Weekly RiskDAO report 13 Jan ‘23

All referenced data can be accessed directly on the Gearbox dashboard by RiskDAO.

Summary

Ethereum’s Shanghai hard fork is planned to go live at the end of Q1 2023, which will enable withdrawal of staked Ethereums. Withdrawals will improve the staking experience on Ethereum and the economics of liquid staking protocols. The associated protocol tokens (eg $LIDO, $RPL) have been at the center of crypto market price action and seen significant rallies.

ETH withdrawals from the staking contract will enable price arbitrage between Liquid Staking Derivatives (LSDs) and ETH, thus opening up a substantial market opportunity for lending protocols as traders will look to borrow tokens for these arbitrage trades.

Arbitrum-based, perpetual swap protocol Mycelium got exploited as the ETH-USD price feed went down.

Alameda liquidators got rekt on Aave: Borrow positions worth $72k got liquidated.

Moonwell users suffered $800k in unjustified liquidations as a parameter change (“collateral factor”) by Gauntlet included incorrect decimals. This mistakenly set the “collateral factor” almost to 0 (6400 instead of 0.64e18). The collateral factor determines if a position is eligible for liquidation. A collateral factor near 0 caused the protocol to allow users to be liquidated, who otherwise had well funded accounts, because it effectively discounted the borrowing power enabled by USDC collateral deposits by over 99.9%.

A reimbursement plan is currently discussed, which would include a combination of funds from the reserve fund (which takes 30% of all liquidation profits) and the remainder being paid by Gauntlet (equivalent to the liquidators’ profit).

What’s interesting about the Gauntlet incident: Gauntlet’s approach utilizes a simulation environment for all parameter changes. Yet, in this case the simulation did not catch the mistake as the collateral factors were inserted manually.

Stablecoin monitoring

GUSD liquidity made a noticeable jump from $16m to $27m. sUSD & LUSD stayed relatively stable at $29m and $56m, respectively. sUSD & LUSD backing both improved from 470% to 520% and 240% to 270%, respectively.

DEX liquidity for all other stablecoins remained largely unchanged.

CEX prices for GUSD differ materially from DEX prices (-7.9%). However, CEX liquidity for GUSD remains low.

Pools summary (weekly comparison)

DAI

Total pool size amounts to $18.9m, largely unchanged from the previous week. stkcvxgusd3CRV increased by $0.7m to $3.85m and the amount of yvDAI reduced by $0.7m to $2.3m.

Total debt also remains steady at $15.8m.

The pool remains predominantly collateralized by stablecoins.

USDC

Total pool collateral increased by $2.3m to $31m, with the largest collateral assets being stkcvxcrvPlain3andSUSD ($17.1m) & stkcvxgusd3CRV ($4.9m), followed by stkcvxLUSD3CRV-f ($2.9m).

This pool also remains large backed by stablecoin assets.

WETH

The pool experienced an increase from $23.6m to $26.1m. The largest collateral assets are stETH ($14.2m), stkcvxsteCRV ($6.2M) and yvWETH ($4.2m).

The pool remains collateralized by mostly ETH or staked ETH assets.

CEX prices for stETH deviate sharply from DEX prices with a difference of -21%. stETH however is a typical DEX token with very little CEX liquidity.

WBTC

No material change compared to the previous week. Total collateral pool size stands at $0.12m and is dominated by WBTC ($64k).

wstETH

The existing debt positions have been repaid and there are no credit accounts outstanding.

The existing credit accounts were likely closed due to the rate differential between the native staked ETH yield (4.8%) and the WETH borrow rate (2.9%): Traders rather borrow WETH at 2.9% instead of staked ETH for which the cost of capital is a function of the 4.8% staking yield and borrow rate.

Summary

Current risk parameters are on par with our models.

We continue to closely monitor stablecoin liquidity.

Additional stats, updated daily, are available in our dashboard at https://gearbox.riskdao.org/