PIP 55 - USD++ Revision To Earn Yield On Curve

I’d like to formalize a proposal for a revision to USD++ to include yield.

On principle, a USD based pie should be made up of a number of wrappers for USD to diversify issuer risk. PieDAO’s history signals it views AMMs as a risk worth taking when it comes to diversity, so long as safeguards are in place for a particular asset to collapse.

Why Curve
When it comes to farming, most opportunities are not stable-stable pairs, but stable-volatile. The single asset stable farm opportunities are rare and short lived.

The most consistent source of stable-stable farming rewards is Curve, and their like pegged pools. These pools minimize the risk of IL, while providing strong returns relative to the industry. These returns are often hidden due to the complex nature of Curve’s gauges and the need to have CRV locked as veCRV.

CRV is most effective for boosting rewards when locked for the longest period of time (4 years) and locks can be extended with each deposit. While this reduces the flexibility of using these funds for other means, I think PieDAO benefits from doing so. Curve is the go to venue for like pegged exchange, and therefore the most aligned with a pure like pegged Pie. This can also be used to bootstrap across pools for other like pegged single asset Pies like BTC++ and sETH++ types in the future.

Asset Selection
Most market products out there today farm a single strategy. However the benefits of boost can max out. This makes CRV farming most effective when rebalancing between multiple pools regularly. For smaller farmers, this is cost prohibitive. We are uniquely positioned to seize on this distinction, to maximize gains.

Since these rebalances are based on effective boost, its quite hard to calculate optimal pools for entry. This makes me think it may be best to start with several aligned projects pools to start and see what kind of size we build up, then diversify outward from there.

Here are some options:

  • Targets
    • either
      • high base yield, alignment with meta-governance of Pies.
      • high min Boost CRV output

Screenshot from 2021-01-14 20-26-26

I think to start it would be in our interest to have a fixed 4 pool strategy while liquidity is collected. From there we can assess how we can best position to max boost given the size of the pool and start adding adapters to additional pools. Ideally we would be able to support any curve pool we deem safe and be able to switch assets from pool A to B if B provides better returns.

We can incorporate staking in other protocols when appropriate rather than locking in Curve, to farm other assets while maintaining a base allocation in a Curve pool.

Harvesting
Upon harvesting Curve rewards, a portion will be kept by the DAO for Boost. The rest will be sold back into the buffer pool in the form of USDC to be used to optimize gas for entries and exits, and can be moved into the pools when appropriate.

Bootstrapping
PieDAO does not currently have a large share of CRV to utilize. We should consider purchasing CRV to stake to initialize rewards. Then allow rewards to scale over time as the DAO grows its CRV holdings with each harvest. Targeting high base reward pools reduces the initial impact of lacking Boost power.

Yearn has a Backscratcher vault which serves as a way for participants to permanently transmute CRV into yveCRV which earns a portion of rewards kept by yearn. We could do something similar. A key premise to this is then creating liquidity for those who want to swap out.

References
Boosting your crv rewards [includes rewards formula]
Choosing a curve pool
Gauge Weights
Optimizer [needs login and some navigation to an address so see option]

9 Likes

I think we should find standards for assessing how we manage rebalances.

  1. Max pool weight.
    -Sometimes a single pool may be the best return for a week, but I think we should avoid any 100% allocations.

  2. Minimum pool value, adjusted as needed.

  3. Rebalance frequency.
    -I’m thinking once a week to start

  4. Harvest frequency.
    -To simplify exits thinking daily but def up for suggestions here

The buffer pool could be replaced with an oven for entry and exit.

6 Likes