========================================================================
Reimagining the AMM with 1inch Aqua
Speaker: Tanner Moore — 1inch
ETHGlobal Tokyo 2026 · Workshop · Friday, September 25, 2026
Video: https://www.youtube.com/watch?v=q96vlWoZCd0
Transcribed with aloha by the AI ʻohana 🌺
========================================================================

[00:00] Hello everybody, let's get started.
[00:13] My name is Tanner, I am a technical account manager at 1inch, and today we're going to
[00:19] talk about reimagining the AMM with 1inch Aqua.
[00:25] So today we're going to talk about the current landscape of DEXs, what Aqua is, and how it
[00:30] reimagines AMM design, tools we have built to streamline Aqua development, extending beyond
[00:37] just AMMs with 1inch Aqua, benefits for LPs, traders, and builders, and then we'll wrap
[00:43] up with a brief touch on our 1inch bounties.
[00:47] So let's get started.
[00:49] So we're going to go over an example of a very common way we all think of AMMs today,
[00:53] with a very simple example of Bob the LP, who wants to earn yield on his Ether and USDC.
[01:01] He will find eventually some DEXs that he can deposit his funds into, we'll just say
[01:05] it's a wrapped Ether USDC pool on Uniswap V2, and he will deposit his Ether and USDC
[01:11] into it.
[01:12] The Ether and USDC pool will now hold his funds, and he'll have some sort of redemption
[01:16] to that later.
[01:18] And another person, a trader, will come in on the other side, we'll call her Alice,
[01:22] and she will have Ether, and she will find his pool and say, okay, I want to trade my
[01:26] Ether for his USDC, or the pool's USDC.
[01:30] She sends her Ether into the pool, she gets the USDC back, and she is happy.
[01:36] Bob also gets a small amount of USDC in the form of a fee, which is the whole reason
[01:40] he was depositing his funds in the first place, was to earn a yield on them.
[01:43] And Bob is happy.
[01:45] But as we all know, there are many, many more pools and many more DEXs than just that
[01:49] one in the example.
[01:50] There's Curve with its own LPs, there's SushiSwap with its own LPs, and this list would obviously
[01:54] get very long if we tried to be exhaustive.
[01:56] And there was this famous problem that happened at the beginning of the DeFi era, and that
[02:01] was, where does Alice actually go to trade these tokens?
[02:07] And as many of you know, because a lot of people have been here a while, this was solved
[02:09] a long time ago with aggregators.
[02:11] One inch is one of the leading aggregators.
[02:13] We also have our own competitors as well.
[02:15] And this solved the issue where Alice just goes to some front end, some aggregator,
[02:19] and then all of the complexity gets completely taken away, and they don't have to worry
[02:23] about it anymore.
[02:24] It's the best rate for the token they're trying to swap, and they are very happy.
[02:27] But besides the liquidity just feeling really nice to Alice in this case, the reality is
[02:36] the fragmentation actually was accelerated by this abstraction.
[02:39] Because people didn't really need to solve unifying liquidity because the aggregators
[02:42] were just a layer put on top of it, and it just kind of pushed that problem very,
[02:45] very far down the road, years down the road.
[02:49] Most of these DEXs, not all of them, but most of them, reimplemented very similar
[02:53] complexity, very similar counting, similar settlement.
[02:56] Sometimes they just cloned the contracts, made a small tweak to them, called them audited,
[03:00] and then sent them on their way.
[03:01] And this just led to a lot of hacks and other problems in the future.
[03:04] And fragmentation, as I mentioned earlier, continued to go along with it.
[03:09] So today, we're going to talk about the drawbacks of the system, not from traders'
[03:12] perspectives, but the LPs' perspectives.
[03:15] So when you're an LP, actually choosing where you want to deposit liquidity really, really
[03:20] sucks.
[03:21] You have to basically go through all of the different DEXs that you've heard of and evaluate
[03:25] the security of them, how long they've been around, what audits they have, and you have
[03:29] to try to pick the place that has the best yield with factoring all of those other
[03:32] things into account.
[03:36] The next one is LP participation has a pretty high opportunity cost.
[03:40] If Bob decides that he wants to put his funds on Uniswap, he no longer can earn
[03:44] yield on Dodo, Curve, any other DEX that's available to him.
[03:48] And if he had a pair that was like Aave Ether, he would now no longer have access to
[03:53] Dow Governance, because a lot of the older governance systems just checked the balance
[03:56] of your wallet, and that's how they gave you your weighting for if you participated
[03:59] in the governance side of things.
[04:01] The third is moving funds is just generally high friction.
[04:06] If Bob chooses to go with Uniswap, he probably wants to go to their website and validate
[04:10] that the contracts that he's interacting with are actually the ones they publish.
[04:14] He probably wants to verify the transaction does what he thinks it's going to do through
[04:19] some sort of simulator, like Tenderly, and he definitely wants to check the call data
[04:22] on his hardware wallet before he actually signs a transaction.
[04:25] Once he does all of that, it's inside of Uniswap, and if he wants to pull it out
[04:28] of Uniswap, he has to do somewhat of a reversal of everything I just said, and then
[04:32] he has to start the process all over again with the next DEX that he's going
[04:35] on to.
[04:36] A lot of verification, a lot of tedium, and you need to be very careful so you
[04:38] don't lose your funds.
[04:40] And then obviously rinse and repeat every single time that happens.
[04:44] And fourthly, new projects just continually increased this fragmentation and made it
[04:48] harder to be an LP, and they even materially incentivized it through boosted
[04:52] pools, through token emissions, all trying to fight and fragment that liquidity even
[04:58] further.
[04:59] So one inch looked at the existing system that we all know today, where there is a
[05:04] trader on one side, an LP on the other, and an AMM in the middle, and we asked how
[05:08] can we create a system that follows the mental model that everybody has of AMMs, but
[05:13] actually directly connect these traders to the LPs, both on pushing the funds to the
[05:17] LP and pulling the desired token out of the LP.
[05:22] And that is where one inch Aqua was born.
[05:24] So let's talk about one inch Aqua.
[05:27] The liquidity stays in LPs wallets.
[05:29] You no longer are depositing into this other contract where the funds kind of sit
[05:32] and then you earn fees and then eventually pull that stuff out.
[05:35] It just stays on your wallet the entire time.
[05:37] The liquidity that you do participate in Aqua with is not bound to any specific
[05:41] trading strategy.
[05:42] It actually is more committed but loosely, in the signature sense, but not actually
[05:47] physically bound to it.
[05:49] And settlement from the user's perspective stays atomic.
[05:51] So this means that people who interact with an AMM on Aqua will get a very
[05:57] similar experience to what they are used to already, where you can just go to a
[06:00] block explorer and you can see the transaction where their token leaves the
[06:03] wallet and the one they get comes back.
[06:06] So let's talk about the architecture.
[06:08] Aqua has three main components.
[06:10] We have the Aqua router, Aqua apps, and strategies.
[06:15] The Aqua router is simply an 80-line smart contract and it is responsible for
[06:19] three core functions.
[06:21] LP registrations, so what wallets are actually participating in Aqua.
[06:25] Balances, so how much of each token is committed to each Aqua strategy inside
[06:30] of the system.
[06:31] And swap settlement, so it is the thing that actually has the permissions to pull
[06:35] funds out of a wallet on behalf of an LP and give it to somebody who's
[06:39] interacting with the system.
[06:42] The next is Aqua apps, so these are the actual swap logic implementations.
[06:47] So you can think of constant product, concentrated liquidity, stable swap,
[06:51] and many more.
[06:52] These are all implemented as an Aqua app inside of the system.
[06:57] And we're going to go on to the next part, which is strategies, which is the
[07:02] usage of these Aqua apps.
[07:04] So these are the LP commitments to them.
[07:06] So if you have a constant product and you want to do it with wrapped ether
[07:09] and USDC and one ether and 1500 of USDC, that would be the actual strategy
[07:14] that is tied to that Aqua app.
[07:16] Another example would be stable swap.
[07:18] You have USDT, USDS, and 1000 of each.
[07:22] So if we go back to the example earlier where we went through kind of a very
[07:25] simple LP position and we go through it from the Aqua perspective, it will
[07:32] hopefully make more sense.
[07:33] So now, from Bob's perspective as an LP, we have between him and Alice,
[07:38] the Aqua router, and a constant product AMM Aqua app.
[07:43] The first thing that Bob will do is he will either approve or permit the
[07:46] Aqua router to access his funds, and he will call Aqua.ship.
[07:50] And this is going to have all the information about which tokens he wants
[07:53] to provide, how much of each token he wants to provide, and that will all be
[07:55] registered inside of the Aqua router's contract state.
[07:59] As part of this registration with the state, it will emit some sort of event
[08:03] which will be picked up by an indexer.
[08:05] Anybody can run an indexer, but we will obviously have our own official
[08:08] indexer, and this will also keep track of all of that state for Bob's
[08:12] registration, his token, and his balance.
[08:15] Then when a trader comes by, they can get this information from an indexer.
[08:20] So we'll just say Alice gets this information from the indexer, and she
[08:22] wants to now get a quote to see if this is where she wants to trade.
[08:25] So she will go to the Aqua AMM app, and she will pass in all that
[08:28] information that she got from the indexer about Bob.
[08:30] If she likes what she sees, she will then shove her Ether into this contract.
[08:34] It will verify Bob's commitments against the Aqua router, which is where
[08:37] he actually registered his position.
[08:39] If that comes back green, then it will push her Ether into Bob's wallet.
[08:43] It will pull the USDC out, and it will give it to Alice.
[08:46] Alice gets her tokens.
[08:47] Bob gets his fee.
[08:49] The same effective system exists, except the funds never actually left
[08:53] Bob's wallet until the moment that trade happened.
[08:56] So that's pretty complicated, but what do we actually get from it?
[09:00] Well, now LPs can send and sell committed tokens whenever they want.
[09:03] They don't have to go do the withdrawal from the contract earlier.
[09:07] They don't need to worry about permission to do that.
[09:10] They can just sell it.
[09:11] They can send it.
[09:12] They can send it to a centralized exchange.
[09:14] Whenever they want, there are absolutely no restrictions there.
[09:17] They can now participate in governance again, because voting rights that use
[09:20] your wallet balance now see the balance inside of your wallet.
[09:23] And most importantly, LPs can reuse these tokens for other Aqua apps and
[09:28] strategies, and that's a really big one which we're going to talk about next.
[09:32] So let's pretend Bob didn't just have Ether and USDC.
[09:35] Let's say he had Aave and Tether as well, all in his wallet, and
[09:38] let's say he had $1,000 of each.
[09:41] So $4,000 of total liquidity.
[09:44] Bob can go to the Aqua router, and he can do what we did before in the
[09:46] earlier example, and he can say, hey, I have $1,000 of Ether and $1,000
[09:49] of USDC, and I would like to put it into a constant product strategy.
[09:54] And so that's $2,000.
[09:55] He can do the exact same thing with the full $1,000 of Ether and $1,000
[09:59] of Aave, and he can do it a third time with the full $1,000 of Ether
[10:03] and $1,000 of Tether.
[10:05] So now, even though his actual liquidity is $4,000, he has an
[10:09] effective liquidity of $6,000.
[10:13] So on the other side, Alice can come in and she can say, hey, I
[10:16] want to swap my USDC for Ether against Bob's position.
[10:20] Trader Tony can come in with his Aave, and he can trade that for
[10:23] Ether against the same wallet, Bob's wallet.
[10:25] And a third trader, Craig, can come in with USDT or Tether.
[10:29] I keep using those interchangeably, but they're the
[10:30] same thing.
[10:31] And he can swap that for the Ether as well.
[10:34] Now, you can probably guess that if two large people came in at
[10:39] the exact same time and sucked all the Ether out of Bob's
[10:41] wallet, well, surely one of those would fail, right?
[10:44] And the answer is yes.
[10:46] I mean, he doesn't actually have more than $1,000 of
[10:48] Ether, but if one person comes in and sucks it all out and
[10:52] the other person comes in, yeah, it's going to fail, that
[10:55] probably won't happen a lot in practice because our indexer
[10:57] will check balances and protect against that.
[11:00] But occasionally, this will happen.
[11:01] But we do not think it's going to be a big enough problem to
[11:03] where it makes it a net negative for the ecosystem.
[11:06] So we talked about constant product to the Aqua app, but
[11:09] we could obviously do concentrated
[11:10] liquidity as an Aqua app.
[11:12] We could do stable swap as an Aqua app.
[11:14] There are many Aqua apps, each for the various different
[11:17] types of trading strategies.
[11:18] And as we started to think about all these Aqua apps that
[11:21] would be developed, we realized that there would be a
[11:23] lot of overlap with the custody, with the withdrawals,
[11:26] all the boilerplate that basically is not the price
[11:29] curves, and so we decided to release something at the same
[11:32] time, which we call swap VM, which itself is also an Aqua
[11:36] app, and this is going to try to extract away all of that
[11:40] boilerplate so people can simply focus on the swap
[11:42] strategies themselves.
[11:44] So with swap VM, you can think of this as just a building
[11:48] block system for swap platforms.
[11:51] So you can think of a price curve as a building block,
[11:54] maybe take your restrictions if you want to do like OTC
[11:57] trades or some kind of limit order, fees obviously because
[12:00] you want to incentivize people that are using your
[12:01] trading strategy, and maybe deadlines if you want that to
[12:04] only last a certain amount of time.
[12:07] So if we look at the left side, don't try to read all
[12:09] this code.
[12:09] This is just an example, which I will go over right now.
[12:12] If you look at the left side, you'll see a very simple
[12:15] constant product implementation inside of Aqua in
[12:19] Solidity, because Aqua apps are contracts that you
[12:21] deploy on Ethereum.
[12:22] And this would have very common things that you would
[12:24] expect like swap exact in, swap exact out, all
[12:28] implemented in the traditional way as an Aqua app on
[12:30] Solidity.
[12:31] And on the right side, you would see what it would look
[12:33] like if you did the exact same thing on swap VM.
[12:36] The first thing you'll notice besides just how small
[12:38] it is, is it is also written in JavaScript.
[12:41] And that is because you are no longer deploying smart
[12:44] contracts, you're actually building byte code for the
[12:47] swap VM, because it actually is a swap VM.
[12:50] So in here, you would just initialize the builder, you
[12:52] would say, I would like there to be a fee, I would like
[12:55] to use XYCSWAP, which is just constant product, and then
[12:58] you call build, and you get the byte code for the swap
[13:01] strategy that you're interested in actually providing
[13:03] to other people.
[13:04] So these building blocks under the hood are
[13:06] actually opcodes.
[13:09] There is one additional thing that's worth
[13:11] calling out, which are called extractions.
[13:13] So maybe your swap strategy would like something other
[13:16] than just being isolated to a single contract.
[13:19] And so an example of when you use an extraction is if you
[13:22] wanted maybe oracle data as some part of your strategy.
[13:27] One other thing that's important to call out is we
[13:29] do encourage people for this hackathon to write custom
[13:31] opcodes.
[13:32] This obviously would require you to redeploy the swap VM
[13:35] contract, that's totally OK for the hackathon.
[13:37] The one really important asterisk here is the swap VM
[13:40] contract is the max size that Ethereum allows, so you would
[13:42] have to pull out opcodes you are not using to fit the
[13:45] new ones in.
[13:46] So don't feel like this is required or anything, it's
[13:48] just encouraged if you guys want to get really creative
[13:50] and go beyond just what we're providing with swap VM today.
[13:54] Next, let's talk about how Aqua actually can extend
[13:56] beyond just AMMs.
[13:57] Everything so far that we've talked about is just AMMs, but
[13:59] it doesn't need to be that way.
[14:01] We have had our Aqua bounties present at like maybe
[14:05] four or five hackathons so far.
[14:07] And we've had people build prediction markets on Aqua,
[14:10] privacy pools, flash loans.
[14:12] We've seen a lot of really creative things, so we want to
[14:14] stress that you guys can do that as well.
[14:16] So if we go back to our earlier example of a constant
[14:18] product Aqua app, let's just shift this a little bit.
[14:21] So it becomes a flash loan Aqua app.
[14:23] So the bottom right is where the Aqua app used to be.
[14:26] We now have flash loans and we can pretend that Bob
[14:29] provided a, committed to a strategy where he provides
[14:32] USDC to that same flash loan app.
[14:34] Alice would come in, she would send the borrow amount
[14:37] and the payload for the flash loan.
[14:39] It would then go to the Aqua router and it would verify
[14:42] the commitments of Bob from earlier.
[14:45] Once that's all verified, it would send the USDC out to
[14:48] the flash loan Aqua app and it would execute that
[14:51] payload and then send the USDC back to Bob's wallet
[14:54] with some sort of fee that Alice paid for as well.
[14:59] So we've talked about.
[15:00] Flash loans briefly now we talked about constant product a lot for most of the earlier part of the presentation
[15:04] But obviously this can keep extending beyond that to maybe stable swaps with three or four coins inside of it
[15:09] And the most important part about this ecosystem is all of this can exist in parallel using the same funds from a single wallet
[15:17] So what is the impact of this on the greater DeFi market?
[15:22] Well for LPs this means that you can have a higher utilization of funds or just simply put better yields
[15:28] for DeFi users this means that there are better prices just in general for DeFi apps and
[15:34] for builders
[15:34] This means that you have a fast path to production and just users in general because as more and more people are using this
[15:41] Ecosystem for DeFi it's going to be easier to get more and more eyeballs on it to push your apps and get people to
[15:47] Use it sooner. So lastly we'll talk briefly about the bounties
[15:52] We have five thousand dollars total for building a brand new aqua app on one inch
[15:56] That's twenty five hundred first fifteen hundred for second and one thousand dollars for third
[16:00] And then our second track is extending it on an existing
[16:04] existing app to include aqua and this one is two thousand dollars total with fifteen hundred dollars for first and five hundred dollars for second
[16:11] So that's everything for me. This is a QR code that will take you straight to our bounties page
[16:16] You can easily find this on the eath global website
[16:18] But just in case you want to grab it now that'll just take you straight there and that is everything for me
[16:22] So thanks guys
[16:28] I think we might have a few minutes for questions. Yeah, so if anybody has any questions, I'm happy to answer some yes in the back
[16:45] You can interact with a deployed aqua app just as like a sub call it shouldn't be any problem
[16:52] Is that is that kind of what you're you're thinking like can you just call into it from like any from a call?
[16:56] That's already existing inside of a contract like it like a sub call inside of a call inside your contract
[17:02] Say that one more time
[17:15] You mean for the commitments like the the
[17:21] You'd yeah, I would think so. Yeah. Yeah
[17:26] Yeah, I'll try to answer that better
[17:28] Maybe you can explain to me just outside in the hall and then maybe we can talk about it more so I can get a better idea
[17:32] Of of what you're trying to build. Yeah, sorry
[17:37] Any other questions
[17:40] Yes
[17:44] Did you ask if you can do custom opcodes on soft VM? Yes, you can it is the the contract itself is already at the max size
[17:52] Allowed by aetherium
[17:52] So find opcodes that you don't want to use and pull those out and then that'll make room for the custom opcodes that you're that
[17:58] You're creating. Okay, cool. Thanks guys