There is not a single simulated number on this page. 13 USD bonds, 6 issuers, $28.06B face value — terms copied line by line from SEC 10-Q / 8-K filings, benchmarked against the US Treasury curve for the pricing day. Where we cannot be real, we leave it empty rather than fill it with a placeholder.
Coupon, maturity, size and ranking are taken line by line from SEC filings; issue spread = coupon minus the Treasury yield interpolated to the matching tenor on the pricing date. This is the entire basis of NCSX-IS, and anyone with the same two public sources can replicate it.
Sector reference comes from ICE BofA option-adjusted spreads published via FRED — daily, real, cross-checkable: US high yield 3.40%, BB 2.03%, B 3.62%, CCC 13.16% (2026-11-03).
Registered funds must report every holding with its fair value to the SEC each quarter (Form N-PORT). We aggregate those filings and take the cross-fund median to back out a spread: 5 bonds currently carry observable marks, latest period 2026-08-31, covering $8.6B of face value. This is the price real managers carry their books at — not our estimate.
Daily traded price and day-over-day spread change require licensed FINRA TRACE data. Until that is in place those fields simply do not exist in NCSX — we do not publish a simulated series to fill the gap.
Sorted by size. A * marks a pricing date known only to the month; the benchmark still uses the actual trading-day curve from that month, so the issue spread is approximate rather than day-exact.
| Issuer | Coupon | Maturity | Size | Rank | Priced | UST | Issue spread | Source |
|---|---|---|---|---|---|---|---|---|
| Galaxy Digital GLXY | 9.875% | 2031-07-15 | $3.51B | Senior secured | 2026-07-24* | 4.43% | 545 bp | SEC ↗ |
| Core Scientific CORZ | 7.750% | 2031-05-01 | $3.30B | Senior secured | 2026-05-06* | 3.99% | 376 bp | SEC ↗ |
| TeraWulf WULF | 7.750% | 2030-11-15 | $3.20B | Senior secured | 2025-10-30* | 3.72% | 403 bp | SEC ↗ |
| CoreWeave CRWV | 9.750% | 2031-10-01 | $2.75B | Senior unsecured | 2026-04-15* | 3.94% | 581 bp | SEC ↗ |
| Applied Digital APLD | 9.250% | 2030-12-01 | $2.35B | Senior secured | 2025-11-14* | 3.74% | 551 bp | SEC ↗ |
| Applied Digital APLD | 6.750% | 2031-03-01 | $2.15B | Senior secured | 2026-03-04* | 3.67% | 308 bp | SEC ↗ |
| CoreWeave CRWV | 9.250% | 2030-06-01 | $2.00B | Senior unsecured | 2025-05-19* | 4.07% | 518 bp | SEC ↗ + |
| Cipher Mining CIFR | 6.125% | 2031-03-01 | $2.00B | Senior secured | 2026-02-05* | 3.75% | 238 bp | SEC ↗ |
| CoreWeave CRWV | 9.000% | 2031-02-01 | $1.75B | Senior unsecured | 2025-07-22 | 3.94% | 506 bp | SEC ↗ |
| Applied Digital APLD | 7.000% | 2031-06-15 | $1.59B | Senior secured | 2026-06-16 | 4.16% | 284 bp | SEC ↗ |
| Cipher Mining CIFR | 7.125% | 2030-12-01 | $1.40B | Senior secured | 2025-11-06* | 3.70% | 343 bp | SEC ↗ |
| CoreWeave CRWV | 9.625% | 2032-07-15 | $1.25B | Senior unsecured | 2026-06-18* | 4.29% | 534 bp | SEC ↗ |
| Cipher Mining CIFR | 6.000% | 2031-06-15 | $0.81B | Senior secured | 2026-06-09* | 4.26% | 174 bp | SEC ↗ |
US registered funds must report every holding at fair value to the SEC each quarter (Form N-PORT). Aggregating all filings for the same bond and taking the median, then backing out a yield and subtracting the Treasury benchmark, gives the price licensed managers actually mark their books at — not our model output.
| Issuer | Bond | Period | Price (median) | Implied spread | Funds reporting |
|---|---|---|---|---|---|
| CoreWeave CRWV | 9.750% 2031-10-01 | 2026-08-31 | 89.43 | 817 bp | 2 |
| CoreWeave CRWV | 9.250% 2030-06-01 | 2026-08-31 | 92.89 | 709 bp | 1 |
| CoreWeave CRWV | 9.000% 2031-02-01 | 2026-07-31 | 90.38 | 740 bp | 4 |
| CoreWeave CRWV | 9.625% 2032-07-15 | 2026-08-31 | 88.30 | 790 bp | 1 |
| Cipher Mining CIFR | 6.000% 2031-06-15 | 2026-08-31 | 97.37 | 215 bp | 11 |
The coverage gap is itself information: CoreWeave bonds are widely held by registered funds, so third-party marks are abundant; Applied Digital / TeraWulf / Cipher secured notes sit mostly with private-credit funds that do not file N-PORT, so no public mark exists — that debt is buy-and-hold project finance by design. Prices are percent of par; yields solved semi-annually; the period is each fund's own fiscal quarter-end.
Official daily par yield curve from the US Treasury, latest 2026-11-04. Issue spreads are computed by linear interpolation on this curve.
ICE BofA option-adjusted spreads, latest 2026-11-03, with the 30-day change. This is the "what similar bonds are worth" yardstick.
Yes. Click "SEC ↗" on any row and you land on the original quarterly report where the issuer discloses coupon, maturity and size. For the Treasury leg, open the Treasury curve page and pick the same date. Between them sits one linear interpolation, with the formula published.
Because there is exactly one lawful route to per-bond end-of-day traded prices: a FINRA TRACE licence. Until that is signed we publish only what we can genuinely compute — issue spreads and terms — instead of inventing a daily series.
Different definition. The Composite claimed to be a secondary-trade spread but was, without a licence, a demonstration series. NCSX-IS measures issuance-time spread only: narrower, but every value can be independently replicated. The old series is marked deprecated and is no longer published as an index.
Because that is everything currently verifiable down to a specific SEC filing. Most of the AI-infrastructure complex (Lambda, Crusoe, Together and others) has no public USD bonds, and miners pivoting to AI mostly issue convertible notes, which are not straight credit. The sample expands quarter by quarter as new deals price.
Licensing: deal terms come from public SEC filings and the benchmark from public US Treasury data, both freely citable; FRED series require attribution to FRED / ICE BofA. Index construction and publication are independent; nothing here is investment advice or a credit rating.
API: /api/v1/real/universe.json · index.json · curve.json · sector.json