Asset risk assessment: The Stablecoin DCHF anchored to the Swiss franc and its governing body Moneta DAO

Protocol Owner Multi-Signature originally seeded 100 million tokens on September 23, 2022. The documentation lists the contracts used for torrent distribution, but many of them have been deprecated with the Moneta DAO. Moneta documentation should be updated to reflect new decisions regarding financial management and distribution. The chart below shows the amount of money that has been returned to the owner since the initial allocation.

We also built a Moneta Dune dashboard showing on-chain transactions signed by the protocol owner, see "MON Collections and Flows."

The figure below shows the actual distribution of MON as of April 17, 2023. Most of the money has been transferred to the new Treasury and Grizzly.Fi multi-signature. The latter provided $1 million worth of seed money to Moneta DAO, priced in USDC, CRV and LIT. Owner multiple Issuance is still a regular source of funding for Curve measurement incentives, stability pool incentives, and airdrops.

Contract address:

Uniswap MON pool  (tx) -- MON/WETH UniV3 Pool  

UniV3 rewards vest-- MON/WETH Pool LP Award

Team vest-- Original team assignment  

lockedMON-- New team member @elnilz;

Treasury multi-sig-- The vault created with Moneta DAO.

Grizzly.Fi multi-sig-- Raised $1 million in seed funding when Moneta DAO was founded

Moneta DAO: Deployer-- Giveaway event hosted by Moneta Team  

Curve Gauge Rewards-- DCHF Pool bonus, weekly allocation  

CommunityIssuance-- Stability pool excitation, monthly distribution  

MON airdrop-- GHNY, LQTY and Grizzly Freezer airdrop; Token acquisition is done through an airdrop contract, which distributes funds from the owner through multiple signings

Treasury Multi-Signature currently holds about 50% of all the tokens. Of the circulating supplies, most are in MONStaking contracts (17m) and Uniswap pools (3m). Since all distributions were made before the DAO was established and without community involvement, questions could be raised about airdrops. The Grizzly.Fi community has received the most tokens (11.17m), and it is worth noting that the core team members are also the cofounders of Grizzly. This overlap can be a cause for concern because most MON circulation supplies come from airdrops, and the team's overall MON allocation is obscured by this airdrop strategy.

Market and practicality

Over 90% of circulating DCHFS are in WBTC/ETH stability pools and Curve DCHF/3CRV pools. UniV3 DCHF/USDC pools are also in the top tier, although only accounting for ~ 2% of supply. As of April 2023, the highest addresses are as follows:

Currently, DCHF is traded on two major DEX, including Uniswap and Curve, and does not appear to be traded on any CEX. The vast majority of liquidity exists in the Curve pool.

Curve has historically led in DEX transactions, although Uniswap has caught up in recent months. The following two charts show daily selling and buying, respectively:

TVL and ununlocked DCHF declined after their initial release in September 2022, and adoption seems to have been relatively flat since November 2022.

DCHF has historically traded at a premium to its CHF anchor pricing, and currently trades at a sizable premium of about 1.04 to 1. This is not entirely surprising, as there is no easily available CHF available for arbitrage, and the Liquity mechanism allows considerable deviation from anchor pricing. The rise is due to 110% MCR. 1.1 will create an immediate arbitrage opportunity for users to borrow at the maximum amount borrowed and sell at a profit. Liquity sees this as a price ceiling. Historically, LUSD has also been priced above anchor pricing due to its conservative mortgage selection (even if arbitrageurs may choose not to step in when prices are above anchor pricing), and systematically loose soft anchor pricing ranges. A rising depeg is less of a concern for users, but could raise questions about stablecoin practicality.

Risk vector

Smart contract risk

DCHF is a fork of Vesta Finance, which in turn is a fork of the Liquity decentralized Stablecoin protocol. Although Liquity is a well-audited agreement, engineers from Vesta have warned against forking the projects because their code is messy and unoptimized and may not have been updated to new features. Despite this warning, DCHF is still a fork of Vesta and seems to have made no effort to follow up on the changes in the original codebase, possibly due to their use of non-upgradable contracts.

audit

DeFi franc Version 1.0 was audited twice by CertiK and General IT, which was conducted by a foreign company. Translation is required in the browser to decrypt the audit content. Two key issues were identified, including:

1. Possible reentrant attack: This vulnerability has been found for StabilityPool and ETHTransferScript contracts. The auditor recommended the use of the "ReentrancyGuard" library, which the team appears to have resolved by committing the changes to the private repository.

2. Incorrect Ether value: This bug appears in the PriceFeed and PriceFeedV2 contracts and involves converting "int256" to "uint256", which may cause arithmetic overflow. The team has considered the problem, but it is not clear if it has been fully resolved.

Four other problems were identified, including:

1.Violation of ERC-20 standards: This problem has been solved.

2. Not enough Gas to complete all loop iterations (DoS)This problem has been taken into account.

3. May be blocked in the balance of the contractTokenThis problem has been taken into account.

4. wrongTokenpriceThis problem has been taken into account.

The CertiK audit report showed that there were no unresolved issues and critical issues with the project. Here are some examples of issues that might be associated with additional risk vectors beyond smart contracts:

GLOBAL - 01 | concentration related risks: Within the AdminContract, the _owner role has extensive authority over several functions, including affecting borrowing/redemption fees, adding or removing mortgage assets to the stability pool, transferring funds from the stability pool, changing lock-in periods, or worse, changing the AdminContract itself, including but not limited to:

In addition, AdminContract has the right to upgrade multiple contracts without community consensus, including SortedTroves, StabilityPool, StabilityPoolManager, and TroveManager. This opens up the possibility for an attacker to change the contract implementation and bleed the Token.

MOT - 01 | initialTokenDistribution (mitigated): Another major issue in the centralization/privileged risk category involves sending MON to the _treasurySig address when the contract is deployed. This means that MON can be assigned without community consensus. To mitigate this risk, the team has documented their Token economics and how MON is allocated.

GLOBAL - 02 | insufficient storage space (resolved): Multiple contracts are scalable, including AdminContract, StabilityPoolManager, and TroveManager. This means that there needs to be a storage gap to allow new states to be added without compromising existing deployments. In response, CertiK's report noted that the team had revised the design and opted for a non-upgradable contract. Although non-upgradable contracts protect against future changes that might affect the protocol, the system still has the option to update. Protocol owners can still add NewCollateral via AdminContract, or they can update the parameters via DFrancParameters contract.

DPK - 01 | insufficient input validation (solve): In the dfrancParameters contract, the setMCR() and setCCR() functions have a small contract risk of causing the user's MCR to be greater than the CCR, which would essentially prevent the recovery mode from being activated. This is not changed in the deployed code.

Audit recommendation

The team followed CertiK's recommendation, assigning privileged roles to the multi-sign wallet and transitioning to the DAO governance/voting module. The Vault Multi Sign and Protocol Owner Multi Sign (control AdminContract) addresses can be found here and here.

CertiK also recommends implementing a time-lock contract with a 48-hour delay for privileged operations. The team has considered this, but has opted not to surrender ownership to a time-lock contract for now, in order to maintain flexibility. The permanent solution could be to relinquish ownership or remove the ability of the owner role to affect the mortgaged assets from the stability pool, transfer funds, change the lock-up period, or change the AdminContract. This would bring DCHF's trust assumptions in line with Liquity, but would sacrifice flexibility, since new mortgage types and price predictors cannot be changed.

Another solution is to introduce fully chain-based governance, controlled by the Token holder's DAO. The transition to the DAO is still in its early stages of development, with very limited community activities outside of the team. Questions may be raised about the fairness of the MON allocation, given the overlap between core Treasury multiple signers and Grizzly.Fi airdrop recipients. The team has set up a multi-sign proxy address and transferred ownership of the AdminContract to that address, providing transparency (proof of transfer) to users regarding Token allocation. This provides a reference to help users check if MON is assigned as advertised.

Centralization and custodial risk

The 4/6 multi-signature ownership of the AdminContract (and all associated system contracts) gives it extensive control over key protocol functions, raising concerns about escrow risk. The team chose a non-upgradable contract, which means that owner multiple signers must add new collateral via AdminContract or update existing collateral parameters via DFrancParameters contract. The multiple signature can:

Although the centralization risks identified in the audit have been mitigated to some extent, the core team still has plenty of prerogatives to exercise.

The team has expanded the multiple signature threshold from 3/6 to 4/6 and added Curve team members (Mimaklas). They also transitioned to the DAO, which is still in early development. For now, the core team is still necessary for the effective operation of the DAO.

To reduce the risk of hosting, the recommended next step is to transfer ownership of all features that are not time-bound to a time-lock contract. Multiple signatures without time locks should be reserved for emergency operations (suspend MON pledge/suspend DCHF casting). The long-term goal should be to fully adopt on-chain governance or choose to minimize governance by relinquishing ownership of the relevant contracts. An access control list of system contracts is compiled on this Google form.

Control risk

Here is an overview of important MON holders:

Grizzly.Fi hatched DeFi franc and Moneta DAO. The team and advisers were allocated 10% of the tokens (10 million MON), while Grizzly.Fi was given $1M of assets (9174,312 MON) to fund preparation (source).

There are two entities in the Moneta DAO governance: Token House and Delegate House. As discussed above, Moneta DAO Token holders will vote on strategic protocol decisions, including allocation of DAO Vault funds, key risk parameters, and protocol escalation. However, in the early stages of development, it was apparent that there was a lack of a clear process for prominent community members to take on representative and licensed roles within the DAO.

In addition, the allocation of MON takes place prior to the formation of DAOs, without community involvement. Given that there is no clear representation process, we can assume that the current Treasury multiple signers and agreement owner agents are the actual representatives. Although the representative carries out the wishes of the Token holder, the relationship is non-binding and trust-based.

While Token holders will have a voice in strategic agreement decisions, the scope of governance remains uncertain. Clear governance processes and clear communication of representative plans would be helpful.

Predictor risk

DCHF uses Chainlink as its pricing provider. Chainlink is a highly regarded provider of seers, although DCHF doesn't have a backup seer. The fetchPrice function includes a soundness check that allows a maximum deviation of 50% between updates that would otherwise have accepted the last good price. CHF may be priced inaccurately due to malfunctions of the predictive machine. This can have a negative impact on the liquidation and redemption mechanisms and ultimately lead to potential loss of funds.

Liquity, by contrast, uses Chainlink as its primary predictor and Tellor as a backup. If the primary predictor fails to update or produces a value outside the range of the specified guardrail, it will algorithmically reject the primary predictor. This strategy is designed to ensure that price offers can autonomously respond to black Swan events without manual intervention. On the other hand, DCHF policies may require more active monitoring and manual updating in case the predictor fails.

Decoupling risk

The DeFi franc Agreement currently has 6,694,401 DCHFS issued with a value of $1.15 for a total value of approximately $7.6 million and holds approximately $17 million in collateral (a combination of ETH and wBTC). Redemptions are backed by sufficient collateral.

To control the supply of DCHFS, the agreement uses borrowing and redemption charges. With 110 per cent collateral providing a floor and ceiling, arbitrageurs have an incentive to keep prices stable. The agreement does not rely on external AMMs for supply expansion or contraction, and the burn/draw amount is executed through the onchain peg mechanism.

As a Liquity fork - a stablecoin that has stood the test - bankruptcy risk is low for now. Liquity does, however, have long periods of unanchoring. History of $1. By maximizing decentralization, it sacrifices scalability and tends to have significantly different soft anchoring than centralized stablecoins.

The collateral chosen by Moneta (ETH/WBTC) is also quite conservative. We believe the biggest risk to depeg going forward is the possibility of mismanagement of governance. If a low-quality collateral type is added to the agreement (for example, Token MON, the local governance of the agreement, or a complex derivative asset), this may increase the chance of failure or manipulation of the underlying collateral, causing the system to fail, and thus decoupling.

Users (borrowers) should keep in mind external factors, such as currency exchange rate risks in traditional financial markets. DeFi reflects a lag in this price change, taking into account the depreciation of the Swiss franc against the US dollar. In the case of the de-anchoring of the DCHF to the CHF, assuming that the borrower converts the funds into dollars after borrowing, then the premise of the re-anchoring assumption is destroyed and the borrower has the opportunity to repay the loan at a cheaper rate. In the other case, when the DCHF is unanchored by more than CHF 1 (but less than CHF 1.10), the situation is different. Since there is no direct arbitrage of DCHF and CHF to balance anchoring in this case, the only dynamic that could cause DCHF to re-anchor back to 1 is the need to supply DCHF to DeFi.

LlamaRisk evaluation criteria

Centralizing factor

1. Is it possible for a single entity to defraud its users?

Risk is reduced by transferring ownership privileges to multiple signatures. Currently, the multi-signature has four signer Settings, including members of the Curve team. But the multi-signer may still take actions that affect users' funds.

2. Can the project continue if the team disappears?

The protocol can work in the absence of the team, but will not be able to update systems or respond to emergencies. Despite the recent launch of DAOs, there is still no direct way to give Token holders direct control over the protocol.

Economic factor

1. Does the viability of the project depend on additional incentives?

The protocol relies on its own incentives to drive liquidity to its stability pool and DEX such as Curve. It does not depend on incentives other than as a guiding tool.

2. If the demand drops to 0 tomorrow, can all users get compensation?

Yes, about $17 million of ETH and wBTC collateral backed $7.6 million of DCHF value. Users can redeem their savings tomorrow.

Safety factor

1. Did the audit reveal any worrying signs?

The audit did not find any critical issues, but flagged a number of significant issues that were partially or fully addressed. The main problem is the excessive authority of the AdminContract, which has been transferred to multi-signers. Although the DCHF protocol uses non-upgradable contracts, it does have a way to update the system through AdminContract or parameters through the DFrancParameters contract.

Risk panel recommendation

We have three recommendations that will benefit the DCHF ecosystem and community:

1. Update the documentation to clarify the strategic role of the MON holder in governing the DCHF agreement

2. Clarify the perks/permissions that Vault/Protocol multi-signatures continue to have in all relevant system contracts, and provide a clear roadmap for decentralization.

3. Create a policy for acceptable forms of collateral and a process for adding collateral types that the Moneta DAO is willing to support.

Overall, we believe that the decentralization challenges faced by the Moneta DAO are typical for the early stages of DeFi protocol deployment. While there are certainly risks and uncertainties, we think their team has done their best to allay those concerns. We hope they will continue to exercise good judgment and be cautious when adding any new types of collateral. We support the team's proposal to add Curve gauge to the DCHF/3CRV pool.

Original link