Proposal To Remove Stable Collateral From IPT Rewards

[PaperImperium]

Summary

This proposal removes vaults using wrapped stablecoins from IPT rewards.

Motivation

IPT borrower rewards are substantial, and meant to distribute protocol control to those who use the protocol. The recent introduction of yield-bearing stablecoin collateral has resulted in large amounts of CHAI entering the protocol as collateral. While this collateral is thought to be quite safe - and it is desirable for users to borrow against it - it skews where IPT rewards will go.

This is because this type of collateral it lends itself to a recursive borrowing strategy where a user could place a stable collateral in their vault, borrow USDi, convert to USDC, acquire more stable collateral, and repeat. A recursive borrowing strategy that allows high leverage to farm IPT rewards is probably not in the best interest of Interest Protocol or IPT holders.

Specification

Borrower rewards will be disabled for any vault posting collateral that consists of a $1 face value stablecoin, or yield-bearing derivative of a $1 face value stablecoin.

This exclusion from rewards applies to both present and future collaterals. At time of writing, only CHAI meets this definition. For the avoidance of doubt, the following hypothetical collaterals would also result in vault exclusion: cUSDC, USDP, GUSD-USDC Uniswap LP position, OUSG, tokenized money market funds, tokenized treasury ETFs, and any other stable-value instrument pegged to a $1 face value, whether or not it accrues yield.

This proposal would be in effect no earlier than April 30, 2023.

[PaperImperium]

Note that this proposal would not disable the use of stable collateral. It would simply mean different vaults would need to be used if a borrower wishes to both use stable collateral and collect borrowing rewards vs other collaterals.

This is because Interest Protocol uses cross-collateralization, so there is no way to simply and fairly “subtract” borrow rewards for USDi borrowed against a stable collateral if there is other collateral also inside the vault.

[nekosan]

Borrowing rate is ticking up, so for IP to underwrite anymore loans, the protocol will need more USDC deposits. So integrating with Yearn’s USDC vault is a top priority for IP.

Last time this was discussed, 2M TVL (and maybe 1M of borrowing?) was the target for IP to hit before the discussion can be taken more seriously.

And we are close-ish to that target, so I was under the impression that recursive borrowing of yield-bearing stable collateral would be a good way to get us across the finish line. So incentivizing such a behavior would not be the worst of ideas.

Now, I am all for removing incentives for recursive borrowing as it’s not the most natural of activities when incentivized. But I am curious what the impetus for putting up the proposal now is? Because this could have been posted when Chai had been first voted on a month ago.

I understand chai / cUSDC etc can be good tools to allow someone to lever up and arb rates especially if USDi borrow rate is low. But absent of a strategy to actually acquire the USDC and keep the rates stable/low, I’m not not sure how this should take precedence? Maybe the yearn integration can happen without hitting the $2m TVL? hoping you can help me see the larger picture

[PaperImperium]

@nekosan

But I am curious what the impetus for putting up the proposal now is? Because this could have been posted when Chai had been first voted on a month ago.

If I’m being honest, it’s that I didn’t think about how stable collateral would interact with the IPT rewards program when I made the CHAI proposal. Maybe others did, but I definitely didn’t.

Compound V3 has significant liquidity mining going on, and looking at how they spend all their rewards just to lock stables in place without much else is what made me scratch my head about the same thing happening on IP.

@nekosan

I understand chai / cUSDC etc can be good tools to allow someone to lever up and arb rates especially if USDi borrow rate is low. But absent of a strategy to actually acquire the USDC and keep the rates stable/low, I’m not not sure how this should take precedence?

Very good points.

CHAI does seem to have gotten IP out of the “need more borrowers” stage and back into the “need more USDC” stage. As the protocol grows, these two needs will crop up over and over. My personal thought is to adopt a barbell approach to supported assets.

Some, like fairly safe, yield-bearing stables, become attractive to arbitrage rates when IP’s are low. They are probably low risk to the protocol’s solvency, but are only attractive as a means to either 1) arbitrage rates, or 2) utilize idle yield-bearing wrappers (such as DAI in the DSR via CHAI, which has no other use at the moment)

Others are more exotic and can’t be easily found other place so when rates are high, they can still provide demand. When it first was added to the protocol, cbETH was an example of this. It carries yield, but at the time was difficult to find a market to borrow against it.

How do the IPT rewards fit into this? My own thoughts are still developing, but it seems like the “safe” trade makes less sense to incentivize if the rewards are meant to bring in stickier usage of the protocol, because recursive stable carry trades are quick to lever up and unwind. I suppose an argument could be made that it gets more USDi sloshing around the system + begins to grow the protocol surplus, neither of which are bad things.

Anyway, will wait for others to weigh in, but that’s what’s going on in my own head about this, and seeing IPTman’s post about rewards reform got me to just pull the trigger and get this up for discussion.

[nekosan]

Ya fair points. I agree it doesn’t make sense to reward vaults with low liquidation risk unless we are desperate for borrowing demand. Though it’s a bit annoying to not receive rewards if you have a stable vault open alongside a risky one. But that’s not a huge issue right now, and if people start complaining, we can revisit.