OIP-2 Protocol owned funds + drip all yield

[DanielVF]

We can move the functionality of the dripper into the OUSD vault itself, then with only one more storage slot, have protocol owned funds. These would allow OUSD to take a portion of funds into risker strategies, increase long term yield, and have a buffer for reallocation transaction slippage.

Current Setup

The dripper funds from reward token sales are allocated in two pieces [ pending | unlocked ]. Over time, funds drip over from pending to unlocked. After a collect, this drip rate is reset to a rate that would move all funds over a one week time period.

The vault funds are also in two portions [ backing | unrebased ]. On a rebase, all 10% of un-rebased funds are given to the buyback contract, and the 90% is rebased as balance increases to all holders.

Future setup

The vault would have [ backing funds | protocol owned value | pending]. All funds would be generating yield. All could possibly be allocated into strategies. On a rebase, we would use the dripper style calculations to take a portion of the pending funds (depending on the drip rate) and split it. Perhaps 10% to the buyback, 10% to protocol owned value, and 80% to OUSD holders. Rebases can only have positive effects, as they do right now.

We would have a strategist action that could reduce the protocol owned value accounting value. This could be used as a part of funding reallocations, or compensate for slippage / losses on riskier strategies.

Effects

• Short term APY would be lower if we enabled collecting protocol owned value
• Long term APY would be higher as there's a pool of funds earning on behalf of everyone
• All yield would be smoothed including reward tokens sales, redeem fees yield events from flashloans, gifts to the contract
• Periods of extra high yield, such as liquidity campaigns could build long term APY.
• Could blunt blunt attacks that target rebasing or strategy valuation. By spreading out the apparent gain of funds over a long time period, this gives us time to react, and operates on a scale outside flashloans, and short term gains would be under the rebase fee.

[franckc]

I like the advantages this provides.

On the other hand, wouldn't that add some complexity to the vault in terms of keeping track of what funds can be allocated to strategies or for withdrawals vs funds that are part of the dripping pool?

[DanielVF]

@franckc This should only affects the interface with OUSD. We just hold back some funds from being rebased. The rest of the vault works as normal. This also means that all funds (even those held back by the built in dripper) can be at work all the time in strategies.

[franckc]

Ah got it, thanks for the extra details... Speaking of funds held back from rebase, I wonder if we should evaluate having an architecture that allows the protocol to set aside some portion of the yield to invest in risky but potentially very high return new strategies. This way we would still guarantee the stability of OUSD (since those funds are not rebased) while enabling investments that could help boost the OUSD APY.
It's probably a discussion we should have out of this OIP but I thought I would bring it up here since this may influence the way we think at the smart contract changes for this OIP.