The Bitcoin Credit Stack
rates are percent a year, continuously compounded · figures as of 2026-10-15 · read 2026-10-16 04:01 UTC
Market › BTC → ETH

BTC → ETH

1 class served in eth

BTC collateral, ETH borrowed: every class that prices borrowing in ETH, each in its own table. The classes form their rates differently, so they sit side by side and are not averaged into one rate.

As of 2026-10-15 · weighting: borrow by borrowed_usd, lend by supplied_usd; spread computed per pool-day then weighted, never as the difference of the two aggregates

The tenor is an averaging window, not a maturity. For the open-term classes (algorithmic, minted): every constituent pool is open-term floating and no borrower agreed a term, so a rising curve means rates were higher further back — not a term premium. For auction and posted card, each point is the mean of rates agreed for their own terms over the window; for corporate, the mean of the instruments' yields. The window is neither a loan's term nor an instrument's maturity.

Denominations are different products. btc and usd are different products and are never aggregated. btc = the borrower receives bitcoin; usd = a stablecoin is borrowed against bitcoin collateral.

On the collateral side. The curve keys its series by denomination, not by collateral, so this page cannot filter to BTC-collateralised pools — instead each point carries a collateral_coverage split showing how much of the book behind it is BTC, other, or mixed pooled. Where the mixed share is large the rate is not a BTC-collateral rate, and reading it as one would be the error this note exists to prevent.

BTC → ETH borrow rate — each class, over time
Collecting — no per-venue rate history is served for this class
The cross-venue surface has a daily series and the financing curve is a current cross-section, but neither carries a rate line per venue. The class and its constituents cannot be drawn against time until one exists.
Each class's current cross-section is below, by lookback window. Drawing it on a time axis would say 8 averaging windows were twelve months of history.
fill route: runbook_DI1_data_ingestion_coordination_v1.md
Window (days)BorrowLendSpreadPoolsVenuesPool debt (all borrowers)Supplied
71.95%1.16%0.60%9450$6.64bn$9.31bn
141.98%1.19%0.59%9450$6.90bn$9.56bn
302.01%1.21%0.59%9450$6.94bn$9.53bn
602.05%1.24%0.59%9450$6.64bn$9.04bn
902.06%1.24%0.60%9450$6.05bn$8.25bn
1802.19%1.31%0.61%9450$5.24bn$7.12bn
3652.24%1.37%0.63%9450$7.21bn$9.61bn
7302.45%1.57%0.66%9450$7.88bn$10.14bn
Collateral coverage per window
WindowBTCOtherMixed pooled
7$563k$329m$6.31bn
14$576k$340m$6.55bn
30$557k$375m$6.57bn
60$568k$386m$6.25bn
90$502k$367m$5.68bn
180$2.1m$356m$4.88bn
365$3.7m$433m$6.78bn
730$8.7m$396m$7.47bn

Ask for this by machine

REST
GET https://api.bitcoincreditstack.com/v1/market/financing-curve
MCP
get_credit_state()

The same figures, the same names and dates. Connect your AI →

What this pair costs for your size and tenor is a question for your own client.Connect your AI ↗