[PAXG] Collateral Onboarding Oracle Assessment (MIP10c3-SP16)


MIP10c3-SP#: 16
Author(s): Niklas Kunkel (@NiklasKunkel)
Type: Process Component
Oracle Team Name: Green
Status: RFC
Date Proposed: 2020-11-27
Date Ratified: <yyyy-mm-dd>



This Oracle would provide the PAXG/USD price as part of the collateral onboarding process for PAXG.

Oracle Data Model

|    Source    |  Asset Pair   |Quorum | Feed Model  | Oracle Model |
| :----------- | :------------ | :---: | :---------: | :----------: |
|    Binance   |   PAXG/USDT   |   13  |    Median   |    Median    |
|    Bitthumb  |   PAXG/USDT   | 
|     Gemini   |    PAXG/USD   |
|     Kraken   |    PAXG/USD   |
|    Uniswap   |    PAXG/ETH   |

Oracle Supporting Data Model(s)

ETH/USD (canonical)

|    Source     |  Asset Pair   |  Feed Model  |
| :------------ | :------------ | :----------: | 
|   Binance     |    ETH/USD    |    Median    |
|   Bitfinex    |    ETH/USDT   |              |
|   Bitstamp    |    ETH/USD    |              |
|   Coinbase    |    ETH/USD    |              |
|   Gemini      |    ETH/USD    |              |
|   Kraken      |    ETH/USD    |              |

USDT/USD (canonical)

|    Source     |  Asset Pair   |  Feed Model  |
| :------------ | :------------ | :----------: | 
|   Binance     |    BTC/USDT   |    Median    |
|   BitFinex    |    USDT/USD   |              |
|   FTX         |    ETH/USDT   |              |
|   Huobi       |    ETH/USDT   |              |
|   Kraken      |    USDT/USD   |              |
|   OKEx        |    BTC/USDT   |              |

Oracle Address

  • Medianizer - Mainnet TBD
  • Oracle Security Module (OSM) - Mainnet TBD

Supported Tools

Remaining Work

  • Deploy and configure Medianizer and Oracle Security Module smart contracts to Mainnet
  • Push new relayer configuration with PAXG/USD Mainnet Medianizer smart contract address


The chosen Data Model utilizes a mix of USDT, ETH, and direct-to-USD pairings to limit exposure to quote manipulation.

The PAXG token is a little bit problematic from an Oracle risk perspective, and this should be accounted for in the debt ceiling if Maker Governance chooses to onboard PAXG as collateral in the Maker Protocol. The issue stems from the volume being exceedingly small. The usual method of protecting from market manipulation is sample a diverse array of exchanges. However, in the case of PAXG, since the liquidity is so low, it becomes much more feasible for an attacker to manipulate enough markets at the same time to execute an Oracle attack. Therefore if Maker Governance insists on adding PAXG as a collateral-type the Oracle Team recommends an initial debt ceiling of not more than 1,000,000 in order to cap losses from a potential Oracle attack.


Thank you for the Assessment Nick! I think the entire community would agree with your recommendation of no more than 1M DC for PAXG. I do think that with time Paxos will get more institutional interest in PAXG (Paxos is out there raising funds), especially if things go haywire for the U.S. economy. So, hopefully we will see more volume. It could also be that PAXG users just hodl, and not much more you can do with the asset. Hopefully in the near future use the Maker Vaults and use their asset for more exciting things, than just holding exposure to Gold.

Nice catch. I really think a Maker Vault could increase the PAXG usability.

It’s like WBTC, there is no need if you can’t use it in DeFi. I’m not sure 1M is enough to foster that. But having an OSM attack is sure to destroy everything.

Let’s start with that and find solutions to this problem if there is demand.

Was there a reason you can’t just use the price of gold as the feed? Assuming there’s not an easy way for Keepers to settle in cash?


:point_up: this. Wouldn’t the global gold market pretty much be the best in class in terms of tamper resistant price feeds?

The price of gold is not 24/24 7/7. But that would cover most of the time.

Let’s not forget that Paxos is the one powering PayPal to help normies get into ETH & BTC–here’s a Vid of how that all works under the hood:

At 7:27 an internal view of the addresses that hold Gold for PAXG Unallocated_Gold

Forgive the perhaps naive question, but what was the reason for this tweak of peramaters compared to MIP10c3-SP11, which appears on Git.Hub but not on the forum?