Distribute 100% of yield via the Dripper

OUSD’s yield can be categorized into a handful of different types.

1. Base yield (e.g. interest from Compound or Aave, trading fees from Curve)
2. Harvested incentives (e.g. COMP, CRV, or CVX rewards)
3. Fees (e.g. 0.25% on each OUSD redemption)
4. Arbitrage (e.g. depositing USDC into a Curve pool when USDT is off-peg)

Historically, the majority of OUSD’s yield has come from farming the reward tokens in category 2. At times when OUSD’s strategy allocation is heavily weighted to Convex, for example, the base yield is very low in relation to the harvested incentives. This caused the yield to be sporadic and gave users a poor experience. An OUSD holder would observe a very low APY for several days and then suddenly receive a windfall rebase whenever someone triggered a harvest.

In February of this year, we launched the Dripper to smooth out yield generated from harvested incentives. This contract receives USDT proceeds from each harvest transaction and then unlocks them linearly over a 7-day period. The OUSD vault collects the unlocked yield periodically, and the drip rate is recalculated each time it does.

This innovative design has now been battle-tested and proven its effectiveness. We should fully utilize this feature and funnel all yield through the Dripper.

There are three main reasons to do this.

1. OUSD’s yield is still too sporadic. Just within the past month, we’ve seen two individual transactions occur outside of OUSD’s strategies and contribute substantially to the APY. We often receive questions about these spikes in the yield and they raise concerns for anyone who is skeptical about the source of our yield. In the wake of unsustainable schemes like Anchor and black-box CeFi platforms, people are rightfully concerned when they see inexplicable, unpredictable outcomes. We can smooth these out so that they blend in with OUSD’s more consistent yield sources and raise fewer red flags.

2. OUSD is safer when the yield is delayed. If an attacker were to somehow trick OUSD into thinking that it had a massive windfall yield using a flash loan, funds could be drained immediately. Spreading all yield out over a reasonably long period of time makes this sort of attack unprofitable.

3. Calculating a real-time, current yield would become feasible. We have always advertised a historical APY and struggled to consistently choose a time period that is representative of the near future. At times when the yield spikes on a biweekly basis, the 7-day trailing average is underreporting the APY. Longer trailing averages, such as the 365-day, often don’t reflect changes in market factors that drive yields up or down across the industry. Even the 30-day trailing average currently includes a one-time arbitrage event that we’re unlikely to experience again in the near future. Users want to know what yield they are likely to receive by continuing to hold OUSD. While we can’t forecast the future, we do know how much yield has accrued during the most recent Ethereum block as well as how much the Dripper is currently streaming. If all yield is smoothed out over several days, then we can much more accurately depict what the APY “is right now” and it won’t be missing recent outlier transactions.

Additionally, the drip duration is configurable. If we wanted to smooth the yield even further and enable the protocol to retain value for longer, a future proposal could extend this variable.

OUSD’s success depends on its ability to consistently deliver a secure and sustainable yield. Fully utilizing the Dripper is another step toward delivering on this promise and a continued improvement to the overall user experience.

Dripper smart contract https://etherscan.io/address/origindripper.eth#code

Contract source code https://github.com/OriginProtocol/origin-dollar/blob/master/contracts/contracts/harvest/Dripper.sol