Allocate all but 25k DAI to the DSR

Summary

Move approximately 2,236,125 DAI from the Morpho Aave and Morpho Compound strategies to the DSR strategy. Leave 25,000 DAI in the Flux strategy.

Background

After the unanimous approval of the Maker DSR proposal, the strategy was deployed and a test amount of 25,000 DAI was allocated to it. After about 36 hours, it earned eight dollars. 🙂

Recommendation

The DSR is currently and will likely continue to produce higher returns than our other DAI strategies in the near term. By moving the DAI in Morpho to our DSR strategy, we will increase our APY by an estimated 30-83bp according to current rates.

Analysis

While the DSR strategy is currently earning a reported 8% APY, the current utilization (25.8%) is expected to trigger a reduction to an EDSR of 5.58%. Rune is also proposing that this rate be lowered to 5.00%, which will fall further to 4.15% once utilization reaches 35%. While we can't predict the future, it's important that we consider the sustainability of current rates so that we're not constantly reallocating to chase the higher yield.

Alternatives

Strategy	Current APY	30-day Avg APY*
Morpho Aave	3.19%	5.81%
Morpho Compound	3.40%	3.88%
Flux	3.97%	3.99%

* per DefiLlama

Current EDSR schedule

Utilization	APY
0% - 20%	8.00%
20% - 35%	5.58%
35% - 50%	4.15%
>50%	3.19%

Proposed EDSR schedule

Utilization	APY
0% - 35%	5.00%
35% - 50%	4.15%
>50%	3.19%

Why not ape in

One might argue that we should deploy all of OUSD's collateral to the DSR strategy. Here's why my proposal stops short of doing this:

1. Keeping 25k DAI in the Flux strategy allows us to verify that the strategy continues behaving as intended. Allocating these funds would not observably improve the APY and the gas cost of moving the funds would be greater than any short-term increase in yield.

2. It would likely be cost-prohibitive to swap the USDT to DAI unless the DSR premium survives for a while. Swapping 14,779,784 USDT for DAI would cost 19,536 stablecoin units. Even in the most optimistic scenario where DAI is earning double what the USDC strategies are earning (8% vs 4%), this would take nearly two weeks to break even. Due to the initial debt from swapping, OUSD would not generate any yield for at least one week.

3. Swapping USDC for DAI is much more feasible, but doing so now would incur unnecessary costs that we can avoid once collateral swaps are enabled for OUSD. A vault upgrade will soon allow us to swap collateral using 1inch. This will unlock 1:1 exchanges of DAI and USDC via the Maker PSM. Until then, we're relegated to using Curve's 3pool.

Note

Over time, the yield spread is likely to narrow across our various DAI and USDC strategies. As long as the DSR produces a considerable premium over Morpho and Flux strategies, we should lean into it due to the relatively low-risk nature of the protocol. But as rates converge, we should return to our other strategies that have uncapped potential.