Yield from the bedrock.

Bitcoin yield attribution on Solana. Every quoted rate split by who pays it.

CA
1EaGq2PtdsxhZicYsrMXcVMLXMdyqpbUNQJjfZBpump

_{The full address, never shortened. Check it against the mint before you act on it -- a truncated address is not an address, and every impersonation of a launch relies on the middle being hidden.}

A BTC yield product quotes one APY. That number hides the only question worth asking: who is paying it, and will they still be paying it next quarter?

We measured the Solana BTC yield landscape on 2026-08-15 using public endpoints only, before writing the program. Four results changed the design.

About $75.4M of BTC sits in these markets earning nothing, because interest is paid by borrowers and there are no borrowers. Kamino's cbBTC reserve is 3.2% utilised.

"Deposit BTC, earn lending interest" is not a seam that exists on this chain today.

Traders genuinely pay these. But the figure is fee revenue before divergence loss, and DefiLlama reports il7d as null for every one of these pools. The number quoted elsewhere is a gross figure presented as a net one.

We estimate the loss and subtract it. Where we cannot, the seam is marked ilUnknown and no net figure is given. We do not invent the estimate.

All 94 BTC-related pools, 647 days of history, not one with a non-zero reward APY.

This is a measurement, not missing data: the same snapshot finds 15 pools with apyReward > 0 elsewhere on Solana. The collector works. The number is zero.

That zero is the most useful thing on our dashboard. When a competitor advertises a double-digit BTC yield, one of three things is true: fee revenue is shown without divergence loss deducted, a points programme has been converted into an APY, or leverage is folded into the headline. Separating the three is the entire product.

GMTrade's BTC-USDC vault reported 214.828% on $1.71M. Its source is trader losses.

On a chart it looks like fee revenue. In character it is the opposite: it depends on someone else continuing to lose, and it inverts when they stop.

sustainable    trading fees, borrow interest -- money an outside user actually paid
emissions      protocol token emissions      -- money the issuer printed
counterparty   the other side's losses       -- money a trader lost

Most attribution models have two categories. The measurements say there are three.

The specifications cite the source lines that enforce them, so a claim in the docs can be checked against the program without taking either on trust.

Assets are identified by mint, never by symbol. Solana carries two different tokens called WBTC with eight decimals each, and one of them has $46K of liquidity.

Spot rates are never rendered. The Orca cbBTC-USDC history contains a single day printing 74,187% -- a low-TVL calculation artefact -- so display uses a seven-day value or a ninety-day median.

A rate on capacity nobody can use is not a seam. Zeus Bitcoin Market USDC advertised 104.6% against $10,927 of capacity, and sits below our TVL floor.

Points programmes are never converted to an APY. Attaching a price to unissued points relabels emissions as fee revenue, so that yield is absent from our figures rather than estimated into them. We would rather understate.

Basis seams are unsupported and say so. Drift and Zeta both return 403 without authentication, and Drift's BTC-PERP has not processed a funding update since 2026-04-01 while holding 250 BTC of open interest. Open interest is not liveness. The category is declared missing rather than quietly dropped.

Every asset routed is a wrapped or bridged claim on bitcoin held by a third party. It is not a coin on the Bitcoin base chain, and this codebase does not describe it as one.

Principal is at risk. Deposits are not bank deposits, carry no insurance, and can be impaired by a venue exploit, an issuer failure or a custody failure. Redemption is a claim on open positions and the queue lengthens under stress.

The program builds and its local test suite passes. It is not deployed to mainnet and it has not been audited.