Bitcoin's dormant capital — estimated at over $1.5 trillion in holdings that generate no native yield — is now the target of a two-front competition. Centralized exchanges and native DeFi protocols are both building infrastructure to capture Bitcoin deposits, but with fundamentally different arch...
"Many bitcoin holders on Kraken have made it clear they want simple, safe ways to earn on the bitcoin they already plan to hold. Bitcoin Vault is built for that mindset." — John Zettler, GM of Payward Services and Head of Kraken Earn Products
Bitcoin's dormant capital — estimated at over $1.5 trillion in holdings that generate no native yield — is now the target of a two-front competition. Centralized exchanges and native DeFi protocols are both building infrastructure to capture Bitcoin deposits, but with fundamentally different architectures, risk profiles, and economic models.
On one side, Kraken launched Bitcoin Vault on May 26, offering up to 2.5% APY through Veda-powered vaults deploying into Aave, Morpho, and Tydro. Coinbase has originated $1.2 billion in Bitcoin-backed USDC loans via Morpho using cbBTC as collateral. On the other, Babylon Protocol holds $4.8 billion in TVL with 56,853 natively staked BTC. Lombard Finance controls 60% of the Bitcoin liquid staking market with $1.5 billion in TVL. Stacks reports $545 million in sBTC deposits.
The combined BTCfi sector holds approximately $5–7 billion in TVL as of May 2026, down 23% from its October 2025 peak of $9.1 billion. The contest between CeFi-mediated yield and native Bitcoin DeFi protocols will determine who controls the economic value extracted from idle BTC — and who bears the risk when smart contracts fail.
Bitcoin DeFi TVL stood at approximately $5.06 billion as of May 24, 2026, according to DefiLlama, recovering from a temporary drawdown. The sector peaked at $9.1 billion in October 2025 and remains 23% below that level.
The top five BTCfi protocols by TVL:
| Protocol | TVL | Mechanism | Custodial Model | |----------|-----|-----------|-----------------| | Babylon Protocol | $4.8B (56,853 BTC) | Native BTC staking | Non-custodial, no wrapping | | Solv Protocol | $2.15–2.8B (19,456+ BTC) | SolvBTC tokenization | Bridge-dependent | | Lombard Finance | $1.05–1.5B | LBTC liquid staking | Babylon-backed | | Stacks (sBTC) | $545M | 1:1 BTC-backed token | Signer network | | Coinbase (cbBTC on Morpho) | $1.4B collateralized | Wrapped BTC lending | Coinbase custody |
The data shows concentration risk: Babylon alone controls over 80% of native Bitcoin staking TVL. Solv Protocol operates across 11 chains with 325 integrated projects and 1.2 million users but generates only $41 in daily protocol revenue against $2.8 billion in TVL — a ratio that raises questions about sustainable fee generation.
Launched May 26, 2026, Kraken's Bitcoin Vault offers BTC-denominated rewards at up to 2.5% APY. The product architecture chains three intermediaries: Veda handles strategy execution, Sentora manages risk, and Chaos Labs provides monitoring. User BTC is deployed into Aave, Morpho, and Tydro — all Ethereum-based lending protocols.
Kraken's broader DeFi Earn product has accumulated $240 million in assets since its January 2026 launch. The Bitcoin Vault charges a 25% performance fee on rewards. Withdrawals can be requested at any time, but may take up to five days during constrained liquidity periods.
The yield source is straightforward: BTC is wrapped and lent to borrowers on DeFi protocols. Borrowers pay interest; that interest, minus the 25% performance fee, flows to depositors. The yield is real but the chain of custody is long — user funds pass through Kraken, a wrapping layer, a strategy executor (Veda), a risk manager (Sentora), and at least one DeFi protocol before generating returns.
Coinbase has taken a different approach. Since launching cbBTC-backed USDC lending through Morpho in April 2025, the program has originated over $1.2 billion in loans, with $800 million currently active and $1.4 billion in cbBTC collateral posted. The borrowing rate sits at approximately 6% — below many competitors.
cbBTC is a 1:1 Bitcoin-backed ERC-20 token held in Coinbase custody. When a user borrows, their BTC is converted to cbBTC and deposited into Morpho. The process is reversible at no fee. Coinbase functions as both custodian and distribution layer, while Morpho provides the lending market infrastructure.
The key distinction: Coinbase's model turns Bitcoin into collateral for borrowing, not a yield-bearing deposit. Yield accrues to USDC lenders on Morpho, not to Bitcoin depositors. BTC holders retain price exposure but pay 6% for USDC liquidity rather than earning yield.
Babylon is the largest native Bitcoin staking protocol by a significant margin. Its $4.8 billion in TVL represents 56,853 BTC staked without wrapping or bridging. BTC holders lock their Bitcoin directly to secure proof-of-stake chains, earning rewards denominated in the BABY token (launched January 2026) and potentially in the native tokens of the chains they secure.
The protocol's architecture avoids the custody risk inherent in wrapped Bitcoin models. Holders retain private key control. However, a vulnerability disclosed in May 2026 by pseudonymous researcher GrumpyLaurie55348 revealed a flaw in Babylon's BLS vote extension scheme. The bug could allow malicious validators to omit the block hash field during vote extensions, causing runtime panics during epoch boundaries. The vulnerability has not been actively exploited, but its disclosure underscores the nascent state of the codebase securing nearly $5 billion in Bitcoin.
Babylon received $15 million from a16z crypto in January 2026 and has partnered with Aave Labs to bring Bitcoin-backed lending to Aave v4.
Lombard's LBTC token represents 60% of the Bitcoin liquid staking market, with $1.05–1.5 billion in TVL. The token reached $1 billion in TVL in 92 days — the fastest-growing yield-bearing token on record. Lombard is integrated with over 70 DeFi protocols, including Aave and Morpho, meaning LBTC can be used as collateral, liquidity provision, or yield input across the DeFi stack.
Lombard sits atop Babylon's staking layer: users deposit BTC into Babylon through Lombard and receive LBTC, a liquid representation of their staked position. This creates a dependency chain — LBTC's value relies on Babylon's security, which in turn depends on the correctness of its staking contracts.
Stacks reported $545 million in sBTC TVL as of Q1 2026, with the deposit cap fully removed. The network's DeFi ecosystem shows $121 million in actively deployed capital across protocols: Zest Protocol ($75.9 million), Granite ($26 million), and StackingDAO ($20 million).
The March 2026 SIP-039 upgrade boosted DeFi capacity by 30x. Fireblocks integrated Stacks in early 2026, giving 1,800 institutional clients access to Bitcoin DeFi. Circle's USDCx went live on Stacks — the only Bitcoin L2 in Circle's xReserve pilot program.
sBTC uses a signer network to manage the Bitcoin peg. The system relies on a distributed set of signers rather than a single custodian, positioning it between fully centralized wrapping (cbBTC) and fully non-custodial staking (Babylon).
Solv Protocol holds $2.15–2.8 billion in TVL across 11 chains with 19,456 BTC in reserves. SolvBTC serves as a 1:1 tokenized Bitcoin for cross-chain DeFi usage. In March 2026, a security vulnerability was exploited 22 times, resulting in the theft of 38 SolvBTC ($2.7 million). The team committed to reimbursing affected users.
The protocol generates $41 in daily revenue against billions in TVL — a figure that makes economic sustainability questionable without token-based subsidies or future fee increases.
Each model carries distinct risk vectors:
CeFi-Mediated Yield (Kraken, Coinbase)
Native BTCfi Protocols (Babylon, Lombard, Stacks, Solv)
The April 2026 Kelp DAO exploit serves as a stress test for the entire Bitcoin yield ecosystem. An attacker drained 116,500 rsETH ($292 million) from Kelp DAO's LayerZero bridge — the largest DeFi exploit of 2026. The attacker then deposited approximately 90,000 rsETH into Aave as collateral, borrowing $190 million in ETH and other assets.
The result: Aave absorbed approximately $196 million in bad debt. Its TVL dropped from $48.5 billion to $30.7 billion in a single day as depositors withdrew. A recovery initiative called "DeFi United" — backed by Consensys, Lido, EtherFi, and others — raised $160 million of the $200 million needed to cover losses.
This matters for Bitcoin yield products because both Kraken Bitcoin Vault and Coinbase's cbBTC program deploy into Aave. Had this exploit occurred with BTC-denominated deposits in the vault, Kraken users would have been exposed to the same bad-debt contagion. The 25% performance fee Kraken charges does not cover principal losses from protocol-level exploits.
The Kelp DAO incident demonstrates that CeFi yield products do not eliminate DeFi risk — they repackage it. The user interface is simpler, but the underlying exposure is identical: smart contract risk on Aave, Morpho, and other protocols, plus additional intermediary risk from the wrapping and custody layers.
The yield on Bitcoin — whether 1% or 2.5% — must originate from an economic activity. In lending protocols, borrowers pay interest. In staking, secured chains pay validators. The question is how much of that yield reaches the BTC holder versus the intermediaries.
Kraken's fee structure: 25% performance fee on rewards. On a 2.5% gross APY, the user receives approximately 1.875% net. Veda, Sentora, and Chaos Labs extract additional fees from the protocol layer. The total intermediary take likely exceeds 30% of gross yield.
Coinbase's structure: No direct yield to BTC holders. Borrowers pay ~6% on USDC loans; cbBTC lenders on Morpho earn a share of that interest. Coinbase earns through custodial services and the lending spread.
Babylon: Yield comes from securing PoS chains, paid in BABY tokens and native chain tokens. No intermediary fee layer — but the yield is token-denominated, not BTC-denominated, introducing conversion risk.
Lombard: Takes a spread between Babylon staking rewards and LBTC yield passed to holders. Exact fee structure varies by integration.
Stacks: sBTC holders can earn BTC-denominated yield through Proof of Transfer (~3.5% via Zest Protocol's dual stacking). This is one of the few models where yield is paid in BTC from Bitcoin mining rewards, not from a secondary token.
The economic value chain reveals a pattern consistent with the broader blockchain ecosystem: multiple intermediaries extract fees at each layer, and the end user's net yield is a fraction of the gross economic activity their capital enables.
The competition for Bitcoin yield is a proxy war over who controls the economic rent from $1.5 trillion in idle BTC. CeFi exchanges offer convenience and familiar interfaces but layer multiple intermediaries — each extracting fees and introducing failure points — on top of the same DeFi protocols that native BTCfi projects use directly. Native protocols offer more direct access but carry immature codebases, disclosed vulnerabilities, and token-dependency risks.
Neither model has solved the fundamental tension: Bitcoin was designed as a bearer asset with no native yield. Every yield source requires converting BTC into a different instrument — wrapped, staked, lent, or tokenized — each transformation adding risk layers that did not exist in the original protocol.
The Kelp DAO exploit demonstrated that the current DeFi infrastructure cannot yet absorb a single large-scale failure without systemic contagion. As both CeFi and DeFi players funnel more Bitcoin into these systems, the scale of potential losses grows proportionally. The sector's $5–7 billion in TVL is small relative to Bitcoin's total market capitalization, but the infrastructure supporting it is being stress-tested in real time.
For BTC holders evaluating yield options, the relevant metric is not the advertised APY — it is the risk-adjusted return after accounting for smart contract exposure, intermediary fees, custody risk, and the probability-weighted cost of a protocol failure. On that basis, the data suggests that Bitcoin yield products remain in an experimental phase, regardless of whether they are offered by a centralized exchange or a native protocol.