DSS 2025 Panel: Financial Security Risks of DeFi

Transcript

Hello. Hello everyone. How are we doing? >> Good. Exciting times in DeFi. I think this uh panel is more relevant than any time in the history of DeFi. We've got really uh great uh panelists and would love for them to introduce themselves before we kind of delve into uh financial risk in DeFi. Yes, sure. Sebastian, co-founder of Steos Financial. We are providing earn product for stable coins and the world defy. We are creating on moro kamino and also managing the grove subdow of sky.

>> Yeah, I'm monet supply head of strategy at spark. Uh we're one of the largest onchain capital allocators in DeFi uh within the sky ecosystem. >> Hey everyone, I'm Matt um working with Allay Labs. We're a risk and uh data firm working with protocols prim primarily lending protocols uh across SPM EVM and uh all sorts of exciting chains. >> Yeah. Hey from B protocol also risk dow over the last few years we are consulting in markets and now also curating vaults on mor and a bit also on oiler.

>> Hello I'm Remy. So I have two hats. I'm the co-founder of lagundet finance which is a permissionless vault infrastructure for curators or asset managers and also a co-founder of nine summits which is an asset manager and curator on morular and silo. >> Fantastic. So we've got like asset managers u and lending protocols and like almost the risk assessment tools and we've got we've got it all. I think uh so I'm sure you know that the October 10th uh crash of crypto has brought a lot of like contagion that spread across DeFi and that brought down some projects and now

the debate is about basically risk in in fin like in DeFi well who's managing risk so I think the first probably the question to everyone in the panel is uh this shift in defi risk the the risk landscape how do you see it in 2025 five. Yeah, I can take this one. I think people learn that when you go on something called high yield uh and there is a high higher yield there is risk. So sadly it's sad that we have to learn it this way but uh yeah I think people now realize that chasing the highest yield is not always the best option and I hope that people will start to thinking more about the risk side on the risk return profile.

Yeah, I think that's that's a great point and it's also uh it's getting more challenging for people to assess risk because you know previously there was maybe a few lending markets like you work with maker you work with a you work with compound now there's you know even dozens of risk curators they all have different approaches different um methods of risk management so um the job of a user is getting a bit more demanding um and I think we haven't really caught up with giving them the information they need in a in a digestible format.

>> Yeah. One of the really nice things about DeFi in 2025 is that unlike in previous years where um as Monet just mentioned there's just a few um opportunities for users to kind of dive into DeFi, uh now there's a lot of different opportunities whether there are different vaults with different um kind of risk and and reward uh opportunities there. And the challenge really is making sure that nobody's taking on much more risk than they intend to for for the given yield. So kind of making sure that the risk and the reward stays balanced. It's um something I think the ecosystem as a whole has a lot of room to improve and kind of um helping users figure out where to to place themselves on that

spectrum. >> Yeah. So definitely more for then also Oiler and others introduced you know like a new concept of risk curators and kind of delegating risk management to others maybe also to the users to some extent specifically in the last year or so ever since it's we saw that you know most of the yield actually moved offchain so a lot of offchain components were introduced arguably with higher risk but you know with the way onchain things are going it's also kind of debatable you know where eventually most of the risk is other than that I think you know like the trend maybe of 26 will be not only risk curator but actual you know like

fund managers in the kind of flavor of midas and others which you know never really existed until now I think what we've noticed in the last two years is uh the externalization of risk especially in landing protocols so there is now like two model where the DAO and or protocol will would uh uh handle the risk and the model introduced by Morpho three years ago where uh the externalization of the risk to curators or risk manager asset manager whatever name you want to call them. So I think it's uh interesting to see how uh for maybe for user now there is uh they have

to make a choice for the protocol and also for their uh curator. So it's becoming more complex to assess a situation and to to look at the risk and uh when you allocate to any anything on chain it's uh you have to know your curator very well I would think >> and I think uh probably in 2025 we've seen the rise of these yield bearing stable coins and basically underlying you know token uh tokenized strategies and uh and uh all of a sudden I think 2025 people have stopped asking like well how do these assets work like are they permissioned are they permissionless back in the day I

remember like two or three years ago everyone said like well you know u the makers uh die was just permissionless you could redeem it and all of a sudden today in 2025 just recently people discovered well I'm trying to redeem this stable coin for USDC but there isn't any where did it go so I think a lot of these questions probably we'll try to answer in this uh in this panel But the honestly like for just the the the average DeFi user or really the asset manager the question is always like what is the risk surface in DeFi?

Uh probably like uh Sebastian you allocate a lot of capital in DeFi what are we looking at exactly? What's the risk surface if you can break it down to to to us? >> Yeah so is uh involved in orienting up to $3 billion of uh capital. So that's a lot and the risk is very different depending on what you do. We are locatedating for instance in T bills. I know in DeFi a lot of people are saying well T bills those are risky those are not liquid T bills are government money market fund are the risk-free yield of trfy at least and now it's coming on on defy. So that's quite easy. It's not instantly instant liquidity but you will

have it the same the next day. So that's fine. Then you have the lending protocol. So here you need to understand what is the collateral and we see depending on the collateral you can have crypto collaterals you can have hedge funds you can have a yield bearing stable coin and you need to understand where the yield is coming from and for those hedge fund strategies like stream finance well you need to understand how they are generating the return because if you don't understand how they are generating the return you cannot understand what are the risk and it's very simple if you don't understand don't touch it I think that's still a rule that works But then users uh might withdraw from the vault because you don't have competitive yield and all of a sudden everyone in DeFi is generating 80% and

you're generating 12%. So maybe it's maybe I ask you also allocate a lot of capital and what do you do in the for example when your vaults like underperform in comparison because now you're taking less risk for example how do you manage this? So you need courage to give user a lower yield uh knowing that you handle the risk correctly and allocate to the right uh collateras and the right assets. So I think a lot of yieldbearing stable coin are in fact uh just vaults with asset managers allocating uh stable coin on chain or offchain sometimes. So um what should be the industry standard is uh when it's

allocated on chain that uh you disclose all the addresses where the you so user can track what you're doing and the transparency that go with it. So it's very uh and when I go to different protocol it's for me sometimes it's hard to just find the final address where the money is allocated to find the octave or the debunk bundle link with all the addresses and sometimes the money is even allocated offchain so then you cannot track anything transparently or you have to rely on uh at a station uh so sometimes it's fine for example like itina has of chain allocation and uh So if you trust the team, if you trust uh operational risk that they carry, uh

it's maybe fine to allocate, but you have to bear in mind that there is no u instant transparency on what they're doing. Uh so there is still a risk, but uh when there is yield, there is always risk. So uh it's just a matter of finding the right balance as a an onchain allocator. >> Uh Yon, would you like would you like to add anything to Uh yeah, nothing much to add. No, I mean like the the reason was like obvious competition, you know, like over I think I think eventually not everyone went full digen but by the way even those who didn't went full dig hit you know regardless like the competition was very obvious

but also you could see some emergence of you know like branding and you know so like there was this feeling that maybe users don't know anything but eventually most of the TVL did go to you know like more recognized brands even to something like a which you know by far has the most TVL so you know like saying that users don't understand anything is not correctly true but yeah there there are like many users who might not really fully understand >> um Mon I have a question so spark you've one of the leading lending protocols I think you've also always question or like consider whitelisting assets and you know and now with this new wave of

DeFi and uh what kind of how do you assess risk in general before like deciding to either considering an asset or if maybe the community asking you like hey why can this asset be like the borrow in the pool how do you look at at risk yeah we we get an incredible amount of inbounds um there's you know an explosion of different issuers a lot of them have interesting strategies but the transparency is often been a bit lacking. So I think that's kind of like a threshold issue for us like if we can't uh get like good transparency into how your strategy works. We have no idea what the sort of like dependency graph looks like. you know, as an example, um the deusdable coin, I think people

assumed initially it was just, you know, cash equivalents and then basis trade, but then suddenly they were allocating to a bunch of different stuff and and um if you don't have that transparency aspect, you're not going to find out about that until too late. Um, so I think that's a really important threshold issue. But then just trying to like look through each of these strategies, each of these tokenized like hedge funds to see what they're doing and making sure that you're comfortable with each component. Um so it's it's quite a bit of leg work but you know I think uh maintaining that discipline is is really important and it doesn't pay off immediately but over time this is how you build user trust and um you know basically just making sure that you're

you're doing everything right and and being patient. Yeah. So obviously risk is is has has become significantly uh serious in in today's DeFi uh with all these assets and all these uh and off-chain operations as well and probably like no one has any any purview over them and uh so now the question is well who's responsible for managing risk on we've got here like some of course like the the risk managers they by the way it's a it's It's a confusing terminology like you know because I don't know do vault managers or asset managers actually manage risk is it part of the responsibility? So probably the

question is who's responsible for managing risk uh on behalf of of of users or even the LPS or whatever. So maybe um you know Seb you can >> Yeah sure. So there is a strong question to be asked if creator are managing risk and in which way they are responsible if something goes bad. Uh I think it's a fair question and the industry didn't decide exactly yet. When we are creating a vault we are creating something with a specific risk appetite. It doesn't mean that we are saying it's risk-f free or anything like that. It's just that we are targeting a risk appetite because we are assuming that some people want to have such a risk appetite. It doesn't

mean we are comfortable with the risk appetite. Obviously, we don't have the riskiest vault because I want to be able to sleep at night. That's my threshold. But for instance, I always say and I said it publicly on Twitter many times for most people, they should only use prime product, not the one. Now, when we created the one, we didn't have it at the start. before they were live. Most of people in my company were asking me how they can get access to the product before it's live. So that shows you that people are always a bit degenerate and actually most of the treasury most of the STO treasury in is in those high product. So it's difficult to to have a clear answer. I would say the fact that

the creator is investing in their own product is showing that they are comfortable with the risk but doesn't mean that it's a good risk for you. You need to ask yourself first what is the risk you are comfortable with. >> I think that's a really good point and I think a lot of users maybe are not honest with themselves of what is their actual risk appetite. They want high yield but they're not comfortable with like you know occasional charge offs or or uh decline in value of the strategy. So that's an important thing that in a sense users are partly responsible for for managing their own risk.

>> Matt um you also allocate capital you also have access to lots and lot you have your own risk assessment um tools and you have uh so whose responsibility is it and uh what do you think that uh people like uh like like you can do more for the users to understand risk? >> Yeah. So I think that you know as we all have kind of been talking about there's um even the same you know curator might have a variety of vaults with different risk levels and it's really um not something that the industry has really figured out how to actually communicate what these risks are and quantify them.

So although it's very easy to look at either the instantaneous yield of a vault or a market and um you know or historical yields for that matter um it's a little bit more difficult to communicate these risks and there's several risks right there's liquidity risks you know are you going to be able to withdraw all your all your assets um at the click of a button whenever you want there's also you know some sort of um charge off risk or insolveny risk in these protocols too and I think you know even if you're investing in you know a bond fund or something like that if you're kind of um you know more on the high yield side just in traditional finance right there's going to be some um insolvenies potentially in there and that's just the cost of business that's not necessarily a a horrible thing um

and I think that something that a lot of us um you know are working on now is kind of how can we better communicate that to users so they can pick so they can choose by themselves as we saw the other day you know there was um you know high utilization in a lot of vaults in a lot of markets It's um if a user is not comfortable with that, you know, maybe needing to wait 12 hours to to withdraw their liquidity, um that's something that they absolutely should consider before entering some of these positions. And it's not really something I think um most of these same people that were, you know, a little upset about it had thought about prior. So I think um you know kind of better illustrating all of these risks and um you know helping users figure out which they're

comfortable with is is really important something that we're all um trying to work on. There seems to right now the debate is well what about governance? What about like you know protocols governance managing risk? There's another other thought uh uh about hiring professional services like of course there's a lot of them uh and some Dows already work with them. So maybe the question to your own how do you see this uh playing out in in DeFi governance token holders versus teams hiring like you know professional services. Yeah, I I think eventually, you know, like the vision we had is that the end of the day, you'll have community or token holders that will kind of have to decide

on the level on the risk while you will have professionals that will have to report, you know, like the risk level. H I think it's perfectly fine for certain communities, you know, to decide they want maybe a risk that is twice as much as other, you know, for others to be more conservatives. And yeah, it might not be realistic, you know, to to expect to token holders to fully understand the risk, but but once the risk is reported by professionals, I think it's eventually up to the community to decide, you know, whether today they want higher or lower risk level than yesterday.

>> Yeah. And um when you look about risk and who manages it, uh there is also a question. So whether it's service provider of DAO or externalization of risk through creator there is a last responsibility of the protocol to actually how the data is show and what are the option for end user on the front end. So for example in many lending protocol um being a curator is permissionless. So anyone could do it but ultimately the responsibility of the of the protocol itself to uh list it on their main front end. uh so uh protocol still have to do curation of curator in the end uh a little bit or at least like

QC them or make a few see that they operate correctly. So uh there is uh a little segmentation about who who will show what the user have access to in the end. I think uh DeFi today is just for really the DJ the the individual who's willing to take massive massive risks and obviously a lot of users just do not do due diligence and probably sometimes because the tools do not exist or maybe very hard to to actually find and sometimes like uh information is very difficult. So who knew like a certain oracle was

controlled by a centralized entity for example or like you could not redeem this redeem that but I think that's uh probably the responsibility of uh of entire DeFi if we want to see DeFi expand want to get like the millions and probably like hundreds of millions of users using it. You can't tell someone crossing from Tati that well sorry like you you know someone run away with your money because of this or they lost on a centralized exchange. And now kind of bring me to the stream case like stream finance uh and alexir recently I think that's probably like the consequence of the crash October crash lost money and and thousands and thousands of users like I personally deal with this every day and trying to help users with the

users of of silo finance recover their funds that's our number one top top priority because and uh now maybe someone uh can walk us through what happened exactly like why stream and are like both kind of like the you know uh >> maybe I can uh pop in a little bit. I've looked looked into this stream situation recently but uh stream was doing a few things that kind of amplified the risk of their strategy and and had it created contagion essentially with other related products. So first of all they were um delegating or investing their user funds uh fully offchain in a very non-transparent manner. um they've disclosed that they lost 93 million in one of their strategies, but it's still

actually not disclosed who was running the strategy. What was the strategy? I've heard some people say that it was like a volatility selling or like some sort of option strategy, which is it's not the typical, you know, sort of delta neutral strategy that people might be expecting from one of these these off-chain assets. Um the other thing they were doing though is they were um minting their own token and then posting it as collateral across various lending markets to essentially draw in like leveraged capital. Not people who are intending to invest into XUSD but just people in various different vaults across the space and there was a bit of like sort of crosscontamination between XUSD and DEUSD and various other funds

that were kind of mutually investing into each other. Um so all of that means that you know when they lose a certain amount of money um the the sort of flow of losses spreads very wide and it also gets levered up. Um so yeah there was there was many different things that were kind of wrong with stream and um created a lot of challenges that that people didn't expect to happen >> and uh until recently until probably like uh few weeks ago no one talked about this. Why did DeFi fail to spot those like red flags?

>> So actually someone talked about it on Twitter and gave some warning two weeks before uh CBB. So uh it's a good idea when someone gives some fud about a project to just withdraw everything and assess later as he said. Um so there were a few people also took beyond other like uh so just pointing finger at uh when you looked at the detail of the strategy it was a um big leverage uh crazy else factor across the board but that was that was only for the transparent part. Uh what I didn't know is also that they were allocating on offchain and uh yeah to me the the story doesn't make any sense. uh we

allocated to an external manager and he lost 100% of 93 million. It's very it feels I think there is more to be learned about these stories that just uh what we know uh just just now and um yeah just looking back at uh so I actually met those people uh in real life uh in April and there were young people uh doing uh crazy uh Den and I was like am I am I getting old in D5? Am I doing very very conservative grandfather strategies? is this the new uh defi and I was like okay this is not for me but uh yeah it's crazy how um things turn out but yeah if there is a

thud on something it's better to just withdraw and assess later >> yeah I mean if and oh I was just going to say I think um the idea that no one spotted this is also not quite correct I think that um a lot of people right would see this and maybe they're asked to allocate you know some of their vault to this product or or invest There's there's a million things, right? But I think um you know, when you kind of see a lot of these assets come, you know, through your desk, as they say, right? Uh a lot of them are not a good fit. And that doesn't mean you necessarily have to write like a big tweet thread about, you know, how this is a bad asset or something. I think the standard process is just to say like, "Hey, this asset is

not for me." and move on because there still are a ton of assets that are allocating um you know user funds more or less offchain and um they're not particularly assets that you know I want to deal with right but there's still some out there I'm not saying there's like more assets like this that are going to have a huge uh you know contagion and and you know causes much excitement but um just because no one kind of like wrote a blog post on it I think you know I think people did did see it it's just you know >> it's it's very hard to say no to high yield extremely hard especially if if others are shilling high yield on Twitter you just feel like you're missing out so a lot of users do it so Sebastian what what's your take on the situation you haven't allocated any

money to you did not allocate any money to >> no we we didn't so we didn't made a due on stream finance uh at all because I mean it was just I mean obvious one DM are you an regulated hedge fund with a ton of leverage answer was yes I would say okay not for me But to your point, why isn't there so many warning and some people are seeing on Twitter and so on, but there is some risk, but there also some risk on other projects and we were not and it's just a probability game. Why did stream finance blew up? It could have not blew up at all and it could have worked for 10 years or 20 years or until infinity. So if you are coding it

then you have you are having on crypto Twitter you will have a lot of people fighting for you. there are other bad project or project that can blow up every time you get money to a random guy on the internet and expecting yield. I mean you are taking a risk and there is plenty of project some that are quite big. I remember with my supply when there was anor taluna we were trying to find out how to short it but we didn't say publicly that it's a bad thing because that would be suicide on crypto twitter. So the reality is that we cannot speak up because we want to be friends with everyone because even if uh stream finance was taking a lot of risk, maybe you fake it and at some point you

make it and that would they would end up being a regulated stable coin that was all fine. So you cannot cry against a wolf too early. So I would that's my my take. Looks like they faked it all the way. It's a so I so recently people started asking questions about like okay the oracle why it's still a dollar like in some lending protocols. Well, it's uh it's and then started digging into the source like how do you really read the like how do you value a certain asset and then all of a sudden we discovered that not all of us but I mean some people discovered that well it's navbased so net asset value based uh and no one there was no third party to

really uh verify how much money they held versus what's the backing of the stable coin. So the question is um you're on to you like with the what kind of tools what what's the minimum threshold that actually should defy enforce at least in somehow on protocols to actually have at least a minimum you know uh um transparency. You know, I I think you know, nervous oracles are not necessarily pure evil. You know, even for sectis and, you know, other assets, they're even the kind of the best practice ones. Specifically with stream finance and elixir, you know, maybe a year from now after there might be some partial recovery, we will

know if in retrospect if it would have been better to have, you know, like a n based or other. I think you know like it's it's mostly concerning and it's mostly challenging when you use it in a permissionless framework such as moro maybe oiler I don't know rather than frameworks like a where in a you know you do have kind of the liberty to change the oracle if you want while in ma maybe technically it is possible but already require some coordination and you know like around curators to begin with which is not very realistic so yeah I'm Uh I'm not sure you know sometimes it will be better

sometimes it will be worse especially for low liquidity stable coins if you use market based pricing you know we saw this incident with Moonwell on some st that use market based pricing and it was priced like 3,000x you know higher than the real price so yeah like if it if it is possible and this is where maybe Dow can perform better and you have people who can take decisions at emergency and change it maybe it's better but otherwise is, you know, it's part of the risk. >> Yeah. And I think like, you know, oracles are just hard in general, even if they're, you know, cryptonnative assets, like even something like Bitcoin, right? To have an oracle for that. Maybe that's the easiest one, right? But it's still it's still an endeavor to kind of implement that correctly. And for some of these um RWA

based tokens, um even if they're treasuries, right, you're dealing with stuff that's offchain. Uh maybe you're dealing with, you know, a variety of of durations. And I think there's a lot of room here for exploration. Um, you could imagine maybe having something like third party addestations being published on chain. Um, all sorts of things like that that I think could really bring the the risk of, you know, some funny business uh down a bit. But I I think the design space is actually pretty pretty large here.

>> Yeah. Um when I have a question. So I think DeFi has evolved from a place where everything has to be decentralized, permissionless to accepting some level of centralization and some less transparency. I think uh we've gotten a little bit lazy and now the recent uh it seems that we've accepted too much of compromise in DeFi. Um what's uh for like for example the current uh some of the current assets with that have massive operations offchain uh how do we how do users and protocols uh can can how how can they navigate risk of of such assets?

>> Yeah, it's it's a major challenge um and it's something that we uh we sort of struggle with at Spark. So we previously we've had some exposure to Athena not any longer um but we also work with Maple. We're looking into other sort of like hybrid onchain offchain products and and services. Um I think there's a huge opportunity for like real-time attestations um and making them provable. So like ZK proof uh TE um a lot of these technologies can definitely help. Um but I also think that part of the reason that we have so much of this off-chain uh assets coming in is because DeFi isn't really feature complete yet.

Um there's people who want to do fixed rate lending as an example. It's a huge um priority for institutions and they're willing to pay a significant premium to borrow fixed rate versus variable rate. Um so you know in the current condition where fixed rate lending isn't really uh available in DeFi or onchain. Um of course there's this huge incentive to you know work with with Maple or maybe do it yourself and and lend through custodians. Um I think Morpho V2 and and kind of new infrastructure is going to solve that. And then similarly with perpetuals, we're seeing like an explosion of um you know sort of increasingly sophisticated per protocols. So uh in the future you could theoretically see a world where all these delta neutral strategies are

happening in a much more secure transparent environment. And I think that's um you know in the interim we're probably needing to do more of this sort of verification but uh I think we'll move more of the the actual activity on chain in a kind of a safer environment. Remy, would you like to add uh to uh to that? >> Uh yeah, I think like RWA coming on chain is something that uh is just starting and uh it's going to be it's going to be huge. Back to the valuation question. Uh I think for tokenized hedge fund uh so general purpose vaults uh as opposed to uh lending vaults um the NAV so the the valuation is

usually handled by the asset manager himself um which can be problematic but also when you're hedge fund manager uh you look for specialized third party provider that compute nav in defi there's not a uh of options and uh and service provider who are experienced and who are doing it uh right. So I think it's very uh we're just starting and maybe the space is going to professional professionalize. So uh this tokenized fund can have more uh independent third party uh life provider that can give a a real value and ultimately that would be

uh better for the users. Okay. So, there's no uh risk-free yield in in DeFi. That's it's a it's a myth. >> Tivos. Yeah. I think like probably D5 is under a lot of pressure. If you really want to make like four or five% a lot of users just won't use D5 just go use stratfi and just you eliminate at least the smart contract risk and it's a really quite quite challenging. So everyone in D5 just wants more and more and more. Um so now what's how do we build really just more transparent and more resilient DeFi? Like what do we need? Do we need just the tools? Do we need to centralize uh risk uh management with curators? Um

what the what what do we have what do we need to do? Yeah, I think transparency and explaining what for instance creator are doing, what is that montal model for each product is quite important. For instance, history obviously what we do for vault is fully on chain. We have some dashboard but you can always say well it's difficult to understand. We didn't uh publish a lot at the beginning mainly for regulatory reasons because when you publishing anything you are taking a risk I don't write regulation but that's how it works. So now we have published to our risk framework. So people can understand how we are thinking about risk what we are analyzing what we are adding on new collateral we will publish a bit more thanks to the fact that now there is a

more friendly regulation in the US but I think it's mainly on disclosing disclosing disclosing and make it simple for to for people to understand for instance it will be and although having some counter parties that you can trust I mean trust is still quite important I remember when there was USDC that was on boarded at Meador. Everyone was like, "We cannot trust Circle. It's offchain. Come on guys. Now it's regulated by Genius. It's a big public company. Maybe we can start to trust them." And it's the same people that didn't trust Circuit 5 years ago are probably the one willing to give money on the internet to stream finance or YUSD people that are

who knows who they are and they have no expertise, no regulation, nothing. So for instance we added a fasinara MF1 as collateral and people are saying oh it's offchain it's complicated there is no transparency I mean yes that's correct you have less transparency it's more difficult to have the prospectus because you need to be accredited because there is a lot of regulation making sure you never get the right information but it's a regulated entity that is managing five billion that is doing this business for 15 years. How could it be more risky this one that random people on a website that you don't know who they are? So it's very simple at the end of the day.

There is ins of what you can trust. If Blackhawk is launching a product, no one will read the prospectus. No one will ask any question. If you see Blackhawk, you will invest. But Blackhawk has risky product as well. So you just need to find proxies of knowing when it's safe and when it's not safe. Lindy effect is something it's not 100% but it's still something to to see. uh stack house obviously at the beginning you could have thought well who are those people we are now here for two years on moru you might say well maybe we know one thing or two and we try to be more and more transparent and know we are operating stuff same for everyone but if someone new comes up and you are

giving them 1 million and you don't know who they are I mean at some point just it's common sense at the end of the day what would you like to add >> um yeah I Think one thing that maybe people uh are considering more in light of the balancer hack is although you know we're we're defi native we're not comfortable with these off-chain uh assets where maybe we have a bit less transparency we have to trust a counterparty um you still face some pretty pretty large um sort of 0 to1 risks operating on chain even for protocols that have been around for a while. So, uh, you just need to kind of weigh like, uh, the security risk, the smart contract security risk

versus the counterparty risk. And there's not, uh, it's not like necessarily always better to be fully on chain. Sometimes it's okay to trust a counterparty if they've they've proven themselves and they're existing within a a rigorous sort of like legal and regulatory framework. And, uh, what kind of tools do you think we uh, we need to provide? So, actually, let me let me back back up. um a minute and we've talked about the responsibility but we haven't really talked about the responsibility of users themselves like uh it's information yes is difficult to to to kind of collect and really sometimes understand but this is an unregulated space it's extremely convoluted information is is available

everywhere but then it's still user's responsibility to do it so why can't defy come up with like say for example like maybe a a some sort of like organization that could provide insights about this new projects coming up and kind of like flagging these and and even like you know uh giving limited access. I think Euron uh has >> done something similar to that. >> Yeah. So it might not be popular to say it in this venue but to some extent AI you know it's like a huge buzz word that everyone likes to throw and yeah definitely it's not perfect but you know one thing that this curator sourcing doesn't do it doesn't give you any personalized risk you know kind of advices or managements

and we say ah you know definitely it's not perfect but for some users it's probably better for than nothing you know and it can help you maybe convey the risk more and maybe understand the risk more again like from not sure if it's like a legal solution and you know from regulation point of view etc but I think at this point it can definitely outperform many of the kind of existing defa users >> um so I think first there is no uh risk-free at all there is always some risk when there is some uh some yield and one of the problem that user face is uh there is no normalization of the way

we display uh yield or IPRs in the space. So some projects or some vaults uh some fun display projected APR which doesn't make any sense. Um some of them display uh the instance uh APR which doesn't make any sense also and some of them display 7-day moving 30 days uh one year and since inception so at least what we could do as an industry is find some kind of normalization around how to uh show historical uh yield uh net net yield. Uh there is uh usually shares and price per share. So maybe uh someone could find a way to uh to show all

vaults on all products and just show uh how they compare to each other in a normalized way. So I think some uh volt aggregator already kind of are trying to do uh to do that. So it's a good first step but I think as an industry it would be great if we could like normalize the way at least the way we show APR and about the the risk montal model uh it's bit more difficult uh so I know there are some risk notation agency that are coming to defy so I don't like this too too much because it remains some bad memories about the supreme crisis where everybody would give very good uh uh ratings uh

there also there is can be conflict of interest and this uh notation agency can charge the curator and because they are the one paying them then conflict of interest so I'm not I don't like too much uh the notation agency personally but I think we could normalize IPR to to begin with >> okay now to conclude this uh one last thing I'm going to ask each of you to provide one sentence your vision of safer defi Uh I would say it's better product, less concentration risk. Let's just try to do the add safe stuff and not go chasing the highest APY even if everyone is asking you to put the crisis collateral

to have a better IPY. >> I would love to see much more clarity on like the risk surface of each product. So, as a user, you should be able to see what particular asset drops for XYZ assets that you're exposed to in this product are going to cause you to begin losing money. Um, it's not something that exists today, but I think uh that'll allow users not just to look at like a letter grade or like a rating, but they can actually express their own risk preferences and and what things they're comfortable or not comfortable with. >> My sentence or just two words would be more transparency. I think that when users are depositing into a market vault, whatever it may be, um kind of better understanding what counterparty

risks they're taking, um as Monet just said, what qualitative risks and volatility risks they're taking, I think that's super important. And I think, um as the front ends currently are, I think it's a bit difficult for people to get that on on first glance. The the data is usually there, the information is usually there, but it requires a little bit more more digging on their part. >> Yeah. So transparency for sure but one thing that maybe also missing you know like in the kind of standard security audits etc maybe also feature in complexity and not only you know transparency and uh risk.

So I think to to finish with a positive note that um we just we talked a lot about the risk about the dark side of DeFi about people taking loss but there is still some incred incredible opportunity for risk adjusted return in D5 that you don't find anywhere else and if you want to have access to those yield come reach out. >> I think we have to go back to the days where we uh when when we used to say verify do not trust. I think that's the way I think we've trusted uh too much. All right, that was it. Thank you so much everyone. Give it to our uh Thank you very much. Thanks.