How To Think About Restaking Economics with Tarun Chitra

Transcript

how team should be valuing cryptoeconomic security can use the same Capital at a million Services I'm sure you've spoken to a lot of these lrts are they prepared to like welcome to Igan lay unlocked an interactive educational Journey Through the Igan lir ecosystem created by technical Founders and builders for the entire crypto Community a special thank you to our partners all lir polymer authentic skate and lrange for helping to make this happen our goal is to raise the collective knowledge of the igen lair ABS ecosystem and unpack the technical designs of the top teams in space welcome to ig layer unlocked welcome back to aan layer

unlocked today we have an episode with tun of Gauntlet to discuss a lot of the recent research that Gauntlet has been doing in the reaking space uh to run studied mathematics in physics at Cornell um so he's got a way bigger brain than us and has also been involved with robot Ventures um doing a lot of early stage investing congrats on the recent fundraise there um and yeah man welcome on great to be back well you know I know this is a special edition but yeah each time has been a special edition the modular March into the aen layer one of these days we'll just have

have a regular old normal Edition and uh and and have some fun I mean that should be your Halloween costume right it's like a special it's like it's like a special onetime thing like we're wearing a costume we have a normal podcast for one day yeah oh man well let's jump right into kind of the the core questions that Robbie and I had we've been doing these presentations now for several weeks really have opened our minds to a lot of the different uh strengths and weaknesses of I layers designed some of the Innovative AVS use cases um and so we were doing some research into gauntlets uh actual research itself and we found a piece that you recently put

out titled how much should you pay for reaking security which we thought was particularly interesting that it discussed kind of like how smart allocations to abss is very important um ABS is needing to provide certain amount of rewards in eth term to kind of help direct where these lrts allocate um what we're trying to understand is what is the correct mental model for valuing cryptoeconomic security and how should we be thinking about uh how team should be valuing cryptoeconomic security yeah so um it's a great question maybe I'll give a little bit of History first of like how we kind of got to that result

and then um you know from there we can kind of talk about what it means so in I think I want to say October 2023 or maybe September I was talking with Sham and I was talking about this idea that hey how do people think about ABS allocations how should they think about you know which ABS they're in how much uh stake they allocate to each one and it was kind of clear at that point that like one important thing in the white paper was there are some conditions on allocation that are based on security and so uh sort of ironically the source of a lot of the new research

on recking has come from appendix B of the Ian layer white paper which is only like two pages but it's actually the most detailed part about this idea that like how you choose avss impact security like how node operators decide which avss they choose and how they choose the sizes is actually important to the Global Security of the network and you can think of that as sort of a result it's say something that's local like me a node operator how I choose something uh converts to a global property like there's security across all the avss as long as everyone follows some particular set of rules and so those rules in general in crypto those rules

need to be incentivized right like they need to be there needs to be some reason for people to rationally follow them which means it sort of has to be profitable to do that and so they gave sort of an initial kind of condition on what maybe those rules look like and then over time I had actually done some simulations based on some some old papers I had written in 2019 and 2020 on how you should think about security when you have liquid staking tokens and and lsts um and I was trying to modify them to say something about reaking in the sense of like hey you know there's a sense in which you may want to allocate

to ethereum Stak avs1 avs2 av3 how do you determine that in sort of like some rational way um and in the numerical simulations I had there was sort of this interesting thing where like people who had static allocations so they picked the same AVS and they never adjusted had sort of worse outcomes for themselves and they also had worse outcomes for security and security here sort of means there's a slashing event that occurs that slashing event Burns some stake and by burning that stake it actually potentially if the the

validator was shared across multiple Services burn reduces the security of another Network so I think at the time last year and this is kind of how fast this stuff moves like last year everyone was worried about the sort of like pre-launch pre- anything being live everyone was worried about this idea that there's these like cascading attacks where like an AVS gets attacked and then it reduces ethereum security right so like an AVS getting attacked Burns a bunch of stake now e the amount staked in ethereum is lower so it reduces the security can that you know affect the L1 and another question you might ask is imagine I got slashed in one service can it affect the security of another service and so

that was a question that from this part of igen L appendix B in the white paper it became clear that actually you have to consider not just how much a service affects dl1 but also how it affects other services because a node operator could be M could be operating multiple services and if they're operating the same service and let's suppose they got dosed or they went offline or they got slashed in some way then they affect multiple services at the same time so there's some risk that there's correlation between them and so I sort of had these numerical results but I was like not you know I didn't think they amounted to like a full

theoretical thing and and that was around the time that Gauntlet we started doing a lot of reaking analysis and and monitoring the live networks um but in February of this year I was at an academic conference and I was talking to a couple people about reaking and they were like we have this paper we have this result uh about cascading attacks and so cascading attacks against reaking networks are something where one service you have an operator who's operating multiple Services they get slashed on one service it reduces the security of another service which now allows someone to attack that service which then might

make another service attackable dot dot dot right so it's it's it's sort of like in in in lending onchain lending there's this notion of liquidation Cascade or in perpetuals exchanges this is the kind of reaking analogy now the interesting thing about reaking versus lending is it doesn't really depend on an oracle price as much um and so it's a it's it has some extra security guarantees that are kind of different than pure lending or perpetuals and so that was sort of one thing you have to include in the model so these people who are Tim Ruff garden and navine derisa uh they they wrote this paper um and they sort of came up with some very hard constraints that

improve upon appendix B of the aular paper to show when you don't have these cascading effects so like if there is enough stake like at each service if there's sufficiently enough stake relative to how much profit you could get from attacking it then uh you don't have any of these attacks so the first sort of research that showed there is a a sort of a notion of security here in which you can kind of Define what it means to be economically secure now their initial paper was extremely pessimistic so like the results said you have to be very over collateral so this is sort of like what you see in decentralized stable coins like maker or Sky although now maybe they're going

back to being maker down but whatever uh where you have to be over collateralized in order to get some security anes and and it sort of is is is interesting because there's some El now what uh what we were thinking of in this paper of how much do you need to pay for reaking security is their paper assumed there's no incentives there they assume sort of people were operating Al purely altruistically or purely randomly like you had no you had no understanding of like why they were operating a service and so what we said is but in reality people are operating a service running a node operator because they're getting token rewards points future rewards some type of fees something where they're making right

like that's just the Baseline and if you assume that then maybe you can be less collateralized like maybe you can actually have a a safer network with less collateral and um what we showed was something that's like as long as there's not too much overlap so like say I have two avss as long as the sets of node operators and the amount of stake they have don't overlap too much then you can give them AVS Rewards or fees at a certain rate such that these cascading attacks don't occur so it says like with incentives you can actually lower the collateralization a lot and so what that meant is it tells you something about how AVS should think about designing

their incentives based on how much their node operators overlap with other services so there's kind of this tradeoff right there's I get a lot of capital efficiency as a node operator if I can use the same Capital at a million services but then I create a ton of risk and so you know the the way to have no risk is a node operator only validates one service but that's not efficient right and so there's kind of this in between and and what our paper shows is like there is how how do you compute what that in between is that's safe um it's still somewhat pessimistic so like I'm working on another paper that's like giving a a better B but hopefully that gives you I know it's like a long story

but that hopefully gives you yeah I it does and I think it also says something about like how you know you start pretty conservative but then you can kind of hone in on like more accurate and hopefully more optimistic results I have a question about um the way abss ought to think about how much apy or fees or Rewards or incentives they should give to the stakers and those who are committing Economic Security is this something that AVS is should coordinate with other avss on and we should have a globally coordinated system for valuing this or is this methodology something that a particular single ABS can use to Value Economic Security for its own stake yeah that's a good question

so by the way in in response partially to these papers partially to other things partially to stuff that AVS was wanted igler created a new security model where instead of it just being purely the node operators choosing stake the uh Services could demand something of the following which is like suppose I wanted to put a 100 eth into one AVS or service sorry sometime I I like calling them just service because it's just like simpler in my head than ABS they are same thing uh and and uh sometimes you you know you want to put 100 e into a service the service has no

the original model the service didn't have any ability to guarantee that they could slash you so like what I mean by that is suppose you put the 100 e and you're validating service a and then you're also validating service B with the same uh thing now suppose service a wants to slash you because you did something and the other operators agree on it but now suppose service B does sort of an meev thing where it slashes you first now service a can't slash anything so it can't it can't cause any penalty so like you you don't get any penalty for doing the bad action and so there's sort of this idea that uh service a has no guarantee on how much it can slash right but the new model for a security which has this

concept of unique stake it says that hey if I give 100 intuitively like the math and the code don't exactly do this but I think this is a broad a simpler way of of intuitively understanding is like a service is a contract and the cont say if you delegate 100 e to me 10% of that eth is reserved and only I can slash it which means that I can always slash at least 10 as a service as and so what that means is that it allows you to kind of have some amount of isolation and and we've seen this in lending and perpetuals markets where inevitably like the pooled model works really well for efficiency but then is hard riskwise and then you you you kind of have a hybrid model between fully

pooled and isolated where like only one asset can be borrowed or L right and and you can kind of view this unique sake thing as hey I am contributing my Capital to an AVS to earn yield to to operate it to make it work uh and in exchange the AVS gets to guarantee that can slash at least some fraction of my St and pays me for using right so this unique stake thing is actually uh quite important because it actually means that it lowers the cost quite a bit for the service so sorry I this is a longwinded answer to to your question but it's actually quite important because it actually it's a parameter the service gets to choose so the service can kind of think of like

hey I'm a really I I really need to be able to in the worst case slash completely the user then I say I have a really high unique s like maybe I'm a rollup sequencer where the damage that can be done by an invalid sequence is arbitrarily large on the other hand you you could consider an AVS this maybe more like an oracle where it's like okay well if I only move the price by 1% then the damage I can cause is 1% times the current e price or something like like you have some way of valuing it and for those you might say ah I would rather be more Capital efficient and have more shared state right and so I think because of this and this is inevitable right each AVS will have its own demands

for security versus not security um or versus like less security rather not not security uh and because of this they will have their own parameters that to adjust that are their own which are like how much stake unique stake do I get now the more unique stake I choose it's not free it means that I'm gonna have to sort of incentivize people more such that if if if I'm saying like your eth is only earning from me and no other service some fraction of it right so there's like how much do I want to pay for that is is guaranteed to be service to me and how much am I willing to pay for that's like potentially shared amongst any Services right and so I think the answer to your question is

like they have to think about both they have to think about how much they need at the bare minimum and how much they're willing to have shared and and I think that's I think the inevitable conclusion of of blockchain systems is that you have this hybrid of of shared and not shared and then and based on your use case you adjust it and that like the best of all worlds right in the long ways yeah we recently had um Mike from ethery on a podcast as well as uh a couple others in the LRT space and we were kind of going back and forth about this topic in particular um this kind of allocation to avss and kind of flipping it more towards lrts rather than how

avss should be thinking about it because I think that that's important right that that that there's specific use cases that can uh allocate a a a wider percentage of this total stake to be Slash versus kind of not you know there's super there's there's very financialized use cases as you mentioned like a sequencer something that is crucial for a large amount of assets and then you know perhaps we have something a little bit less important that can kind of uh narrow down that amount and and you know that customizability makes it makes it particularly flexible for these services and then kind of flipping it to the lrts what's interesting what's going on with lrts right now and The Operators is that one of these two

parties has to be proficient at some sort of financial modeling to be able to allocate and adjust risk forecasting and these types of of things in a traditionally like engineering heavy kind of role like operators are like similar to like a validator who runs and lrts are probably more financially savvy so kind of coming to coming to this this state of LRT allocating their stake across abss um how how can lrts use proper risk modeling to prevent or minimize any sort of systemic risk and kind of like how how are you thinking about the way that these lrts should

allocate stake across avss yeah great question so you know the papers between rough Gardens paper and my paper they kind of get at this idea that like there is some in a very over collateralized version you have no systemic risk so we know we know like when systemic risk exists and when it doesn't exist but then of course you want to find the efficient Frontier right like you want to be at like no systemic risk but also efficient Capital like high returns effectively right and that's the that's the goldilock state I like to think about reaking it as sort of in between pure staking and operation and defi right it

like has some aspects of defi but it has a lot of guarantees that defi doesn't have right defi has to rely on oracles it has to rely on liquidity in a certain way it has to rely on a bunch of other things I think reaking dependence on those same things is generally a lot lower and part of the reason for that is you're using the same unit of account at every service like you're reaking East every now obviously you can restake multiple tokens and whatever but realistically the largest portion is going to be still so there there's sort of a sense in which you can you can kind of you have it's like it's like lending or perpetuals but like you you are bounding the risk a little bit and that's why I think you see this you know

duality of man thing of like the finance side versus the pure Tech side right like it it is actually like in the middle and that's personally why I find it the most interesting but I like to think of you know if if we do some trafi analogies first before we get back to the lrts uh I like the think of you know the normal token Market is like closer to equities and like lending protocols like a morpo compound those are like stock loans like I borrow against my stock I get some cash maybe I get liquidated because of stock drops right those are very volatile loans that they generally have low predictability

because there's not really like Revenue right like I'm borrowing against a stock but if a stock doesn't have dividends then there's no real Revenue so I'm not borrowing against like future income I'm borrowing against exactly what's there so I have to be conservative on the other hand most loans in the normal world are borrowing against future income right like a mortgage is borrowing against my future income to pay for this house now and then I pay it over time right uh and most most Credit in the world is is not like that now stakers have the in most interesting form of credit they have a sort of cash flow coming in they have the blocker words that are somewhat predictable now they're not that predictable in the

sense of like hey someone could just add a bunch of e to the pool dilute my return and now I have fewer block rewards they could they could add you know the fees might go down because the l2s are taking all the fees or what whatever whatever the reason is um so they they still have variant but they're much more predictable right like they and and and that's why you should really think of you know basic staking as like the bond market like there's some you know the default conditions are getting slashed but it's like generally a lot more predictable and and less you you so you can be less conservative on lending that's the the other side of this is like PE the biggest lending Market in

the world is the repo market for treasuries uh which is like tens of trillions of dollars uh you know that exists daily so it's like it's like huge so there's sort of this natural analogy then that like reaking and I know people don't like thinking about the rehypothecation aspect because like you have a lot more safeguards on chain like the L1 contract and slash you immediately it's not like rehypothecation in the real world where like if someone [ __ ] up I have to go sue you right it's like it is different in that way but there is a sense in which the payoff to the node operator looks a lot like getting leverage on my stake but in a predictable way without

going to a or without going to something where I'm taking a lot more relative risk because I I don't know what the borrower quality is when I think about the AVS as a borrower their quality is much cleaner to assess it's much more about the kind of credit rating of the ABS what the historical fees look like what their token looks like all of these things right so because of that it's like really like the bond market except you a lot more protection because you do all these things that you don't need a legal system right that's the beauty of crypto you can have all these conditions for getting slashed and rewarded so I think if you take that lens that AVS is are like the bond market there's sort of a sense in which ABS need to

like generate predictable revenue and I think one of the reasons the rollup restake rollups and the decentralized sequencers are at least from a naive standpoint Point quite attractive as AVS is is we know they have regular Revenue that you can then use to kind of decide how to think about this for and so my suspicion is that the ABS Market will unfurl in that it will start very bimodal well there'll be like this huge set of companies that is or avss companies that are building ABS that is very predictable Revenue like things people know have Revenue orles reings whatever and then there's

going to be this like whole like pure VC thing of like AI models and and whatever and then over time people will learn a lot about them as they're live and they'll learn about the reward profiles they'll learn about the slashing and you'll slowly get more of your normal distribution gaan thing of like things that generate a lot of Revenue things that generate very spiky Revenue maybe might die and you get kind of the whole bond market and because of that you need to think of the risk reward of the market in that way I think a little more than than like pure tokens where it's like more like Equity where you really you really have to be more conservative and I actually think the interesting thing is you can be more aggressive on the efficiency

side in reaking than you can in lending and perpetuals and that's that's kind of why I it captivates me but also is like why I'm spend so much time trying to write the math for this is like showing that that's true is is still it's it's like an you you intuitively feel it but you don't have the exact kind of math yet and like I think we're like you know to your point earlier of like you start really conservative and you get more and more efficient it feels like there's more steps you can take in reaking efficiency wise and you can in Def that I mean that that's a great parallel to the bond market um and so I mean like I'm sure you've spoken to a lot of these lrts are they prepared to like like do they have the forgot I didn't even

really answer your question about the lrts who's I I think the thing is like I think the lrts are like bond funds or honestly and like people will hate that say this but they're like they look a little bit like cdos from like the financial crisis that like bundle a bunch of mortgages and securitize them I it's it's very similar I will say the naming of the thing LRT squared is almost two on the nose because there's a thing called the CDO squared from the financial crisis that blew up very poorly that like it reminds me of but maybe it's because I'm old and everyone else is 10 years younger than me or whatever but uh I uh I I think there's kind of this like this kind of funny thing that it's like yeah the LRT is

gonna have to be sophisticated and like everyone knows all the funds that blew up in the financial crisis but no one talks about all the funds that made tons of money and like figured out how to manage the risk afterwards and like it's not like the CDs Market went to zero it's still like 10 trillion do market so like my point is like you know you have these collapses and crashes but from those you're able to learn how to make it very sustainable and I think lrts are going to have the same Reckoning where someone is going to blow up they're going to have chosen only the riskiest things and and we saw this in defi summer where after Ur was successful their five million like yield aggregator clones and the only thing they

differentiated on was like how far on the risk curve they're going to go like what new protocol were going to integrate it also just blew up yeah cream cream Finance is one that comes to mind on the lending side their entire Edge was like you could just borrow against anything Al destroyed them over that yeah like ironically yeah uh yeah yeah so like my point is like there's inevitably going to be the ones who blow right like but you can't control that and I think the beauty of the decentralized world and transparent world is you can see all their positions in trafi the CDO Market CDs Market you can't really see those positions so like one reason I think the financial crisis was actually worse in

some ways and this is my opinion there's some data suggesting this but it's I would say it's mixed um but it's just my so I'm dis disclaiming that is when you think about a Lo a mortgage that defaults it's not like and imagine there's a bank that securitized a bunch of mortgages and then sold it to investors so like investors are like own this portfolio of like five mortgages now suppose one of those mortgages someone doesn't pay for a few months you don't have like it's not like the security like the thing the investor owns immediately prices the default like

the the thing defaults then it it like is in foreclosure and then maybe it goes to court and then eventually after it's actually you know somehow the tile changes then the security suddenly gets repriced right it's like it's whereas in crypto it's like immediate right like if there's a default event atomicity and composability mean it's like already at the end like every possible person who depends on that state transition has already gotten it right that's a very different world than the world that we see in the tradire world and that's also a reason why you can make these things safer because you can put all these safeguards when you know they're Atomic and composable and you're not like waiting 5 years for some [ __ ] random

government municipal government to tell you this house is in foreclosure yeah it's a much more Dynamic world right and so when these lrts are thinking about reallocating and I think there's allocating when everything is fun and good and like very normal equilibrium circumstances and then there's reallocating in the case of some some weird thing happening and so in that case everything moves much more quickly because they're all everything is programmatic and so what kind of programmatic controls do you anticipate these lrts or even like you know these guys that are just R running igen pods

with their own eth and not in in lrts what kind of programmatic controls could they have or should they have to reallocate risk in in sort of a Black Swan event yeah that's a good question so I think a couple things are one being being cognizant of the overlap between operators of the same service like will they all get slashed at the same time will it be correlated in the same way that you had this kind of like correlation event the financial crisis like I sorry I keep referencing the financial crisis but like it's just a very easy analogy to make of like the thumbnail is going to be to rued reaking financial crisis no that's that's not what I'm trying to say just like well

the funniest thing is I wrote so a long slightly long story but why I got into thinking a lot about reaking before it was live like kind of when you know I mean I I invest was one of the First Investors but I that was more because because like I think sham had a really good idea but I think the way it was presented in the beginning looked a little bit different than reaking and when it kind of once it was kind of more like reaking the the number one thing that I've been thinking about is I wrote this paper in 2019 and 2020 with Alex Evans from Bane about uh how to model LST risk and the interesting thing is we we made this

model and showed these properties which basically showed that like an LST is sort of in a lot of regimes very similar to a mortgage back security it like has a lot of properties of mortgage back secur you can think of the mortgages as each node operator's non-fungible position of locked eth in in this case and the LSC is this basket of them and like if if all of them default at the same time they'll get slash at the same time your LC will be could be worth less than what you think um and so I had sort of been like oh this reaking thing has to be kind of credit default swap bond market like because it like if the lsts were like that then like this has to have some flavor and it doesn't have to

be exactly the same right like again the main important thing is that in a programmable money World these covenants in alone like oh if this condition happens you're defaulted or if this condition happens we pay you in the real world those don't happen like instantly at the right time they happen kind of Slowly by osmosis and between 9 to five or whatever right but the idea that I those happen immediately means that now I can put in conditions in the contract that are like if these two events happen at the same time do X so if I get slashed in a and slash in B then do why and this is the type of thing an LRT can do in response in a way that you kind of can't

do in a in the normal world it's like the two event the correlation is truly there like but you don't see it because it's not transparent and it's not obvious to you and I think blockchains allow you to actually have that be extremely transparent and so you can now put all these extra conditions or like if these two things happen reallocate move my e do whatever right and I think that our initial paper kind of was built around showing something like that uh and and the title of that paper was like what proof of State can learn from mortgage back Securities which is kind of similar to the thinger said but like I don't want to like raise the alarm Bells I want to make more be like hey those SEC things are like maligned in my opinion I think like credit default Ops

it's still a big [ __ ] Market it's not like it went to zero right uh and it's because it's useful and people get value out of it it's like a better form of insurance and if you look at a lot of the stuff in reaking like stake sure a lot of the insurance sounding things within sing they have a lot of that in it except because it's programmable you can do a lot better and like we should be thinking about how you can do all that stuff better and like that's to me that's sort of that's sort of how I I view this is like I can write these I can write code this is if these two events happen together like get the [ __ ] [ __ ] out of this other AVS right like I can't really do that in trafi right now

it's like it's like too many disperate systems they might not tell me the state of everything immediately it's just like you don't get this this atomicity stuff is like extremely important to risk management which I think tragi people don't have they hope for the legal system to make up for lack of atomicity I mean for for all those advantages would you say that it's just a matter of time before this whole reverse repo Market 10 trillion doll credit bond market is onchain I would love I would love to see that I mean I I actually think like the credit default SW Market itself should probably like onchain makes way more sense than this bilateral party [ __ ] that we have um it's also a market that I bet you would

be 10x the size if it was Global so it's a market that's extremely local uh mortgages and credit default Ops and bonds are very tied to whatever country they're issued in and we live in a world of the internet bonds should not be restricted to one country or one place liquidity shouldn't be restricted and like all crypto has shown you is that you give people the advantage to to not have to think about that you can get a huge amount of growth and I generally think like that that's going to be true for this stuff too and that that's why reing is exciting because like if there are real services that are not rollups that generate a ton of

Revenue I that that's like an international company that anyone in the world can buy the bonds of that that's kind of what you're you know like you can't do that very easily I mean the bond market is even harder to trade than the equity market right so like Access wise yeah I think that that's that goes back to your previous point and I wanted to kind of double click on that as far as the revenue of rollups I had a really good presentation with Emirate alt layer and that was the thing that particularly like stuck out to me we we we've asked a ton of these different AVS projects or services like basically where's the demand going to come from where's the yield going to come from who's going to kind of uh you know make this work and um the the decentralized sequencing

answer was was definitely one of the best um as far as a clear product Market fit in today's world but then kind of to flip that on its head if these um AVS or if these Services all launch tokens which to me is still kind of fuzzy because my initial unique value proposition of igen layer and reaking is like you can bootstrap your own service or your own network without having to kind of launch a token and launch this um you know proof of stake token and go go down that kind of path which is it's clearly not going to be the case because these these These abss are funded by VCS there's uh liquidity tokens are the way so this kind of brings me to like a

two-part question which is one in the immediate term do are is the revenue really going to matter if there's token incentives to kind of uh and token emissions that can be basically forecasted as revenue and two how are we thinking about the value of ABS tokens and service based tokens compared to like a L2 um like like yeah like what's your mental model there for the these tokens yeah they're still yet to be launched you know I apologize for always thinking about everything in terms of like Basic Finance but the math always makes more sense to me and I would say the interesting thing is like in normal

companies you have derivative instruments like safes for raising money that are effectively bonds that deliver a interest rate in equity right so it's like a say this derivative and I pay I get a discount and a valuation and based on the next valuation get a you know I I get some guarantees but my interest rate and and arguably my principle depending on how you look at it is effectively paid in equity I think the interesting thing about tokens is people flip that they like they they have like a an equity product that is like converting itself to a bond over time right it's like I'm paying you these blocker words until I get fees and then it becomes a

regular thing right so think about ethereum it's like I'm paying you an equity in ethereum Block rewards to participant the network with the belief that there's future fees and those fees will be regular predictable dot dot dot like some some kind of understanding of them and then if that's true then I become a bond right which is like if you think about eth it like had this reverse transition of what a safe like a kind of a company would raise early money on these convertible note type of things it's sort of the had that as well recently if if you see the chart they did full emissions for like the first two years or three years that was like the main source of Revenue and now it's those have stopped and it's pretty much

fees and I think I think I mean like capitalism is capitalism like you don't [ __ ] make fees you're not surviving a long time like I don't I don't care what mecoin [ __ ] tokenomics thing you do in the long run like in the short run you can do stuff but I think eventually you have to be able to convert Equity to bond and like that is kind of an interesting thing about crypto because if you have programmatic money it's much easier to adj just how you emit your Equity as a function of your fees right like EIP 1559 is a great example of that and because of that programmatic thing where I can control that and I can do it continuously as opposed to these kind of like in the normal stock market I have to like I can't like just like oh today

I'm I'm gonna emit 1.375 shares of my company and sell you know it's like that's not happen it's like I have to do a tender offer and and and all this other stuff right like but the idea that I can prog prog these this thing until I reach a point of sustainability until I reach a point of fees and I can do experiments with how this kind of like Equity like thing eventually converts to a bond like thing this is why you should never view tokens as pure Securities because like they kind of have this amorphous chameleon nature where they they go between these things and I think reaking will be the same thing uh I just think you know having been in crypto for a while sometimes you just don't see where that

comes from right like you don't see where the new asset creation or fee creation if I split them into two comes from right like defi summer was sort of a weird one where like asset creation was correlated with fee creation because like every new asset created more trading or lending volume which created more fees and you had this feedback loop which is I think what really grew it then whereas like l1s you don't quite have that l2s you definitely don't have that um but I think reaking is actually something that looks a lot like Defi and a lot of the reason I think reaking you know you know I don't have too much to add to this our l2's parasitic to l1's debate but I will say one thing I think

reaking improves the monetary premium and the moneyness of the staking asset because you if I look at the rest of the world the countries that have the most liquidity for finding new yield opportunities have generally the largest Bond markets right so it's like the US I can do tons of borrowing against my bonds and reinvesting and other things whereas I look at countries that like can't do that like turkey where it's like oh the bond pays me 50% in the local currency but people still don't want it you know it's like it's like clearly there's like something going going wrong right like and and sort sort of there is this

natural thing in which how much liquidity you're able to attract kind of determines your your success and so if you're a proof of stake Network and you want to cement yourself as money the ability for people to borrow in different ways against their stake and use it for new asset creation that inevitably becomes a way of getting more and more liquidity and you become more and more like money so I think for proof of stake assets reaking actually it almost ironically because given how much bitcoiners would hate this but it makes it more like money some like gives it more of the money in it's property and I think that is one of these Financial things that like I've

been trying to figure out how to write the math for and like get a more real like result for but I I I it's just intuitively that sticks out to me is like rollups don't necessarily accrate value you know like I'm sure there's me and other stuff that that that does and there is of course the like they're use if they use the same base asset that there is some some premium but I think it's much smaller than the reaking premium and I think in the long term the AVS is that figure out how to correctly turn this we start with incentives as equity and then convert to like hey we earn a la fees that's going to win in the long run and whatever we see right

now will likely not be the one that is like the best with them and and that that's the be that's the joy of crypto right it's like you're always like finding something you're crazily surprised that's able to generate all that Revenue yeah yeah I think that's that's a a really good way to put it right so right now I think we'll start to see like we'll start to see a lot of these avss eventually and and the services launch their own tokens um they will probably be a wide cast net from lrts based on the different risk parameters that we spoke about today and then kind of con consolidating back into that normal distribution curve that you kind of spoke on in the future as the ones that actually generate fees um you know kind of yeah stick earn staying

power and then this is all bullish for eth eth's money eth the asset uh is a big a big beneficiary of this I just have one last question and I don't know if Rob does I know that they'll running low on time um there's a notion of we should be able to restake anything uh we should restake Bitcoin we should restake tether we should restake near we we're going to restake aan and slashing we haven't even talked about slashing today we could do a whole episode of slashing um what are your what are your thoughts about this idea of being able to restake different types of collateral how should we be thinking about like like the risks there and the idea I think it it's sort of dual to this LRT question of AVS selection in the sense that like the

types of collateral accepted kind of changes the credit quality of the AVS in a lot of ways um under slashing events and and other cases so like for instance usdc staking or usdt staking seems fine right like all you're doing is lowering the volatility of rewards you're also making the like you're making the equity like investment much more like a pure bond-like investment if you imagine those were yield bearing stable coins it's almost like having a portfolio of stocks and bonds right so in that case it actually makes a ton of sense like maybe the person reaking doesn't want to take as much risk and like this is a way of hedging for them and it's a way of hedging that's very

intuitive and easy to calculate um kind of the risk you're taking right because like adding I just push back on that though yeah because because if if you have like offchain asset like like eth as an Asset rwa Risk right like that's like that that is like you can be minced to zero instantly whereas eth is likely not going to drop 80% in a given hour that is true that is true I that it's a very that's a that's a very fair point that I I'm ignoring the the uh The Blacklist and rug RIS we could res it might be it might be safer than usdc or something like Frack or die just something totally on chain I I basically

wanted to start with like okay something that lowers volatility yeah yeah yeah it's like kind of a reasonable thing something that increases volatility but doesn't increase the rewards doesn't help you with security right because it doesn't make it more likely that node operators are going to stay and keep their locked in so there's clearly some threshold of asset quality between like lower volatility and minus the offchain risk let's pretend it's like die or something so something where we have more confidence that it's that lower volatility lower reward but it's sort of like I'm a more conservative reaker higher volatility higher reward and then higher

volatility lower reward because higher volatility lower reward can easily happen right like imagine that uh navs pays rewards as a function of the eth value of what you're staking now imagine I stake shitcoin 5 and shitcoin 5 like it's e value goes up because it just was launched on pump. fun on Bas equivalent and it looks like I'm staking a ton and like the first reward thing I get a a million e of reward but then imagine that it crashes 90% now I get right like clearly there's there's some that have this this negative cost so there is clearly this threshold

um and I once again will uh bring everything back to thinking about the bond market in that there if if I had Turkish L and I go to to an American company and I say hey would you allow me to buy your Bonds in Turkish L I'm not going to convert them to dollars as the principal I'm going to give you Turkish the company will say all right I'm giving you insane discount or insane premium like you have to pay way more right to accommodate for that that's like the same exact thing right and this is historically why bonds are very localized to their markets now obviously governments regulate them but

also it's like the people who are accepting it have to want to accept the principal currency currency that they're they're taking the initial principle I think the internet Bond you have to be because trustless there's it's going to be it's going to look different but that's a new set of cool challenges and and and obviously there's going to be some that have horrible credit quality great but I think like in the long run Vault operators and lrts and stuff like that like this is this what are you paying them a premium for it's like this right like yeah and then and then I'm curious then how you see igen token fitting into this model of risk versus reward effectively or

volatility versus uh staking yield or yield in general I think there's there's you know Ian is probably One Step Beyond reaking in all the financial analogies I make which is why I think it's like it's it's thinking a little more Galaxy brain than than even the way I'm thinking of things because I I try to always think again everything I try to turn everything into like boring but like simple Finance stuff I think this the long-term goal of enan layer at large is verifiable cloud right like notion of cloud services I can verify anything or everything that's going on inevitably that has some things that are let's say subjective decisions

amongst node operators and maybe you have to say the word inter subjective somewhere to to to reason about in that world I sort of view Ian token as the thing that is the excess value created by having this full verifiable Cloud over just the bonds right and the value igen itself as a token is like that the that Delta now I think the verifiable cloud stuff you know obviously we have proven out if not product Market fit insane demand for the bond market right like there's no doubt that all these people who had a ton of assets St St and we're like I want more yield and I want

it in some way that's predictable and lower variance than like lending it they they flooded in with 15 billion right like that that is representing the reaking aspect but the verifiable cloud aspect has this resolution dispute resolution aspect on top of just the bonds right um dispute resolution is very hard in the normal World it comes down to Legal cont so one of the most difficult dispute resolutions and and I'm I'm saying this one because this one is an inter subjective dispute and even though it's a TR trafi boring World dispute it is it is actually meets the definition of Interest where there was a

fork which is Argentina sold a bunch of bonds in the US forget exactly the year uh and then over time kind of synthetically defaulted didn't whatever and a bunch of hedge funds namely Elliot management and a few other people bought a lot of those bonds at a huge discount and then the Argentine government and the US government or US court system negotiated on bankruptcy with the with the of those like default of those bonds with holder and these hedge funds who had like some minority percentage

um so less less than 50% by a lot uh held out they said like we're not taking the settlement like like Argentina the country agreed with most of the bond holders and these these people held out and they held out for like 10 years and then they eventually got paid out and the court system here but in Argentina this was viewed as illegal like this was like not a real thing but the problem is they issued the bonds in the US they're custody in the US the US was the L1 that was managing that and Argentina the L1 was like [ __ ] you we don't recognize this and then 10 years later they eventually came to some agreement so Argentina forked and then they merged again and you could kind of

view this dispute as like the reason you need this inter subjective thing for the verifiable that's that's like the finance answer is that the bond market could actually exist without this resolution if you have very clear default events like clear slashing events like you didn't send a block in 5 seconds and everyone gave time stamps and showed you didn't but for these more unclear ones which are like the Argentina thing I I I I think you you do need to sort of you do need this extra layer and that's what igen token values now the question is we need abss live that are getting in disputes over you know for this to work so yes it's very forward it's like in my mind it's like the corporate bond market

existed for decades before we started getting all these like 1980s onwards inter subjective disputes that were like which kind of resemble the forking so again I apologize for my my always bringing everything back to financial history but I do think there are analogies to this inter subjective stuff that exists yeah yeah epic conversation through I mean we we touched on kind of how how to Value cryptoeconomic security um smart allocations to ABS is ABS yields revenues kind of the how we'll see the power law play out there LRT is allocating stake to avss how how kind of

you and Gauntlet are thinking about that and all the parallels to the the cdos and kind of bonds throughout the entire conversation tokens valuations all I can say is uh this has been a banger and thanks for coming on thanks for having me very I appreciate your financial analogies I think they're stronger than just knowledge I think they're they set they're precedents well I just think it's like easy to reason about you know it's like because I feel like there's always new words in crypto and if you reduce them to Old words then you're like oh I know these old words mean this yes red reduction spoken like a true mathematician it's just easier thank you guys thanks so much