Launch ETH Bonds

luxophyBBZG

Summary:

Olympus should implement ETH bonds to supplement DAI bonds to diversify away from DAI as the sole reserve asset and to reduce DAI supply bottleneck on Ohm supply growth. The Protocol can leverage the new ETH to initiate Maker CDP positions to mint DAI in a capital-efficient manner to back new sOhm while gaining an additional source of revenue.

Background:

Currently each newly minted sOhm is backed by 1 DAI worth of risk-free value. Through DAI bond this process is 100% efficient as the risk-free value of 1 DAI is 1 DAI. However, DAI bonds are often absorbed by current Ohm stakers. They unstake and sell their Ohm, swap to Dai via the Sushiswap LP pool, and rebond. While this internal bond rollover is not detrimental, it merely helps the Protocol realize the DAI liquidity that it had all along. The Protocol could have simply removed that DAI from the pool itself and minted the sOhm directly.

More over, by tying the supply of Ohm to the supply of DAI, the total supply of Ohm is capped to maximum number of Dai in circulation (~$4.5B). Olympus is already in the Top 120 for largest DAI wallet. We can only grow to the extent that the Protocol can continue to absorb Dai from the market. However, most of the DAI float is already locked in various vault contracts. In reality, the amount of available DAI is much smaller than even the circulating supply. Therefore, Olympus will likely experience a DAI supply bottleneck sooner rather than later.

Abstract:

Under the ETH bond proposal, the Bonder would use ETH instead of DAI to bond. We can be sure that we’re taking in new capital instead of realized SLP liquidity. The other bonding mechanics are the same i.e. the Bonder will get a certain amount of Ohm at a discounted price which will vest over 5-days. The Bonder would save gas by not having to swap out of ETH to DAI to bond. This reduces frictions and gains capital efficiency.

The Protocol will treat the ETH differently than DAI. Instead of discounting the ETH heavily like it would for a Sushi LP token, the Protocol would send the collected ETH on a regular basis to a DAO-controlled Maker Vault to generate or add to a collateralized debt position (CDP) to mint new DAI directly. Currently the Max Collateralization Ratio for ETH in a Make vault is 150%. If the market price of ETH were at $3000, Olympus Maker Vault could mint up to $2000 DAI. This new DAI could back the minting of new sOhm in the Total Supply for distribution. Alternatively, the new DAI could be partially deployed elsewhere such as a Yearn vault to earn yield.

Other Considerations:

Stability Fee: The Stability fee for Maker Dai vault is 5.5%. The fee gets added to the debt position. The Protocol could choose to pay using income generated from the minted DAI, or burn the excess Dai on a regular basis to maintain the original collateralization ratio. Alternatively, the Protocol can simply mint at a low collateralization ratio and allow the debt to build at a sustainable rate.

Liquidation Risk: In the worst-case scenario where the DAO’s ETH collateral is liquidated, it could recover the collateral by using a Keeper bot to bid at par value using DAI from its own Treasury. In an alternative scenario, the DAO could simply allow the collateral to be liquidated to reduce price exposure and debt. The DAO still owns up to 60%-100% of the current ETH value in the form of DAI and had no price exposure to ETH during the downturn that would have otherwise forced removal of sOhm from the supply to maintain backing.

Smart Contract Risk: ETH collateral in Maker vaults are subject to external smart contract risk (while the minted Dai is not). However, the Maker DAO runs a world-class protocol that is battle-tested and draws from years of experience. The smart contract risk would not be substantially different from simple liquidation risk.

Motivation:

To realize its vision as a true multi-asset crypto-backed currency, Olympus needs to diversify away from DAI and exert greater control over its own destiny. This proposal only addresses ETH vaults, but a potential Maker integration would open up a whole new world of possibilities through the creation of a multi-asset CDP that can be tailored to meet our monetary policy objectives. New Ohm bonds can be created as-needed to absorb any new asset that is compatible with Maker CDP vaults (UNI, wBTC, LINK, PAXOS, YFI, AAVE, COMP etc). This should trigger a flood of new capital. In effect Olympus would abandon the inefficient RFV ratio for minting and use the Maker vault collateralization ratio instead as the baseline. By boot-strapping off Maker vault contracts, the DAO will be free to focus on monetary policy and growth instead of building infrastructure from the ground up. If the new ETH bond were paired with a lowering of the APY, the public will view Ohm in a different light and the Olympus Protocol will attain a whole new level of legitimacy relative to its peers.

FOR: Olympus should implement ETH bonds and manage the ETH via a Maker CDP Position

AGAINST: Olympus should not implement ETH bonds

Cartesius

Giphy - Adventure Time Mind Blown GIF

domPablo

Giga brained. I'm in.

G_MON33Y

Love this idea hope to see this fleshed out some more on discord. 🔥

Zeus

Really cool idea and something we can definitely do. Two things that stand out to me though:

1) the new capital point is kinda mum, since there's nothing that stops someone from selling OHM for DAI, swapping the DAI to ETH, and then bonding the ETH (but this is no worse than what can happen now).
2) It might not be wise to start buying ETH when its up massively already; the risk/reward profile gets worse the higher price goes. This is a timing issue, who knows what ETH is looks like in a few weeks when we could have this ready, but we should certainly be cognizant of not only what we are buying but what price we are buying it at. If we were to start this right now, we'd need to do very low collateral ratios, and we still need to realize that even though we have discounted its value on our books, we do still lose money if price goes down (we should be aware of cost in addition to book value).

Nonetheless, super cool proposal that I'm looking forward to discussing!

Chochy

Zeus Eth with continue to grow in MC and its always good to integrate with the big daddy train we ride on.

luxophy

Zeus If you swap your Ohm for ETH you would likely be doing it from a non DAO controlled liquidity pool because it would be silly to go from Ohm into DAI and then to ETH just to bond and pay slippage twice. If the ETH didn't come from Sushi pool (through DAI swap) it's likely to be new capital if the total liquidity in the SLP is unchanged.

With respect to the high price of ETH, price risk can be managed by staggering your ETH transfers so that you're not sending all your ETH and minting immediately. The DAO can target 40% total CR ratio relative to the 30 EMA price of ETH upon every CDP vault transfer and not the current market price of ETH so you can mint more or less accordingly.

Chochy

Big Brain! Lets do it frens

TheCommenter

Hell ya man 🔥🔥🔥🔥🔥🔥🔥🔥
🦊

badgerpawz

Hi frend, interesting proposal. I disagree with some of the assumptions however:

"We can be sure that we’re taking in new capital instead of realized SLP liquidity."
Nothing to stop existing stakers from removing OHM to sell to ETH to take advantage of bond discounts, it would just be a single multihop tx. I purchased my first OHM with ETH through Sushi.

"The Bonder would save gas by not having to swap out of ETH to DAI to bond. This reduces frictions and gains capital efficiency."
People with ETH that want to bond can just swap ETH to DAI and do so, its marginally quicker and definitely less gas than removing staked OHM to sell to DAI and then purchase a bond, which you suggest is where the current bonding is occurring.

Im wary of adding an extremely volatile asset into the treasury - unless we were to do that a lot closer to the bottom of the cycle. Also if we are then using it to create CDPs and DAI anyway, how can these bonds be valued competitively against DAI bonds that have 100% efficiency?

Will follow the discussion in Discord and hope to learn things, cheers

badgerpawz

badgerpawz Hey Zeus made some of the same points while I was typing - Do I get a cookie or something?

balotelli45

Good idea to start thinking about what other assets can be held by the Treasury. Few thoughts I have:

More management needed from the team. Are they ready to engage in that atm?
Would be good to have some data freak do some modelization on different price crash scenarios and how it would impact the Treasury
I agree that having an ETH bond would help attract a lot more people to Olympus. It would also help to distance from the OHM = DAI meme.
How does stability fees volatility affects all of this? Or are the rates fixed?

Just throwing something out there as we are talking about Maker; What about using our SLP as collateral to mint DAI against? Could also be applicable to an ETH-OHM LP in the future as some have expressed the desire to have ETH denominated liquidity. Not sure what are Maker requirements though, I know they are a bit more conservative.

Dudemyguy

Hard no. If we are going to do anything with what can be used to buy ohm from the protocol, it should be adding more stables to the basket rather than this. I get it, giga-brained blah blah, but, it adds all of these non-elegant factors to the protocol and increases risk. OR if we are going to add ETH to the basket, then we should just create a safe OHM minting ratio for ETH (similar to the SLPs).

unbanksy33

Thank you for writing a thorough proposal. While I agree with the need for OlympusDAO treasury to diversify its assets, leapfrogging other stable assets (USDC, USDT and smaller cap projects) by introducing ETH doesn't seem right. I also feel the motivations outlined do not require introducing a volatile asset like ETH. Below I propose alternative solutions:

DAI circulating supply is capped - instead of introducing ETH, focus on stablecoins like USDT, USDC and smaller cap projects like FRAX, RAI and BANK. If the concern is that USDC/USDT are too centralized, then the contract can leverage Curve to convert directly to DAI or any other decentralized stablecoin.
Capital efficiency - if the issue is that Bonders convert OHM to DAI and use OHM-DAI SLP in the process, then wouldn't introducing new stable pairs resolve that? ETH wouldn't meaningfully change anything here.
"Reinventing the wheel" - imo, introducing volatile assets into the treasury shouldn't be the first solution in optimizing development resources. I also don't think the RFV mechanism is as inefficient as you propose it is.
Unlocking growth - I, too, have a sense that adding ETH will bring new community members. But so will partnerships with FRAX, RAI and BANK. These communities, imo, are more like-minded to us and already bought into the vision of algo stablecoins. And consider FEI: if ETH pair will bring in the same people that aped into FEI-TRIBE, do we really want them?

I also believe that introducing volatile assets makes sense as a middle-game strategy, not an opening strategy. Keep in mind the $3.8M treasury RFV was built on OHM-DAI alone. The goal now should be replicating this success sustainably to grow RFV to $10M, $20M and beyond. Reaching these milestones will unlock more hype, more growth, more treasury RFV to get us to the next level.

Dudemyguy

unbanksy33 This. This is just plain correct.

CMill

unbanksy33 Love the idea of these other projects before ETH. We would probably need to add a conversation chart at staking so people understand that 1 RAI = ~ 3 DAI, et al.

luxophyBBZG

unbanksy33 Thank you for your thoughts. Based on my smooth brain understanding of Zeus' original vision, Olympus should not tie itself to dollars coins at all in the long run. So USDT and USDC are out as are other knock-offs. How will Ohm be a USD-alternative primitive without exposure to and backing from ETH and BTC? Why would anyone pay this huge premium over intrinsic if it's just another dollar coin that does not hold its purchasing power relative to ETH and BTC? 1 Ohm is backed by 1 Dai, but we can inflate the total supply in accordance with the reserve that we use so that if you're staked, your purchasing power remains the same--but the same purchasing power relative to what? The USD or something else?

If you don't believe what I'm saying, please see Zeus tweet reply for example:

https://twitter.com/DiciCapital/status/1378799309324443648

Or this DAO Medium post:

https://olympusdao.medium.com/introducing-olympusdao-a-true-digital-currency-protocol-648c00c572d2

"How are we addressing this problem? With dollar coins (imo incorrectly dubbed “stablecoins”). At the time of writing, there are $38 billion in USD tokens circulating. They have become the primary trading pairs in crypto markets and the most popular assets in DeFi. There’s a strange irony to the fact that the most utilized cryptocurrency is really just a digitized dollar. While functional stablecoins may achieve a stable USD value, that does not mean they’re stable in purchasing power. Their real value changes just like dollars in a bank account, and that value is heavily reliant on the policies of the Federal Reserve and US government, and on the US economy."

"OHM will be most successful if it is positioned as a USD-alternative primitive for DeFi. We will work hard to get OHM integrated into money markets, DEX pairs, and yield farming protocols. We will also encourage its use as an alternative to holding USD. Eventually, you’ll be able to quote prices in OHM with a reasonable expectation that they’ll remain the same in the future. At that point, we hope to see people flee to OHM as a safe haven during risk-off periods rather than USD. We will do as much as we can to realize this, but ultimately it is a mindset and belief system (like any money or currency) that will rely on the support of you, the community."

I take Zeus at his word.

In a very real sense, Ohm is already tied to the majors as the DAI that back each Ohm was minted from ETH or similar in the first place. As far as I know, USDT/USDC can't be minted from anything other than good ole $USD. Seeing another USD coin drowning in a crowded Curve pool excites no one. You're merely sinking together. DAI was initially only mintable by ETH but when it kept selling at a premium over $1 USD, Maker capitulated and allowed all these other USD coins to come in and dilute the backing. We can do better.

Let's cut out the middleman and take control of our own monetary destiny. Let's declare independence from the Federal Reserve and mint our own backing from our own reserve CDP vault. I refuse to believe that backing Ohm with USD coins will take us where we want to go as a crypto primitive stable currency.

unbanksy33

luxophyBBZG We're in agreement on the long-term vision but disagree on the timeline.

I agree that US government introduces a system failure that will have an impact on all USD pegged stablecoins. I also agree that we should introduce alternative cryptocurrencies, like ETH and BTC, to Olympus treasury.

However, the timeline in which this makes sense is not now. I don't believe such a cataclysmic event will happen within the next 1-2 years which is the timeline we're working against. And I strongly believe OlympusDAO can grow to $1B market (or $20M Treasury RFV as another metric) within next 6 months on just a reserve of stables. When this happens, the treasury will have more stability and avenues to take on new assets, like ETH and BTC.

JaLa

Thanks for posting. Really well thought out.

My concern is less with running out of DAI and more with DAI remaining our sole backing. This proposal does not change our reserve asset. DAI is minted against ETH in a Maker CDP and it is still DAI that backs the minting of each OHM. We realize upside in ETH to the extent we mint more DAI against it in the CDP but eth isn’t our backing. This point is not a deal breaker for me, just something we should consider – do we want our protocol fully dependent on Maker? Its not just smart contract risk, I foresee governance risk, as well.

What’s really cool about this proposal is that we can benefit from the upside in any asset without realizing any of the downside. By marking down the value of the ETH to an RFV equal to the amount of DAI we are willing mint, we maintain upside in ETH while the treasury doesn’t suffer if the CDP is liquidated. This where I disagree with criticism that this introduces too much volatility (and with Zeus’s point that the price of ETH is high right now). We can only benefit from volatility – its like a free call option on ETH with no expiry (maybe some long-dated expiry if we don’t service the debt). That is extremely valuable.

Regarding debt service - I see a very simple solution. Given that we will constantly be adding ETH to vault, we can just mint new DAI only in an amount that brings us back to our target debt ratio. It’s essentially paying for the interest with the ETH we bring in instead of with actual cash (DAI) from another source.

luxophyBBZG

JaLa Thank you, JaLa. We are on the same page with respect to the debt:

With respect to the high price of ETH, price risk can be managed by staggering your ETH transfers so that you're not sending all your ETH and minting immediately. The DAO can target 40% total CR ratio relative to the 30 EMA price of ETH upon every CDP vault transfer and not the current market price of ETH so you can mint more or less accordingly."

With respect to your first point: DAI is not the backing. In my proposal, the reserve asset is the real backing and the DAI is merely a convenient unit of account. When 1 ETH was priced at $300 USD I could mint $200 DAI. Now if 1 ETH is priced at $3000, using that same ETH I can now mint $2000 DAI or 10 times as much Ohm for the same unit of reserve asset (reflecting the devaluation of DAI/USD relative to ETH). If I were a (3,3) Ohmie my purchasing power would be preserved because that inflation is dripped to me as a reward to maintain my constant share of the total Ohm supply.

Let's say you do an ETH bond using ETH at $4000 DAI per ETH. For 1 ETH you get 10 OHM in 5 days ($400 per discounted Ohm). The Protocol can go and mint 2000 DAI to back sOhm (200% CR). Worst-case scenario ETH drops 25% to $3000 DAI per ETH (150% CR vs your 2000 DAI). Vault gets liquidated and bots take the ETH. DAO still has 2000 DAI and no ETH. What's the problem? The CDP got liquidated not Ohmies. And that's assuming the DAO does nothing to hedge price risk by depositing more ETH e.g. send more ETH every week to maintain a target CR ratio in the CDP.

In an ideal case scenario, ETH doubles to $8000 per ea. DAO can mint an additional 2000 DAI (maintain the original 200% CR in the CDP). We didn't take in any new assets and we just doubled the sOhm that was initially minted. This means more rewards for everyone! If you had just taken in 4000 USDC you'd only ever be able to mint 4000 Ohm no matter what happens to USDC. And by definition you are now SHORT ETH. Of course leverage works the other way around too but your downside is limited by the CR ratio (e.g. the CDP gets liquidated). Your upside is UNLIMITED.

In effect, the DAI in the Treasury right now is ALREADY liquidated because it was minted from ETH (or similar) at a certain CR and we don't hold the ETH backing anymore (i.e. liquidated). What I'm proposing is that we control the upside too by holding the collateral ourselves.

el_dan0

unbanksy33 You think this will still happen in concert with lowering the APY? We need to increase the runway larger to inspire confidence. Why not include both those stables you mentioned and ETH? I don't think this is a binary choice between the two