Continuous Clearing Auction | Dingyue Liu - Uniswap Labs

Welcome back everyone. Next up we have a talk on continuous clearing auctions. Uh and we're going to be hearing from Kite who's research scientist at UNIS Swap Labs. Please join me in welcoming Kite to the stage.

Thank you.

All right. Okay. Um hello guys. Uh my name is Kite. I'm a research scientist at UNOP labs.

Uh very happy to talk about CCA with you guys today. Um where we are exploring a new mechanism for continuous price discovery and fair onchain participations. Um so we see a lot of new and low liquidity tokens are looking to build liquidities. These is right now uh primarily occurring on centralized exchanges um with extractive fees. We often see centralized exchanges can charge upwards of up to 10% of token supply to work with them.

You often need to work with professional market makers. Initial liquidity is often provided by these sophisticated market makers uh who can dump at any time. These are usually involve some OPAC agreements um which is often most of the time unclear of the details behind the launch um biasing towards the insiders and you would you would have think that the OPAC markets with extractive middleman's issues could be solved with onchain solutions. Uh at least that's the hope. However, we do see that the shortcoming with these there's a lot of shortcomings with what we see currently on the markets of these onchain liquidity bootstrapping models.

Starting with airdrops um I I I guess the first thing when I talked about airdrops is always bot far bot farmings. um a lot of issues wi with that there's empirical academic papers that studies a lot of major airdrop events and what we see is that um more than twothirds of the tokens through airdrop is being sold immediately after claim uh which creates these immediate dumping issues and then we don't really have price discovery with airdropping events. Um moving to the next level we can have some sort of fixed price cells. The main issues with fixed price is how do you actually determine that fixed price? Um if you set the price too high, you have under subscription issues.

If you set the price too low, you have over subscription issues. Then the issue uh then the problem becomes how do you actually allocate when you have over subscriptions. Um depends on how exactly you want to run that. Um there there mine creates priority races that mine creates gas walls. Then on the next layer you might think oh we can do some sort of auctions which we can try to do this price discovery.

Um the overall issues with auctions is that you have these strategic waiting. Um the informations no one really have incentive to share informations to review informations until last minute. And then we have these like more uh cryptonative bounding curves which are um really exposed to me the exploitations which which are exploring to uh timing sensitivities. So the common thread of audi uh is that the existing mechanism that we see today really reward timing and sophistications over genuine valuations uh which creates opportunity for manipulations and we don't really see uh immediate sustainable liquidity right after um the launch. So what we imagine a better onchain approach should really prioritize fairness.

What we mean that is that people should really compete on valuations but not timing or technical sophistications. It should be transparent. We should have real-time visibility into this price discovery process. Um ideally we have credible price discovery that the final price whatever this final price we determine on chain is going to reflect actual demand without manipulations. We want it to we want there to be onchain liquidity for the secondary market.

We want it to be decentralized. uh we don't want there to be gatekeepers. So this is what we claim that UNISOP's CCA the continuous clearing auction protocol delivers. So what is CCA? If I have to explain it in one sentence I would say CCA is a continuous time uniform price auction where supply is released over time on a schedule.

At every moment we try to clear the demand against that supply at one price for everyone. um very hard to understand if within one sentence but let me give you an example. So suppose we're running an auction that say the token team is supplying 1,000 tokens over the full block auction. So this is so each block we have 250 tokens to uh available. So this is the uh supply is going to uh as a schedule part.

I think a really good mental model to have is imagining we have this swimming pool that we we we need to fill water in that we have this in whatever shape of the swimming pool we need to fill water in. Uh and then this a thousand tokens over four block just tells you what the shape of the swimming pool is. Um all right and then now we have some users coming in and then we the user says oh I have some budgets that I'm willing to allocate and I have some willingness to pay uh for for each marginal unit. And suppose we have this Alice that says I I have $800 to spend. For each tokens, I'm not going to spend more than $3 on them.

What the protocol is going to do is the protocol is going to spread across all future auction blocks. Um so the 800 is going to split into four different $200 budgets uh for all the future remainings. How do we determine how to split that is based on the um supply schedule, right? is in a supply schedule we say in the next four blocks it's equally equivalent uh for all four blocks so we split the 800 equally across the four blocks for Alice going back to the analogy of the swimming pool you can imagine if you have a bucket of water and then you want to pour into the swimming pool what happen is that the water will be level uh will be leveled throughout the swimming pool that's basically saying that we want to do the same thing once some someone comes with a bucket of water and they want to pour into the swimming swimming pool. Uh we just spread across all the future periods.

Um all right. And then so so we split uh Alice bit into four blocks. Let's zoom in into this first block. What happened is that in this f first block um Alice demand becomes I want to spend $200. So you can actually draw this demand curve for Alice.

Um and and you can specify at every single point how much does Alice want. Um you can expect that if the price is above four uh $3, Alice doesn't want anything. If the price is below, then Alice wants some allocation, some tokens. You can do this for every single participants in the in this market. Uh suppose you have a Alice and then you have a Bob.

Um then you have you can aggregate all these different demands from um each individual users. So the protocol is going to do that. The protocol is going to define Alice and Bob's ask in this block and then you can aggregate everyone's ask. Then you have your market overall demand. Um hopefully at this point it triggers some econ 101 demand and supply curve uh memories.

Um but basically we now have the demand curve. We we know how many how how much tokens people want um at any at at this block. What we also know is that we know that we have 250 tokens we want to sell in this block. So what what what can we do with this supply and demand curve is that we can find that one specific price that is going to clear all the tokens I want and then such that it really f uh it really uh fills up what people's demand is. Um then this is that $1.

4. So if we sell at $1.4 we can sell 250 tokens to the people who want who would like to have it. So the protocol is going to do this for every single block during the auction. Um so so suppose we have a five-day auction.

We're just going to have every single block we're going to try to do this exercise and then we're trying to find a price discovery uh happening. Um so we already done this uh so we launched CCA last um December and we do already see some actions. Uh for example here uh what I'm plotting on the accesses is day. Um, so we go from December 1st, 2nd, all the way to December 6th, this fiveday auctions. And on the y- axis is that clear in price.

It's that 1.4 that we found in each block. And I'm just plotting it continuously uh along each different block. So each block we try to find find a price. And then this is that entire price discovery history uh of that auction.

and we were very happy that Aztec is our um first partner and very luckily they're right right in front of us. U so so what what what we see in the Aztec CCA um is that 99% of the users contributed under 100K. So we really uh see this like wide distribution um of the tokens and then in in in total we see more than 1700 uh 17,000 different biders uh with more than 20,000 different bits across time and uh in in total at the time of the east prize they raised more than 60 million in this CCA process. Um so so so so that's what we see in in the past uh in the past December. Uh and then we've envisioned many more to come.

And then if you have any questions or if you want to learn more um there's a white paper on the website. There's also a unisoft front end available uh for you to discover u different CC auctions that are happening in real time. Um okay, thank you guys.