# Most oracle tokens are useless, despite providing critical blockchain infrastruc - Ugur Mersinlioglu

- Channel: [ETH Warsaw](https://streameth.org/eth-warsaw)
- Date: 2025-10-07
- Duration: 26:57
- Watch: https://streameth.org/watch/yt-e9RfpyN-6pc
- YouTube: https://www.youtube.com/watch?v=e9RfpyN-6pc

## Description

Everybody knows about oracles and the service they provide. If you don't know about them initially, most of the times you will learn through some type of exploit that caused people to lose significant money, like in the case with Avi & Mango Markets. But despite providing one of the most critical pieces of infrastructure to blockchains, oracle tokens seem to heavily underperform. In this presentation, i want to look into what causes this discrepancy, the typical oracle business model, and why a revolution is needed in order to revitalise the space.


🧜🏻‍♀️ ETHWarsaw is a series of educational and entertaining events for an active community of blockchain builders, developers and enthusiasts with focus on Ethereum-related tech. Once a year, we organize a large conference and hackathon for the community in the center of the Polish capital with speakers from the best web3 projects and participants from all over the world.

Follow ETHWarsaw on social media for the latest updates!
X (Twitter): https://twitter.com/ETHWarsaw
LinkedIn: https://www.linkedin.com/company/ethwarsaw
Telegram chat: https://t.me/joinethwarsaw
See you all at our events in Warsaw 🙌🏻

## Transcript

welcome back to another exciting talk here we're going to be talking about most Oracle tokens are useless despite providing critical infrastructure uh and blockchain infrastructure and why and now I'll be giving you to ogre and he will take over perfect thanks for the intro uh it's a small room so or like a compact room I like to have this interactive so if you have a question just immediately shoot I'm happy to answer it right in the spot because sometimes you forget it towards the end uh so I'm might break some hearts today I'm already sorry for that uh but I'm essentially going to be talking about what was already introduced uh why most Oracle tokens are useless despite providing one of the most critical pieces of infrastructure in the space and uh before I start uh that which oracles do people here know if you want to just shout a name I'm fine I guess we have to pth Chronicle what else any anything else yeah great uh that's a lot what would you say the the success of such a project okay interesting any other guesses what defines the success for an oracle project VC money VC money good guess good guess uh typically what people do is they look at how other like let's say dabs uh defi Landing dabs or something like that measure themselves and what you typically see is graphs like this that show how much value is secured by them like how much people trust them by putting their money into it and in a similar fact Session One Core metric that has established itself is uh something called TVs so how much do these oracles secure with the data they provide and if you look at the space uh for the sake of not fighting my own bags and potentially other bags I've taken all names away so you're going to be seeing charts without any association to the project uh if you look at the TVs charts just from this year uh they're pretty much going in a certain direction so you see at the beginning of the year and it went up a lot for this project again increase movement up here again uh insane movement up so the the natural thing you would expect now is it appears like people trust these oracles they the value that they secure goes up that means either the projects that they secure are super successful and they grow in tvl or it's a mass of customers coming in and driving the TVs up uh now from that you would kind of think that these projects are super successful and that should somehow translate to positive movement on the market right and uh that is essentially a question that I wanted to ask here so if TVs which is like the one number one core metric that everybody measures Oracle success at currently is the success metric and the charts that I've previously show you are up only so why do price charts look like this or like this or like this and this is essentially in the same time frame and what people typically like to tell me well this this is kind of bad because the whole Market is down these are price charts against BTC or eth so like in related in accordance to the market these tokens are basically underperforming heavily especially if you take into consideration that their usage is going up drastically so any guesses why why that is and again shout your answer if you if you have one token unlocks token unlocks that's a good one yeah but uh token unlocks you also have other projects that have these but they still go up if that makes sense any other guesses the token has no actual intrinsic value within the protocol itself yes and uh that is a good one and uh essentially you would expect that the token is somehow used in paying for the service right but one of the core problems in the Oracle space and I can go into why that is is that providing data generates no money absolutely zero money I know that a lot of people think you you're an oracle you provide a service which is data and then you get money for it well you don't uh from the very early days oracles have some been something that has been subsidized heavily uh by projects like chain link to get a lot of users and you have now a lot of dabs that have gotten used to this state that they never pay for this service so when your new Oracle project coming into the space what you generally do is you provide the service for free in order to gain adoption because you cannot charge for something that the industry leader is giving away for free what people typically tell me is well this is untrue I've seen chain link Forum posts where the chains are paying a certain amount and that is essentially a myth that I want to bust here yes you have chains sometimes paying the fees to get oracles onto your network however those fees are basically to cover the gas costs and sometimes to cover the node operators in no way shape or form does that money even in a smallest bit justify the multi-million dollar valuations of these projects because the money doesn't flow to the Token in any way it is there to operate on train so they make no money price goes down so what is the what is the Playbook here and there are essentially three steps to currently run a successful Oracle project uh step number one is you climb the defi Lama ladder defi Lama has a beautiful TVs chart that shows you how much is each Oracle securing and typically what happens nowadays is you have these Oracle projects competing for customers that are not paying and they're offering them bribes in the form of their own token to get them over and this like the the amounts that are being paid nowadays are literally getting ridiculous just to get like a big customer over because you want to clim that ladder you want to show people how much you're securing after you've accomplished that you Market heavily how much you're securing and while you do that you pray for green candles like you pray very heavily for green candles because there is no mechanism to actually get more money in you're basically hoping that this metric makes people think you're used and your like your token is meaningful and then you do what and Oracle projects are not the only ones uh essentially guilty of this it's kind of all around the space but what you do next is you sell your token you you need to somehow pay all of your employees you need to justify all of the expenses so you have something that has no intrinsic value people think it has value but you just dump it relentlessly on the people and again uh this is not magic I can show you price gra graphs from several years ago and there's a cur current Trend that shows and um unfortunately this is essentially the current meta and the Oracle space just to summarize again clim the ladder bribe people if you need to Market heavily how much you're securing sell sell sell sell sell and essentially this is uh a raise to zero and people would think well this is not an issue because at the end of the day somebody is going to come out a winner right like all of the projects are going to go bust and the last one standing is winning right yeah you would think that but the problem is you're not only competing against current Oracle projects you're competing against anyone in the future that is going to raise money and create another useless token to sell and compete with you so you cannot actually win this game because everybody is just going to zero so you need to do something else than provide data it's pretty obvious like providing data generates zero money sorry need water and there are multiple ways to do this I'm not going to say there is the only way that API 3 has decided on the space is intelligent enough to come up with numerous other ways uh we personally have decided to concentrate on something called Oracle extractable value and uh it's a subset of meev that is related to Oracle transactions and I'll try to summarize that in a very non-technical way but I have to be a bit technical if you have questions again for the newcomers just scream them in I take them immediately not at the end or also at the end so essentially Oracle extractable value describes the value that can be extracted after an oracle updates or if an Oracle doesn't update when it should have and generally when it comes to oracles Oracle updates are very important but not each update has the same intrinsic value sometimes an update allows for a lot more opportunity than others and I'm going to show this on lending protocols so every one of you probably has used AA or compound or numerous of their Forks in one way or another you deposit some collateral you can borrow a certain amount after that and what defines what your collateral is worth is typically an oracle so it tells you you put an eth in I think we're at 2,400 right now and then you're able to take let's say a $1,200 Loan in usdc after you take out a loan you're typically presented with a health Factor as long as that Health factor is above one you're all good your loan is fine the moment your collateral drops and your health Factor goes below one your going to face something called the liquidation typically Landing protocols don't liquidate themselves what they do is they offer a certain percentage of your collateral as incentive for others to come in and essentially liquidate you and that can range I have written down here 5 to 10% I've actually Dove a bit deeper and saw some insane numbers like 25% so when you get liquidated somebody gets a 25% discount on your collateral to liquidate you and how this typically goes is that this entire flow is started by first of all the Oracle updates after the Oracle has updated your health Factor might drop below one what then happens is that anybody can come in and typically a lot of more sophisticated players Searchers M Searchers in the space see that you're now liquidable and right now here I'm going to focus on eth mainnet what happens on mainnet then is you have these people compete in Block space auctions for your liquidations so essentially they're bribing block builders in order to have their transaction included first and at the end of the day you get liquidated the Searcher gets the money and minus whatever he pays to the validated auction now the interesting part in this entire scenario is how much are Searchers earning when they do this and there's a pretty nice report from a company called monoceros the book is called uh the M book and they have a line in there that focuses on liquidations which says liquidations are highly competitive as the strategy is simple to perform this means that it is common to see liquidations paying up to 99.9% of their revenue to the validator a lot of talk let's see what that means here on the right side you see the biggest Liquidation on AA this year for $10.4 million roughly a month ago actually exactly a month ago uh the reward for this liquidation was $82,400 the profit that the Searcher took home is $87 because he bribed 82,83 so uh just to illustrate that this number is not a madeup number these people are giving up 99.9% of what they're being offered coming back to the sentence that I had before Oracle update are important but that not each update is equal in value the problem with oracles nowadays is that they don't look at what their updates enable they just update blindly they don't care what happens on train afterwards even if that update would enable such a liquidation like I've shown before so this current status quo just to summarize uh is kind of insane that on a $10 million liquidation you give 800k in incentives it is a especially insane if you see that the people liquidating give 99% of that away and it is even more insane when you look at the fact that oracles technically hold all the power to not allow this so what we thought about is why don't we take the auctions that happen on the box Bas level and conduct the auctions at the Oracle level so instead of blindly updating we take the competition that happens and make people fight for the data that that even allows for the liquidation and that is in effect what we build uh we have a solution which is called the oev network it is a layer two solution that essentially sells Oracle data to the highest bidder it operates on our data feeds so imagine the bottom side here you have a bunch of providers like Co geeko and others and they maintain the data feed on a chain like mantle eusd on mantle now the important thing here is that this data is slightly delayed imagine it like a regular Highway you're slow you get to your destination but a bit slower than usual now the same Oracle nodes sell real-time data on the oev network this is what you have to imagine as your Express highway so you pay to get faster to your destination and now Searchers can effectively get there quicker if they pay for it and whatever they make here we give to the dab and take a small cut so now effectively the Searcher doesn't necessarily have to wait for these Oracle nodes to perform the update they can take the data and have a 30- second advantage and since they are the only ones effectively winning they can decide when the liquidation happens and this is now quite neat because we can essentially take the competition that happens after an oracle update and take it for the Oracle update and at the end this kind of just means that instead of this insane amount going to block Builders uh it goes to the Oracle auction host which is us and I'll get to the split that we do after um how big is the opportunity when you look at the issue uh over the past I think the data I have is from the four years or so AA compound and Venus together have lost roughly 300 million to oev uh it is kind of harsh to say lost because because technically it is not Ava paying this out it is how much users are being overcharged for using the platform and essentially our business model here is simple uh we run auctions for Oracle transactions we return 80% of whatever we make back to the dab and we take a 20% cut so now we will provide a service for every $4 that we give to the dab we make $1 they are happy they earn money we are happy because we earn money actual money not token dumps and uh that's kind of the summary I also wanted to illustrate the numbers for just this this year because talking about the last four years is kind of abstract so if you look at just 2024 from January until basically when I finished the presentation two days ago uh Venus has paid out 6 million to Liquidators AA 57 million compound 13 million even smaller dabs like banki which is like the biggest Landing protocol on avac has paid out over 1.4 million moonwell on Bas I believe uh roughly closing in on a million and this is money that these dabs can now have additionally and I unfortunately forgot to put this on the slide there was this nice nuke that we had roughly a month ago in the market uh where things went from 65k down to 55 and uh there there was a post that stunny made uh that Ava made $6 million in two days because of the liquidations that happened what nobody mentioned is that in the same two days they paid 23 million in liquidation incentives so essentially we have Landing protocols celebrating that they're making 6 million in profit while handing out four times as much in incentives to Liquidators and OE is a chance for them to recapture significantly more money like it is much more bullish for your token to say you recaptured 29 million in profits than to say well we incentivized people for 23 million and actually only made 6 million and I think that's about it so I'm uger again strategy lead at API 3 we build oracles but profitable wo that was that was a really really interesting talk we have lots of time for questions does anyone have have any questions yes so if I understood correctly the price update is predictable in the same way as in ch so if there's a change by 1% or after some period of time yes yeah we operate uh price feeds on the respective chain uh in the deviation or heartbeat basis so essentially this is configurable we have a market that you can choose you can have it at 1% or 0.25% whatever you need uh the only important factor here is that this is slightly delayed right but then if I don't pay you I could probably still predict when you are going to cost a new price yeah you can do that with almost any push Oracle like chck as well correct but the thing is uh you can get away with that yes but only if the update that might happen doesn't have any value so let's say let's say you see that in 30 seconds we're going to update because this is delayed and you know in 30 seconds our update is coming because the 1% change has been done and that update will cause a liquidation that has 100K in rewards you can wait and hope that nobody else act essentially takes the highway because they're going to have an advantage over you for 30 seconds so that currently what you described is how any Oracle up Works chain link updates or people wait for instance on train link and then after that they compete and you can do the same thing here you can like wait and hope that nobody takes the shortcut and front runs you and all they have to do is pay for it and how much they pay depends on the competition here so if you're the only person integrated here you can bid a dollar and have a 30-second advantage over everyone does that make sense yeah I'm just wondering whether it's still not more profitable to post just spun transactions in the first um instruction to check that price uh and so for many short transactions until I hit whether that going be more profitable than bitting in your ocean for for the exact time you won't get the liquidation because the person before you like if if I if you wait here and I take the shortcut I can bundle the liquidation with my Oracle update and you will get nothing because I decide when the Oracle updates sorry so you operate on on Main net no everywhere on every chain I mean almost every evm chain yeah yes so then so the chains where you don't bundle where there are no bundles there's no flash bus or anything that that still work uh yeah it would still work it's centralized sequencers so one thing that I might have forgotten to answer here is or say here is that each stab reads a specific uh through a specific proxy and we run auctions for each specific proxy so Ava and compound are reading different uh data feeds they at the bottom follow the same base data data feed that is delayed but if you win over here you win the right to update AA you don't win the right to update compound and then you essentially the auctioneer takes whatever you want and signs it so only your address on that specific chain on that specific proxy is able to update so you can't just take that message I think what you mean is why can't I just take that message once somebody has one and relay it to other chains well because like it's kind of bundled to the chain specific proxy to your specific address so you can't reuse the data and you can also only update by paying so kind of buy the Monopoly to trigger the update uh for 30 seconds yes any other questions yeah are you the only or that's providing price data uh no like PRI price data in itself yeah I mean you introdu the 30 second today but what your compe uh you can use the competitor the competitor doesn't pay you but would you then get the data on time or yeah I mean the thing is you would only be delayed if there's no intrinsic value in updating faster like if there is a liquidation uh you will rarely see a 30-second delay because people are just going to take the shortcut and update immediately so the effective uh difference delay that you might have is like we're talking two or 3 seconds here for the guy to get the data from the L2 and post it on whatever train you're on but like this is essentially worst case scenario and this is what these oracles do they maintain a price so if there's no bids or nothing the price is still maintained but in case there is an opportunity which there is a bunch uh people can always pay to get instant realtime data and that is why this also essentially says that real time data is sold here you're not bound to any you can have the newest data on any chain that you want uh however fast you want to push it there and there just one more last question in the corner can I ask two questions yes uh so the first one is uh um couldn't you just just need the layer for delaying the data from orles and auctioning them yes not provid yes Yuma does that they have a Sol if you heard of them uh they have a solution called oval uh the problem is you're hoping that the Oracle plays along Yuma announced that they are essentially delaying chain link updates and then just selling the right to unlock the chain link update for to the highest bidder and there are two issues with that the first is you can only sell after chain link has updated which means you're always waiting for chain link that is not the case here you can buy any price you want always as long as it's hit and the second thing is and the more important is you're hoping that the Oracle plays along and Yuma announced that they did this and chain link came out like two days later on the whole thing and telling anyone that integrates this that they're not liable for anything going down and almost nobody uses oval now so you're kind of trusting that the underlying Oracle is going to finance your operation basically and then you take the only profitable like the only venue for oracles to make profit you're going to build on top and I'm just going to go out on the wh and say they're going to Blacklist you okay aren any any comp that uh yeah I mean the PFF guys have also build something but maybe you can hit me up after and I don't build a OE comparison live on stage the second question would be if we wanted the uh the person being liquidated to capture the value instead of us then example you could give that back to the user and then if we want them to pay only the actual cost which is like $80 in in in your model it would be $ 16,8 yeah yeah essentially what would happen is like the the search would still take home $80 or whatever because that's how high the competition was you get back this insane amount and what you as a project then do is up to you you could give that back to the user I had several interesting conversations with this technically a good way like I prefer giving it to the user because it drops liquidation costs from 10 like 5 to 25% to practically nothing but uh one argument I had that I had with the oiler uler guys is uh people that let them M themselves get liquidated kind of endanger the The Landing protocol and that is why they need to be punished that's what he said so whatever you like essentially we just offer you a venue to get that money back what you do after that is up to you users probably will prefer a liquid like a landing protocol that penalizes them less token holders of AA or compound might prefer if that money is used to buy back their tokens and make price go up so it's up to the projects decide what they do effectively at the end it's five times optimization I yeah I mean this is the optimization that's the amount people lost just in these last eight months awesome thank you much this was an amazing talk really [Applause]