[TBTC] Proposal for Collateral Onboarding - tBTC

Bringing Bitcoin (and Bitcoiners) to Maker

Hello everyone, I’m Matt Luongo, the project lead for tBTC.

We’ve built a Bitcoin sidechain to Ethereum, meant to give Bitcoin the full superpowers of Ethereum and bring additional collateral to DeFi.

While I believe BTC can be great collateral for DAI, more importantly, I believe a native BTC on-ramp into the ecosystem can grow the protocol’s user base. An on-ramp into DAI and Ethereum that avoids centralized gatekeepers and unnecessary third parties is the best way to introduce Bitcoiners to the Ethereum ethos en masse and bridge the two cultures.

I’m excited to share what we’ve been working on for the past year and a half, tailored to the usecase of MCD. I’m looking forward to feedback from the community on how we can best serve the protocol, strengthen the peg, and work together to build our new financial system.


  1. Who is the interested party for this collateral application?

Matt Luongo of Thesis and the Keep project is interested in submitting the following tBTC collateral application.

  1. Provide a brief high-level overview of the project, with a focus on the applying collateral token.

tBTC is an ERC-20 pegged 1:1 to Bitcoin. Instead of relying on a single custodian like wBTC or federation of custodians like Liquid, tBTC is supported by a decentralized bonded multi-federated peg model. The system is a supply peg designed so that tBTC can be minted or redeemed for BTC at any time without 3rd party friction.

Each time a Bitcoin is deposited into the system, the underlying Keep network’s random beacon selects a new random set for signers to form a federation. This signers set is bonded 150% the value of the deposit in ETH. The signers generate a key, BTC is sent, and after 6 block confirmations, depositors can post a BTC SPV proof to the Ethereum chain and mint TBTC. Redemption is the reverse: TBTC is sent to the system along with a small fee (5 bps) to pay the signers and once received, BTC is released and the TBTC is burned.

  1. Provide a brief history of the project.

The Keep team grew out of Fold, a popular Bitcoin app first launched in 2014. Keep has been building a system for decentralized signing and custody since 2017, and in late 2018 started building tBTC as the first app to launch on the Keep network.

The tBTC technical spec was released in August 2019 by Keep, Summa, and the Cross-Chain Group. tBTC has been live on the Ropsten testnet since January 2020, and will launch live on the Ethereum mainnet on May 11th, 2020.

  1. Link the whitepaper, documentation portals, and source code for the system(s) that interact with the proposed collateral, and all relevant Ethereum addresses. If the system is complex, schematic(s) are especially appreciated.

Technical spec: https://docs.keep.network/tbtc/
Developer documentation: https://tbtc.network/developers/
Ropsten demo dApp to mint TBTC from testnet BTC: http://dapp.test.tbtc.network
tBTC contracts: http://github.com/keep-network/tbtc
JS library to build on the dApps atop tBTC: https://github.com/keep-network/tbtc.js
SPV relays: https://github.com/summa-tx/relays
Bitcoin SPV library underlying the relays: https://github.com/summa-tx/bitcoin-spv

  1. Link any available audits of the project. Both procedural and smart contract focused audits.

tBTC and Keep underwent a 6 week audit by ConsenSys Diligence, and is entering into an additional audit with Trail of Bits.

ConsenSys audit report: https://diligence.consensys.net/audits/2020/02/thesis-tbtc-and-keep/

For context, check out our head of engineering’s summary of the results, and one of our auditor’s reflections on the system design and process.

We’ll share the additional audit results from Trail of Bits when they’re available.

  1. Link to any active communities relating to your project.

Keep & tBTC Discord: https://discord.gg/wYezN7v
Cross-Chain Group Telegram: https://t.me/crosschaingroup

  1. How is the applying collateral type currently used?

At launch tBTC will be used as BTC-pegged collater in a number of DeFi apps like Compound, Nuo, Set, and Aave. tBTC allows native BTC to be used directly in these projects, without any redirects or special places to mint or burn.

  1. Does one organization bear legal responsibility for the collateral? What jurisdiction does that organization reside in?

tBTC as a protocol enables any KEEP holder to act as a partial custodian of the underlying BTC. No single organization bears legal responsibility for the collateral.

  1. Where does exchange for the asset occur?

After May 11th 2020, tBTC will be listed on OasisDEX, Uniswap, Radar Relay, Loopring, and Kyber, among other exchanges and DeFi projects. tBTC can also be directly minted and redeemed using the BTC and ETH chains.

  1. (Optional) Has your project obtained any legal opinions or memoranda regarding the regulatory standing of the token or an explanation of the same from the perspective of any jurisdiction? If so, those materials should be provided for community review.

It has, but providing private legal memos risks waiving attorney-client privilege. Intuitively, our understanding is that as a commodity-backed token, TBTC appears to be favorably treated in the US and similar jurisdictions.

Additional Consideration

We recognize that tBTC hasn’t existed for as long as candidate collateral like WBTC, and of course agree that confidence in security and market behaviour that comes with time must be a factor in approving assessing collateral MIPs.

Because we believe that a non-custodial model for Bitcoin collateral better fits the risk profile of Maker’s design, we’re proposing an additional guarantee on tBTC’s safety as collateral.

If tBTC is accepted as collateral, the Thesis team will additionally escrow $15M in KEEP (~12.5% of the supply) for the first 6 months to further backstop tBTC collateral for a debt ceiling up to 10M DAI. We’ll also pursue additional insurance for that period covering up to 500K DAI.


I would definitely prefer tBTC to wBTC


Thanks for posting @mhluongo!

The main reservation I have is the risk of tBTC having problems if the ETH/BTC price drops. This means tBTC vaults would be vulnerable to 3 types of price action (instead of 1 for typical vault):

  • BTC price drop
  • ETH price drop
  • BTC price increase

Great point, though I’m not sure it plays out the way you’d expect. ETH drops relative to USD don’t matter here, only drops in ETH relative to BTC.

If ETH does drop severely relative to BTC, deposits will be redeemed or signer bonds will be seized. Any drop under 33% in ~3 hours won’t have an opportunity to threaten the peg, and a bigger drop than that won’t impact the peg unless signers are also malicious and aren’t worried about losing their work token.

Happy to put together a spreadsheet to show what I mean if that’s helpful :slightly_smiling_face:

EDIT: To put a finer point on it, tBTC isn’t a synthetic. The same design can work without the ETH if signers are assumed honest or if the work token is valuable enough- the ETH is a “second layer of defense”.


tBTC is the best when it comes to security and decentralization
I hope it can scale

1 Like

Hey @mhluongo - really excited to see this application too. I have read the audits in particular, and like the design. I agree that it seems more in line with Maker’s principles than other types of collateral (though I am all for multiple kinds/approaches).

I have two questions:

  1. I’m not sure you answered @monet-supply’s actual question (s/he did also specify ETH/BTC, not USD). This spreadsheet would certainly help plebs like me.
  1. Why the rush? I’m inclined to agree with some of the Foundation devs that letting the contracts exist for ~3months on mainnet is a good way to provide at least some surety that they won’t get hacked. The CD audit is thorough, but it does show up the inherent complexity in the system, and time is a much better test of this than any human, no matter how gifted. I suppose that the MIP process properly executed is at least 3 months, so that could be a vlaid reason for starting now…

It was a muddled response- I’ll see if I can put a couple clean examples together. My point though is that the peg isn’t “held” by ETHBTC stability like a pure synthetic- it’s held by signer honesty, and the additional ETH helps ensure that.

Just starting the discussion- it’s on the MKR holders to judge risk and timing here, but the sooner we hear concerns, the more time our team has to address them.


This makes sense. However, I’d assume the risk of signer dishonesty would be greater during the launch phase (lower value of work token / stakedrop period where KEEP is not required / fewer independent signers participating), so ETH does seem like an integral part of the system for the time being

1 Like

I’ve been following this project for a while and looking forward to it possibly being included in MKR. We should strongly consider it for inclusion.


The token isn’t launched yet. Is its code available somewhere?

Indeed! Check out https://github.com/keep-network/tbtc


Thanks. I think I found an issue with the code.

Good luck on your launch! Interesting that there could be more than 21m BTC via collateralized synthetics.

1 Like

Thanks for the report! I don’t think it’s an issue- signers have plenty of selfish motivation for courtesy calls- but I’m happy to be proven wrong if you want to demonstrate on Ropsten or give more details.

Appreciate it! A core goal of this project is to prevent “extra” BTC from being put into circulation, but sBTC and co are all over it :slight_smile:


I’m worried about the rush.
You are in such a hurry to start that it can be dangerous. Some aspects of the system’s operation are still unclear. I am concerned about such discussions when the official documentation says one thing and the official Twitter account says another

tBTC is definitely needed for MakerDAO, but you can not rush with someone else’s Bitcoins. Can it be reasonable to start the network first and add tBTC for MakerDAO after some time? Not the other way around.


Just saw this on the Keep twitter: https://blog.keep.network/keep-tbtc-launch-supported-by-more-than-40-industry-partners-ae5a2f0689be

Seems a little misleading with regards to Maker. Some quotes:

Here is the full list of launch partners announced to date:

tBTC integrations:

  • MakerDAO and OasisDEX

I don’t think we’re a launch partner? While it’s difficult to estimate consensus, from what I’ve seen it looks like there is a well-supported undercurrent of ‘let’s wait 3 months to make sure it works as advertised.’

MakerDAO, Compound and Uniswap are among the decentralized platforms with plans to integrate tBTC as their choice for BTC

Again, this is misleading, claiming tBTC is ‘our choice’ of BTC collateral completely disregards the fact that we added wBTC, and that it seems likely we will support multiple BTC-on-ethereum tokens.

Don’t get me wrong, I’m excited to see tBTC added but this sort of blogpost promises a lot of stuff on our behalf, which I don’t like.


Long, that wasn’t our intent at all.

These posts were reviewed by the Foundation, and went out before the community added WBTC outside normal collateral governance. Everyone mentioned in these posts gave sign-off.

I’d never try to speak for the entire community - just the MKR holders and Foundation we’ve been working with. I’m happy to revise to make the distinction between the Foundation and community more clear.

1 Like

Anton, please stop astroturfing. I’ve mentioned above that there is no rush- the Maker community has an established risk function to make this decision. That said, the fact that you’re a Ren partisan harassing us across multiple social media channels is making a good faith discussion difficult.

We’re working to improve our documentation, but the best possible documentation is an audited, open source system. You’re welcome to review the code and docs and open a pull request with improvements- we’d love to get you involved.

Of course, we would never :slight_smile: That’s why we have started a second audit, and will be limiting the dApp to small amounts at launch to protect users. That’s why we’ve gone out of our way to publish risks and a security model that won plaudits from our auditor.

The only way to be confident about a system like this is time and additional collateral. That’s why we’ve offered to backstop the collateral and pursue additional insurance. The Maker risk team and community are in the best place to judge the risks and benefits here

Fair enough. For future reference:
MakerDAO = amorphous blob that is incredibly poorly defined.
Maker Foundation = The Maker Ecosystem Growth Foundation
Maker Community = Amorphous blob - Foundation (Some contention around this too, should the Foundation be considered part of the community?)
MKR Token Holders = People with MKR
Maker Governance = People with MKR that vote + Community members involved in governance.

Cool, I probably need to be complaining elsewhere then.

Regarding revision, that’s your call. I am just worried this sets up expectations which seems unlikely to be met, namely:

  1. That tBTC will be the only BTC token in the Maker Protocol.
  2. That tBTC will be available to borrow against in the Maker Protocol at the launch of tBTC.

Agreed. I’ll take a pass at it with more accurate language and make it clear this isn’t exclusive. Thanks for the catch!

1 Like

I also found the article that Long linked to to be misleading and exaggerated. The same issue goes for Compound; there are no “plans to integrate” before a governance vote can be had, let alone as “their [singular] choice for BTC.” This is simply inaccurate. Price feed approvals and testnets are not mainnet integrations. A public retraction would be appropriate and appreciated.