Currently TLV for USD++ is $38K. While I expect this to grow as yields are captured, this does affect the flexibility of the strategy.
At present, PieVaults are still new, and Dev time is better utilized implementing the more direct strategies that can be broadly applied. To that aim, I would like to target single asset farms. These fall into two categories.
- Yield Farming (Internal Yield, same peg)
- Liquidity Mining (External Yield, distinct peg)
For single asset Yield Farming opportunities we have AAVE, Compound, or Cream for collateralized lending. We also have alpha homora v2* for with leveraged LPs as collateral.
- AAVE supports: DAI, USDC, TUSD, sUSD
- Compound supports: DAI, USDC
- Cream supports: DAI, USDC, sUSD (sUSD has currently 0% rate)
- Alpha supports: DAI, USDC
Due to this, I think adding support for AAVE lending makes sense as a base farming opportunity. Yields are internalized and instantly realized. There are no harvest costs. All assets are supported.
For single asset Liquidity Mining opportunities we have KeeperDAO with lending fees and Rook rewards.
- KeeperDAO supports: DAI, USDC
Liquidity mining programs do tend to have larger revenues, but rewards are not continuous. Some do not have a representative token, such as BarnBridge, which can make them harder to integrate.
Even with purely continuous returns, each asset will grow at a different rate. We will need to determine a point to return to target ratios. Rebalances, like harvests, should not occur too frequently. I believe we should set a minimum divergence to trigger a rebalance event.
The cost of a rebalance scales with the number of assets being rebalanced to. Perhaps a cyclical rebalance where assets growing too fast alternate using returns to bump up the most lagging asset. For instance if TUSD is earning 4% and DAI is earning 30%, then occasionally use DAI return to add to TUSD. However the next cycle, maybe use that to fund sUSD if its now the most behind. This doesn’t perfectly hit target, but keeps us within a threshold, and can be more affordable to do more frequently.
The issue here is frequency of harvesting returns. At current locked value and interest rates it we would earn near $700 a month in external rewards. It would not be worth the gas price price to harvest and sell back into underlying assets.
At $300k locked value, monthly revenue at 30% external rewards for the USDC and DAI portions hits $5k. I feel this would be a fair size to compound at monthly epochs. As we grow we can compress those harvest cycles, but I believe USD++ makes sense as a long term asset
We can host more frequent harvests if we allow that to supply to a single asset in the pool. If we need to supply to each asset to balance distribution, it could increase costs, and raise the floor return needed to trigger.
Framework For External Reward Viability
I think we should determine a framework for assessing when an external reward becomes viable.
We have four assets, of which two tend to have the broadest reward opportunities. Anything external is volatile. This is a stable Pie. What kind of rewards tempt us to hold volatile exposure for an epoch?
Hope the below polls will help us determine what kind of thresholds we should target.
- Yay (more ghosts)
- Nay (find other haunt)
Max Epoch Lenght Acceptable Between Harvests (weeks)
Minimum harvest target per epoch ($1000)
Minimum Yield to consider external rewards (APR %)
For information on recent resolution between Cream and Alpha on hack see here