• General
  • Launch an Operator-Free Cooler Loan Clearinghouse

Background: This is a follow up proposal to Create a Cooler Loan Clearing House posted January 4. 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:

  1. Loans are extended in DAI, to gOHM collateral

  2. Loans are extended to any gOHM collateral

  3. Loans are extended at a ratio of 3,000 DAI per gOHM

  4. Loans are extended at an annualised interest rate of 0.5%

  5. Loans are extended for one year at maximum

  6. Loans can be rolled over at the same terms into the future

The proposed system is permissionless, immutable, and would depend only on the treasury-level permission to access reserves. The Operator component of the previous proposal was a point of weakness, and this improves it with an open format.

Motivations: This is the best use of the Olympus treasury at this point in time when optimising for the protection of network value and the privatisation of risk and reward from asset deployment activities. The expectation is that this proposal has the following effects:

  1. Reduces OHM-collateralised debts with third parties by 40-100%, saving the network an estimated $325,000-$365,000 annually (interest currently flowing from Olympus to other protocols like Frax, Vesta) and grows the treasury by up to ~$900,000 per year on a interest-receivable basis.

  2. Provides significant liquidity (~$175m in the case of full subscription) to holders.

  3. Minimises discount-to-backing.

  4. Stems the continued contraction in supply.

Considerations: Subscription will reduce the liquidity of the treasury as raw assets are lent to gOHM holders. Activity regarding treasury management would be curtailed as a result.

Assumed State Post-Implementation: OHM is infinitely liquid relative to itself if its market value remains similar to where it has been for the past year. Holders would be able to both hedge against industry risks (another event like the USDC depeg, for example) and macro risks, and maximise asset productivity on an individual level, should they choose to do so. Supply would stabilise.

One can argue that this proposal eliminates all risk seen in the network today and allows work to proceed in a positive-or-neutral (but not negative) manner into the future. This comes at the cost of treasury liquidity, and the flexibility it provides activities today.

Next Steps:

  1. Cooler is re-audited with a new, open Clearinghouse.

  2. Volatile treasury assets are consolidated into stable coins.

  3. Loan terms are finalised.

  4. DAO plans for consolidation in scenarios like full subscription (in which treasury liquidity would fall to $10-20m).

Post-Note: The following has been carefully crafted so as not to, or to minimally, interfere with network features like cross-chain, RBS, and BLE. The motivating concerns reside primarily in forces outside of this community’s control and a negative outlook on the macro-economic environment in general, and the crypto industry in specific. This proposal is a conservative route where our community is derisked while the developments of the network (which work, for the most part) can still see their day in the sun.

On Loan Interest Rates: The thought process on the interest rate is twofold. First, the lending facility is open to all, with enough liquidity for any interested OHM holder; as a result, the higher the rate, the higher the cost essentially imposed onto yourselves. Second, the higher the interest rate, the higher the likelihood of default. 0.5% is seen as a solid middle ground in which cost is real enough to ensure non-default is purposeful, while low enough that it is not prohibitive. Little value is seen in a term structure in which most/all supply simply defaults and is destroyed upon loan expiration. That said, anywhere from 0.33% to 0.67% is reasonable and accomplishes the same result.

Where does interest go? Under the structure proposed, interest is a receivable until a loan is repaid and collateral is reclaimed, at which point it is contributed to the treasury and utilised for i.e. RBS.

Temp Check

Gm. Thanks for following up on the previous proposal and for sharing the audit details.

Removing the operator seems like it significantly reduces the complexity and overhead of the loan process. In the future if Olympus were to move away from DAI, would Cooler change what token the loans are extended in, perhaps to ETH? I know that ETH is more volatile, but it is good to be able to adjust as needed if the crypto landscape changes. How would this impact loans that are already outstanding? Just wait until they mature and then re-issue as a v2 market, or temporarily have two concurrent loan markets as the v1 market is sunset?

If there is a one year maximum but also a rollover this is essentially a perpetual loan, right? Do you envision gOHM collateral retaining voting power and if so do you see potential risks with essentially 9,9ing loans at the cost of the interest rate to obtain this governance power?

I agree that it makes sense to have users be able to take on the risk/reward. It sounds like there should be a dash showing data of total loan volume, number of participants, interest receivable, interest received, defaults, etc. Thank you for explaining the rationale behind the interest rate and where it winds up. Looking forward to the discussions around the proposed terms and next steps

    This is music to my ears - of course the reserve asset of defi should have the lowest borrowing rate!

    Interesting to see the update! And agree that a model without an operator makes more sense. Few thoughts though:

    - Considering the staking rate reduction plan I would rather see loans against OHM, especially with the proposed loan terms (at the end of which gOHM will effectively have been phased out already).

    - I understand the argumentation on the interest rate but I think undercutting the market to this extend is a bad idea for two reasons. First, if this is implemented at scale it will limit the adoption of OHM as an asset in external lending markets since it will be impossible for others to compete against this subsidized rate. Second, 0.5% is below today's baseline yield strategy for DAI (DSR @1%) while it carries significantly more risk for the treasury. Considering those two points I think it makes more sense to target a rate closer to today's market rate (4%-6%) which is less risky and carries more upside for the protocol.

    - Margin of safety: the proposal of 3,000 DAI per gOHM is actually above the liquid backing of OHM today which, at the time of writing, is $2997. If the protocol were to lend at scale at these terms I think it makes sense to lower this to $2,500/gOHM to guarantee a margin of safety - which would also be more in line with the Vendor/Myso pilot deposits. Maybe worth mentioned as well, is that there is a risk here that this programme becomes an RFV vehicle to exit above liquid backing (which itself is already higher than the market price). I don't think that the protocol should encourage that and/or open up that opportunity.

    - The sizing ask is very significant and would prevent any other allocation or strategies that the treasury team has been working on (it would also invalidate the new treasury framework). Plus it would concentrate all of the assets into a single audited, but not battle-tested, contract. I think it would be much more prudent to do a pilot deposit similar to TAP-21 and evaluate the results.

    - "Volatile treasury assets are consolidated into stable coins." This seems to go against recent community votes to favour a more volatile backing over stablecoins. Would you mind expanding on this point?

      Shpadoinkal so reviewing this I think I wouldn't want to see such a large size for this market without significant treasury growth. Basically caps any upside in OHM

      0xFelix I completely agree with all of Felix's comments here.

      Things I would like to see altered:

      • OHM as opposed to gOHM as the collateral option.

      • 0.5% interest is far too low to be worthwhile for the protocol. In my opinion, 3-4% is a great rate for both the protocol and users.

      • Capital should be spread across multiple markets as opposed to a large majority being on one platform.

      Are there also any distinctions between Cooler Loans and Vendor Finance? They seem like carbon copies of one another with Vendor obviously being launched first. I know one distinction being that Vendor take their own fee, which is paid by the borrowers and so this shouldn't deter Olympus from lending on there, does Cooler have any loans (assumption is no as nothing is mentioned in the proposal) or other features that differentiate itself from Vendor? Just seems strange to me that there will be essentially 2 of the "exact" same protocols.

      Lastly, are there any updates on UI/UX, branding of Cooler Loans etc.? Would be nice to visualize things as well.

        0xFelix to answer these:

        1. I suggest gOHM because it has a delegate() function which allows borrowers to participate in governance, and OHM does not. If gOHM is phased out it would make no difference between the two, besides the marginal gas cost of staking/unstaking.
        2. My calculations from 25 tracked protocol addresses shows ~$3,142 backing per gOHM ($188.5m / 60k circ), so it does not come out above RFV (though it is at the slightly-lower-than-reality dashboard figure). I can share these addresses if there is not a publicly posted list already (help, anyone?). Keep in mind that CVX position has unlocked and AURA position unlocks in coming weeks. My opinion is that, within some margin (i.e. 5%), the closer to backing these are, the better. A topic for debate, I suppose.
        3. I see diminished treasury management as a positive. I think that if individuals want to go out and hold defi coins and even ETH, they have every right to, but it is detrimental when done on a protocol level. I think this applies to your fourth point as well. Again, a topic for debate. When it comes to contract security (putting aside that the Cooler protocol seems extraordinarily simple), this would follow a second audit. As to the question of sizing, I struggle to see how you can gate the accessible liquidity without arbitrarily placing holders into different classes of "get loan"/"don't get loan". I think all or none should be entitled to access -- this was the issue I saw with the Operator in the initial proposal.

        hOHMwardbound new loan origination could be prevented on the DAI set by removing the treasury permission to access DAI; then a new clearinghouse could be set up lending ETH. I would see that as the way to transition from one to the other. Rollover would persist though, yes. I don't think having both concurrently would work well but to be honest, it is hard to conceptualise and I've mostly put it aside.

        Joel33 the proposal does prioritise holders over the protocol itself, but it does so knowing that it benefits all holders equally. I see this as the same dynamic in which staking emissions benefit holders more than the protocol. That is why targeting some fed-funds related rate does not make sense to me. Regarding the difference between Vendor and Cooler, I see it as having a strict loan-to-collateral versus a loan-to-value. Vendor uses an oracle to determine price, then lends a percentage of that price. Cooler lends an amount per collateral token. Vendor might be better for other tokens but in this case, with the concept of backing, I see Cooler as superior. There is also the matter of needless expense on fees, which you point out.

          Thanks for advancing this proposal @nicnombre

          I am (very) supportive in the form put forward and believe this is overdue. My main feedback is around the LTV level as outlined below.

          Fundamentally OHM exists to be a highly capital efficient defi native currency for its holders. I believe this functionality leverages the unique advantages of Olympus relative to other third party borrow / lend platforms.

          Re the interest rate: I think a nominal rate of 0.33% or 0.5% is appropriate in the context of the service provided to OHM holders and the risk profile. I believe this is FAR less risky to the protocol than depositing DAI in Maker DSR for 1%. That is because Olympus can always deal with the collateral by taking and removing from the supply. This is what makes the facility unique.

          Re gOHM: I still think gOHM is the appropriate collateral because we do not have certainty over how the base staking rate will be resolved in the future (particularly if bonding resumes in a material fashion). It is also beneficial for the protocol to have protection in this regard - i.e. the collateral is not diluted away by the staking rate.

          Re gOHM collateral value or loan LTV: I think there is some merit in having a buffer here given we are not 100% stables in the treasury and at the extreme this could effectively be a wind-down of the treasury (i.e. if all holders participated). My proposal is to set the max borrow at 85-90% of backing per gOHM. This creates value for non-participants in the event of default (essentially an inverse bond at 10% premium) and / or it creates an incentive for the borrower to repay their loan and unlock their more valuable collateral.

          Most importantly, let's press forward with this with some urgency.

          0xFelix

          The idea that anyone should pay above 1% for borrowing against the reserve currency of defi is absurd - the token is backed

          We will continue to be poorly regarded as an asset where the protocol allows our holders to be feasted upon for ridiculous borrowing premiums which form the very business model of a number of projects

          A low borrowing rate is essential to establishing OHM as a defi reserve currency - this or a similar proposal needs to be progressed urgently

          The treasury team framework will still be required even if this proposal passes in full

            Mark11 The interest rate in lending markets is not set based on whether an asset is "backed" or not. I'm just pointing out that it will not be possible for any third party to compete against this highly subsidized rate. That has nothing to do with the backing composition of OHM/gOHM, but rather with the economics of running a lending market in DeFi which has certain expenses.

            By doing this we will be massively reducing the desirability of listing OHM in any third party lending markets and that is something that should be taken into account. We can say that OHM is the "reserve currency of defi" but if it's not actually used and or listed in DeFi applications, how true is that statement?

              nicnombre Just seen the recent proposal wrt DSR rate increasing to 3.33%, which deters me even more from a 0.5% rate.

              Vendor is also loan-to-collateral via it's "Lend Ratio". This can either be determined by 25/50/75% of the current asset price or more specifically set at a # of tokens per collateral (i.e. 2,500 DAI per gOHM). See image below. They also don't charge any fees to the lender (i.e. Olympus); however, they do charge a fee to the borrower (i.e. Ohmies). These are my reasonings why I say they are carbon copies, which seems awkward imo.

                Mark11 The cost to borrow the world's reserve currency is ~ 5% and OHM as the reserve currency of defi is still only a goal. An interest rate of 0.5% exposes the protocol to too much leverage. We all remember what happened last time cheap liquidity was granted to users who do not fully understand leverage.

                I'd argue that adoption is more important than cheap liquidity for OHM to be successful and we do not help ourselves by undercutting other protocols and taking away the profitability of an OHM integration.

                  0xFelix since these have a defined loan-to-collateral, it does not mean third party loans cannot exist, it just means they need to happen at a higher loan-to-collateral. Let's use Vesta as an example -- those loans would make sense again (under proposal terms) if gOHM trades above $4500; you would now be able to borrow more than 3,000 VSTA per gOHM put down. I don't think it makes any sense to give loans and interest to third party lenders right now because the protocol takes on all the risk for them with RBS, and they just cash checks. You can make an argument for opportunity cost but all of the current options are protocols minting stablecoins to provide the loans, so that concept is kind of null.

                  Joel33 that is not really true -- according to their docs, Vendor takes 10% of interest (which is paid to the protocol) and 3% of the defaulted collateral (also paid to the protocol). Charging the additional interest fee up front does not really mean its not taken from the lender, and the default fee could be a huge cost.

                  notSHAFT I think the biggest differences between this and Rari (9,9) days is that these loans do not have liquidations and they are not written as a percentage of market value. What happened back then was: borrow -> buy -> borrow more (since price is higher) -> buy -> etc until liquidated -> sell -> liquidated more -> sell -> etc. Neither side of that recursive loop exists here.

                  hOHMwardbound considering the replies talking about timeline, could we get some discussion on dating a preliminary vote to move forward?

                    nicnombre Ah, the docs will be outdated now as they recently launched a "V2" and so those fees have been superseded. They now use a 0.3% flat fee, paid by the borrowers. So the lender (i.e. Olympus) gets 100% of both interest and any defaulted collateral. Should of clarified this in my other comments to avoid this confusion.

                    nicnombre 5 days rfc? It's a proposal with a lot of potential implications and we have the holiday weekend to contend with for discussion. Let's plan a community call, when works for you?

                      @nicnombre so does this mean that we end up being somewhat pegged to backing? Does this ultimately stop any sort of price appreciation?

                        nicnombre Thanks for the clarification. I do believe something like cooler loans is needed to create a rate arb and bring down the cost to collateralized OHM, but by going as far south as 0.5% we will nonetheless release animal spirits and undercut ourselves as a cost. That doesn't make a lot of sense to me. An interest rate of 3.3% seems more appropriate IMO.

                        It would be beneficial for you to talk with the team/community so we can be aligned on more conservative subscription terms before this is moved to an OIP. As things stand it is a bit of an overreach. My suggestion would be for you to bring 3 subscription terms to Discord and we can discuss and vote on what makes the most sense.