Whale Summary + Plan

Most of you probably know there is a borrower with 6m SOL collateral who is looking to borrow 150m+ in stable coins. He had his entire position on Solend, but has recently started moving some of it into Mango and other protocols. We will have some additional risks If the borrower deposits his collateral into Mango and max borrows USDC. Here are possible scenarios:

  1. Liquidators are unable to liquidate the SOL collateral fast enough before the account goes negative equity and causes losses to the insurance fund then socialized losses to USDC depositors.
  2. A decent portion of the whale’s SOL is borrowed such that liquidators cannot immediately sell the SOL deposits they take over from the whale. They can use SOL-PERP to hedge up to a point. But eventually liquidators may run out of margin capacity to continue liquidating.
  3. USDC utilization may go to 100% which will prevent people from going spot margin long and/or withdrawing their USDC deposits.

I believe (1) is our primary risk. If the scenario comes to pass, we have 10m USDC in the Mango v3 insurance fund that will be used to pay off losses. The DAO also has an additional $60m+ in treasury assets that could be used to help pay off losses or provide buffer liquidity, but that should not be expected.

I believe (2) is unlikely to be a big issue but if we see SOL borrows go up, we should start contacting market makers and liquidators to set up the infrastructure so they can source SOL from centralized exchanges. This already exists to some extent.

(3) can be compensated for with a higher max rate in the USDC pool.

While we have a decent automated solution for this kind of problem in v4, unfortunately we only have manual DAO tools available to us in v3. It’s my opinion that the current USDC rate curve does not sufficiently account for the additional risks posed by one large borrower who has a single collateral at a single liq price.

I propose increasing the USDC max rate from the existing 150% to 300% APR. The USDC pool plays an important role in Mango and we should make sure the rate curve accurately reflects prevailing risk conditions. We want to sufficiently compensate USDC depositors for the risk they’re taking and discourage borrowers from going to 100% utilization. That said, the whale is talking to Solend devs and is looking to spread out the risk across multiple platforms so I don’t think we need to overreact. Max is proposing an additional change to reduce optimal util from 70% to 20%. I’ll let Max defend his proposal.

DAO votes take 3 days so the plan is to propose the rate change now and plan to reject if the situation has changed substantially 3 days from now. I’ll also encourage market makers and liquidators to have the capital and CEX infrastructure ready in case things get out of hand. There’s a great deal of upside in governing proactively, but in a way that’s consistent with long term stability and DeFi ideals. This is what the DAO is for.

8 Likes

Agreed on 1 & 3.

wrt 2) I’d like to propose a shift in optimal util to 30% (not 20%) as well as an increase of the optimal borrow rate to 6% (approx. staking rate). This is to account for the unusual size of the largest position on mango which is rn 1.2M SOL of 1.75M SOL or around 68% of the SOL deposited.

In case we try to liquidate this position we might run out of SOL in the mango group’s vaults, liquidators would effectively exchange USDC balance for SOL liability that can’t be sold, they would need to sell off-exchange or on perps to not incur price exposure, which frankly most of them are not equipped to.

There is 460k SOL lent out rn, and i expect this ratio to increase, as rates traders flock in and borrow to supply liquidity to markets that offer more competitive rates.

4 Likes

Very prudent. Thank you for proactively taking steps to reduce risk.

I understand the problem, but I think it’s very ad hoc. If the whale withdraws do we put up another proposal to move the optimal util back? The people who borrowed SOL to do an MSOL arb or SOL-PERP funding arb will now suddenly have to pay a lot more in interest.

Also, I do think the SOL liquidations can still proceed. The max interest rate will still be high for SOL. We just have to line up liquidity providers who have the ability to hedge on CEX or who have excess SOL to deposit.

I’m amenable to raising the SOL rate a little bit or reducing the otpimal util a little bit. But this seems too much of a kneejerk reaction.

2 Likes