Unlocking DeFi for risk-averse participants

Transcript

thank you hi is it working okay um hi everyone good morning um thank you for coming uh today I'll be speaking about the risk management in D5 and so I named the presentation D5 is the Dark Forest I really like that analogy it's you probably all heard of it it's very used quite often so it introduces this like concept of a Dark Forest where basically detection means elimination in in the context of risk management I think it's very

um like close to the reality where people constantly search for hacks for weak points in the smart contracts in once there's a detection certainly someone will exploit those uh and so it's no secret that ethereum blockchain is like every other blockchain is a very adversarial landscape not just as a financial World in like D5 uh just basically like a PVP game but then also technologically it is a Battleground um and so I will not talk today about financial risks um those are like Market risk volatility prices go against you liquidity risks basically are when you cannot like

liquidate assets fast enough and uh so you have like slippage and all that you can carry losses through that that's all Financial risks credit risk we know there's a bad debt uh accrued on a lot of like lending protocols there and there so then a more overlooked category is non-financial risk so systemic risk the one that we like to ignore often but we know like the whole blockchain can be like shut down and a lot of people can carry losses through that then there's a regulatory legal race risk and basically you know like the team of the project can just get like sued they shut down uh we know what happened to certain names but then my focus today is on the operational risks

so that's where all the exploits hacks kind of like technical failures and shortfall events come in so I like I like this chart um I think we can see that um basically every three days there is like an exploit uh happening and the losses even though like the risk rating kind of goes down but then uh there's still billions of dollars like lost to all these exploits and hacks and most people are always found surprised and kind of like that's something you don't expect uh happening to you even though like we all are well aware of those risks but then all the people that suffered big losses due to hacks

um you know feel the emotional pain of being just dumbfound but then it's kind of like normal to everyone so how do you hedge those risks I want to start with kind of the basic fundamentals so the level zero of your risk management would be don't put all your ads in one basket that's kind of uh I think everyone heard of it and everyone knows about it so if you toss everything kind of like on one chain on one protocol uh it's getting hacked you're then you can upgrade your strategy um so the very simple logic is basically to have a portfolio share kind of like exposure to a single protocol less than

your weighted average API what that means is like you know if your weighted average is the apy is say 20 over the year you generated 20 if you spread it across five protocols and one of them getting hacked um in a year considering you know there are a lot of assumptions here but considering that uh you carry the yield throughout the year and then at the end of the year one of the protocols getting hacked you're at least kind of coming to a zero level so you didn't lose anything but then as I said there are a lot of like assumptions and then you can step up your game and try to analyze probabilities so you would try to predict um you know like the frequency of

exploits what's the probability of a single protocol getting exploited and then based off of that you can model out your portfolio and like different exposures and then see if you're actually generating you know API that can most likely make you money but it's all probabilistic in statistical and that's basically kind of like the challenge that the whole default Insurance space takes on right so instead of trying to do all that modeling all that statistical analysis yourself uh you can ask the smart guys and buy a product for them that sounds very easy and simple but of course the Smart Guys on the back end deal with a lot of challenges um with that and so by using a product

you're virtually trusting um that someone did risk management someone did statistics and came up with those like numbers that's sort of like one thing so you go to the conventional coverage policy you check availability right because not everything is insured or offered or there might be not enough liquidity to cover you so then you choose the cover amount duration and pay the premium those products are very similar to what you get for insurance like on the car or something else at least like the ux tries to mimic that and then

there's events of a loss you get hacked so pretty much like all those products follow um the similar pathway you submit your claim somewhere to someone and then you have other people vote whether like token holders or some committee um or somebody came up with like a trigger ahead of time to State you know like what's the what's the case but then regardless um you don't really like deterministically get the payout if you suffered the loss by determinism I mean like you know same input carry the same output so you kind of like know what the outcome is um so with that whole process there are common denominators as you notice like

there is probabilistic underwriting so when you like check availability you see how the liquidity spread across different products and then all the Define insurance protocols they kind of generally mimic the insurance companies and like in traditional Finance or um like just like any traditional industry for insurance basically recreating that insurance company like on chain and trying to like decentralize or like automate certain parts and components so you have probabilistic underwriting because um no one decided to like back the risks like one to one certain people tried but it didn't really like scale or work and then another problem with that is discretionary voting so you are at the

mercy of some token holders voting and usually here you have kind of like incentives clashes uh because people voting are the ones that are going to pay you out you know I don't want to pay you out you know so I would vote against you um and then that's kind of like a problem in the theme that has repeated of course like as a startups we all want to do our best and you know be like worthy of our users and the customers but then you still run into a lot of these like gray areas because even though everyone individually is nice but then there can be like conflicts all right so

the philosophy of D5 versus tradify in you know like what most protocols like did well and scaled is like in D5 you have trustless protocols you have deterministic outcome right so you're not really dependent on other people's opinions and that's what we take pride in so then let's look at an established vertical of lending in traffic it's under collateralized or uncollaterized all the traffic people say oh how efficient it is and like great and but it's all probabilistic because basically you know you give out more money than you cover but it all works at a grant like a big scale uh with like analytics

and all that but in D5 The Lending is overcollaterized and inefficient and on paper uh you know like when I first like heard about lending in D5 like you know four five years ago I was like who the hell is going to use this because like it doesn't really like make sense like what's the point of meat like taking the credit against like my own money that I have like more but then we see that there's a lot of use cases for that and it does deliver value it's very useful and is you know attractive like a lot of liquidity so the feature that I like there is that it's deterministic um both lenders borrowers they all know

of the outcome so um it is like programmed of course like they're like Market forces and like you can get like liquidated but everything is you know set kind of like in stone um and that's really what drives a lot of like adoption and utility um and so taking that analogy onto the defy Insurance base um you know everyone went recreating traditional um insurance company basically on chain but we um at solace you know didn't really see any product I think it's like no secret that defy Insurance hasn't really you know sprouted or blossomed uh well there's like very little liquidity uh

Never Enough capacity for like the people who wanted or the products just like don't really um work well to the point that you don't want to use them you'd rather sort of like try to uh spread the liquidity yourself or just take the risk I mean like we're all here during the bear market so I think well we're pretty comfortable with the risks there in um and so we thought okay like can we make Insurance like deterministic uh and have completely trustless like no voting like no need for the claims and stuff like that and the lending was a good analogy so we thought okay lending uh you borrow a token you're basically shorting that token right in Insurance

you're shorting the operational risk so we started thinking along those lines of just like what if you could use lending and basically collateralize the loan uh with like some very risky asset and then borrow liquidity against it right it's basically like you're getting the payout um immediately at the time of you know like using the product right so it's kind of like buying car insurance but the insurer paying you the car's worth of money right away so that's kind of like the new idea that no one has like really tested um and so we're really excited about like you know seeing that hypothesis working out uh because basically you can hedge the risk uh with just like simple

lending when you collateralize uh your loan you just deposit your let's say like vulture like yarn token that you know you know they can get hacked or maybe some like yield farming like LSD token like aura or like pendals like PT tokens right like they do carry a lot of underlying risk of the protocol like if you're familiar with like Pendle uh principle tokens uh they basically carry like the fixed yield um on them and but underneath there's like Pendle there's Aura balancer and then the staking protocol itself like you know anchor for example or like rocket pool so but by using that as collateral borrowing it against it now um you can do different things so that's

basically kind of like the pathway um how you get there you get let's say like Ethan to frax you get the the fraxis uh use up into like pendants you get like the BT token now you carry like the stack of four protocols risks um you know like that's a lot a lot of people use that and you know of course like it's all about lsd5 now like the narrative is like really strong um but yeah they're a lot of risks so if you can use that as collateral and borrow ether against it uh now you have two options uh you can actually you know go more d-gen um you know like lever up or just invest in like a Json strategy um you can also like that adjacent strategy can be lower on the risk

profile and so basically you are hedging like some of the risk and uh you can do a lot cooler things but at the same time you know if you just want to sit with the eth in your wallet I mean it's probably going to be very expensive and not really attractive so you do want to generate some yield on the side through like a safe environment if possible but that kind of like opens the door for doing ensure what insurance does but without any uncertainty like probabilities and so the people who underwrite those pools right so if you just like set um a single pool of like two tokens one collateral and one Supply and borrow uh then it doesn't carry any risk of like

you know these like bigger Landing markets so like for example on Ave or compound the risk is pulled right all the tokens in the market are carrying each other's risk but if you have just like two token pools everything is spread and then people who are supplying um Eve to it they're the ones who need to think harder about this but if there's liquidity that's how you can like ensure and it is trustless it is deterministic uh you get liquidity before anything happens and so in the exploit case right let's say your um like PT token for example like lost its value due to whatever reason um then you just keep the borrowed Ethan don't

pay your loan um so that's as simple as that there's no claims you don't need to rely on anyone um and so that's uh that's what we're doing at Solace uh we're gonna launch like in August and start driving more liquidity onto like this risk on tokens um and we'll see how it plays out but I'm super excited about like this approach I don't think anyone really tested it but the beauty of it that okay uh the demanded insurance is kind of like sparse it's hard to predict it's hard to like diversify but then I believe that if the tool has multiple use cases and can also facilitate you

know degenes and like leveraged farming or moral kind of like Risk on strategies that can actually open up the door for more risk-averse users to also like partake so thank you maxi [Applause] oh question okay uh I think this is the time for questions if anyone yeah what's up I mean people who want to generate

interest I guess on their eat um yeah I mean like it's hard to predict uh but we see that generally people are comfortable um these Landing I mean like we know like some PT tokens have already been listed on other platforms and and then also like for example like stake teeth you know like have been added to a lot of the markets but it is like a risky asset it's not just like the token so for example like eth or like rap teeth or like any tokens like they carry just the market risk but no sort of like technological risk on the protocol of the chain uh but then pretty much like every other asset does and so

um I think I know like starting from like the most popular like there will be demand of course like the interest will be higher um in the economics of it um you know like we'll see but it is definitely should like generate higher interest on Ethan like a lot of other products um so I think it should work out I mean like I I don't know the future really it's like really hard to model um but yeah we'll see I think it's like really worth uh testing out um but yeah thank you and I'll talk to

you [Music]