Minimizing Risks in Staking

Transcript

welcome everyone we're so talking about staking today and especially minimizing risks in staking I have invited a great panel and let's go through introductions and everyone at the beginning will explain one important term to understand staking because I don't know where you all stand some of you might just be interested in hearing more about staking some of you might be experts and wanting to learn more about specific risks we're here for both of you and I would ask if you have any questions that are important to the moment just jump in and ask and if there are more general questions that don't fit the moment uh we also have a few

more Q&A minutes planned at the end my name is Marcus perria and I've been in crypto security and products for nine years professionally by now helping protocols and Banks and startups figure out the right way forward in getting crypto business models together in a regulated way making sure things are safe things are scalable and that there's a proper product in place good let me get started yanis introduce yourself and help people understand what is taking actually hello everyone I'm yanis I'm a senior engineer at chain security so

chain security is a smart contract auditing firm uh we are actually one of the ogs we've been auditing smart contracts since 2017 and we've been auditing like the blue chips of the defi world uh from Lio maker uh you name it and so uh yeah for staking we have also done a lot of Integrations with uh we have AED Integrations with various staking providers so what is staking uh so very simply staking is locking your assets your ether your know whatever native assets you have in order to get some yield and this sounds kind of a Pon scheme if you describe it like this so what we're really doing here is

you you lock your ether in order to participate in the validation of the network and this makes the network secure and that's why you get uh rewarded so yeah I guess that's the simplest way to define staking thank you very much Arena tell us about non-custodial staking hi my name is Arina I lead corporate strategy at all nodes all nodes is an infr node provider um we work with various protocols we currently support about 80 of them and our clients span from exchanges wallets custodians stops dexes foundations um the list goes on really so we are a non-custodial sticking provider which means that we don't have access to your

private keys and our clients are not locked in so we don't lock them into a contract um they can they're free to go whenever they want they can download their validator and withdrawal Keys uh from the dashboard when they want to and we don't keep that from them so they are free agents we are just a provider we're providing them the avenue to stake um but we don't keep C custody of their keys wonderful I would love for Eve to also introduce the term slashing all right first introduce myself my name is Eve Holstein with Bitcoin Swiss responsible for custody and staking we're like a full-fledged uh financial services provider uh doing brokerage

custody staking lending payments and asset management and I think the main difference here is that we do only custodial staking so it's the contrary of what what what arena just said so if somebody wants to participate in staking vs they have to have custody with us have to be an on board declined um and regarding slashing um as Yiannis just said um staking is when you provide your assets in order to participate in the validation of a network and slashing generally is a way to prevent bad behavior the rules are different for each network but typically when you behave badly or try to cheat uh you get

punished by the network and typically this means uh you lose some of your assets um and that's an economic incentive to behave properly thank you very much rul last but not least help us understand reaking thank you very much uh well first I introduce myself my name is rul I'm one of the co-founders of the nectar Network nectar network is a multi-layer Bing network uh under which we are covering liin and also Reen uh everything based on DVD which is the validation technology certain by the basics we are not a provider we produce software so that people can use that over in a permissionless manner and

regarding liquen and ven liquen I will Define it as the way of depositing e getting a receip token so you have a receip token that you can use in other platforms or with in other means and Ving is leveraging the security of the etherum network in this case uh to be used H to secure other things so in my mind the easiest way to describe it most probably is if you want to bootstop a network and you need notes you can tap in in the security of ethereum and use the notes that are already secure in the theum network do your own network excellent when we speak about cyber security the first thing that

comes to mind is usually private keys in the when you speak about cyber security in the context of crypto and blockchain of course and private keys in a context of staking are involved on on multiple layers we'll start with the private keys that I as the holder of the assets have when I want to stake how do I secure these assets what are the risks that come with staking to me as the owner of the assets I can take that uh so yeah whoever owns the private key uh for an address they can make changes on behalf of this address so the

biggest risk if you lose your key then you cannot make changes right and the question now is how do you secure uh your private keys and actually there are many ways I mean there are many simple ways like never letting your private Keys touch anything that you know can leak it or then you can go to you know for more involved approaches like having multi6 for example what is this it's like smart contracts that in that hold the assets in order to make some uh to execute a transaction multiple people have to sign and in this way you can actually prevent from you know having a single point of failure which is

essentially uh the biggest risk uh in this case and then I think one of the most recent advances that you know people haven't really uh realized uh its potential yet is uh abstract account abstraction which is kind of allows people to it it allows for for example for social recovery so people don't really create transactions they don't have to sign them but they create what we call an intent and then another party is responsible for creating the transaction on their behalf I don't know if someone else has to add

something if there's nothing to add um the second aspect where there are private Keys involved is at the validators that are run as an infrastructure by companies such as Arenas and EES what are the risks that exist there when it comes to private Keys yes I mean as Yannis mentioned the risks are about losing your private keys so when you're in a non-custodial situation like us the client is responsible for managing their private keys so if they lose them I mean that kind of responsibility is on them right um when it's a custodial solution like Eaves it's completely opposite so it's just about you know as a client you manage

your approach you figure out what do you prefer doing do you prefer trusting a provider um whether it's a custodial seeking provider or a custodian um or do you prefer kind of you know doing the direct route and working with a non-custodial provider maybe to add on that um the validator in most protocols it's always a little bit different in each protocol but generally um the validator uses the private key to actually sign the blogs with the transactions so there's two main cases that come to me to my mind either I lose the key the validator is no longer able to operate in most protocols this means that potentially slashing or other penalties could then

happen but the assets are typically safe because you can usually unstake them still uh even though the validator isn't running uh the more dangerous part is operationally if you suddenly for usually human error set up the same key twice or multiple times which means that the network believes your cheating and trying to double sign transactions which usually gets punished the most hard way the network can do typically with brutal slashing and uh in order to protect those risks uh you typically apply strong process controls for i6i principles you often times Outsource the um the keys from the actual validator to a

centralized um signing device so that it cannot be there more than one time uh or you use new technologies like DVT distributed validated technology which also ensures that you cannot have the key multiple times running yeah maybe to dig a little bit deeper on DBT as you were mentioning because it's actually is kind of the core base of our protocol uh the way that we are doing it is actually um in basic terms it splits up the key between different participants so no a single actor can control the key and everything is controlled by the protocol itself so actually sorry can you explain what is DVT I mean what do the letters DVT stand for sure H this validat validator technology and actually is a way to

validate or to become an ethereum operator in this case or an eum validator in this case h in a distributed manner so instead of doing being just myself with 32 e and my computer validating or my whatever equipment I have validating the blocks what I'm doing is getting together with other people in a permissionless manner and uh we are splitting that key between the different participants so we are having what we call Key shares which is actually a kind of a piece of that key we can say we all come together and we validate so everything is controlled by the protocol itself and uh it's not a single or validating the blocks and controlling everything so DVT is addressing the risk

of double signing Eve mentioned is this also um addressing the risk of availability if a validator goes down or is not on the network Yes actually for the if you are using DVT and I have to say this at least in my mind is the most resilient way to to perform validation duties um in H one of the Clusters at least twoth thirds of the of the cluster needs to be down in order to uh to get as last so that makes kind of a big difference at the end of the day what you are doing is and I think it was Yan is the one who was mentioning like one of the biggest

risk is Central is the single having single point of failures you're avoiding having a single point of failure because you are leveraging the power of a network I hope it makes sense right anything else that came to mind in this topic of really hardcore crypto private key type topics all right then we move over to speaking about protocol specific risks because when I look around many people who are start to be interested in staking they just look at what staking protocol gives the biggest yield and there's a long list starting with 20 or more per when it's about L1 protocols

going down to 3 to 1% and less and of course these different needs come with different risks because each protocol has a slightly different way to implement staking and we have slashing that Eve and few people already mentioned which is implemented differently some protocols don't have any slashing other protocols only slash the potential rewards and then finally there are protocols that even slash the full stake that is at stake so that you can lose everything that you use for staking and that's certainly worth considering when you pick a protocol for

staking uh then um well I I'd like to open up the round you maybe what are things that your clients look at when they pick a protocol for staking I mean I think it also depends how the client identifies risks when it comes to choosing a protocol right for some of the priorities going to be liquidity for some of the priori is going to be the um the slashing risk someone's going to be the amount as you said the yield so it kind of depends um obviously the risk associated with different protocols as some protocols have slashing some don't um the UN bonding the bonding period will vary I mean we can't tell our clients which protocol to choose but we can present them with the facts we can tell them that you know with this kind of protocol you will have this kind of

bonding period you will have this amount of of uh commissions um etc etc so it it's up to the client to decide what kind of risk appetite they have um it's up to the client to decide also um you know how much liquidity they need or working capital they need um can they wait two days for un bonding can they wait 21 days for un bonding or do they need it you know instantaneously so it's it it's really depend on on how they operate their internal structure whether they're staking their treasury or they're staking their clients assets um but yes Protocols are different they're not all the same they're noten made the same and that's why different protocols will work will work for different uh clients can I just spend a few words on

what is a bonding period of course so the bonding period is a time that is needed to actually um Bond your um to unbond your stake so when you want to withdraw your assets um you then have to wait a certain period of time before they become available to you uh yeah I can add to add to that so yeah uh there as I mentioned there are also Solutions on chain like Lio like rocket pool where people can uh stake their assets and earn some yield and there there are still a lot of risks and uh I can just name a couple of them uh for example H stakers should always ask is

the the technology or the smart contracts in my case mature enough to you know to sustain the risk that they're undertaking uh another question people should ask is whether this protocol is centralized enough or not for example people uh stake with Lio but Li is an entity and it's I think it's a legal entity even though it's a DA so you know it's again we end up in a single point of failure situation right because if Lio is forced to take a decision then all all the people that have staked uh their assets are essentially forced to follow this decision even though they might not want

to and this is very important and you know sometimes when we are focused you know on the yield and the profit you know we tend to forget uh these kinds of risks and I think to add to the centralization aspect is we also have permission permissionless networks um with the permission Network you have a bit less of centralization there's limited amount of validators who are offering the validation service on the network so things also up to the client to decide if they want to go with a network that's permissionless or permissioned maybe one more thing um regarding how the clients select uh the protocols is since we are custodial we're a financial service provider so we have a lot of obligations on risk

disclosure so typically if a client wants to stake with us he has to sign an additional let's say Annex to contracts that can three four pages with like all the risks that are protocol specific per protocol which can be a little bit intimidating sometimes but uh we really have to like show them all the foreseeable risks um and the magnitude of them and where he can actually lose a significant portion of of his assets um but then on the other hand we're probably a lot more restrictive than um uh all noes for example is with the selection of what do we even offer to our clients because we also have like regulatory risks that we

have to take into account um so that the protocol is not potentially security token uh etc etc so we already do some filtering and are only allowed to offer um let's say a selection of protocols where some of the risks have been eliminated by us and there are some risks you mentioned smart contract risks there's a protocol polygon Matic for example that uh is actually staking the smart contract on ethereum it's like a chain four or five cascaded smart contracts and sometimes can get really really difficult to understand what are the actual risks so we really dig deep into that risk um analysis before we offer something you want to add anything sure

um no maybe well first of all I have to admire the way that you can reduce the complexity to your clients when you have to explain them something like a polygon and um and secondly um yes coming back a little bit to what H jiannis was saying again um I think one of the factors that is also important to take into account when we are speaking about risk when a Shen and within the whole ecosystem is a also the market dynamics risk at the end of the day H you were talking about Lio as an entity whether is aada or not but one of the biggest risk that we are facing at different levels also at the ring level is the number of participants

that are within the market H when at the end of the day this is a infrastructure market so each participant within the market is offering security to the whole ecosystem and if we are speaking just about ethereum when you have some participants of between two participants that can cute and control more than 50% of the market you ending the the security of the whole network so we should be incentivizing Market diversity and also taking into account from a risk analysis perspective what is the market concentration within each different network that we are approaching right we indiscriminately spoke about protocols taking protocols like L1 L2 protocols in the blockchain

space and also more smart contract based staking protocols and there's probably also a risk difference in between the two one that's baked into a base protocol and the other one that is added as a smart contract on top and I would ask you honest if you have anything to add when you audit these kind of smart contracts or maybe you've also audit the base layer protocol what are kind of things you specifically look for I mean usually the issues uh relate to you know to the underlying logic and how well it complies to the specification and for example uh when we're talking about staking on ethereum

uh you know the staking is very well specified uh but actually when people Implement solutions that you know make use of staking they make mistakes because exactly they haven't fully understood the specification well enough so uh yeah one of the things uh we check is uh this and then of course then there are always blind spots I mean it's code that it's been written by humans and yeah humans unfortunately make mistakes so yeah we always have to take care of this kind of issues all right anything to add yeah um maybe I'm jump a little bit on I mean

on the next points or not but I just wanted to say that I on on the sense that J was mentioning and what ifs and in has just said I really want to kind of to pay my respect to especially to if and inen because I think that some something that everybody needs to understand is the enormous complexity of explaining especially then that have to deal with third party clients and third party that are trying to get into the market even if they are specialized within a virtual assets or crypto assets the enormous tax that that they have in explaining that to someone who may not be that familiar with the market

in a previous life I used to engage with a with with investment I mean firms and investment terms and I remember that just to explain a plain vanilla investment to a prospective investor is it was I mean 50 page summaries documents and everything so the way that they have to deal with this complex structures and complex infrastructur is I mean the challenge that they have and the way that they are handling that is impressive I think the fund starts is when the client understands what vanilla staking is and then you have the reaking liquid staking all the other additions adding on and they're like we thought we understood this but we don't now that was actually a nice Segway over

rul thank you to the next uh part today but let me take a step back many protocols especially ethereum have it in their ethos to encourage self staking basically coming from the be your own bank and and not trust any intermediaries uh ethos that's very heavy in crypto but of course not all protocols have staking requirements in a way that make it easy for just anyone to be their own staking providers and that's why we see many third party providers out there such as all noes and Bitcoins with and when you as a person or as a person working in a company want to pick such a

service provider for your staking needs what are the kind of Frameworks that are in place to do this in a way that's repeatable that's understandable how do people assess the quality and security of third party staking providers you want to have a go Eve yeah I would say there's some some Basics uh that you maybe just look at your potential provider how long it's been around uh what's customer base do some secondary research is there any bad news about them Etc but then more recently there has been like a secondary industry that has developed around this there are websites like staking rewards stakes.

network and Beacon chain for example that start listing staking providers rating them um some of them like staking rewards start with certifications for those providers so that's that's one way how to how you can identify the maybe the more reputable ones um but then if you go one layer down um as a provider ourselves we have to manage operational risks as well and it's always easier if you can follow a framework and there Lio has been really really helpful there's been a group of Lio operators and then let's say entities close to them like us supporting that created a framework called the duck framework which is

basically a risk assessment a mitigation framework uh building on the typical standards in Risk Management um and helping basic basically players like us to structure the way how we how we want to make sure we reduce the risks and then there's others players other players like aluvial which is a liquid staking provider for compliant um liquid staking which means kyc liquid staking they work together with price falh house to create the framework Etc so there's more and more developing in this area where on the one hand the client can leverage them to select a vendor and the vendor like us can use it to improve their

maturity and eventually get audited on that H how do you actually do this in practice Arena I mean for us I mean also we suggest that clients look at the history as Eve said performance uptime of validators um how they structure their product um look at their Tex tack look at their you know if it's Hardware is it software is it a cloud stor I mean there there multiple points you have to analyze when deciding who to work with and again look at what is a priority to you um look what your addictions are covered if you are in the states or you're in Europe or in Asia um look where the hardware is set up to

minimize latency to improve uptime to improve the performance um multiple factors go into account let's switch to perspective maybe and look at how do all nodes instill trust in their clients that use Frameworks or don't use Frameworks establish Frameworks what are the key points that you bring to the table I mean instilling trust comes from obviously providing a good service a good product and listening to your clients and I think we base a lot of our product development on feedback is quite simple um we talk to our clients both on the retail and institutional side because we have two segments um we listen to what they're saying what they're demanding what they're asking what problems they're encountering um some of the things that you know we

offer which I think is unique to all nodes is running our own data centers uh we have multiple data centers set up across the world where we don't rent them we don't lease them we actually physically own them um that you know adds to the security aspect to the Privacy aspect um we have more control over the data centers and you know really specializing in the custom Solutions um I mean yes saking is one of our products and we have a lot of other node in for products but focusing on staking um I think that having unique things that differentiate yourself from other providers uh comes and gives you that leg up right maybe back to Eve how do you because we tackle the challenge um oh it's mainly we we're we're profiting from the history and the

reputation right now um but we are working on now implementing those Frameworks that I just mentioned we're not there yet um and then eventually want to get audited and and then add um slashing insurance as one example but uh in our opinion it makes more sense after we have been audited naturally can negotiate better right um so that's the way we want to no longer rely on on History because clients get more professional as well so while the name was good enough until maybe a year or two ago uh nowadays the clients have Auditors they want to see something Etc so we also have to go with the flow and uh and get more formally um

mature right anything to add from you Yan or all right yeah we have more topics to go through that's perfect and actually I would love to dive a little bit deeper into liquid staking and reaking um rul only touched on this very lightly at the beginning I asked him to cut it short but it's quite a complex topic and I don't think uh that short intro did it full justice so let's have another go slightly deeper slightly more time ra I think there's a quite a few people in the audience who would love to understand and walk away from this today saying now I know what raking is I I don't really know if I know what rein is fully I mean no I mean it's it's

a relatively new thing at the end of the day rein H just to give you an idea there isn't yet uh a fully functional ER agent in the market I mean at the end of the day the lead with within re has been aan leer and even a leer hasn't implemented yet slashing and all the functionalities that they were promising they are en rolling gradually just to give you a sense of what is the the status of of the technology at this stage and um what is uh basically Ren ER I in order to put it in simple terms uh in a second what you are doing is secur a network through value depositing your e and your crypto assets within that

Network in order to secure that you are not behaving in a way that you shouldn't behave H re in simple terms is just okay that asset that I'm putting in this network to secure that I'm behaving in certain way I'm going to reuse that asset to secure other things so I have other things again later call call them abss I think is a a nice term as well and we use them as well uh what is AVS stand for AVS stands for active validation services and actually is actually broad term H an ABS in my mind is anyone who wants to Bo strap a new network because it can be almost anything um and uh anyone that wants to

drop a new network instead of okay it's that setting its own set of incentives or well it sets its own set of sets of incentives but instead of starting from zero trying to attract people with their computers and with the equipment in order to start securing that new network H what these abs are doing is okay let's try to reuse H the people the the the agents that are securing other networks in this case ethereum and the but there is also reing in in Bitcoin by the way um and these people what these abs is what they are doing is stop in into that security so what they are saying is okay I'm

going to give you these incentives if you start performing additional duties to R my network and if you don't perform the way that I want and you start behaving on a different way maybe because you are focusing just on the other network and you are performing the task I'm going to slash you and to penalize you in the network where you have those assets deposit that is extremely positive from an economic point of view because you are having the same asset securing different things so obviously the economics H increase is some somehow leveraging the the value of the asset that you have in that initial network but it also poses certain risk because you can create Cascade effects so if you

have multiple avss and everything starts to fail and they they start lashing The Operators within the primary chain actually they will be penalized that that chain will be affected by something that at the end of the day is early end to end of course people would love to understand what raking is because they want to assess whether the risks that come with the high yield of risk taking is worth the risks that they take and maybe that's uh something we can spend a few more minutes on at before we come to the Q&A at the end uh I want to point uh something else first uh so yeah we're talking about the immediate risks of raking and indeed

this leveraging this cascading effects is a big risk but I think there's a bigger risk we don't really talk about and as Marcus mentioned uh before uh the ethereum ethos is do it yourself right and unfortunately it looks like with reaking uh the barrier or the threshold the technological threshold that someone has to overcome in order to participate in staking is higher and this means it's even harder for someone for a simple you know person to do to to use their ether to get the maximum guil and what happens then then people will turn to various services in order you know to get the

maximum yield out of The Ether and this will eventually uh lead to centralization or could potentially lead to centralization and then you end up you know totally contradicting you know to the ethos of ethereum uh which is focus on sensorship Resistance because if you have only a few players who really you know participate in the network because they do the staking and they do all they participate in all avas then essentially you end up in an ethereum which is you know with consists of way less players than what we are used to and I think this is a big risk we are not really talking about today all right so there's not only risks to people who stake and service

provider who do staking but also to the ecosystem at the protocol which is that can be encouraged uh by certain developments right exactly sure so who are what kind of risk appetite is needed to do risk taking today who are people doing risk taking um I mean from the let's divide into two sides from the seconding side obviously anybody who wants to get a higher real a higher yield and they have a a higher risk profile because at the end of the day even in basic term I think everybody can understand that you are posing yourself not just to the risk of one chain but to

several agents that can't I mean produces Lin and can Dage your I mean that that asset that you're seeking and um on the operator side on the not operator side I think maybe if an yeah I'm going to ask that that's a good Segway so your clients know your clients Eve are these the kind of Risk Takers who want to participate reaking today I think it's really early to tell I think right now obviously the first clients we're going to be adopting reaking is a retail segment on our side um they have a bit of a larger risk appetite um they are willing to maybe kind of play around and see how it works um institutions on our end what we're saying are curious

obviously they're observing but they are not interacting with the product just yet they're waiting to see how um you know how it's going to work but I think that institutions are usually the last ones to adopt a product that is new uh we have some very crypto native clients also the institutions are typically Wales that just turned into institutions um they're really open for risk but we are not so we we just observe the whole topic because I think nobody really understands all that cascaded risks that you guys have mentioned and maybe we'll look into it a little bit closer towards the end of the year but until now we observe thank you very much thank you all for for

attending I would like to advertise the cyber security working group of the crypto valy Association that is meeting monthly and having these kind of conversations every month we have an expert come in and present a topic that they're working on or that they researched to share information within the community everyone and anyone is welcome to join whether that's deep experts who can contribute and challenge each other and learn from each other or newcomers and people who would like to learn and challenge their knowledge and become more knowledgeable in the space please sign up through the CVA the crypto Val association website or approach one of us directly uh we also have a telegram

chat where we share relevant information keep up to date on Hacks security developments changes in Risk profiles of protocols so that's a good way to stay in touch with the topic and now we have not too much time left for questions but let me see if there's appetite for one or two questions please noncustodial staking provider which is of course much bigger so I have a very general question for actually maybe Anda do you think that if there is over time a concentration on the bigger staking providers providers that

the vast majority of theing percent even though the risk is of the From perspective to a bit of synchonization risk in terms of the eum do you think you can get as much Market as you like there it's a very loaded no no it's a very loaded question so you're asking there like many steps there I'm trying to keep track so you're asking pretty much what's going to happen if let's say a validator has a larger market share of the stake right I mean at the end of the day if the client is coming to us and they're saying that you know we want to bring you our assets to stake um we're not going to say no no to them but we

have before taken measures where we've seen that we are now getting close to a larger market share and we've increased for example our uh our fees in order to discourage clients uh retail clients especially from coming to us and maybe going to another validator now that does not happen very often with then that once or twice um in our kind of History um at the end of the day we believe that it's up to the protocol to make sure that you know there is the centralization and as I mentioned with either permission or permissionless way I mean at the end of the day for us we are providing a service we're are serving customers and we are providing an Avenue to staking um unless I don't know you guys do it differently no um and we're we're we're way too small so

uh I mean we're number four in custodial business but still we're way too small um so looking at ethereum for example there's about a million validators time 32 e that's a lot of money um so it gets dangerous the moment you have 33% of the network in the case of ethereum that would fail or be manipulated because from that moment on you can prevent finalization of the network um there's the biggest one around is uh coinbase with 11 12% but around Lio this discussion has actually sparked last year quite significantly because Lio was uh above 30% um and there was some fears that

even though you have around 30 providers within Lio um in theory um Lio themselves control what those providers are doing because they control the smart contract so they could be a certain attack Vector um that has been very much addressed by the commun I would say and that there's awareness for that and and it's like almost a self contr control so Lio has been really attacked that they have to take care um other protocols may be a lot more concentrated where this risk is bigger and I think for us also I mean we are wary when we are providing a white label solution to other validators for example um if we have multiple validators on our Network um that's when

we are paying attention to I mean how much of the share that we have thank you R for the great question question apologies everyone else there's no more time for more questions with the time that these people take and answering them but we are going to be around join us for lunch outside and ask more questions we're happy to talk and please join the cyber security working group [Applause]