Maximilian Roszko, Curve Finance | Stablecoin Market Design After Launch

My name is Maximilian. I work as a business developer at Curve Finance, which is a protocol for launching on-chain markets stablecoins like kind pairs, but also crypto and FX. And on the BD side, I get a lot of questions about how to launch a market, how to launch a stablecoin pool. So, I wanted to go over the fairly basic questions we get and sort of have like a mini workshop on it. It's fairly basic.

It can be helpful for anyone who's stablecoin asset issuer, even if you're established or if you're a new one. Some tips and tricks I'll go over. So, the main questions we typically get is like how much TVL should you use or focus on like what should you pair it with, how much incentive should you do, and generally just how to grow. And for us to answer these questions as Curve Finance, we need to know what the market is for. Like, what are you actually trying to build?

If you don't know this, we would recommend that you focus on usage-first approach. Who would actual first users you trying to support? What activity are they going to do and repeat? And sort of what liquidity depth do you need to support this activity? Because you can buy TVL, you can rent it, you can buy the incentives, um but you can't like buy product-market fit.

So, you should be testing for real users as soon as possible, not overspend on incentives or market makers, instead having something that can inform your business model. So, how much TVL? It's actually you don't actually need that much TVL to support stablecoin swaps. As an example here, if you want to support a 100K swap less than half a percent of slippage, You only need like half a million of TVL. If you have a balanced pool, uh a 100k swap is going to clear that easily, uh even at pretty standard concentration levels.

And if you have even more extreme concentrations, we don't recommend it for small pools, but you can get away even with just 200k 250k worth of TVL. So, very little TVL is actually needed to support large swaps. Um And if you have a lot of smaller swaps, like it's perfectly fine. Um Here's a a pool is a very good example on gnosis as a euro pool. Uh they have pretty modest TVL.

It's just 200k uh euros to different USD stable coins. But it's it can easily do more than three times its own TVL in volume. Uh and it's very low slippage for these users. Uh it's rebalancing very good. Um this is like a dream pool.

If you have a business case that and you don't you typically don't need that much. Um Here's another example. You can have massive flow as well. Like hundreds of millions worth of uh uh volume, and you don't need more than you can have less than 10 million TVL. Um So, you don't need to aim for super high supply values.

Uh and that's going to save you a lot of money uh when bootstrapping markets. So, how do you reach the TVL target? Well, if you're in the case where you can provide your own liquidity, that's great. Uh that simplifies the equation. Um but if you need to attract external liquidity, you typically are going to have to do in some kind of incentives or having a lot of flow.

Cuz the LP return for stablecoin pools, uh LPs typically expect at least 5%. They might be a bit lower these days in a bear market, but still 5% has been like a pretty good benchmark for a long time. Um So, you have swaps, you have incentives. Swap fees are typically not going to be enough if you don't have a lot of flow. So, then you have to pay for it uh to get your TVL target.

Um so, if your target is tell me 10 million TVL, like you need to support large swap people doing 1 million trades, institutions, any like uh big players. Well, if you if you need to attract that amount, say the swap fees account for 2% of the APY, you still need to top up 3%, and 3% for 10 million, that's 300K per year you have to provide. So, that's very expensive. Um Some players still do this, uh but can get a bit extreme. PayPal, obviously, they're a huge huge institution.

So, for them it's not a problem. I don't know exactly how much incentives they're giving, um but it seems to be attracting about 50 million worth of TVL at around 5% uh LP returns. So, maybe they're paying 2 and 1/2 million per year. Uh this can also just be like a shorter campaign, but if they're going to keep doing this, this is an ongoing cost they have to keep uh keep paying. Um so, that's why flow is super important.

Um if you have uh well, so users, they they care about uh when they do a transactions, they care about the slippage and the fees. And typically, slippage is much worse than the swap fees. A lot of asset issuers, they put a very low fee uh to be competitive, but that can hurt you because um if you have low fees, then the flow will pay less to the LPs. So, then you have to attract more liquidity to lower slippage. Uh So don't recommend setting very low fees.

Low fees is usually not the problem, it's slippage. For that you need liquidity, you need depth. Uh and that you can uh subsidize by having fees slightly higher. Um it's something you need to balance when modeling this. Um flow is even more crucial for FX uh markets because if you have an FX market uh you want to keep track of an off-chain rate then you need to move the liquidity around as the off-chain rate moves.

Moving liquidity around has a cost. Uh we've done simulations. This can range anywhere from 3 to 10% per year of the pool's TVL that you would have to be paying all the time. If you have no flow then you have still have to keep paying this to sort of have a pool that runs uh smoothly, keeps track of it off-chain rate. So that can get very expensive.

But so here's again the Euro pool analysis. They have They have this Gnosis Gnosis Pay integration. So they get a lot of retail swaps, people swapping 500 euros to USD 200 like it's lot of real user swapping. This is more than enough to pay for all this rebalancing cost. It even gives an LP yield.

Uh it's In the last like 3 months it's been around 6% I believe. Uh if you have if you have an actual business case, you have corridors, people uh using your FX market then you don't have to worry about this cost. The fees are going to take care of that. Um and another thing about sort of having being able to pay off the cost of uh the TVL. Uh lot of issuers, they have this instinct we should pair with the most liquid stablecoin, uh Tether Circle, but the problem is that they will usually not provide co-incentives.

So, then you'll be able you'll be the one covering that cost, and you'll basically be paying Circle or Tether to make their asset more liquid. Um and they might be your competitors, so it's really not the greatest thing. Um a better approach is finding a partner who can provide support in reaching your TVL target. Uh that can be from some kind of underlying yield on the asset, it can be they're helping with incentives, or there's like LP demand of holding like this particular pair, or if that asset helps with flow. Um so, an example here with Frax, there can be other uh partners also doing this, but Frax is a good example because they provide yield uh on the Frax portion Frax USC portion of your pool.

So, if you have a 10 million balance pool, uh so it's 50/50, then they provide about 3.4% uh in a year, that's 170k per year that you'll get um just because you partner with them instead of USC. So, it's a big cost savings. It can significantly reduce the cost of reaching your TVL target. Uh so, yeah, I run this through quite quickly, but basically I'm at my end now.

Uh you should design markets in a way um that they're they're they're sustainable. Uh liquidity is expensive. Yes, if you have a lot of flow, uh that can help pay for it, but you should probably have some kind of problem you're solving for your users um that comes with some kind of flow that helps pay for it, and you should only be using incentives to sort of reach the TVL target if that's what you're missing. Don't just overspend on incentives to try to reach some kind of adoption because that's not going to help. Um Yes.

That's it. I'm done. Questions if anyone has any. Or we can move on. Are we good?

Okay. Yeah. Go ahead.

For the company you work for, what is the final What do you have there?

Yeah.

to this thing.

Yes, he's asking what value we bring as Curve. Curve, we're an infrastructure provider. We provide uh the contracts to launch these pools, very efficient DEX pools. Um and we do and we advise and we we also have some kind of We do have some co-incentive thing that we can do with our Curve token emissions. Um but generally, we yeah.

All of our contracts, they're integrated with all the DEX aggregators. So, if you deploy a Curve pool, uh typically, any DEX aggregator will route through that pool. So, it makes it easier uh than launching your own DEX pool. And uh yeah. Hm?

I have a question. When you increase the alpha factor won't you maybe risking some impermanent loss?

Uh when you increase the concentration?

Yeah. The the alpha factor, yeah.

Yeah. Uh So, the concentration uh parameter determines uh how the pool treats a stablecoin pool. If you have a very high concentration, yes, what can happen is that the pool gets imbalanced. You have more of one asset over the other because they're not always going to trade exactly one-to-one. So, that's why we generally recommend with conservative values in the beginning and then you can ramp it up.

Uh but for stablecoin pools, you don't you don't get impermanent loss. Uh that happens more with um more crypto markets, yeah. Yeah.

I'm an LP and I'm a human.

Yeah.

Have you seen less humans over time and more AI bots?

Um we we do extremely little little user tracking. Um so I I actually don't know. Uh we talked to a lot of projects building sort of AI um AI bots to do these things, but there's there's always been arbitrage bots ever since back in the day. People have been doing that um throughout almost since we launched, so an arbitrage is extremely common and that's that's always done by some kind of bot. Um so yeah, maybe that's increasing.

Maybe, but it's also maybe a bear market and a lot of people are just going YOLO and swapping things. Um but I can imagine more and more it's going to be more um bots. I'm not sure. Yeah. Mhm.

What is your basis for running trade

Is it the U bases is a separate uh project by our founder. Uh it uses Curve technology to do impermanent loss-free Bitcoin to USD pools. It has its own And feature It has some kind of lending mechanism and leverage to sort of help do this permanent loss thing. Yeah, they're one of our biggest users now you could say because we have a couple of 100 million TVL Bitcoin pools. We're growing the Ethereum pools as well.

So we're extremely We have one of the like deepest Bitcoin pools now thanks to you basis. Um And we're very competitive with Uniswap in that regard.