A Roadmap to WBTC Decentralization
Today, WBTC’s custody model has migrated from a 2-of-3 multisig with a single US custodian to a new multi-jurisdiction model.
While the change has been controversial, moving from a single point of failure could mark the beginning of a new stage for WBTC.
In this proposal, I outline how Threshold might work together with BitGo, BitGlobal, and the greater WBTC ecosystem to further decentralize WBTC, maintaining its healthy lead on other BTC tokens across DeFi.
Background
WBTC is the largest Bitcoin-backed ERC-20 token on Ethereum, and tBTC’s largest competitor.
Founded by BitGo and a number of early DeFi enthusiasts in 2019, WBTC has been the de-facto choice for DeFi users who want exposure to BTC across Ethereum, Arbitrum, Optimism, and other popular chains. WBTC makes use of a network of merchants to mint and burn the token as users deposit BTC, but until recently, it only relied on a single BTC custodian — BitGo.
In 2020, tBTC was launched as a decentralized alternative to WBTC. tBTC proved to be a robust Bitcoin bridge, but struggled to scale to meet demand. tBTC v2 was built to address these issues, and since its launch, has securely bridged over $1.2B in BTC across DeFi.
On August 9th, BitGo announced a plan to change how the BTC behind the WBTC token is custodied. This plan led to serious concerns across many DeFi protocols, some of which BitGo was able to successfully address.
Last month, MacLane proposed a merger between WBTC and tBTC, including a high-level plan to migrate WBTC users to tBTC.
What’s changed?
Since MacLane’s proposal last month, there have been some major market developments.
- The trend of exchange-backed Bitcoin has taken off with the introduction of cbBTC, with Kraken and a number of other exchanges rumored to be planning their own offerings. While the wider access to BTC across DeFi is positive, many are worried about “paper Bitcoin” flooding the market.
- Chaos Labs, BGD Labs, and Llama Risk all released positive risk reviews of tBTC.
- GMX has launched a tBTC market, and is growing steadily on Arbitrum.
- tBTC and cbBTC have both been onboarded as collateral on Aave, ending their sole reliance on WBTC.
- cbBTC is being onboarded on Sky (nee MakerDAO), ending their sole reliance on WBTC.
We’ve also been party to multiple discussions with BitGo, BitGlobal, Aave, Sky, and a slew of other DeFi protocols who want to ensure safe, neutral access to BTC for their users — without relying on a single point of failure.
Alignment
After discussing the plans for WBTC’s custody with nearly everyone involved, I’ve come to believe that despite the plan’s reception across the space, the recent changes to WBTC have been proposed in good faith — and that ultimately, we all have more in common than not.
- We all believe deeply in the power of censorship-resistant non-government money.
- We all believe Bitcoin is bigger than a single jurisdiction, and vital to the growth of DeFi.
- We all believe in transparent custody — after all, WBTC and tBTC are the only major BTC-backed tokens with Proof of Reserves today.
So, how can we get aligned? We can align our interests, and we can align our outcomes.
Aligned Outcome: a secure, decentralized, neutral, cross-jurisdiction Bitcoin-backed token for use across DeFi.
Aligned Interests: ownership and governance participation in the network behind that token, structured so that we can all grow sustainably together.
Here’s how I think we can align all parties to build that shared vision, combining the scale of WBTC and the decentralized network behind tBTC.
Outcome
Today, BitGo has expanded the WBTC signer set from one legal entity to three — BitGlobal will have one key, BitGo Singapore will have another, and BitGo in the US will have a third. This plan better spreads risk across geographies, and depending on legal details, might prevent unilateral seizure of funds from 1 of the 3 relevant jurisdictions.
Paraphrasing Jameson Lopp, an early BitGo engineer and Bitcoin security expert, censorship resistance can be measured by the number of doors you need to kick down to censor a system.
Adding 2 more distributed custodians to a 2-of-3 multisig backing WBTC doubles the censorship resistance of WBTC. It’s a strong step forward.
I propose we continue BitGo’s plans to decentralize WBTC, expanding the signing set from 3 to hundreds via the Threshold network, and establishing BitGo and BitGlobal as major participants and partners.
Interests
Similar to MacLane’s original proposal, I propose a surgical T mint to better align our interests and welcome new stakers to the network.
Here’s an overview of the economics terms of the proposal.
- 16% T mint
- 10% is sold by the DAO to stakeholders
- The sale is conducted in tranches, incentivizing earlier participation, and prioritizing existing DeFi protocols and stakeholes.
- $20M of the proceeds are converted to tBTC, and granted to BitGo with performance-based vesting
- 5% is set aside to align BitGo and BitGlobal as Threshold network participants
- 2.5% is granted to BitGo, with performance-based vesting
- 2.5% is granted to BitGlobal, with performance-based vesting
- 1% is set aside to align other WBTC stakeholders, including the WBTC DAO and merchants.
- 10% is sold by the DAO to stakeholders
- BitGo and BitGlobal are immediately delegated 40M T each to join the Threshold network, and after 90 days of staking, to be considered for the beta staker program
- All WBTC DAO members are offered a delegation of 10M T each to join the Threshold network, and after 90 days of staking, to be considered for the beta staker program
- All WBTC merchants and exchanges are offered a delegation of 5M T each to join the Threshold network, and after 90 days of staking, to be considered for the beta staker program
Decentralizing WBTC
Rather than an upfront merger, the idea here is to slowly grow the signer set of WBTC, avoiding making changes to related contracts, and maintaining today’s market structure.
Phase 1 — BitGo and BitGlobal join the Threshold network (1 month)
In the first phase, the DAO delegates 40M T to both BitGo and BitGlobal to run nodes, enabling both firms to join the Threshold network.
Both firms will be able to take part in the decentralized custody backing tBTC, earning T emissions, while getting to know the Threshold network — and further decentralizing custody in the network.
In addition, all WBTC DAO members and merchants with the operational capability will be offered smaller delegations, further aligning interests and strengthening the network.
Phase 2 - Establish a fee-free decentralization path for WBTC holders (3-12 months)
Next, we’ll establish a cost-effective path for WBTC holders to decentralize their custody.
- The WBTC custodian will waive all burn fees for holders wishing to decentralize custody.
- As confidence builds in the new system, a contract to atomically mint tBTC from WBTC can be deployed.
- Finally, we’ll introduce a vending machine for smooth 2-way conversion between the two tokens.
Phase 3 - Decentralize WBTC operational overhead (12+ months)
At this point in the roadmap, tBTC and WBTC will be largely interchangeable, with two different minting schemes.
In the final phase, merchants will migrate to decentralized minting, and the WBTC token contract ownership can be transferred to the Threshold DAO or to the 0x0 address. Any remaining BTC held by the WBTC custodian can be migrated to the threshold wallets backing tBTC, fully decentralizing custody.
Together, we’ll have achieved a widely used, decentralized BTC-backed token with easy onboarding for retail and institutions alike.
Milestone-based vesting
To align all parties to this roadmap, we turn to the T and tBTC vesting schedules. As BTC is migrated to the larger signer set, the grants for BitGo and BitGlobal will vest, leaving them free to stake in Threshold or sell.
| BTC migrated | T unlocked | tBTC unlocked |
|---|---|---|
| 5k BTC | 20% | 20% |
| 25k BTC | 20% | 20% |
| 75k BTC | 20% | 20% |
| 100k BTC | 20% | 20% |
| Full migration | 20% | 20% |
Hardening tBTC’s custody
Today, BitGo custodies over $9B backing WBTC alone, and far more including the rest of their customers.
We’ve been building tBTC for that scale and beyond for the past 5 years. The model has proven robust, but security is never “done”. We have a lot to learn from BitGo’s defense-in-depth practices, and I’m excited for the opportunity to work together to secure the network.
Below, I’ve shared a few planned technical upgrades to increase tBTC’s resilience.
Phase 1 — Increase the signer set
There are 258 distributed nodes running tBTC today, with over 3B T staked.
These numbers are great, but they don’t tell the whole story. The nodes participating in custody are limited by the Beta Staker program, a temporary, permissioned list of stakers who are chosen to custody the newly deposited BTC backing tBTC.
Today, there are 35 beta stakers as approved by TIP-067. The easiest way to immediately harden the network is to grow that number, inviting other reputable organizations to stake on Threshold.
In addition to BitGo and BitGlobal, delegating to all operational WBTC DAO members, merchants, and exchanges could bring participation to over 100 reputable nodes.
Phase 2 — Upgrade to Schnorr / ROAST
The Beta Staker program exists to address a denial-of-service flaw in tBTC’s implementation of threshold ECDSA, often known as GG18.
ECDSA signatures are incredibly popular in the crypocurrency space, and are the default signature scheme used by both Bitcoin and Ethereum wallets. Unfortunately, it’s a difficult scheme to build a secure threshold implementation — it wasn’t designed to be easily decentralized.
ECDSA was likely chosen by Satoshi because the best alternative — the Schnorr signature scheme — was patented at the time. If history had played out differently, we might not use ECDSA signatures at all!
Schnorr, on the other hand, simplifies threshold signature construction. In 2021, Schnorr signatures were activated as part of the Taproot upgrade on Bitcoin, and in 2022, the advent of ROAST enabled an attributable, DoS-resistant decentralized signing protocol.
Upgrading to Schnorr/ROAST will enable larger, performant wallets — think 1000+ members — as well as allow the complete retirement of the Beta Staker program. To see progress and performance benchmarks, check out our initial public implementation.
Phase 3 — BitVM2 vaults
The Taproot soft-fork didn’t just bring us Schnorr signatures. Taproot also extended Bitcoin Script, allowing for more expressive off-chain smart contracts.
Last year, building on Taproot, Robin Linus came up with a challenge-response protcol that he dubbed BitVM. BitVM allows fraud proofs to be “settled” on Bitcoin L1, similar to how fraud proofs on optimistic rollups like Arbitrum are settled on Ethereum. Instead of relying on an honest majority assumption, like today’s tBTC, BitVM only requires a single honest participant (1-of-n).
Over the past year, there have been a number of improvements on the BitVM construction, including BitVMX and BitVM2. Unfortunately, BitVM-based bridging still relies on specialized bridge operators to deposit BTC. Individuals won’t be able to bridge funds easily themselves, instead depending on large holders.
This experience is a huge impediment to adoption. Sometimes, you have to look backward to move forward.
We plan to integrate BitVM vaults directly into the tBTC bridge, moving to a two-tiered custody model remniscent of cold / hot wallet schemes on centralized exchanges. End users will maintain a simple deposit experience, with the additional security benefit of “unilateral exit” for the majority of bridged BTC.
This effort is still under active research.
Improving economic security
The tBTC v2 security model is an honest majority with forward security. This model lessens the need for additional economic security, and enables serious capital efficiency.
Still, economic security is vital to building trust as we onboard the rest of Bitcoin to DeFi.
Funding the DAO
For tBTC to maintain a forward-looking security posture, the DAO needs a robust, independent treasury.
After MacLane’s proposal last month, a number of major DeFi protocols and other stakeholders reached out to fund an independent tBTC. Interest at the time was over $33M, and has grown since then.
I propose the DAO form a funding working group to capture this interest, conducting a sale of up to 10% of the token supply — nearly 2/3 of the proposed 16% mint. The working group should have discretion to identify and dilligence values-aligned buyers, as well as structure the sale in a way that’s beneficial for both the treasury and tBTC’s utility.
A revamped economic model
The long-term success of tBTC depends on the long-term success of the Threshold network at large.
Today, tBTC contributes to Threshold’s success via mint and burn fees. While the velocity of the tBTC bridge is far higher than WBTC, there’s a real concern that it won’t be sustainable without additional value capture.
For that reason, I propose the DAO form an economic model working group to reconsider how T and tBTC interoperate, with a plan to deliver a recommendation to the DAO by early next year.
Next steps
- Solicit input from stakeholders — BitGo, BitGlobal, the WBTC DAO, large WBTC holders, and large venues like Aave and Sky. This is going to take time, but I’m confident that with more feedback, we can get everyone aligned.
- Identify candidates for a funding working group
- Identify candidates for an economic model working group
Edit
Cleaned up language that confused multiple reviewers, and clarified token amounts.
