xToken Strategic Update: Lending, L2 & Leverage

We’ve been discussing xToken Lending in our community for a while now. The value proposition is clear. Instead of using your vanilla DeFi tokens as collateral on alternative lending protocols where they’re likely to earn a yield somewhere in the ballpark of 0%, you should be able to deposit the xAsset version – earning 5% to 15% – as collateral and take out a loan against your yield-earning investment position. Obviously, we believe this is strictly better than earning nothing on your collateral.

There’s a small problem, though. With gas prices at current levels, the switching costs for borrowers are extremely high. Even for whales, the incremental yield may not be worth the transaction costs of unwinding from a competing lending protocol, minting and depositing the xAsset and then taking out a loan.

Given the current climate – where a simple token transfer costs $50 – we’ve made the decision to launch Lending on Arbitrum. This does not preclude a future deployment on L1 and definitely does not rule out future deployments on other L2s. We’re going to play it by ear, but for now we’ll be focusing our efforts on an Arbitrum instance. This focus on Arbitrum extends beyond our lending protocol, as we’ll discuss further below.

Why Arbitrum? At this point in time, Arbitrum seems to be the Solidity-compatible rollup chain that is closest to stability. (This could of course change in the future.) Additionally, there is an ecosystem developing on Arbitrum that fits nicely with our current and planned products. We already have a test version of xU3LP deployed on Arbitrum and we are in the early stages of building our liquidity strategies for Perpetual V2 and Futureswap V4, both on Arbitrum.

As should be clear now, we’re making a broad-based investment in Arbitrum. By the end of this year, we should have an active lending market, liquidity strategies, several composable 3x leverage tokens and our upcoming LM Terminal product all live on Arbitrum.

Back to Lending. We won’t immediately be able to list L1 xAssets on L2. L2 to L1 exit liquidity is still challenged for any asset without substantial cross-chain liquidity; however, we will be researching potential methods for approaching this. That said, we’re excited to list vanilla liquid assets like ETH, WBTC, LINK, SPELL and others initially. Then we’ll list xToken liquidity strategies like xU3LP, xPERPLP, xFSTLP and others. Then we’ll list staking strategies native to L2 and eventually those native to L1.

While empowering our users to borrow easily against their collateral is an important part of the xToken Lending mission, those who follow our project closely may recall that there is another major motivation for building our own lending protocol. Our long term vision for xToken is for a fully vertically integrated protocol, where we have internal access to the key DeFi primitives that are essential to the ecosystem.

Why do we need internal access to a lending protocol? Take our upcoming 3x tokens as an example. We want to be able to safely offer xETH3x or xSPELL3x for example (3x exposure to long ETH or long SPELL) without risk of liquidation. With our own lending protocol, we can protect our own strategies from scamwicks, centralized oracle failures (Lending uses Chainlink price feeds) and the like. We can experiment with increased leverage without dependency risk. (Clever readers may protest that this is simply transferring risk from 3x holders to lenders. Our answer is yes but no. But we’re going to have to save this discussion for an independent blog post.)

Having our own lending protocol as a playground for internal strategies comes with another benefit as well. Instead of paying out interest to external lending protocols, we can keep that value within our own ecosystem, including the slice to be earned by XTK holders.

Wen Arbitrum? Wen Lending?

We’re currently re-wiring xToken Market and xtoken-js to work across chains, while wrapping up work on our Lending UI. We’re also continuing to build out the robust infrastructure required to maintain a lending protocol. We’re targeting a full Arbitrum launch around Thanksgiving.


Great work team! Always striving for progress.

1 Like

Excited to see this live

1 Like

excellent write up

the economics and popularity of ETH L1 mixed with the technical readiness and liquidity already on and coming to arb L2 make this strategic move very compelling and close to needed, frankly

i look forward to the future discussion about how building our own lending product reduces our risk of liquidation on leveraged products while at the same time not putting lenders in more existential danger; im sure a lot of lenders/market makers would like to hear you out on that

I imagine this take won’t be popular, but I would urge everyone to consider not rolling a custom lending solution in house. The recent CREAM exploit makes clear the dangers of running a lending platform, especially one that handles tokens more complex than a basic ERC20 gov token such as yield bearing or wrapped tokens. Even AAVE is now scrambling to address the potential implications of this.

Now that Rari Capital’s permissionless Fuse pool creation is live, there are major advantages to cooperating with another entity on xToken lending and little friction to prevent this. All of the benefits mentioned to xToken lending besides avoiding the Rari platform fee can be accomplished through a Fuse pool. This platform fee is a small price to pay to reduce chances of a major exploit occurring, and participating in Fuse will likely make it easier to bootstrap liquidity in the market as opposed to an isolated offering.

Can also consider using multiple pools for risk isolation, such as a pool that only the DAO can borrow from when creating leveraged products that has special liquidation rules. The team at Rari are also working on useful tools for cross layer communication such as Nova which could be beneficial here.

This isn’t intended as a Rari shill, as I see the merits for xToken to operate a unique lending market. Of course using Fuse doesn’t preclude a problem occurring, but having less novel code and more eyes on the system can’t hurt. Just something to think about carefully as there are already a lot of moving parts in the XToken ecosystem with more to come.

Totally agreed on the choice of Arbitrum for layer 2 strategy and excited to see XToken become a major player for managed liquidity and staking there.