Twyne at $0TVL Hour (Winner): Programmable Credit on Top of Existing Lending Markets

Hey everybody, big fan of Polaris. Everything that I'm going to talk about right now would also be buildable on top of them. So happy to look into that. We are building programmable credit on top of existing lending markets. I mean all lending markets are programmable but we look at the loans as objects and we try to transform them into new risk branches to just build new DeFi primitives.

This is the uh leverage profile of the A protocol. You can see that there's a $12 billion sitting at 0% LTV taking no leverage at all. And on the far right there is the emote. They are leveraging up 66 billion dollars and you can see these two clusters. Everybody gets the same access but everybody's using the product wildly different.

So what we are proposing is to create a marketplace for idle credit. You can see the collateral of the lender and its credit capacity. On the right side there's a borrower. They are approaching their depth limits. And what we built is a credit vault in between to allow the lender's credit to flow over to the credit of the borrower and the borrower can now borrow more on the AR protocol.

The whole flow also goes back with the yield and everybody just uses their capital in a more efficient way. How does it work? It is just a simple accounting trick. It is one position and we have two perspectives on the position right now. You can see, okay, you can barely see it, but there's a $80 debt position and $100 of collateral.

It's an 80% LTV and Twine gives you 90%. In the same the same position is also on is on the AV protocol, but AV sees the composite position of the creditor and the lender. In this case, they have a 50% liquidation LTV and the AVA LTV is the composite position which is $40. If the price now moves against the user, he gets liquidatable on the Twine protocol, but AVA is still a healthy position and everything is kept together by invariants that make sure that the math always holds. We even can tune the buffer between keeping the positions healthy or unhealthy or making them liquidatable at the same time.

Um yeah, I mean risks usually arise in lending markets when liquidations happen. We have two liquidation layers. The first one is a twine liquidation. If that happens, creditors are always protected. And the second liquidation is the AR liquidation kicks in.

Here the outcome depends on the efficiency of the liquidation. We have a clear loss absorption order. The borrower who's taking the leverage, they get hits first. If then any losses remain, the creditor absorbing them and a is completely segregated from all the things happening on the twine layer. And even if we have different credit vaults, different borrowers are isolated from each other.

So they don't inherit any risk from other users. We try to analyze the CLP losses. If all the liquidations would be are the fallback liquidations pushing ETH liquidational TV to 94% you'll get annualized losses of 0.16%. If you as a creditor earn more than that then it's a positive EV play for you to be a creditor and yeah I mean taking on that risk is helps a lot because a small liquidation boost goes a really long way.

You can here see if you would have an if you would boost the liquidation LTV from 95 to 98%, it's just a 3% increase, but you'll get 30 more loops out of your strategy, which is the aggressive way of using Twin Protocol, but you can also use it offensively as liquidation protection. We ran a simulation to see how liquidated user positions would have evolved over the next 30 days if they had a 10% buffer on top of what a gives them. The result is 94% of the position would have survived. Twine has this crazy efficient way of bootstrapping new markets. This is how the position would look like if you boosted it from 95 to 98.

There's $1 in credit needed which is which enables $32 in leverage borrowing capacity on the AV protocol. So all these $32 can now use all the borrowable tokens from A and they just come for free because we're building on top of them. The other nice effect is that the $32 if they would pay 1% in yield as a cost for the credit that yield would be multiplied as well which is going back to the creditor as 32% in yield. So the bootstrapping mechanism is just crazy efficient compared to setting up a new morph for market. Um what makes it also better is you can choose your liquidation LTV.

You can keep it at 95. you can go to 96, 97, 98 and everything in between. The risk and interest rates and the supply caps, they are always accounted for. So the higher the liquidation LTV you choose, the higher the risk premium is that you're paying. And on the other side, the higher the liquidation LTV is, the smaller the supply cap is that you can use.

So we naturally penalize the high risk takers while also allowing people to choose the full spectrum of the liquidation device. Let's zoom out a bit. Lending evolution. I mean, holy Um, that's one big AR pool. Everybody shares liquidity and shares the risk.

Moro uh innovated on that by segregating pools. You have stronger risk um contamination, but you fragment liquidity. And twine is just layers on top of the existing markets can be built on both of them where you can share liquidity but still isolate the risk. So where where are we now? We just launched a new PT senior USD looping vault that would be yielding 30% on a with our boost you get twice as much.

On the plot you can see where we launched the vault. Since then, the TVL 4xed in the last 10 days and yeah, still at zero TVL basically and looking to now grow these vaults and yeah, I mean QR codes if you want to follow us and yeah, see what we're up to. That's it. Thank you. If AI or Oiler came along and just copied your features and functions, what would happen to Twine or how would you keep it defensible?

I mean, yes, you can say that for any DEFA protocol, I mean they can build it themselves. Our usual like thinking about it is that you're building a network and the bigger your network is, the bigger your network effects are. So if you build one network on AA, one on Moro and one on Oiler and you start connecting the three of them, you build like a mode that a will never follow because they will not aggregate other lending markets. So our pathway of escaping that is is to be on top of all the landing markets and then be the aggregation layer for all of them. kind of like AMMs were aggregated by just routing through protocols and then the newer intentbased systems started to batch the users and just use the network effects of these users to provide a better service and we'll do the same for for lending uh okay this this uh idle borrowing capacity on a um like it it serves also like this risk buffer purpose right on all

yes

are you somehow like influencing it or removing that that function

I mean depends how you look at it I mean risk buffer in terms of the protocol yes the LP they seem to be not taking on risk because they cannot get liquidated in the case of an exploit they have all their money in the protocol so they get hit the hardest And our answer to that is if you are leaving all your money in the protocol, you should utilize all its value and get paid for it, which you're right now just like you're just wasting your capacity. So redelegating that and being properly paid for the risk that you're taking is for the LP the better answer for a I mean we we have to basically we are constrained by the logic, right? We cannot like our our composite position always follows the rules of our protocol. So if they make vaults more conservative, we basically need to follow that path. So as long as they set the parameters the way they do, our logic like always flows up and our creditors take on the additional risk like they just get hit first.

Okay, thanks. Can it overall lower the collateralization ratio of a as people use more of uh of their credit capability and then liquidation becomes more correlated basically.

I mean can it lower the collateralization ratio? No, we use the underlying collateralization ratio. What we I mean we increase the utilization of the lending market in the best case. So all this idle capacity is just you can think about it like a load balancer like there's some idle value somewhere someone else in the system needs it so just create a pathway so the credit can flow there and for liquidations I mean we built a liquidity v2 style liquidation if you are a lender with a big balance sheet with borrowing capacity you can simply absorb and inherit liquidated users positions so you're not dependent on dex liquidity and you can essentially farm the people that are taking too much leverage.

Um about the security parameters, are there curators or are users just selecting their own?

As of now, we I mean the Twine protocol is set in one vault, but I mean we can launch all the curators at any time. Like we are built on top of the oil ABC EVK stack like we get the corators for free essentially but we want to first I mean improve the use case grow some balls and then risk corators can be onboarded on top of that but the users so we set the max liquidation LTV and the users can then choose whatever liquidation LTV they want from a to the max

can can I get my question in um is this across all blockchains or just one chain Like are you dealing with fragmented liquidity or

to start with on one chain on one lending market? The grand vision is to then like be on one chain on all the lending markets and then the obvious next step is to just connect it also crosschain through all lending markets. But yeah, there needs to be some single entry point to all the lending markets that and if you are have the network that generates more value compared to the underlying, you should over time get more and more flows and just grow the overall network.