Liquidity unleashed: A research driven analysis of post Shanghai liquid staking derivatives

Transcript

[Music] [Applause] [Music] my name is mingan I'm from University of Chicago um quick intal I'm a quantitative researcher um I do mostly math and data research in uh Defi and tokenomics and for the past summer I've been a research uh intern uh Fellowship whatever you call it at nethermind which is one of the uh ethereum development and research companies and this is partly what I worked on at nine um before nine I was a research analyst at polygon um where I did more defi

research and I'm happy to to talk about like lscs because they've been very popular um in in Defi and a lot of people are doing research into it but uh there are a lot of risk factors I I think think is a that requires communication um with relevant shareholders in the ecosystem so I'm happy to to share about my research here all right so uh quick agenda so I I'll do a brief intro of ethereum staking and uh liquid staking um for those of you who are not too familiar with this emerging field and uh for part two I'm going to provide like an

economic framework and and financial framework for analyzing uh L's and especially do a risk analysis um for liquid STI and derivatives and uh third if we have time uh I want to walk through a case study using rocket pool which is the uh second or or maybe third largest U liquid staking protocol today so uh let's get into it um quick disclaimer not Financial advice uh I have no Financial ties and uh views are my own and I will be sharing some research tools and uh sources throughout this presentation if you're interested you can do your own research uh according to those sources so very very brief history of

ethereum staking uh it's only been a year on mainnet since the merge so people know about the merge right so essentially ethereum switched from proof of work to proof of stake um and so now anyone can stake 32 e on mainnet and C rewards as a validator um we we just heard um another talk on this so uh earlier this year there's the so-called uh Shanghai capella upgrade some people call it the chaella upgrade and this was a huge deal for ethereum staking because uh it allows you to not only stake uh on on Beacon chain uh which is the where the staking contract lives uh it also allows you to withdraw that stake so it

allows you to quit as a validator essentially or or withdraw your rewards before chaella it it was not possible so as you can imagine this opens up a lot of opportunity for you know potential investors and and stakers who want to come into the space right without fearing they're just going to lose the money forever their money going to you know get stuck on on the Chain forever right so uh we we've seen a lot of growth after this chapella upgrade that's why I'm partially seeing that uh it's a big deal and I'll I'll give you more data on this later on now uh I want to like kind of solidify the numbers of ethereum staking uh typically we say it's just like a

3.5 um% APR or or maybe 3.5 to four is a more accurate range right and uh but but slightly more nuanced than that so there are two parts uh of uh staking aprs two components so um one component is the consensus L rewards which deals with everything from you know uh block proposal attestation so anything that has anything to do with uh a consensus itself right so um you know uh it runs on the consensus client um basically and the other components is execution a rewards so uh you have first of all transaction fees gas fees that people are paying for their transactions uh

which was modified by 1559 and uh there's the other big part which is me which I'm not going to talk about today but you know it it's a big part so uh on the chart you can see uh the APR breakdown from Lido's Doom page so uh it's a pretty good representation of where we are right now so uh the orange section is consensus day rewards uh you know since it deals with consensus it's fairly stable and it's not really you know responsive to to any Market changes um you know reasonably but the other part which is execution rewards is quite uh well it varies a lot um from day to day since you know the

transactions the nature of the transactions vary a lot day-to-day uh so this is something to keep in mind all right more data um currently there is approximately 22% of total e Supply that is St uh it's not a huge number compared to like some of the other l1s um they have like crazy staking numbers um but ethereum is more used so it makes sense um the breakdown of of sources um of all east states the first of um the largest section is currently liquid staking at 37% which is huge um and notice another

like very big chunk like around 20% is centralized exchanges um and some centralized exchanges also issue liquid staking tokens as well so the the actual number of lsds market share is probably around around 50% I think um give or take all right now we mentioned liquid staking let's do a formal inol so lsds or leg GRE derivative liquid staking tokens where we call it uh are erc20 tokens that represent the East tokens that are Stak uh in the staken contract of ethereum so here's a very like basic model of what most liquid sing protocols look like so

users will deposit their e with the protocol which is typically a a pool of some sort uh and the pool would delegate uh 32 like batches of 32 e to the node operator um the node operator runs the validator software and they use that 32 e to run the software and uh they stake that e on on Beacon chain so that's how it works uh in return the protocol will issue uh lsds those erc20 tokens back to the user um for their you know contribution to the um e staken pool and when the user gets the erc20 token they

can do whatever they want with it right they can trade it on defi um like stake it somewhere else um not not stake it collateralized uh somewhere else to you know get a uh I don't know a loan from a a um they can collateralize their trading portfolio on like a Perpetual exchange whatever they want with it right so that's that's how it works and and by trading it uh on amm or or any other application other people will have access to this liqu sticking derivative too I'll explain later that it's a really interesting mechanism that they can do it so uh benefits of liquid saking so

first of all there's the return that you can get right um and the second benefit which I think is a a bigger deal is that uh most LSS are retail friendly meaning that you don't need that 32 E capital to to stake your ethereum and get the staking reward so you can stake anywhere between I think for for most protocols it's like minimum of 0.1 East or maybe even lower sometimes um so it really allows a lot of retail users to enter the staking space and contribute to you know ethereum consensus which I think is really good thing for the ecosystem and the third uh the third thing is uh well

as a name suggests lsds are liquid right so ethereum staking itself is is a lockup mechanism um so you can't really access that until you you know do the uh and I explain how exit works but you can't really exit until you um you go through the entire queue of exit and withdraw and blah blah blah so liquid sing derivatives solves that um and the use case is a kind of mention some of it right you can use it in lending you can use it uh to cize your trading portfolio etc etc uh those are the I guess the most popular use cases right now people are exploring other options like uh LSD

backed stable coins and stuff like that but those are still like in the idea phase or some of working progress so these two are probably the the most common use case right now um and uh I I didn't mention this but those thing derivatives are redeemable for e at any time we will see uh when that is is not the case and why that will be a big problem but for now let's let's assume that you know LC is redeemable um and uh the last point I want to mention is is is a really interesting mechanism that's kind of unique to LC's compared to other def5 tokens so LC's a cure rewards automatically um different protocols

have different you know ways to implement this kind of automatic accural so um but but in general like holding uh LC tokens are is equivalent to like staking e in the protocol itself so uh when I say different protocols I mean for example Lido um Lio has a oneon-one peg of their stake ethereum with ethereum itself but how are they going to like get the reward to to users right they like literally airdrop the rewards um to the user balance um pretty much they update the balance uh to reflect the the number of rewards um every every

certain time period they have uh some other protocols like rocket pool implement this differently so rocket uh r e is there LD um the the exchange rate of r e and e is not fixed it's not packed um it actually increases um based on like how how much time has passed uh since you stake um in rocket pool so uh but but in general like all lsds Ure rewards automatically um in this sense now uh I can of mention this so the data shows that from April which is uh when the chaella upgrade happened to uh I think this is end of August right

we've seen huge growth um for the LC sector it's currently the number one D5 sector and Lio is currently the largest D5 protocol uh right now and uh it's like a I think I counted is like a 40% increase uh over like five months which is kind of crazy um and as you can see here this is from def5 Lama by the way um the blue chunk which is kind of dominating market share um You probably guess that it's Lido um so this is the current like landscape of the of the market all right now I want to Pivot into you know some more economic analysis of LC protocols um and like LC

as an asset as well so here's my here's my own mental model I think of LC's as similar to Banks so this this might sound a little bizarre at first but let me explain so um the economic definition of banks um they're in the business of doing the following they're the financial intermediaries that create liquidity through the following two mechanism first they gather liquid deposits uh from customers that's what we do like with banks right we we deposit with them and we deposit cash with them so that's liquid uh that's a liquid liability and uh what do they do then they then invest those liquid assets um

in in presumably bigger chunks um into a liquid investment projects so some of the common Bank Investments include like government bonds uh corporate you know corporate uh loans and mortgages right those kind of things which is less liquid but generates higher return uh and that's their assets and isn't that what exactly LC Protocols are doing right they gather liquid E from the um from the depositors and hold them m as liabilities right and then they invest those e in bigger chunks namely 32 e into ethereum staking and that's the asset they're

holding so here's here's my mental model for LC's um typically um the good thing about this is that it allows a lot of you know research insights from trefi um and even some some cases web 2 um to like Enlighten the research of this very emerg in very new field uh of lsds um I I'll mention some of them later on but here's the model so the natural question to ask um that's what I did when I when I like kind of thought about this model is how liquid are LC's really right when we talk about like Banks cre liquidity LC

protocols created liquidity how much cre uh liquidity did they actually create right so I use um this metric um I actually use two metrics uh this one is called amihud uh it's from the finance literature I don't want to go into it but uh essentially it measures the liquidity of of an asset and it it's really nice to have this kind of quantitative measure because you can compare all of the LC's together um in instead of the you know just the depth of liquidity measure that we typically use in quote in in defi um so then start the blue line is uh ethereum itself and uh the other ones are uh lsds so it's kind of interesting like there's not too

big of a discrepancy right uh obviously LC's are still less liquid than eth itself I think we can all agree on that but it's actually not uh it's actually pretty close um and uh we we also see like a kind of liquidity hierarchy on this chart so R that's rocket pool is at the bottom uh a little bit above it is coinbase and then Lio is kind of on the top uh which I think like is consistent with most people's like mental models of you know how how the competive landscape is right and here's a similar chart um but this one is more like highly varied

through time this another measure by BPW 2011 but you know basically the same Char just with different fluctuations um it's interesting that like coinbase actually you know flippen uh in a sense uh e itself in some time period back in April and May which to me is really interesting to see um but I don't really have a good explanation for this um it's it's part of my ongoing Research into liid staking currently all right now let's let's go back um about this you know Bank model for liquid staking derivatives um when

we talk about banks in economics the the most relevant Theory uh I kind of alluded to this but is is Diamond diig um if you don't know them they are the Nobel Prize winners for last year and they are also you Chicago professors uh sorry um diamond is a professor at you Chicago de is not um and their Theory specifically says about bankr so bankr as the name suggests is when you have a bunch of depositors a bunch of users run to the bank to get their funds out um to you know redeem their funds because they fear that the bank may not be able to give them their funds back if they

they're late for example so they're going to line up in front of the bank um and and you know yell or or whatever um to get their funds back and I I started to think well is that the same is the same thing possible for you know LC's in this case if they function so similarly as Banks and the very disturbing answer is yes um so Bank RS are typically triggered by two things they they can be triggered by you know a sou increase in demand for liquidity um in our case there could be you know a liquidity shortage for example we we might heard of the the curve event uh I think it was

what July right it was July when a bunch of liquidity pools were drained uh and liquidity was um on a shortage right so that could trigger uh a bankr and another you know reason for why you want e instead of LC is just the price dropped um and you want to just get out of this EAS investment entirely right that's a reasonable reason and there's also like regulatory concerns like uh if say the L um sorry the the SEC comes in the room and say guys uh we now declare liquid staking derivatives as Securities because they acue rewards

and blah blah blah right what do you do with them I think many people will want to dump LDS uh in that case so that's a concern too right and uh another thing that could trigger a bank run is the expectation of insolvency uh this is more serious but it it could also theoretically happen to uh lsds so you know I think we've seen the past is all kinds of bugs and hacks and exploits happening to protocols and when protocols lose that chunk of fund they may not be able to pay back um their their the polies so that's a real risk in in defi specifically and we've seen a lot of those so I think I think

by 2023 people won't you know just just very um idealistically assume it wouldn't happen right because it will um and the other possibility is that there's large scale slashing uh or uh penalty of validators slashing happens when a validator does something that the system doesn't like so basically um like uh if you if you heard AUM consensus in any way you'll get slashed um that's the short explanation for it right so these can happen so what happens during uh a bankr what happens if you have a like a line

of customers line up at your door wanting to withdraw the EAS right there are two main methods um um uh of how those customers can get the funds back so the first method is to direct Redemption from the protocol like we see in charts right and second method is actually through dexes they can trade uh the LC's back to East uh for like amm pools um aggregators whatever they want yeah and what happens after these run out right then the protocol will have to withdraw from those Investments AKA they have to withdraw their their funds from the beon chain and that's very hard to

do because there's currently a turn limit of I think they just updated to like 11 or 12 um like last week literally so the exq only allows like a very small percentage of validators exiting every day so it will create a um a que um in the execute and after the execute you have to go through another thing called the withdraw queue uh which is how you get your funds actually back into your wallet and this is um the same cue as uh the partial withdrawals so the people who want to only get their um their rewards back um but is but full withdraws are processed much slower um than partial withdrawals and this is

data from rated currently the withdraw queue is 7 days and exit Q is very short because nobody is really exiting right now um but you can imagine like in a bankr scenario right where everyone want to exit they're going to queue up they're going to queue up and similar to like what the activation queue is currently at which is 24 days um and they have less rews so in that case what do they do right what the protocol do they could pause the withdrawals which I know it's it's not pretty looking but they can do it um Lido currently has a so-call bunker mode that you know they they will activate in

in the case of you know large slashing and and some other concerns right and they could also choose to sell protocol Li uh Equity um so for example most protocols will have governance token they could choose to sell that from their treasury but beware this might cause what we call a self-fulfilling prophecy of insolvency right so in def5 people call it death spiral so when you when like imagine a bunch of D DG see your selling your Equity to meet the withdraw demands it's it's a very bad look um so people will dump on you and you know bad things will happen even if like initially it was just a small like

gap between liquidity right so this may cause insolvency um what's missing here is actually um what banks do uh why banks are safer right and that is insurance so we don't have that kind of like FDIC Insurance mechanism in defi um so like this is the only like tools in their toolbox um practically um I I will skip the systemic risk we're kind of uh out of time here I I want to just um end here and um ask if you guys have any

questions questions questions for our speaker I don't have a question per se but could you talk to that first point a little bit more about another mind and Lio collaborating to address that centralization or is that too deep of a topic to I I would definitely be able to be happy to chat about this uh afterwards it's a pretty big project that we just finished for Lio um it it take us like a few months to do that and I've been on this Fellowship for three months and worked on it pretty hard and and pretty deep uh in into a lot of

those fields um but this is something that is still work in progress and um I think a lot of people are criticizing Lio in in the right way um and this is a part of a solution that we are addressing those criticisms yeah so so I'll be I'll be happy chat about it afterwards okay all right thank you all so much um for listening and um I I'll be you know at the event until end of the day and I'll be happy to do chat whenever all right thank you