Impermanent Loss - Angel of Death for AMMs

Transcript

hi everyone uh very excited to be here today so impermanent loss is probably one of the biggest issues in d5 right now and today i would like to deep dive a little bit more around it so notably i want to show you why it's a big issue for more widespread defy adoption i want also to give you more details about how it impacts your portfolio and how you can optimize your your farming strategies to avoid it i want to give you also more visibility on the options that you have to remove it and what are the tradeoff associated with it and the last part is i would like to

show you what a new permanent plus free world would mean for us so as for myself i'm a serial pastor i'm the co-founder of swap uh protocol we are a new amm that strongly reduces impermanent loss that uh just launched on polygon and i spent the last year basically studying impermanent laws and its darkest secrets and so that's why i want to share them with you today um so first i'm my conviction is that if that implement loss is basically a key issue for uh more widespread defy adoption why is that first impermanent loss is really scary

and acts as a deterrent for new users when they want to start using defile um amms traditional mms don't display it it's like very sneaky it's super hard to understand and it's a very bad um signal for retail traders that's the first part which is more like a kind of feeling um the the biggest part of course is that it's a huge problem for lps automated market makers like uni swap have enabled any type of investor to benefit from market making strategies which in c5 or tradfy are only for high net worth individuals that lp into uh funds like citadel securities or winter mute

however those strategies are basically not profitable and that's a big issue a study made by bancor on univ3 showed that approximately 50 percent of lps in univ3 were at a loss if you translate that into absolute capital losses it becomes crazy on just fifty percent of the pool during five months it has been estimated that 260 million dollars have been lost due to impermanent loss on univ3 so if you just round up to uh like the whole defy an am ecosystem you can take billions of dollars every year of losses

and the worst is it's not just a problem for lps it's also a problem for traders and that's why also we really decided to tackle this issue with swap basically when you think about it the trading cost is a function of two things first you have the price impact which is the depth of liquidity that you can find in the amm and second you have the trading fees so first trading fees they have to be super high in order to compensate the lps for the impermanent loss they are undertaking so trading fees are usually around 30 basis points which is huge for uh for volatile assets then the price impact basically a lot of

people are refraining from providing liquidity to amms so price impact remains a huge issue when you want to trade assets on chain so that's why overall i think that important loss is a huge adoption issue for d5 so maybe let's dive a little bit more around what is it and how it affects you concretely as an lp so that you can further optimize your farming strategies so what is the improvement loss exactly basically amm implements a kind of market making strategy which doesn't include any external information so what they do is that they determine the price of the assets just based on the internal demand that they have the more demand for an asset the more

they sell and the more the price increase the problem is that essentially it means that they are always short in terms of inventory versus their initial inventory the best asset the asset that performs the best and they are always long the worst assets the asset that performs the worse um and that's how they undergo like strong capital losses when the prices of the assets start to decorate so what are the consequences concretely maybe let's take an example try to adapt what we've seen from the previous cycle and see what it would mean in our current setup so during the previous cycle the east price approximately did like a 30x

improvement during the bottom from the the 88 right so let's suppose that is is worth one thousand dollars in our scenario if you start helping now on a traditional amm so you will get six percent of losses if the price doubles it's almost the case right now i think uh this week um you get twenty percent of losses if the price goes back to ath and then if we have this the same uh bull run as we did during the last uh like cycle essentially you will lose 65 percent of your assets 65 loss of your assets means that the amm needs to perform a 300 percent return from the fees so that's huge

um so you're going to tell me okay uh impermanent loss kind of sucks when you want to be an lp and the price rises but you're hedged when the price goes down but that's not the case unfortunately when the price goes down actually the same scenario happens because the banner set becomes the good asset so in our case when if price rises the good asset is ease and the data sets is die let's say in the in the case where the price of his decrease is is a bad asset so you accumulate ease but at a bad price um and you you like you you reduce your diet position so the

same thing happens and if uh macro is very bad then we start to have a very huge bear market and the movement is in on the other side things are going to get uh very bad uh for you um so um what are the solutions to be honest all of them have huge trade-offs actually so first one thing that you can do is you can go for assets that are like you can choose okay i don't want to help you because there are too much too many risks so you're going to have or lido for instance but risk i mean the lack of risk has a price right you get lower returns essentially the other thing that you could do would

be to lp into other types of amms like stable amms the problem is that you don't get exposure to the underlying assets so if you're bullish on ethereum uh basically you're going to get return on your stable but you're not going to get return on your eth then if you really want to lp the solutions are not so great first you can spend a lot of time trying to analyze the position uh adjust it in real time you're going to spend a lot of money on gas and you're going to take very directional bets which means that um essentially it's an active strategy and it's not fulfilling the value of being able to provide a passive market making strategy the other thing that you can do is to go into a

concentrated liquidity manager so uh a product like arrakis or gamma which basically manages your position on university for you um the main issue with those approaches is that they have not demonstrated alpha yet um then you can go for more creative solutions one of them for instance is to hedge your your like lp position into an amm so bancor used to provide that now they have uh they went back on on this you can do this also with open for instance the big issue is insurance is essentially that um you like imagine just you have a house

insurance always has a cost so you have a house and you don't want people to steal your stuff right you don't want any rubber to come in your house right so one one solution is basically to close your door close your windows when you go out put a lock on your door another solution is just to keep your doors open your door open your windows open and buy a home insurance well that's exactly the same with uh hedging your impermanent class it may solve the problem but it's super expensive and it's not the best way to solve it um so the last uh yeah another option is to go for a pmm like dodo

so the problem with solutions like dodo is that they don't actually deal with the another risk that they're creating by removing the stability of the constant formula which is called the inventory risk and which is a a very common risk in finance actually and then we have swap so basically what we do at swap is that we our mission is to enable any type of investor to benefit from passive and profitable market making strategies and um in order to do that basically we have we are building an amm which strongly reduces impermanent loss allows multi assets uh multi-asset pools like balancer which enables you to basically create

index-like positions that nets your returns and that has lower trading fees than traditional amms we are using a dual approach based on oracle-based pricing which enables us to strongly reduce the arbitrage that is faced on traditional automated market makers and um we are also using a mechanism of dynamic spread in order to ensure that we limit the risk of the trades that we are taking um i'd like maybe to step step out and zoom out a little bit and to show you why we would uh defy world without impermanent loss mean uh i think we can

maybe like there are there are like three main use cases that would be enabled first one is profitable market making strategies open for everyone i think it's uh it's something it's basically the purpose of defy to enable strategies investment possibilities that are not available in stratfi for everybody without necessarily any skills the other thing is the reduction of trading fees because you don't have to essentially pay for the implement loss of lps and it will enable us to basically um flip the centralized exchanges where most of the volume happens right now i think the

sec sex dominance is basically 90 uh today and the last part is also you have very interesting institutional use cases so for instance dows when listing their token can hurt yield on top of the liquidity that they provide instead of uh just having to pay for the impermanent loss uh they incur on traditional amms um so yeah thank you very much for your attention uh and don't hesitate to uh to yeah to come and talk with me if you're interested thank you surreal we probably have time for a question if someone's out there would like to ask a question

yes uh you said that impermanent loss was also bad for the traders and not only liquidity providers can you explain how it is bad for the treasures so essentially the problem is that fees on traditional amms have to be super high in order to uh to allow people to trade uh i mean to to compensate the lps for the impermanent loss so what what we've seen basically by studying the data is that um on uni swap basically you have a different fit here for the pools and so for instance uh for volatile assets so what i mean by volatile asset is for instance east versus uh die you you have essentially two main pools

so you have a five basis point pool where the trader pays five basis points and you have a 30 basis point pool essentially people are losing a lot of money on the five basis point pool and the 30 basis point pool is still very risky but uh more acceptable so if you want to create like sustainable lp positions on on on amms you need to have fees that are high and 30 basis point is really high for for an asset as liquid as is okay thank you i think if there's more questions we'd like to take them outside so we can stick to our schedule we have another speaker coming up thank you again syriel thank you you