Leveraged DEX | Fenix Hsu, Teahouse Finance | ETHTaipei 2025

[Music] Um, hi everyone. My name is Phoenix. Um, I'm the CEO of T House. Yes, right there. I feel honored to be here and to be put together with all this awesome uh crypto brands and protocols right there.

Um actually I just uh got back got back from uh Ethereum Denver and I'm still lagging. So most of this talk will be talking about what are the product we're building and some of the language use will be part English. Okay. So without further ado so let's just started it. So, first of all, I'll be really I'm really thankful that you guys are here to listen to my small talk because my this is the most weird title on the on the agenda today is a a leverage deck.

The other people got like the path to the future, the question interability, something like that. And mine is a leverage deck. So, it's really boring. So, yeah. So without let's going to talk about it.

Um and I didn't do a slack. I didn't do the deck and so you guys have to bear with me and let's go ahead and start it. So this is my small notes. The title should be T-Rex a leverage DAX built on top of the liquidity pair. T as tea house Rex the the the dinosaur as the reactive liquidity DAX something like this and what we're building is a spot leverage DAX is a counterpart of the hyperlquid or unis swap or the other perpetual product.

So T house finance if you guys know we started by doing the strategy on top of the unis swap v3 because when unis swap v3 come out I think it's a red bull yeah when unis swap v3 come out so you can set a liquidity range between each liquidity pair and we realize that the older user cannot actually perform well on a liquidity pair so they has to be a professional strategy layer on top of the unis swap v3 so you can be able to make money from providing degr because most of them will be screwed by the IIL implement loss right anybody know what IIL means is oh yeah of course you do over there yeah and we got crushed by IO too because like we tried a lot of stuff like uh you can constantly rebal rebalance your position and your IIL become realized lost. It's not in permanent. It's permanent lost. And once you rebalance that more, you got a lot of like cost fee cost. You got a lot of like position deduction.

You got a lot of going on. So we realize that if you want to provide liquidity with a profit, you have to find a way to actually h somewhere else not on the USB v3. So we keep looking for other solutions. We try to talk with the Binance all of the centralized exchange. We try to go with other per DAX but we realize the overhead is actually too big.

So we realized that oh I use a lot of realize. So we build a product T house uh T-Rex which is the T house reactive DAX which is a smart exchange. So we want to discuss the the relationship between the liquidity pair and the DAX. What does that mean? Usually the liquidity pair on the unis swap v3 provide a price to a perpetual product.

So what's the relationship with them? First there's no relationship. You have a price oracle and you build a perpetual deck. So all your price is coming from link um from link from path. You don't need unis swap v3.

You just you build a standalone casino for your user. You are not actually in the ecosystem, but you make money. That's great. And the other thing is you use the other relationship is you use a price from the liquidity pair which is a price discovery and you use a price as the to as your to your uh perpetual product like hyperrid and GOP. But what happened recently is kind of like you realize that when a small cap token can actually uh scoop out all your work by manipulate the spot price.

Anybody know what happened in the uh hyper liquidity these days? Okay. Yeah. Exactly. So basically the merid of the unis swap is like when you want to list a token you can just create a liquidity pair and you list a token.

Anybody here want to list a token through binance? It shouldn't be here. It shouldn't be in token 2049 right? Yeah. So once you do that there's a price on the US v3 then the your perpetual deex is just using that price to fit into your your perpetual product.

So one in what happens in light uh hyper liquidity is somebody short the perpetual product on hyper liquidity and buy the spot on the liquidity pair. So actually hyper. So what do we do to prevent that problem? Actually you cannot you can overall eventually they solve it by the centralized solution. You just use the price oracle and trace back the other uh trade and everybody take a loss because you can't beat the house.

I think that's the Marie here. The mar is you cannot beat the house. If you beat the house and there's no game and there's no trade, then eventually all the DeFi collapse. But this is what we become. Do we want to become a standalone casino just a house and acquiring player to play and we win that money and to sell GOP or HLP?

Can we build another solution? try to uh solve the impossible triangle these days. There's a new imposs impossible triangle in the D5 strategies in the D5 DAX right now is a you have the leverage the huge leverage what kind of leverage you a trader can go to like that's one your one uh asset of your triangle B is the probability of your staker, your lender, your GP or HL holder and C is the safety of your protocol. Like eventually you realize that you cannot actually certif um to have these three condition all align with your alignment. So How do we do about this triangle?

Can we just take can we just compromise it a little bit a little bit to make a new product? That's what T house is doing because what's the ultimate goal of a DIY goal here? Uh what's the ultimate goal D5 goal here is for every liquidity pair there's a multiple liquidity pair of say Ethereum and USDC and you got a price from multiple of them and you generate you probably take out the top one and the bottom one and you generate your own price oracle from this multiple liquidity pair into your DAX product that's kind of the more the most defi solution to this kind of a question otherwise you going to use a centralized oracle, right? Because you never know the so-called decentralized oracle what their price are generating from. They may be from the same oranges.

You never know. So in order to do that that you use the multiple liquidity pair price to become your price oracle and you cannot because the liquidity on those pair is not that many for meme token. So you cannot require um to let your trader trade like 10x 100x of that token. probably just do like five 5x or three even 3x. So what do we do it?

We actually book the protocol called T-Rex and we help the when the player want to trade say 10x the BTC. The trader deposit $100 into our our protocol. We actually lend $900 to it and they become $1,000 and we buy spot WBDC and we keep it there. So if the PDC price go up, the profit can go to infinity. There's no counterparty over risk over there.

So if the price goes down, we want to know if the the price is going to if it hit the $100 this deposit maybe like 90 we can say 90 $80 whatever we go through a decidation process. So that's basically our protocol. So what's the marina over this pro port pro pol pro pol pro pol pro port pro pol pro pol pro pol pro pol pro port protocol is it actually is like a Lego pace Lego piece you build on top of a swap and you put on top of a lending protocol and you help facil facilitate the entire blockchain transaction there are not a standalone casino because most of the time the swap has an incentive and the lending protocol has an incentive so it actually help user to trade on your leverage decks that's Our product it's actually live today on SE uh which is a L1 SEI uh part there's a booth over there people can try to open up the position and close it up once you do that you get a little present from T house protocol. Yeah. Blockchain transaction.

So, I wouldn't say this product is more uh competitive, but it's more um blended in into the most ecosystem because all the EVM on all the L1 has its own swap and has its own lending protocol. So you build a Lego piece on top of it. Then you stop on there. That's basically what we do. And uh so why why does T House Finance start building this?

Because eventually if you think about what user wants all the traders wants to 100x something right or 100x long 100 short something but they don't need a unis swap unis swap cannot provide this feature for them swap long but why do the other but the other thing is why do unis swap still is six is if you are a protocol and you're issuing a token you have to provide the greedy pair and you have to provide a way your user can 10x long or 10x short your token that's why you need this so eventually we real te house realized that ordinary user do not use the unis swap or liquidity pair provide liquidity to the liquidity pair anymore. Like if I ask you a question, who here is excited about there's a liquidity pair APR is a,000%. may why why isn't that because we all know that I will kills you and we all know it's not going to last but why is it unis swap still exist because it provide a way to actually create you guys to list you guys token and once you list it on unis swap you still need to find a way to let your user 10x or long or 10x show your product because volatility gives your token life and gives you free and gives you the actual uh liquidity provision fees. So go back to THA that was the DAX. So we actually creating a strategy to become a more delta neutral for liquidity provision.

Yeah. So um I actually have some pictures. Yeah. So remember why we're doing this? We're doing this for a protocol because user you can just trade it everywhere, right?

So we're doing this for a protocol. So we're actually provided strategy. We can we can use the DAX to make the liquidity provision on the liquidity pair more delta neutral. And this is our result. uh and this is our result.

The result is really promising and hence we built a lot. So we want to create a world where you when you guys have a token there's a delta neutral strategy to let people provide liquidity into your tokens pair that's going to create a portal's life. It's a lifeline of your protocol because for every protocol coming to T house and ask us about how to do market making on DeFi, they only want to do one thing. How to sell our token quietly and gradually without impacting the p impacting the price. And I and I answer the answer is you have to have enough liquidity over there.

And how do you have the liquidity over there? The when you provide liquidity into it the they has to be delta neutral otherwise people are just get killed by IIO. No matter how much bonus token you give no matter how much a thousand% APRs over there people still getting killed. So that's why we developed this strategy eventually all go backs to how do we grow the entire unis swap v3 then how defy system back when people are willing to provide equity into the pair I bet right now no one's doing this But we find a way and we really want to do it to make it work. So the the final strategy um so results there's a mean reversion strategy based on the uh delta neutral strategy.

The mean reversion is basically we believe something like uh the alcoin season is going to come if you think that way. So the result is like we use uh 10K and we kind of made 30 3K something our API 53% and it casually works and we can actually do the dynamic things is like we can actually accommodate when the protocol want to sell their token pretty much is like you use a delta neutral strategy plus a dentin in your equation because the protocol want to sell on top of it, right? Then you become the it's a new way to actually do the liquidity management. Yeah, that's basically what T house is doing. But there's actually another way to solve that impossible triangle I talked before.

It's called order book. because then then you don't have that that kind of problem anymore, right? So I feel like if you want to solve this problem eventually we'll see the older boot coming back because I think right now the blockchain capability you don't need the like the gas fee on top on and you kind of need a you can actually build a AI agent to help your user to sell their token on the order book right like help me sell this token 10% before the mark above the market price within this 24 hours. You can actually build a lot of AI agents on top of it and it's easier to do it with the orderable way. But like I said, right now we're still focusing on how to make DeFi greater, which is you have to build a Lego pieces on top of each uh established layer ones, layer twos and established uh policles um to become one of them and eventually I well we will together make D5 actually meaningful.

Uh there's a five million left. I want to talk about my trip uh from the Ethereum Denver. Anybody here go to Ethereum Denver this year? How was it? Not really good, right?

Yeah, that's what people are saying. People taking um people take out their trading app and feed they think Ethereum is going down. I don't believe so. I believe most innovative thing and most uh things gonna take off is going to happen on Ethereum. That's why I still believe T house finance has been working with not with working on Ethereum EVM for like four years.

We're still here. So we believe like eventually everything will take off and the Wi-Fi password will actually make sense to everyone. Yeah. Okay. So I we hope really hope next year Ethan Danfare will be really good and follow up the Ethereum there's not going to be another Ethernet.

I don't think so. Right. But Ethan in Asia will be really good too. Thanks everyone. [Applause] Thanks, Phoenix.

Um, Super Crow to see how T House is thinking about Dexis from ecosystems angle. Next [Music]