Mark Spread: see where private credit funds mark the same loan differently
Shows where BDCs mark the same loan differently, from SEC filings. Private credit LPs and analysts pay.
Crowded · Analyst's one-day build, $40M rival raise
Written for: Small startup team
Next step
Call the head of private credit due diligence at an LP consultant such as Albourne or Cambridge Associates. Bring 10 borrowers held by three or more BDCs with their marks side by side, and ask whether they would pay for that file every quarter.
Why now
Two signals arrived together. A practitioner showed he wanted cross-firm loan marks badly enough to build the tool himself. Cognitive Credit raised $40M to grow AI corporate-credit data in the US, which shows budget in the category. The demand is the gap between firms, not the scraping: when one BDC holds a loan at 95 and another at 80, LPs and analysts want to know which is right. A one-person internal tool will not keep up with every filer every quarter. A maintained, entity-resolved dataset will.
What became possible
Coding agents and current models can now pull schedule-of-investments tables out of quarterly filings and match borrowers across funds. A portfolio manager at Benefit Street Partners built a working version that collects loan valuations from SEC filings and charts them in about a day. Extraction is no longer the hard part. The hard parts are matching borrower names across filings and keeping the dataset clean every quarter.
Three ways in
- Mark Spread file
A quarterly dataset of every borrower held by two or more BDCs, with each holder's fair value, cost, non-accrual flag and the spread between marks.
- Customer
- Private credit due-diligence teams at LP consultants and pension funds
- Model
- Annual data licence, $25k–$75k per seat group
- Mark Move alerts
When one holder cuts a loan's mark, alert every other holder's analysts and LPs before those holders file.
- Customer
- Sell-side BDC equity analysts and credit hedge funds
- Model
- Subscription per user
- Fund outlier report
For a given manager, show how often its marks sit above peers on shared loans. LPs can use it before committing capital.
- Customer
- Allocators running manager selection
- Model
- Per-report fee plus annual refresh
Competition
The core product, a dataset of borrowers held by two or more BDCs with each holder's mark and non-accrual flag, is already sold by PitchBook LCD, SOLVE and Bloomberg. Octus and 9fin publish analyses built on the same idea.
Full competition notes (8 competitors)
The core product, a dataset of borrowers held by two or more BDCs with each holder's mark and non-accrual flag, is already sold by PitchBook LCD, SOLVE and Bloomberg. Octus and 9fin publish analyses built on the same idea. The quarterly Mark Spread file has little room to stand out. Mark Move alerts are partly covered: SOLVE claims filings are captured minutes after release and maps co-investors. But no vendor was found selling an explicit alert to every other holder when one lender cuts a mark. The least served idea is the Fund outlier report, a per-manager score of how often and by how much a manager marks above peers on shared loans, packaged for allocators choosing managers. PIMCO and Octus have published one-off research on how far apart marks are, but no dedicated manager-scoring product was found. A new entrant would most plausibly win with a focused, cheaper manager-scoring product for LP consultants, pension funds and wealth platforms that won't pay for terminals. It could add things the big vendors don't seem to offer per manager: a track record of how marks held up against later recoveries, adjustment for the price each lender bought in at, and coverage of private and non-traded BDCs. The underlying data is free in SEC filings and the incumbents could add this feature quickly, so the advantage would have to come from method, accurate matching of borrowers across filings, and distribution, not data access.
- PitchBook LCD BDC dataset
LCD's BDC holdings data, part of PitchBook since June 2024, explicitly lets users find BDCs that co-invest in the same company and compare how various BDCs value the same investments. That is the core Mark Spread file idea.
- SOLVE BDC Data Analytics Workstation
Covers about 162 BDCs and maps co-investors to show how other BDCs mark the same asset. It also reports non-accruals, claims filings are captured minutes after release, delivers by API and feed, and publishes quarterly analyses comparing BDC marks with market prices.
- Bloomberg Private Direct Lending data
Launched in April 2026 from BDC filings and other sources. It merges shared loans into one FIGI-identified record that carries each BDC's mark and non-accrual status.
- Octus (formerly Reorg)
Reconciles marks on shared borrowers across all 174 BDCs and found nonaccrual exposure about 60% higher than reported. Publishes paid research on marks diverging across BDCs.
- 9fin Private Credit and BDC Watchlist
Consolidated quarterly portfolios for 150+ BDCs, plus an AI-built watchlist of loans marked below 90 with a falling trend. A side-by-side mark comparison tool was not confirmed.
- Preqin private credit BDC data
Standardized loan-level BDC holdings across listed, non-traded and private vehicles, with risk analytics, sold to LPs.
- Dakota BDC Holdings
Combines loans across 40+ public BDCs so a borrower's total exposure across lenders is visible. Sold to allocator-facing sales and research users.
- Founderpath software private credit dashboard
Small free dashboard of software loans from BDC filings that tracks markdowns. Narrow in scope.
Checked 2026-10-25
Risks
- Borrower names and tranche descriptions vary across filings, and entity-resolution errors would destroy trust quickly
- Filings are quarterly and lag, so the data shows stale marks
- Incumbent data vendors and Cognitive Credit could add this as a feature
- Buyers like the Benefit Street analyst may keep building it in-house
Strong team fit; open space is the weak point (5).
tech 8 · pull 7 · space 5 · fit 9 · build 8
Exposed
Companies that could benefit or be threatened. News and context, not investment advice.
Momentum of its main topic
Sources
Status history
- 2026-10-25 Flagged as buildable