The Bitcoin Credit Stack
rates are percent a year, continuously compounded · figures as of 2026-10-15 · read 2026-10-16 04:01 UTC
Indicators › Cost of credit › Borrow over lend rate, per class

Borrow over lend rate, per class

intermediation-spread

What borrowers pay over what lenders earn, per class, where both legs come from the same venues: pools (from 2022-09-19) and the desks that post both rates (from 2026-10-07). Not at market level, where the two legs come from different classes; not for minted stablecoins (no lender) or auctions (one price). Recast on 2026-10-09 from a one-venue comparison.

Latest reading
1.42 points
2026-10-15 · derived
Unitspoints (continuously compounded)
Computed at2026-10-15
Historyserved

Methodology descriptive · IX1

What it measures. What borrowers pay over what lenders earn, per class: the class’s dollar-debt-weighted borrow rate minus its supply-weighted lend rate, both continuously compounded.

Where it is shown. Only where both legs come from the same venues: the pools, and the lending desks that post a borrow and a lend rate. Not for the market as a whole, where the lend rate is almost all pools while the borrow rate includes minted and desk loans; not for minted stablecoins, which have no lender; not for auctions, where borrower and lender meet at one price.

Read with care. The wedge is the venues’ reserve and the part of supply not lent out, not a profit figure. Coverage steps on either leg move it. Desk rates are posted, not cleared.

Sample. One credit cycle, from 2022-09-19, so it is observational. A descriptive series: not an index, not a forecast and not a signal. Every published version is kept and can be read as known on any date (as_known_on).

History

Borrow over lend rate, pools

What borrowers pay over what lenders earn across the pools, both legs from the same venues. The desks that post both rates (from 2026-10-07) are in the reading and on the explorer. From 2022-09-19, one credit cycle; weekly (Mondays) plus the latest day.

2026-10-15Pools 1.42 points
0 pts1 pts2 pts3 pts4 ptspoints2023202420252026
Pools

Mondays are drawn here; every day is in the explorer. Open in the explorer →

What this cannot be read as

What borrowers pay over what lenders earn in the algorithmic class: the tree's dollar-debt-weighted borrow rate (weighted by dollar debt for 65 of 98 venues) minus its supply-weighted lend rate (weighted by dollar supply for 60 of 98 venues), both continuously compounded. The wedge is the venues' reserve and the unlent part of supply, not a profit figure. Coverage steps on either leg move it. Observational: one credit cycle, from 2022-09-19.

https://www.bitcoincreditstack.com/indicators/intermediation-spread

Full payload

Every field as the API serves it
idintermediation-spread
nameBorrow over lend rate, per class
measureCX-IX1
questionWhat do borrowers pay over what lenders earn, in each class where both legs come from the same venues (pools; desks posting both)?
headlinecx.class.algorithmic.intermediation_spread
sourceDerived in the vintage store: credit_tree borrow rate − lend rate
series
2 entries
[
  {
    "series_id": "cx.class.algorithmic.intermediation_spread",
    "date": "2026-10-15",
    "value": 1.418383928137,
    "unit": "percent_cc",
    "kind": "derived",
    "coverage_lend": {
      "in": 60,
      "of": 98,
      "markets": {
        "in": 27,
        "of": 27,
        "weight_in_usd": 2820769105.498754,
        "weight_of_usd": 2820769105.498754
      }
    },
    "coverage_borrow": {
      "in": 65,
      "of": 98,
      "markets": {
        "in": 27,
        "of": 27,
        "weight_in_usd": 2520850279.856482,
        "weight_of_usd": 2520850279.856482
      }
    },
    "limitation": "What borrowers pay over what lenders earn in the algorithmic class: the tree's dollar-debt-weighted borrow rate (weighted by dollar debt for 65 of 98 venues) minus its supply-weighted lend rate (weighted by dollar supply for 60 of 98 venues), both continuously compounded. The wedge is the venues' reserve and the unlent part of supply, not a profit figure. Coverage steps on either leg move it. Observational: one credit cycle, from 2022-09-19.",
    "vintage_id": "cx@2026-10-15T08:15:00.589Z"
  },
  {
    "series_id": "cx.class.posted-card.intermediation_spread",
    "date": "2026-10-15",
    "value": 8.246410788005,
    "unit": "percent_cc",
    "kind": "derived",
    "coverage_lend": {
      "in": 2,
      "of": 24
    },
    "coverage_borrow": {
      "in": 5,
      "of": 24
    },
    "limitation": "What borrowers pay over what lenders earn in the posted-card class: the tree's dollar-debt-weighted borrow rate (weighted by dollar debt for 5 of 24 venues; 5 posted rate(s) carry no weight and are not in it) minus its supply-weighted lend rate (weighted by dollar supply for 2 of 24 venues), both continuously compounded. The wedge is the venues' reserve and the unlent part of supply, not a profit figure. Coverage steps on either leg move it. Observational: one credit cycle, from 2022-09-19.",
    "vintage_id": "cx@2026-10-15T08:15:00.589Z"
  }
]
methodology
urlhttps://www.bitcoincreditstack.com/indicators/intermediation-spread
noteChanges to how a measure is derived are recorded in the runbook and every published version is kept (as_known_on).
sampleOne credit cycle, from 2022-09-19: observational. Descriptive, not a forecast or a signal, and not an index.

Where this comes from

REST
GET https://api.bitcoincreditstack.com/v1/market/measure/intermediation-spread
MCP
get_indicator({ id: "intermediation-spread" })
get_indicator({ id: "intermediation-spread", include_history: true })
get_series({ ids: ["cx.class.algorithmic.intermediation_spread", "cx.class.posted-card.intermediation_spread"] })

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

Ask for this indicator's reading and history in your own terms, by connecting.Connect your AI ↗