The Evolution of On-chain Asset Management

Transcript

Okay, great. Well, it's great to be here. Thanks everyone for joining us uh this early. Uh hi Juan. Great to meet you. Before we begin, I'm just going to uh ask Juan to introduce himself uh and tell us a little bit uh about himself. Thank you very much, Muna. Uh so my name is Juan and um the one of the co-founders of Krock. Kro is an algorithmic trading firm. Been around in the space for about eight years. We do all kind of uh financial services using proprietary algorithms.

Great. Uh and for those of you who don't know me, I'm Avanguard. Uh we are um a vault strategist and we are a curator in the DeFi space and I'm very excited to be moderating Juan today. When I first met Juan, he told me that he really believed the future was onchain asset management and uh I knew straight away we were going to be friends after that. Uh so I'm but I'm curious you know sitting in a kind of market making role in CFI I'm really curious to know um what you thought Juan was the first credible form of uh onchain asset management and what led you to take that view.

Um so when we when we started looking into into asset management and chain it did seem like all of the characteristics on how to attract capital were there and that at the same time all of the traditional asset managers were just trying to replicate the format in which traditional you know hedge funds have been successful but they were not actually managing to be successful. So I went on this this this journey on on really meeting a lot of managers and interviewing them and I started to realize that they actually didn't have a lot of money under management which was which was actually very interesting realization where at the same time very inexperienced teams were managing to raise uh a lot of money on chain along a lot has changed since I started the

research um I've I've recently started to update uh my understanding on the TVL on the different vault strategies and it's actually going very past like a lot of the teams that I would say six months ago I I talked with that were starting to build the first you know I would say onchain vaults for asset management are now starting to reach 50 to 100 million AUM uh without much track record to be honest which is very impressive and uh why do you think onchain asset management matters why is it important I mean it's all about access and and and the and I think it kind of like depends on on the idea. So when you look at all

the onchain protocols out there, I would say you in terms of like their their set of beliefs, you actually have like two kind of like types of DeFi. You would have uh everybody that is looking at it as a high volume, low margin business where it's all about like how do you make something very replicable and then take a small piece of that and abstract value from the system. And then you have people that are more in the camp of long-term wealth creation. So if you think about a business like pump pumpf fund, it's a great business, but it's a business that is pretty much conscious that majority of their users are going to lose money, which is absolutely fine.

And even if you go back to like before crypto in the in the space of FX brokerage, majority of the very successful FX brokers were very conscious that majority of the users would lose all their money and they were absolutely fine with that. uh that you have a movement now a lot of DeFi protocols and and builders that that really believe that it's all about long-term wealth creation and and the reality of it is that over the past seven years people have taken a lot of good decisions but a lot of bad decisions as well and I think right now that that combination of still long-term vision and believe belief into the industry but at the same time realization that you need to be better about risk management and and you know diversification of your assets

I think leads now to this idea that DeFi has to be more than financial infrastructure has to be long-term wealth creation and I think that's why this new trend is really is really starting to take off. Was there a pivotal moment for you when you saw maybe trust shifting and uh you know really seeing onchain asset management as a viable alternative to traditional asset management? Was there any one moment that you you saw the shift? E even though I'm not uh the the biggest fan, but I think looking at the the spectacular growth of of what it was Athena and and Morpho, I think we're were very big moments. I think looking at both the capacity for for uh Athena

to to all of a sudden basically have a tokenized fund product and and raise billions of dollars without pretty much any any real track record. uh uh really showcase the power of of you know broad market access that the onchain ecosystem would give. That's interesting. I would agree with that actually because before that you had uh multiple collateral and morpho really helped like fragment the market and uh segment risk let's say which really created nice building blocks for more sophisticated strategies let's say.

Yes. Yeah. Um and um what do you think it will take for the sector to really break into the mainstream? Do you think there are still some elements missing? Um I mean I think that the for it to really the the the assets that are or the type of exposure that is being created needs to be more diversified. So right now in the level of maturity from chain asset management where we are now is is basically two types. We have uh what are called say beta products. So, uh, something that lets you go long on the market with a slight sophistication.

There's not that many of that actually, like there are not that many places where you can in a comfortable way say, I'm going to go long on crypto and I'm going to put my money into this vault. People still just like, let's say, buy Bitcoin and that's it, you know? And then you would have yield. And so yield right now, it's it's basically the idea that you would build something that is principal protected. So you have principal protected or pseudo principal protected. and and that would mean that what people are actually chasing is is a pseudo bond. Then you have all the growth of of Pendle with bond stripping which comes from the concept that theoretically you believe that what you're buying is a bond correct as a principle plus a yield. uh for asset management to actually grow, we need all everything that is in between those two

which is all the spectrum of alpha and actually actually that's the the largest part of the market and and you need that because uh the reality of it is that majority of people do want to get exposure to the long-term success of this market. They just want to do it in a way where they are maybe not facing the same level of risk and volatility and so products that would fit into that category are still needed. And then from the asset assets itself uh we would need to be able to give exposure to more assets. Majority of the products right now are basically Bitcoin products or Ethereum products or Salana products. Um the the tokenization of equities and depth is now bringing that capacity to build more sophisticated asset

management products. I think that if like we could have on chain something where you have like I don't know a combination of equities and bitcoin I think people actually would be pretty attractive. So yeah I would agree with you that most of the products we see right now are yield orientated and that space is uh yeah starting to get a bit crowded. Uh have you seen anything interesting in the middle of the spectrum? You said it's lacking but have you seen any interest anyone doing anything interesting there? I mean there are definitely some some people building some different some different things. there are some some kind of like strategies more geared towards volatility arbitrage where people are using options markets and liquidity pools to do kind of like

mispricing of volatility which is which is an interesting concept and and I think you can you can do it on chain fully which is also al also interesting but then for everything else I think we're still at the stage where you would use the vaults and the tokenization more as a as an access point but to run anything more sophisticated you would still need to do that in Binance or you know key X or something like that. So at that point the the onchain vault is more like a like an interface wrapper to automate fund administration processes which is already great because they're actually very expensive and then from there you would you would still want to do majority of your executions in in a place where you have a large amount of liquidity and variety of assets and um as far as AI is concerned I don't

know if you've been following the trends there with regards to onchain asset management Um, do you see this playing out in a meaningful way? And are there any examples of AI agents starting to outperform, you know, kind of more traditional methods? We'd love to hear your input on that. I mean, I don't know if I'll perform and we'll see about that. But I I'm very active into in in within my my research team. We also look at and try to keep up to date with all of the new developments in in AI trading in general, even more than asset management. I think majority of the innovation in AI trading first comes from prop. I think we're still at the stage where people are not even thinking yet about asset management. They're just thinking how do we make it work like can can you make money with this? Um is it's

definitely possible like like there's no doubt uh even ourselves at at kio we we are allocating resources to allow AI engineers to experiment and try to test different strategies. when it comes down to onchain, it seems that at the moment the the only real tangible use case is concentrated liquidity pool provision. So you would have an agent that understands the restrictions on on a unis swap v3 style pool and then it would optimize for width of range, time in range and I would say kind of like amount of liquidity in range and uh actually because liquidity pool

provision overall compared to like traditional market making where you have an order book is relatively slow. Correct. Like you you only move a range maybe you know, every few hours at at best. Yeah. Um, it is actually very ideal for for a model that just learns and tries to figure it out and fine-tune how to open the range, where to open it, when to get out. Um, I I do think these models are are going to work. There are there are great companies now building them. So you know companies like like Giza or or theoric or you know see zero gravity as well providing infrastructure for these are building use cases for for that and I would say after that then you

will have more sophisticated use cases maybe even funding arbitrage as well run by by agents but not yet. That's really interesting. Um and um are you guys utilizing this in house yet or are you still experimenting or Yeah. So so we we are in contact with all these companies. We we're very close to all of them. Um and and we're trying to figure it out. Uh pilots wi with them to test the technology. Very often we have to to tell them to go slower. You know they're startups. They they want to go fast. They want to like get money from the public and you know go out.

Well, we at the moment our approach is we we tell them that we're happy to test all the AI technologies for trading but only with our our prop capital. We we don't think it's at the stage where we would want to put it in front of clients. And at the moment we're seeing a bit of a split in onchain uh asset management between sort of open vault-like strategies which are verifiable, onchain, transparent, etc., and more proprietary strategies which uh are maybe not open to the public and not as easy to find. How do you see that dynamic playing out over time in in in the theory of of uh trading in general, capacity is the biggest

constraint. Like even if you look at traditional finance trading, the capacities are much larger but they still exist. And not only you have capacity constraints but also you have diminishing returns. So as as your as you accumulate more assets in a strategy then your returns will go down. And so uh even though there is a lot of enthusiasm from a community perspective on openness and and sharing uh as soon as capacity starts to become an issue I think it would evolve towards uh information being closed because the only way that you can protect your capacity is by not allowing other people to understand what you're doing.

Otherwise the capacity would go down very fast because you just basically get copied really really really quickly. Uh we see that in a lot of a lot of the early vaults. This happened as well for example in Salana where a lot of the early vaults have been deployed were very successful in the first few months but it was very evident what they were doing like you could like really decipher it and all of a sudden three four five bolts of other managers would pop up with the exact same strategy and the whole thing like just gets drawn down to you know a yield level that it starts to be just very market average and kind of like just normalizes and so what you would what you want generally is to find a balance between capacity and yield. And so you want to protect your your you know your intellectual

property. Um so I I actually do think the majority of onchain asset management in the future would most probably be in some sort of kind of like you know privacy preserving format when it comes down to trading strategies when you I mean there's there's two types of onchain strategies right there's the there's the automated smart contract type where everything is on chain there's no proprietary element at all and there's the part where the execution happens onchain the the strategy is offchain but I think you're also talking about the Third thing where with privacy where even the execution is privacy is that private issue. So so I think if you if you have like your compute of chain and you just go on the chain you execute um that already helps a lot because it's much harder to

understand where it comes from. But very often the strategies are pretty simple actually. So like people can decipher them very quickly or they can just like follow your trades. Um, so you would need even even further ways to uh offcate your trades um or to or to split them. So, for example, you could have this strategy running on on chain and then you would have like a a third party vault, something like a like a hidden road or or a Fireblocks wallet that also lets you do some trades on some perpetual markets or on the options and then so you're so you you can see a bunch of unchained trades, but then you're holding and managing your risk by hedging it against some options and that you cannot see. So if somebody would just naked replicate what you're doing,

it wouldn't makes any sense for them. So you can like, you know, you can throw them out a bit. Very interesting. Um so um I think um we have time to open up for some questions if uh anyone would like to ask anything. Okay, great. Well, thank you very much Juan. Very exciting times and thanks for joining us. [Applause]