Coming off of the success of the initial stETH proposal (FIP-9), this proposal seeks to increase the investment to stETH by a further 10,000 ETH from the PCV Dripper (currently reserved for the 0.95 Redemption function) into stETH.
Motivation
FIP-9 was conceived as an objective by the PCV Diversification Working Group by @Bruno, @Eswak, @countvidal and @Matthew_Graham, which has been successful in creating revenue for the PCV, generating approximately 2 ETH every 2 days as of this writing.
Separately, the current of FEI remains stable at $1.00 and recent trading is mostly above the $1 mark, buoyed by the rise in ETH pricing. As such, the Redemption function has been sitting idle with over 174k ETH available for redemption to act as a virtual price floor. As the peg reweighting system and the various initiatives to seed FEI liquidity are stable and expanding, the Redemption function does not need such a large reserve of ETH to back up its purpose.
Specification
Therefore, to further leverage the assets of our PCV and to activate idle funds, this Proposal will take 10,000 ETH from the PCV Dripper and deposit into stETH. There should be minimal dev/coding work as the contracts used should be the same as FIP-9. As of this writing, there is approximately 244k ETH in the PCV. a 10k ETH deposit represents just over 4% of the PCV. It will also represent one of the largest investments into stETH by a DAO.
Please refer to the original FIP-9 proposal for an analysis of stETH and its risks and rewards.
The below poll would signal interest in this proposal, with additional no-go and options to either increase/decrease the amount to be invested.
Do you support further investment into stETH?
Yes - per current proposed specification at 10,000 ETH
From a technical perspective, how long would it take for the PCV to totally unwind from all of its stETH position in an emergency? Liquidity is one of most important aspect of PCV assets and I believe if the DAO is to pursue a larger allocation into Lido, it should be contingent on the condition that they can be liquidated in a matter of hours.
Swapping 50,000 stETH out to ETH on Curve would give 49,850 ETH today, i.e. a 0.3% loss, i.e. 3 weeks of yield farming, assuming 5% APR. I think we can size up the proposal ?
The idea of an emergency liquidation kill switch is interesting
Are you imagining this functionality for when FEI is having problems or when LIDO is having problems or when all of defi is having problems?
I think that when LIDO or defi is having major problems that curve liquidity will quickly dry up and even hours will be too long. This would be bad.
But if it’s just a situation where FEI needs to rapidly liquidate stETH then curve seems very nice. And being able to stretch this out over days or weeks would be even cheaper.
As FEI expands its use case, the collateral ratio will most likely go down. The scenario I have envisioned is where ETH suffers a sudden plunge, or a particular event triggers a massive redemption of FEI for ETH. In which case the portions of the PCV staked in Lido/DPI etc would be required to be converted to ETH almost instaneously to cover FEI’s liabilities.
Of course some other catastrophic event involving Beacon chain deployment, Lido, or the wider crypto complex can also cause the protocol to require immediate mass liquidation of non ETH assets.
This is highly unlikely in the short run, as the collateral ratio is very high, and the percent of PCV assets locked away in low liquidity assets are very low. But as we allocate more and more of the PCV to such enterprises, and FEI’s use cases goes up steadily (which can causes collateral ratio to go down, lets say another $1B of ETH is deposited and the corresponding FEI minted, the collateral ratio goes down immediately to 140%, we must remember that the current high collateral ratio windfall we enjoy is mostly a product of painful liquidations by the genesis group who have given up on FEI altogether); We must be mindful of this liquidity crush as an eventuality in the long run.
Commercial banks have suffered this exact scenario through out history. The most fitting analogy for the situation of FEI protocol would be US commercial banks in 1873; to ensure the long term health of our protocol, we must have internal provision, as no external entity exists to alleviate any liquidity crisis.
Technically speaking, a dynamic liquidation “switch” is definitely desirable in the long run. Something that can detect massive market movements and situations that threatens FEI’s liquidity, and then rapidly and automatically exits less liquidity position in favor of more liquid ones. Hopefully market conditions will allow the protocol to gracefully exit over the course of days/weeks, though preempting and designing a mechanism with such foresight is much more difficult.
There is a good point in that we shouldn’t try to concentrate our assets into stETH especially with the discussion about emergency liquidations. Also we should refer back to the PCV Guidelines by @Bruno. As such I have only recommended an additional 10k ETH for a total of 20k or 8% of PCV. I would not suggest to put all of our eggs in one basket and rather utilize other strategies for PCV growth to reduce idiosyncratic risk. Personally I would not allocate more than 30k ETH total (about 12% PCV) from a more conservative standpoint. Finally, if we have a sizeable investment could we ask for prioritized liquidity if we do need to exit the position quickly?
Another question. Do we get any LDO tokens from staking? If so what should we do with these?
I support an increased deployment of 10-20k ETH into stETH. I also think the remaining dripper ETH should be deployed to Compound or Aave as a part of this proposal
We can’t get prioritized liquidity, on a technical level we’d be swapping on Curve as the only liquidity source. I don’t think we get LDO either unless we use the stETH to provide liquidity.
We can add a Guardian-controlled PCV Controller that can immediately sell some or all stETH on Curve. @eswak wrote audited code for the Curve swap and I wrote an audited Controller that we can retool for the guardian.
The same controller can be used to automatically send ETH from any yield bearing deployment back to the stabilizer, making the dripper unnecessary. I think we should aim to fully deploy the dripper in this proposal, putting whatever ETH is left into Aave and Compound