Date of analysis: 2023-10-18
Summary
Following up on the @Anthias.xyz proposal, we assessed the risk implications of listing a Prime ARB market on Notional’s Arbitrum deployment. If the DAO is in favor of listing ARB then we suggest listing the asset with an initial supply cap of $1M (~1,200,000 ARB), a liquidation discount of 8%, and an implied minimum collateral ratio of 160% vs stablecoins. Our asset onboarding review highlighted no significant technical issues to list the asset. We welcome community feedback before this proposal is formalized into a Notional Improvement Proposal (NIP) and goes to a formal governance vote.
Asset Overview
- Asset symbol: ARB
- Underlying Protocol Name: Arbitrum
- Historical daily volume (avg. last 30 days): $115M (CEX and DEX)
- DEX liquidity: $28M
- Market cap: $1.05B
- FDV: $8.2B
- Total supply: 10B tokens
Economic Risk Assessment
Asset ownership & DeFi presence
ARB is currently held by 702K addresses on Arbitrum. The top 100 wallets own 84% of the supply. 35% of the token supply is held in the DAO treasury contract, 6.3% is held in the vesting contract and 5% is held in a multisig for offchain labs investors. Let’s note that the current circulating supply is about 13% of the total ARB supply. As investors unlock their ARB and as the treasury disburses funds the circulating supply will increase which could materially change ARB’s liquidity profile.
Arb is listed as a collateral asset on Compound V3 and Aave V3 Arbitrum deployments. 6.3M ($5M) and 3.6M ($3M) ARB tokens are currently deposited on these protocols respectively.
Underlying protocol KPIs
Arbitrum is a layer 2 EVM compatible chain. It is currently the 4th chain by TVL with $1.7B in TVL. 101,000 wallets interacted with the Arbitrum chain in the last 24 hours according to DeFi Llama.
Token
ARB is the governance token of the Arbitrum chain. ARB token holders can vote on DAO votes or delegate their tokens to other DAO participants.
Historical volatility
ARB historical price vs ETH
Historical volatility analysis
ARB
| Time period | Volatility (last 90 days) | Annualized | Volatility (last year) | Annualized vol. (last year) | Source |
|---|---|---|---|---|---|
| 1h | 0.65% | 60.84% | 0.70% | 65.29% | Chainlink |
| 1d | 3.02% | 57.66% | 4.35% | 83.09% | Chainlink |
ETH
| Time period | Volatility (last 90 days) | Annualized | Volatility (last year) | Annualized vol. (last year) | Source |
|---|---|---|---|---|---|
| 1h | 0.37% | 34.53% | 0.43% | 40.47% | Chainlink |
| 1d | 1.72% | 32.87% | 2.07% | 39.57% | Chainlink |
Historical maximum drawdown analysis
ARB
| Time period | Max drawdown (last 90 days) | Max drawdown (last year) | Max drawup (last 90 days) | Max drawup (last year) |
|---|---|---|---|---|
| 1 hour | -7.24% | -7.24% | 4.51% | 5.52% |
| 1 day | -11.83% | -12.16% | 9.12% | 23.69% |
ETH
| Time period | Max drawdown (last 90 days) | Max drawdown (last year) | Max drawup (last 90 days) | Max drawup (last year) |
|---|---|---|---|---|
| 1 hour | -5.38% | -5.72% | 4.96% | 4.96% |
| 1 day | -10.44% | -10.44% | 7.07% | 7.07% |
ARB has historically been more volatile than ETH. It also had more pronounced drawdowns than ETH’s over the last 90 days making the asset more risky. ARB is also subject to smart contract and governance risks.
Therefore, we propose a 160% minimum ARB/USDC collateral ratio. This would protect the protocol against the worst historical maximum drawdown over a 1 day period. This implies that the protocol would have still been overcollateralized if risky accounts were not being liquidated for a period of 24 hours.
Asset liquidity
ARB to USDC slippage analysis:
| Expected slippage | Trade size |
|---|---|
| 0.5% | $260K |
| 1.0% | $475K |
| 2.0% | $550K |
| 3.0% | $725K |
| 4.0% | $750K |
| 5.0% | $760K |
| 6.0% | $770K |
ARB is listed on multiple DEXes and benefits from deep liquidity with less approximately 2% slippage for a $500K trade. Slippage increases rapidly for trades larger than $700K.
On-chain liquidity breakdown (on Arbitrum):
| DEX | ARB liquidity | Non ARB liquidity |
|---|---|---|
| Balancer | $20.9M | $3.1M |
| Uniswap V3 | $1.4M | $2.8M |
| Total | $22.3M | $5.9M |
We propose setting the ARB supply cap at $1M based on the asset’s on-chain liquidity profile. If demand for ARB is elevated the DAO could increase the supply cap at a later date.
Other notable risks
ARB is subject to rapid changes in liquidity due to its relatively low circulating supply compared to its total supply. Future unlocks could materially change ARB’s liquidity profile. Moreover, ARB is subject to potential technological risks related to the Arbitrum chain. These idiosyncratic risks can be somewhat mitigated by a lower exchange rate haircut (higher minimum collateral ratio).
Security Assessment
As mentioned by @anthias in the initial post, Arbitrum underwent multiple security audits from auditors like ConsenSys Dilligenge and Trail of Bits.
Implementation Review
No specific integration challenges have been identified.
Proposed Risk Parameters
Based on the different analyses presented above, we propose listing ARB with the parameters presented in this spreadsheet. Based on ARB’s historical volatility profile, we propose to list the asset with exchange rate haircut and buffer parameters that imply a minimum collateral ratio of 160% for USDC borrowers. Based on current on-chain liquidity we propose to list the asset with a 8% liquidation discount. We think a 8% discount is sufficient to allow liquidators to liquidate ARB profitably under historical market conditions. Finally, we propose to list the asset with an initial supply cap of $1M (1,200,000 ARB).
We propose the following listing Prime ARB interest rate curve. The proposed interest rate models imply a 2.26% borrow interest rate and 1.23% supply interest rate at the target utilization rate.
Risk Parameters Benchmarking
| Parameter | Aave V3 | Compound V3 | Notional |
|---|---|---|---|
| Liquidation Discount | 10% | 7% | 8% |
| Minimum Collateral Ratio (vs USDC) | 167% | 167% | 160% |
| Supply Cap (USD) | 16,000,000 | 6,500,000 | 1,000,000 |
The proposed liquidation discount is in line with the discount used by Compound and Aave. We propose a more aggressive minimum collateral ratio than Compound and Aave but with a substantially lower supply cap.