Proposal: Add TBTC as collateral in AAVE V2

[ramaruro]

Summary

Joint Proposal by @ramaruro, @Estebank and @benlongstaff

In September 2020 we submitted a proposal to add TBTC (Bitcoin in Ethereum) to AAVE as Collateral for lending and borrowing: Proposal: Add support for TBTC

We want to give some updates on the growth of TBTC and poll the community concerning the listing of TBTC as collateral in AAVE

The following concerns were raised in the previous proposal, these will be addressed below:

• Time being live on mainnet
• Decentralisation of stakers
• Available tooling for insights into the network

Here we also the address the following items:

• Is TBTC a safe and suitable addition for lenders and borrowers and for the whole of the AAVE protocol
• Liquidations
• Deposits that failed to open

Details about TBTC

The tBTC system is developed by the Keep Network (https://keep.network/).

tBTC is

• Decentralized, Trustless, Permissionless
• Fully collateralized
• Open Source, Audited

Each bitcoin deposit is stored in a separate wallet created by the signers. Depositors don’t have to trust custodians, aka signers, because signers deposit a 150% ETH bond of the value of the BTC they hold in custody and have a stake of KEEP tokens that can be slashed if there is collusion.

The permissionless nature can be summed up by

Question - “How many doors would you have to kick down to stop my TBTC being redeemed for BTC?”

Answer - “Good luck finding the doors”

Main Stats about TBTC
This site shows current Supply of TBTC, total Value Locked, Mints and Redeems, as well as Bonded Value (securing the TBTC): https://keepstats.org/

• ~3 Months operation on mainnet with no loss of user funds, signer liquidations will be discussed below
• ~1780 TBTC in circulation
• ~$42M total TBTC value in Ethereum
• ~11,000 TBTC minted
• ~9,200 TBTC redeemed
• ~100,000 ETH bonded
• 174 nodes
• Audited, fully open source smart contracts and dapps.

Addressing Previous Concerns

1) Time live on mainnet

The tBTC system has been operating since September without any problems.

2) Decentralisation of stakers

There are 190 operators registered in the network, around 95 of those would have had their stake purchased on the market. Note that not all of the operators have finished setup on the network which is why this number is higher than the number of nodes listed in the main stats.

3) Available tooling for insights into the network
The tooling built by the community to gain insights into the network has matured significantly since launch.

The main recommended sites are:

https://keepstats.org/ (statistics about the network)
https://allthekeeps.com/ (details about deposits status and operators)

Other Questions Answered

1) Total TBTC Supply:

You can see Statistics of TBTC live on https://keepstats.org/). Currently there are about ~1750 TBTC in circulation ⇒ ~$40M total TBTC value on Ethereum. If 10-12% of this comes to AAVE, i.e. ~$4.5M, that is similar to several other Assets currently available. Also, TBTC will grow significantly from there.

2) Is TBTC a safe and suitable addition for lenders and borrowers for the AAVE protocol ?

Depositor funds are given the highest priority in the network. 1 TBTC token is always backed by 1 BTC

Signers are randomly selected and put down a 150% bond in ETH for any BTC they take custody of.

The economic security model of over collateralisation protects depositors against

• Being unresponsive in returning the deposited BTC
• Signers colluding to steal the BTC deposit
• Under collateralization

Failure to return a user’s BTC results in the signers’ ETH bonds being seized and auctioned off for TBTC.

3) Liquidations:

There have been 39 liquidations of ETH bonds since the network launch. This is the system acting as intended: https://allthekeeps.com/deposits/liquidations

All of the custodial risks are offloaded to the signers instead of the depositors.

Of the liquidations 38 were caused due to the collateralization level falling below 125% and 1 was due to operator error. Whenever a signer has an issue or misbehaves, it is the signer who loses funds. The depositor is always able to seize the signers bonds to recover their BTC.

4) Deposits that failed to open:

There are a large number of deposits that were not funded in time, causing them to entered a failed state. This was a result of operators stress testing the network at the beginning. Deposits that are not funded in time can be notified of timeout to free up the signers ETH bonds for further work.

Polling

Please vote in the poll below to get a sense of community support to then kick-off a potential listing.

• Yes for TBTC listing

• No for TBTC listing

[Peng_Fei]

I am of the opinion that tBTC is still too young to be considered collateral in the AAVE protocol. The experiment itself is a first attempt at providing a decentralized way of bringing Bitcoin to Ethereum, and as such needs to be properly battle tested.

[haave]

@Peng_Fei

needs to be properly battle tested

What would be the criteria for proper battle-testing?

Here it’s been suggested that tBTC needs at least 3 months on mainnet without incidents: Proposal: Add support for TBTC - #37 by ramaruro. Similar expectations (3 months on mainnet) were expressed by Maker governance in response to tBTC application in May. As I understand, tBTC reached this 3-month threshold.

At the moment, Maker seems to be on the way to tBTC integration - it is marked as “Planned”: Collateral Framework Official - Google Sheets
Most probably they will start with a low debt ceiling, to reduce the risk.

[ramaruro]

This is a good point, MakerDAO has been doing a lot of due diligence and should be onboarded soon.
You can see a pretty detailed analysis that they did here on the token and contracts: https://forum.makerdao.com/t/tbtc-erc20-token-smart-contract-technical-assessment/5399

[crew]

the MakerDAO risk team has not delivered their risk assessment of tBTC, which is the biggest test, not the token smart contract implementation review.

it will be interesting to see what they say about risk, because tBTC can lose its 1:1 peg to BTC if the ETH/BTC price is too volatile too fast which happens sometimes, so the collateral isn’t guaranteed to cover the full BTC value, so it’s a more risky asset to list than something like wBTC.

tBTC is also very low liquidity and I don’t think it makes sense to prioritize it over other more liquid tokens that are up for listing on Aave.

[alterkahn]

my vote is “For”, I think the time has come

[benlongstaff]

@crew

tBTC can lose its 1:1 peg to BTC

The goal of TBTC is not a price peg to BTC, it is designed for redeem-ability.

1 TBTC is always backed by 1 BTC

It is expected that liquidity incentives could cause TBTC to have a higher price than BTC

@crew

If the ETH/BTC price is too volatile too fast which happens sometimes, so the collateral isn’t guaranteed to cover the full BTC value, so it’s a more risky asset to list than something like wBTC.

It’s worth noting that the ETH bond and KEEP stakes are used as an additional layer of security to ensure the security of your BTC. The network is quite efficient at liquidating operators that let their Keeps become under collateralised due to the reporter incentive.