Two BTC-aligned chains. Two completely different answers to the question of where your bitcoin actually sits while it earns yield or runs a smart contract.
In Part 3 we looked at Stacks and the sBTC threshold-signer design — a separate chain anchored to Bitcoin with a 15-of-100 signer set custodying the peg. This part compares two systems that take that same problem and pull it in opposite directions. Babylon says: the BTC never leaves Bitcoin. Rootstock says: hand the BTC to a federation and we'll give you a full EVM in return. Same goal — productive bitcoin — almost nothing else in common.
Two systems, one question: where does the BTC live?
Every Bitcoin L2 or BTCFi protocol eventually answers one question: when your BTC is "working", whose hands is it actually in?
- Babylon keeps your BTC in a Taproot UTXO on the Bitcoin chain itself. You alone can spend it after a timelock. The protocol gets economic security from the threat of slashing, not from holding the coins.
- Rootstock takes your BTC, locks it in a 5-of-9 federation multisig (the PowPeg), and mints rBTC on a separate EVM chain. The federation custodies; you get smart contracts.
One is custody-shaped risk. The other is operational/slashing-shaped risk. Neither is strictly better — they answer different questions.
Babylon: self-custodial BTC staking
Babylon was founded in 2022 by David Tse's group at Stanford. Phase 1 mainnet went live on August 22, 2024, and the numbers were absurd from day one. Cap-1 (1,000 BTC) filled in 74 minutes. Cap-2 added another ~23,000 BTC. Cap-3 brought the total over 57,000 BTC staked in Phase 1 alone. Babylon Genesis — Phase 2, the dedicated PoS chain that consumes Babylon's BTC security as the first "Bitcoin Supercharged Network" (BSN) — launched April 10, 2025 at 10:00 UTC.
The mechanism is the clever part. When you stake BTC with Babylon, your coins go into a Taproot output with two spending paths:
- Staker path: after the timelock expires, you (and only you) can spend the UTXO with your own key. Self-custody is preserved.
- Slashing path: spendable by a covenant-emulator committee plus a secret derived from a finality provider's signature, but only to a hardcoded burn address.
That second path is enforced by a primitive called Extractable One-Time Signatures (EOTS). A finality provider signs each block with an EOTS key. If they ever sign two conflicting blocks at the same height — i.e. attack the BSN they secure — the maths of EOTS leak the secret key. Anyone can then construct the slashing transaction and burn the staker's BTC. Sign once: safe. Sign twice: your delegators' BTC goes to a provably unspendable address.
Why Babylon needs a covenant emulator
The honest version: Bitcoin doesn't have native covenants. Proposals like CTV and OP_CAT have been debated for years and as of 2026 are still unresolved. Without covenants, you can't natively express "this UTXO can ONLY be spent to address X under condition Y" at the script level.
Babylon's workaround is the covenant-emulator committee — a multisig that pre-signs the staking and slashing transitions when you stake. Importantly, the committee cannot redirect funds. The script enforces that the only non-staker-path destination is the pre-agreed burn address. They can't steal.
They can censor — refuse to sign your unbonding transaction. That's a real liveness assumption you should price in. Babylon's stated plan is to retire the emulator the moment Bitcoin activates real covenants. Until then, the trust assumption is honest about what it is.
BABY token and the Q1 2026 picture
The BABY token launched alongside Genesis on April 10, 2025. Fixed initial supply of 10 billion, with 8% first-year inflation split between BTC stakers and BABY stakers — so your stacked sats earn yield denominated partly in BABY. Unbonding BABY itself takes ~50 hours. Allocation: 15% community, 18% R&D, 30.5% private investors, 15% team, 3.5% advisors.
TVL hit a peak above $7B in 2025 and then ate a body blow. On April 17, 2025, Lombard Finance and other institutional stakers unstaked $1.26B (14,929 BTC) in a single coordinated wave. Babylon TVL dropped roughly 32% from ~$3.97B to ~$2.68B overnight. By Q1 2026 it had recovered to roughly $3.6B–$4.95B, with around 250 finality providers active across the BSN ecosystem.
The lesson isn't that Babylon broke — the slashing system worked exactly as designed, nobody got slashed, BTC went home. The lesson is that institutional flows dominate, and "self-custodial" doesn't mean "sticky".
Rootstock: the original Bitcoin sidechain
Now flip everything. Rootstock (RSK) has been running mainnet since January 2018 — it is, full stop, the first Bitcoin sidechain. EVM-compatible from day one, before "EVM on Bitcoin" was a tagline. Block time ~30 seconds. Gas token rBTC (Smart Bitcoin), pegged 1:1 to BTC.
Rootstock is merge-mined with Bitcoin. Bitcoin miners include the RSK block hash in their coinbase transaction; the same proof-of-work that secures Bitcoin extends to Rootstock at near-zero marginal cost, and miners earn RSK transaction fees on top of their BTC subsidy. As of 2026, over 80% of Bitcoin's hashrate merge-mines Rootstock. That's the same security argument as the 51% attack economics we covered earlier — except an attack on Rootstock has to fight Bitcoin's hashrate, not a separate, smaller validator set.
PowPeg: the federation that holds the BTC
Merge-mining handles consensus. Custody is a separate problem — and the answer is the PowPeg federation. It's currently a 5-of-9 multisig of independent "pegnatories": Luxor, Earn, Constata, pNetwork, Sovryn, Xapo Bank, BlockVenture, RootstockLabs, and others. Each pegnatory runs a PowHSM — a hardware security module derived from Ledger firmware that enforces peg rules at the chip level. A pegnatory operator can't sign anything outside the protocol's allowed transactions, even if their own infrastructure is fully compromised.
Two recent upgrades matter:
- Reed (Oct 1, 2025) activated SegWit on the PowPeg, cutting peg-out costs by roughly 60%, and raised the pegnatory cap from 9 to 20 (longer-term target: 60).
- PowPeg Composition Change scheduled at Bitcoin block 8,558,281 (~Feb 23, 2026) rotated the federation membership and changed the BTC deposit address. Users had to use the new address; old deposits still work, just slower.
No PowPeg breach since 2018. That's an eight-year track record, which is more than most things in crypto can claim. But it is custody — your BTC sits in those 9 hands while you use rBTC.
Rootstock TVL: small, but real BTCFi
Let's be honest about the size. Rootstock chain TVL in Q2 2026 is roughly $109M. Compare that to Babylon's multi-billion BTC stack and the gap is enormous. But Rootstock isn't competing on staked BTC volume; it's competing on "can you actually run BTC-collateralised DeFi".
The headline ecosystem is Sovryn at around $65M TVL — a vertically integrated stack of Zero (Liquity-fork lending), an AMM DEX, and a margin/perps venue. Sovryn issues DLLR (Sovryn Dollar), an aggregated BTC-collateralised stablecoin whose largest constituent is ZUSD, a 0%-interest overcollateralised stablecoin using Liquity-style mechanics. Around it: Money on Chain (the oldest BTC-collateralised stable, DoC), Tropykus lending markets, and BabelFish as a stablecoin bridge.
This is real BTCFi — small, but composable, with users actually borrowing rBTC against rBTC collateral and minting BTC-backed dollars. The kind of thing that's structurally hard on a non-Turing-complete chain.
Same goal, opposite trust models
Line them up:
| Babylon | Rootstock | |
|---|---|---|
| BTC location | Self-custodial Taproot UTXO on Bitcoin | 5-of-9 PowPeg federation multisig |
| Security model | Slashable economic stake, EOTS double-sign extraction | Bitcoin merge-mining + HSM-enforced custody |
| What you give up | Liveness — committee could censor unbonding | Custody — federation holds the BTC |
| What you get | BTC yield from securing BSNs | EVM smart contracts, rBTC, BTCFi apps |
| Mainnet | Phase 1: Aug 22, 2024 / Genesis: Apr 10, 2025 | January 2018 |
| Approx scale (Q1–Q2 2026) | ~$3.6B–$4.95B TVL | ~$109M chain TVL |
If you want your BTC to stay BTC and earn yield by securing other chains, Babylon is the cleanest answer in the space. If you want to actually deploy BTC-collateralised DeFi today and you're willing to trust a hardware-enforced federation, Rootstock has been doing exactly that for eight years.
A sophisticated user might use both — most of the stack in cold storage (see our guide on holding bitcoin in Zelcore for the self-custody fundamentals), a slice in Babylon for staking yield, a slice in rBTC for active DeFi, and reserves on Lightning for payments. There is no one right answer; there are different risk shapes for different jobs.
Coming up: Part 5
We've now toured the four big architectural answers — BitVM2 / Citrea / Alpen (Part 2), Stacks / sBTC (Part 3), and Babylon / Rootstock (this one). Part 5 is the capstone: a unified framework for picking the right BTC L2 for the right job, what we expect by 2027, and the questions you should ask before parking a meaningful amount of bitcoin anywhere outside cold storage.
