report. each second you can understand like what's happen with money each second. >> So we need to get all asset issuers on chain basically. Yeah. Yes. >> Um maybe um another parts of this coin is um which is your early warning signal of an upcoming credit cascade. Some of the people say observing ws okay if I see the utilization is hitting 100% this is my early warning. and others say that the basis widen for yielding assets.

So what is your early warning sign? >> Yeah, like I said we have a few hundred but uh I would say like actually like utilization is a great one like spikes ideally you catch it before 100% because then you're already too late then something happened and all the liquidity is gone. So how we do it on morpho on our rebalancing uh automations um as soon as a certain threshold gets crossed in the utilization rate we stop allocating to the market and we will start actively pulling funds from the market as well. So for instance with resolve um like let's say utilization spikes above 98% we will start to automatically withdraw as much as possible um like the next block uh from the market. So I think utilization is a

great one in lending markets at least because you see there is a liquidity crunch and like I said before we're okay with spending um like an hour or half an hour in a less yielding position just to avoid any losses for depositors but there's a there's a few you can look at DEX liquidity you can look at minting volumes you can look at like abnormal things like yeah dex um there's all kinds of early warning signs uh you can and should monitor or >> if you pick one. >> Yeah, it's probably utilization rates on Moro at least. Yeah, >> it's interesting because we have been thinking about this a lot and it's it's fascinating because it's part of a risk

assessment but it's also is the notification part that I was mentioning before as in is the job of the oracle to notify when something like this is about to happen or may happen even if it's not an Oracle problem. I would argue it is. I would argue that as part of the data set that you provide, we should evolve into proactive notifications to risk curators for example curators on stuff that is happening and and I'm very happy that Gong and KPK like do a lot of monitoring themselves which is very noble of you

and it's very involved. I'm sure it takes a lot of time. I would like to see this everywhere. uh in applied like as a standard this is just a standard parameter that comes directly alongside the price right or alongside the specific you know information that come directly from the oracle so we're thinking about this and we're thinking about what to monitor and how to monitor it best and what are the vehicles to serve this but it's 100% a problem utilization definitely I would say the biggest indicator at least in certain specific scenarios there are probably 50 to 100 indicators and obviously it becomes very involved to monitor those there's an infrastructure cost and that

there is also you know a riskreward conversation to be had there but as a general statement I do think that a lot of this information should proactively arrive to the consumers so that these automation can be put in place uh in a widespread fashion I add one more uh dex liquidity imbalance uh cuz yeah if uh like something happening usually like somebody knows some insights uh uh what's happening inside uh and that leads to imbalance index liquidity and if you check these numbers uh it can

work as some predictor uh uh and can help to mitigate like some risk for Uh I will actually defer you to the uh urine GitHub uh repo for risk monitoring because I don't want to give the wrong answer for which one we look at the most. But it's all open source so you can check yourself. >> That's a that's a great answer. Um maybe one followup here as well. Gil you mentioned utilization and in defi we have this advantage of a public information so everyone observes the same information at the same time and if bad news are happening like uh

utilization heating 100% everyone is liquidating the same pool age sale depresses the price for the next one so expected shortfall and recovery is super linear and then in the end you have this exit problem. How what what can we do actually here to to solve that? >> Yeah, it depends because like in a lending market um typically there's always funds lent out. So um even if everybody wants to withdraw at the same time that's not possible. So it's kind of it kind of becomes like a race um to be the quickest out of the out of the pool and like yeah we have our solutions how we do this like you can go super

advanced like you can even like start monitoring the meool or work with black block solvers to even get included in the same block. Um but typically like most um like we we saw it last week in action like um let's say liquidity returns in the market um it's a race for the curators to withdraw as soon as possible. We noticed currently like it can be very slow for a lot of parties. I I shared like a tweet last week. Um I think it range between like instant to um about 500 600 seconds. So, it can be like 10 minutes before like an entity reacts to uh liquidity returning, which I believe is absolutely insane. And I

think we're going to see a lot of more um like speed increases and sophistication, especially as more of these stratifi asset managers also moving to the curation space. You have like people like Bitwise and they're also like uh going to be active in this space very soon. Um so for us, it's a lot of the infrastructure and the automation we use. um to detect this and also couple it to actions to act as soon as possible um within like the the capabilities of the blockchain of course. Um but I believe like uh the numbers we're seeing now like I would be very surprised uh that people are this slow uh next year.

>> Okay. Um anyone else want to pick up? >> I hope one of you can answer how to solve a bank run. I don't have the answer. Maybe Mike. >> Um I to be honest I agree. I don't I don't have much to add on this one. I think that um this is a very specific problem that you know curators think about much more than than we do. So I remit the judgment to Gil. >> Okay. Um maybe I I think it will be the last question actually.

Um you mentioned a lot automation. Um and uh it's not a a secret that we are in AI era and we also have like DeFi um welcoming Treadfi and Treadfi is heavily regulated. Um so what do you think if we look in the future who is going to do risk management and um parameter settings? Should will it be AI agents automated things? will be regulators or someone else third parties who will take a marketplace.

>> I love this part of the panel because it's usually the part when I make a fool of myself by giving a wild prediction that never happens. So take this with a huge pinch of salt and I think I'm speaking for all of us just saying this because speculation is speculation but at the end of the day my personal very personal opinion is that there is no near future in which AI can replace risk assessment 100%. >> I'm happy to hear that >> I don't think that will happen. I don't think that should happen more importantly um and the reason is quite simple. Risk assessment is something that as we said has got still a lot of unknowns and AI models are very very

good on working on known data but they're not particularly good at working on unknown data and bridging that gap where and understanding how big the gap is and how important that gap is. So I do think that in the future we will see more AI when it comes to understanding parameters uh potentially when it comes to even identifying the differences between different risk profiles. I do not think I do not hope that there will be ever a switch entirely from human to machine in doing that. The second part of your question and I'm just going to wrap it up very quickly goes back to my previous point which is independence. So the risk assessment should be

independent to me is not a risk assessment if it's not. It's just like the Obama meme of people just like giving the medals to themselves. Uh which doesn't make a lot of sense and and and it's also the the thing that happens in trfire a lot. This is a seven out of 10 because it is a seven out of 10. Um so I think that there needs to be a standard that needs to be by independent or independently verifiable uh with public methodologies and and ways to interpret it that is accessible to the most people possible. Yeah, I think like historically it's been mainly let's let's look at like let's say a or something there the risk parameters have been set by the DAO first but they notice okay this is going too slow so they move to um service

providers like um gauntlet was involved chaos labs etc um I agree like I don't think we should like let AI do this by themselves um but I do see a lot of benefit in using AI tooling to help you speed up the process implement this stuff and better understand it. So to give an example with KPK, we write algorithms ourselves, but we will have like AI um for instance analyze or rebalance data and make suggestions where we could improve some of the parameters like tweak them to avoid I don't know wasting gas for example etc.

Um so I think there's a a real future of let's say uh curators using AI tools um to set this risk parameters etc in a way that's also regulatory um like compliant because that's I think unfortunately that's somewhat the future we're heading in like more of finance is becoming like is getting onchain. I think regulation is good to some extent as well um and we're going to see more of it uh in the future. So yeah, I wouldn't let Claude like run the vaults. Although it could be an interesting experiment with full disclosures to like have an AI agent do all the rebalancing and etc. But yeah, I think we're still a way off u from doing that safely.

>> People trust to people. Yeah, machines to machines. So while agents don't hold most of the money, I guess we uh we are safe. >> Engineer talking about who is holding the money. >> Yeah. Uh I will just uh repeat what was already said by Gil that uh using AI to speed things up I think is a is a good approach in the short term. Uh whether I can predict the long-term future of AI, I'm not going to say anything on that. >> Fair enough. Fair enough.

Okay. Um, I can ask one more question. No, no, no more questions. So, we have to meet again, guys. Okay. Next year. >> Thank you so much. We have to do this again. Thank you. Bye.