- Edited
Background:
This is a proposed configuration for the recent TAP-25 proposal to create a Cooler Loan Clearinghouse. The associated smart contracts underwent an audit with Sherlock; raised issues can be found here, and the repo can be found here.**
Proposal:
Launch a Cooler Loan Clearinghouse with no operator. Loans would be provided with the following smart-contract enforced terms:
Loans are extended in DAI to gOHM collateral
Loans are extended to any gOHM collateral
Loans are extended at a ratio of 99% of Liquid Backing per gOHM at launch. The actual number of DAI that corresponds to 99% of Liquid Backing will be updated concurrently with updates to the RBS
Loans are extended at an annualised interest rate of 33% of the then-current Dai Savings Rate (DSR). Cooler Loan Interest Rate would launch at 1.16% based on the current DSR of 3.49%.
Loans are extended for 33 weeks at maximum
Loans can be rolled over into the future with the same tenor and the updated interest rate if the DSR rate has changed, and updated loan ratio if Liquid Backing has changed
Motivations:
In a recent RFC, ohmzeus delineated the economic rationale for why Cooler Loans represent the best use of the Olympus Treasury to optimise for the protection of network value and the privatisation of risk and reward from asset deployment activities. His reasoning is sound and we have nothing further to add with respect to the underlying motivation. The Temple community wholeheartedly agrees that the Cooler Loan facility can play a key role in the future of Olympus Treasury deployment. It is therefore imperative that Cooler lending terms are configured in a way that balances the desires of the entire Olympus community. Being a key stakeholder in the future of Olympus, the Temple community would like to submit for consideration a sensible Cooler loan configuration that it would enthusiastically support.
Previous Snapshot Results Reveal Voter Divide
From the last round of voting which took a Goldilocks approach to policy, each Cooler loan parameter was put to a vote. This approach resulted in a random and somewhat incoherent set of lending parameters that was championed by no one in particular. This post draws upon the last round of votes (see results for TAP-25 and TAP-27) to find common ground between the most popular options to present a cohesive set of recommendations. We will address the rationale behind our proposed settings one-by-one.
Total Loan Capacity Should Be Driven By Market Demand
While the aim to provide full liquidity to holders is laudable, in practice Olympus would not instantly deploy anything close to the entire Treasury no matter what the vote was (~$175m in the case of full subscription). The notion that we need to provide 69 million or more idle DAI or LUSD into a lending contract immediately after approval is negligent from the perspective of both contract security and capital efficiency. Treasury allocation should not be significantly and irrevocably altered until we have data on how well Cooler loans perform. Let us take a pragmatic approach and start conservatively at 33 million DAI cap and let market demand dictate the next raise. If and when the cap is reached, Olympus will have another Snapshot vote to raise the lending capacity. Further Treasury consolidation should occur on an as-needed basis concomitant with community agreement on a Decrease in volatile holdings and an Increase in stable holdings.
Discount-To-Backing Increases Governance Attack Surface
Minimizing the discount-to-backing for OHM is crucial to reducing the “Honey Pot” threat posed by a potential governance attack alluded to by many in the community including ohmzeus. The liquid backing (LB) of gOHM currently stands at ~3014 DAI while total backing stands even higher if illiquid assets are included. At 99% of LB (or ~2984 DAI at time of writing) per gOHM, the discount to LB would shrink to 1% and reduce the sell pressure on the lower bound of RBS. While the Loan Ratio of 2850 DAI from TAP-25B seems like a more conservative number, when combined with a capacity of 69 million puts far more capital at risk while locking in a generous ~6% discount to backing. The Cooler LTV parameter should be updated periodically at the same cadence as the Range Bound Stability (RBS) parameters and reset to 99% of the then-current LB. To complement the Cooler LTV, the Lower Cushion of RBS should be set to ~100% of LB to help close the governance arb. Cooler LTV that is set to a very modest discount to LB complemented by a far more conservative initial lending cap of 33 million represents a sensible compromise between Loan Capacity and Network Security.
Cooler Interest Rate Should Be Accommodative and Adaptive
Interest rates should reflect macro conditions that affect the cost of credit. While the Olympus community seemingly agrees that we should provide loans at below-market rates to lower the cost of capital of holding gOHM, how low to set the rate remains controversial. Some favor 0.5% fixed due to the high quality gOHM collateral, while others prefer 3.3% to close the spread versus the Dai Saving Rate (DSR). Temple supports a positive spread that will react to market changes while still making Cooler the most attractive credit facility to take out a loan against gOHM collateral. By pegging the interest rate to a fixed fraction of the “risk-free” DAI yield, Cooler lending rates will guarantee best-in-market rates even when macro conditions change. Therefore we recommend setting the Cooler Interest Rate to 1.16%, or 33% of the then-current DSR rate (3.49% at time of writing). Should the DSR rate change, the Cooler Interest Rate would also be updated for new loans, or for old loans that are being rolled over. An interest rate of ~1% should stimulate lending without significantly elevating the risk of user default while ensuring solid Treasury revenue generation.
Cooler Loan Tenor Should Be Long Enough To Encourage Creative Allocation
There is likely no perfect answer for the maximum duration of a Cooler loan. Someone will always think it’s too long while another will think it’s too short. Nevertheless, a tenor of 33 weeks is the halfway point between the two most popular choices of 3 months (13 weeks) and 1 year (52 weeks) and should not arouse strong opposition from either camp. The meme-worthy marketing potential alone makes 33 weeks a strong choice.
Summary
To unite the cOHMunity and rally support from all stakeholders, we deserve ONE cohesive set of parameters that everyone can understand and consider in a simple up-or-down vote. Economic dynamics of Cooler loans are complex and we should not conflate an excess of voter choice with an abundance of foresight.
The Cooler lending facility should be attractive, available, and adaptive. More importantly, it should maximize Optionality for both Borrower and Lender. We believe that the Cooler loan configuration outlined here is bold yet cautious and fully aligned with the spirit of DeFi experimentation embodied by OlympusDAO.
Timeline
This proposal has a tentative vote date of Friday, June 30, given forum rules have been met. As with TAP-25 & TAP-27, it will run one week.
00:00 Tuesday, June 27 posted at the conclusion of the TAP-27 vote
00:00 Saturday, July 1 to snapshot
00:00 Saturday, July 8 snapshot concludes****