yes my question exactly… wouldn’t the insurance system be essentially susceptible if the the rest of the blockchain was?
I’m still very confused what we are insuring against. A list of possible scenarios would be nice.
It’s not as if the insurance pool is held off chain or in a bank vault.
I find it weird that we are using the Luna community pool to subsidise/create an insurance pool for a different token albeit a child (ANC). The comm pool has 63 million luna, so a bit over a Billion USD. It’s a huge sum for $luna holders to transfer over with the benefit primarily for $ANC holders, and of course the new $O3 of whom current $luna holders will get 15% of. Of course there is some overlap and an airdrop… but from a luna perspective I’m not sure there is value there esp when considering the below…
I am also concerned that the double use of previously burned luna returned via seignorage has the effect of increases the genesis supply of luna over 1 billion, as if everything was unwound (ie UST burned to mint Luna) there is genesis luna + seignorage luna on the table = more than 1 bill.
As a standalone token, I think ANC should be responsible for any additional insurance in its own right. It doesn’t have to be a double funding of the entire ANC platform surely? Just seems such a massive tie-up of capital, and if not held on anchor its value is inflated away? And if held on anchor it’s surely subject to the same risks (unknown) that we are supposed insuring against?
@WMoon I certainly agree that double using what we thought was already burned Luna from the community fund to fund an insurance pool can pose a great threat of dilution to the Luna supply, especially if the Ozone contract itself is hacked and all the 1 billion $ of UST is compromised by the hacker, the hacker would attempt to cash out all value of UST by diluting the Luna supply using the native swap mechanism.
That’s why it would be much more prudent to have another native insurance mechanism that is only dilutive to UST in the case of Anchor protocol failing by smart contract bugs/hacks. Such a mechanism can be built at the protocol level, we could add an interface accessible on the menu of Terra Desktop Station called Anchor Insurance, where the in the extremely rare event that Anchor is hacked you can simply deposit your aUST and get back newly issued UST, minus a 1-5% fee of the total deposit to avoid misuse.
I completely agree that insurance costs should be absorbed by ANC holders since it would align incentives. ANC governance holders have more incentive to review the robustness of Anchor’s code if the liability of insuring everyone fell on the supply of ANC tokens. The insurance mechanism here would be to issue new ANC tokens to buyback UST until lost deposits are replenished. However, this would need to happen at the Terra protocol level to be properly enforced since ANC tokenholders would have a conflict of interest in accepting the dilution of ANC to guarantee Anchor deposits.
The only reason this would not work is that ANC is simply not liquid enough and doesn’t have a big enough market cap to be the dilutive solution insurance mechanism, therefore I think that at the Terra protocol level, aUST should always be exchangeable for newly minted UST minus a small fee of the overall deposit to act as Anchor’s insurance. UST liquidity and market cap is much larger and would be able to handle this mechanism perfectly in my opinion, and it might not ever even be used if no problems with Anchor’s code ever come to fruition(bugs, hacks etc).
I think some people are confused about what the idea of what insurance is for, the idea for insurance to guarantee the redeemability of aUST into UST (aUST is an IOU of UST, not UST itself, very important to note).
The only way you can lose aUST in your wallet is if your wallet private key is ever compromised, which insurance does not cover. Your keys are your responsibility. However since aUST holds redeemabilty risk due to the behind the scenes Anchor protocol risk, this is what insurance is trying to solve. What good is it if you hold aUST but can’t withdraw/redeem it into UST due to Anchor code being hacked/bugged?
Wondering if there is a small typo in this part of the proposal:
Shouldn’t insurer_revenue be equal to ‘cover_fee$ * (1- climate_tax%)’. This simplifies to ‘cover_fee$ - climate_tax$’. Original proposal looks to have put a minus sign where a multiplication sign is needed.
Crazy suggestion #1:
Insurees should have option to utilize their deposit as insurance coverage as well (effectively double up as insurers). I did not think thru the mechanics deeply, but overall flow is:
-insuree deposits funds thru CV
-insuree may designate some fraction of his principal/interest-stream as part of the global insurance pool
-insuree gets incremental yield
My goal is to find ways to maintain/incentivize insurance coverage, especially in the case when the Ozone incentives run out, and the target_leverage_ratio/cover_fees dont provide enough yield.
pretty sure you’re correct
I’ll take a stab at this:
$O3 whales don’t deny or accept claims. I think Do imagined a game theoretic model but almost easier to think of it mechanically as a capital structure whereby the Insured Capital has a first priority claim on all the oAssets (capped at value of the Insured Capital Claim…think of claim as min(oAssets, Insured Capital)). There are more oAssets than there are First Priority claims by design because below the First Priority claims are claims by insurers and token holders. So there is no control over the payout to the insured participant, it is automatic given their priority in the cap structure.
oUST does not need a a yield > anchor rate to incentivize insurance. Insurance providers are getting a share of the revenue generated by the oUST assets. If the share of revenues is a larger percent than the share of capital, then the yield to the insurance provider will be higher than the yield on the underlying assets. The insurance provider does not think the risk on the underlying assets is high and so is happy to take more yield for what he perceives is not more risk. The Insurance purchaser, likes the yield of the underlying asset but is afraid it is risky in some way so the insurance purchaser gives up some of the yield to effectively get an over-collateralized position claim on the underlying assets he is interested in. The underlying oAssets can fall in value up to the amount of the over-collateralization without the insurance purchaser incurring a loss. That is the “safety” purchased by the cover_fee$.
I think the incentive for insurance coverage comes from the fact that the insurance provider will get a higher yield than the underlying oAssets due to “leverage”. The leverage comes from the fact that the percentage claim on oAssets revenue will be greater than the % of assets the insurance provider contributed to the oAsset pool.
yep agreed; at genesis, leverage is set to 1, and i think this has to increase (maybe dramatically, as per @unkn0wn6 's example) to keep insurers happy with their yield.
the thinking with my suggestion is to attempt tapping into the insuree funds.
Worth brainstorming what the insurance model can insure.
Off-hand, I can see 2 hazards that can be insured:
Risk of aUST falling in value verses UST. If aUST is the underlying yield generating asset and someone is worried aUST might have a protocol bug that makes is fall in value relative to UST, the Insurance purchaser gets the added protection of a claim on a pool of oAssets that is larger than their claim. So If you invest $100 with insurance and there are $200 of oAssets that fall 50% in value, you still don’t lose money.
There is actually an opportunity to insure UST itself. Say I’m a large institution and want some certainty that a unit of UST won’t be worth less than a unit of Fiat USD. Well if I put my UST in a covered vault and the terms of the insurance contract allow me to get my UST out at the Fiat USD equivalent price, then even if UST fell relative to Fiat USD, the over-colleralization of my claim can cushion a certain percentage of the UST depeg risk. If my UST claim is 80% of the oUST pool, then even if UST depegs 20%, my Insured Claim still gets the full USD Fiat equivalent value. Only a small amount of aUST yield asset is need to pay premiums to the Insurance provider to the Covered Vault.
This might be a real hit in European countries that want a stable store of value for large sums of money without paying banks a fee for holding their money but also some protection against the “cash equivalent” algorithm stable coin depegging from the underlying currency.
I agree with WMoon here. Spending the community pool in this manner would seem to be a violation of the every UST minted is a LUNA burned messaging of the ecosystem. I believe that the community pool initially had 10 million or so LUNA in it that was “valid”, shouldnt the funding come from that limited pool rather than the larger 63 million (of which 53 million should not exist)?
I’ve been thinking about this since the initial proposal of Ozone.
Ozone protects us from smart contract risk. Ozone is a smart contract. What protects us from risk of the Ozone smart contract? Is this not a never ending cycle?
“Provide coverage” paragraph, second bullet:
“overall pool” should be understood as “covered vault”, is that right?
A proposal to create a new vault includes a cover_fee. And I would guess that people would propose vaults with a cover_fee that accounts for the factors you mention (i.e. audits, length of time in the Terra Ecosystem) so it is attractive for insurers and insurees. I don’t think there is any official overarching target cover_fee because each protocol that is covered will be different.
Compare the situation of lending to Anchor directly vs. lending to Anchor through Ozone’s Anchor Vault.
If you are going direct to Anchor and someone finds an exploit in Anchor’s smart contract code, you may lose all your funds.
If you use Anchor through Ozone, you will only lose your funds if someone finds an exploit in both Anchor and Ozone, which is much more unlikely. I think the risk of Ozone and one of the covered protocols getting hacked is pretty independent so you are definitely reducing your risk by going through Ozone.
Or $OOO. Do wants the chemistry connection, this might be a more discoverable way to get it.
This is a bit unfair to say unless any of the economic benefit LUNA holders have already accrued from ANC were to be reversed, and ANC and LUNA exist as separate and distinct protocols…which obviously is not feasible. It’s a symbiotic relationship, a la the moon and gravity? Also the proposal clearly states that Ozone “facilitates levered coverage of technical failure risks in the Terra DeFi ecosystem”… i.e., not only Anchor but other TeFi included.
Is there a better use of the existing community pool than to bootstrap and scale Ozone at inception, especially considering the significant added value Ozone will bring to the overall Terra ecosystem? Isn’t this better than burning the community pool to no other accretive effect, knowing that post-Columbus 5, there will never be an instance like this whereby the community pool builds via seigniorage? Yes, supply will decrease by the burn, but pros outweigh the cons. And of course, Luna benefits from any increased UST adoption with Ozone…
Also, if the community is too concerned of insurer’s locked-up UST being invested into TeFi contracts and there being any correlation or risk of “cross-contamination”, perhaps funds can be invested into a diversified pool of tokens or a basket of yield farming opportunities?
When insuree claims insurance, oUST from the CV should get slashed/burnt right?
Else equilibrium for insurers is just that cover_fee*utilization ratio is constant across all CV and there’s no point in letting insurers choose which CV to stake in.
Thanks @Gresham for your thoughtful responses. I’ll address the second point first as I think it’s part of a bigger question that will be addressed more holistically in question 1.
The difference in our views stem from the difference in how we look at oUST. I’m taking the simpler approach of “which do I prefer based on my risk appetite” (from lowest risk to highest risk)
Following that logic, oUST needs to compensate for the additional risk taken which is why the yield has to be > anchor rate.
For question 1, the reason I bring up the issue of denying claims is due to the section of “Challenge coverage”. From what I understand, coverage for a CV can be challenged by governance - “all claims not yet processed fail, and the UST returned to the insurance pool.”. Once it goes to governance, whales would be able to vote to deny the claim, protecting the value of $O3.
The game theoretic model I believe refers to the claim process, where people would claim their own aUST instead of against Ozone if funds are safe. Because this is a binary decision, there’s no need to have another level of intervention at a claim disbursement level.
However, I guess there needs to be a level of oversight to protect claims against abuse - just like how people might be claiming for unnecessary tests on their medical insurance, driving up the cost for everyone else. The key concern is whether the oversight introduced is subject to abuse itself. The way I read the proposal, the incentives encourage self-preservation of $O3 holders which can destabilize the fundamental purpose of Ozone.
Your thoughts on looking at this as a capital structure is interesting. You’re approaching each CV as a overcollateralised pool of aUST/oUST? Presumably all of it is deposited in Anchor and in the event there is a slashing event/technical issue the aUST can still claim 100% (capital + interest - cover_fee) and oUST gets the residual (capital + interest + cover_fee - loss on slashing)?
I feel that this requires insurance providers to think of themselves as buying “equity” instead of the simple decision of “providing insurance gives me interest+X%”. The difference in risk-reward and structure may make it more complicated to tweak parameters to incentivise behaviour.
In a normal functioning situation where the CVs aren’t underfunded, (D) in your chart does not exist. It only gets called (very slowly since it is dripped in via fees) when real_leverage_ratio > target_leverage_ratio. If a CV gets closed, all aUST holders redeem their money and interest, and the remainder will be equal to insurer’s capital - losses + cover_fee - climate_tax (i.e. exactly what the insurer is meant to receive). There is nothing left for (D) in this case? Doesn’t that effectively remove it from the “CV capital structure”?
Maybe you’re looking at Ozone as an entire “company” in and of itself, and O3 forms the equity, while insurers are HY bonds and insurees are IG bonds? Sorry if I’m misunderstanding you but this is still qutie abstract to me and I’m trying my best to wrap my mind around it!
Another idea: Perhaps we can take a leaf out of the insurance companies’ playbook and allow for a mechanism of “capital injection” by $O3 token holders when a CV gets close to being underfunded (instead of the slow burn/rerouting of fees). This might help with directly tying the performance of each CV with the $O3 value and also boosts confidence in the solvency of each CV.
Yes i think this makes sense, $O3 whales will only come in when a “Challenge coverage” is triggered. Otherwise, oAssets are meant to guarantee that the insured capital is paid out accordingly.