Bletchley BDC Distress League Table

The GI Bletchley BDC League Table ranks 137 Business Development Companies — and other SEC-registered private credit funds filing the same disclosures — by how stressed their loan books look and how honestly they’re marking them against peers.

Rank 1 is the most concerning. Rank 118 is the least concerning.

The data comes from public SEC filings. The analysis is proprietary.

Bletchley is built by General Intelligence Capital. We built it because no one else was doing the one thing LPs most needed: independently cross-checking how different lenders value the same loan.

Every BDC publishes a Schedule of Investments every quarter, listing every loan they hold, what they paid for it, and what they think it’s worth today. These filings are public and free.

But until now, nobody cross-referenced them at scale. When Blackstone, Blue Owl, and Apollo all held the same loan, nobody was asking “are they marking it the same way?” The answer, often, is no.

We extract per-position loan data directly from SEC EDGAR using the structured XBRL feeds the SEC itself publishes. Coverage:

• 137 Business Development Companies (publicly traded, non-traded, and private)
• 41,852 individual loan positions
• 4,125 unique borrowers held by two or more lenders
• Most recent quarterly and annual filings (10-Qs and 10-Ks)

We do not use proprietary data feeds, paid research, or third-party analytics. Everything is derived from filings you can read yourself on SEC.gov.

Every BDC gets three scores from 0 to 100. Higher = more concerning. The composite score is a weighted blend of the three.

Pillar 1 — Portfolio Stress

How stressed does the loan book look in absolute terms? Pillar 1 measures what the BDC itself is telling you. It’s computed entirely from position-level metrics you can verify against the filings.

• Percentage of positions marked below 90¢, 80¢, and 70¢
• Percentage of positions paying PIK (borrower paying with additional debt rather than cash)
• Weighted distress severity across the portfolio

A high Pillar 1 means the BDC is openly reporting a large share of distressed loans in its own filings.

Pillar 2 — Marking Integrity

Are they being honest about their marks relative to peers? For every loan a BDC holds that at least one other BDC also holds, we compare that BDC’s mark to the consensus across peers (excluding their own mark).

The underlying dataset contains 4,125 overlapping borrower cases, including 258 borrowers held by 5 or more different BDCs and 19 borrowers held by 10 or more. Pluralsight is held by 9 BDCs with a 46-point spread.

Important context: at the BDC level, marking differences are modest. Across 9,456 cross-BDC loan comparisons, 82% of marks are within ±2 points of peer consensus. The aggregate BDC-level signal is thin. Pillar 2 flags the outliers — not systemic misrepresentation. At the borrower level, divergences are dramatic and frequent.

BDCs with fewer than 10 shared loans cannot have Pillar 2 computed reliably. These are flagged as “Partial” coverage.

Pillar 3 — Bletchley Pattern Score

What does our AI see that isn’t obvious from single metrics? Pillar 3 is computed from a proprietary pattern engine that has been rigorously validated through a three-way temporal holdout process.

The methodology:

• We started with 35 pattern hypotheses derived from historical BDC loan behaviour
• We split data into training (pre-Q2 2023), validation (Q3 2023 – Q2 2024), and out-of-sample test sets (Q3 2024 onward)
• Each pattern had to pass all four thresholds: precision > 25%, lift > 1.5x, p-value < 0.05, and minimum 10 test-set signals
• 16 patterns survived. 19 were eliminated

Aggregate performance: top pattern precision 78% (versus 19% base rate — a 4.15x lift). Multi-pattern convergence precision: 63%. All surviving patterns: p < 0.01.

Of the 16 validated patterns, 9 run on the SEC’s current structured data feed. The remaining 7 require narrative text parsing from 10-Q filings and will be added in the next product version.

Composite Score

For BDCs with full Pillar 2 data: Composite = (P1 × 0.40) + (P2 × 0.20) + (P3 × 0.40)

For BDCs with insufficient shared-loan data: Composite = ((P1 × 0.50) + (P3 × 0.50)) × 0.75

The 0.75 multiplier on proxied BDCs reflects lower confidence — we cannot independently verify their marks against peers.

Each BDC in the table is labelled either Full or Partial.

• Full — The BDC has at least 10 loans in common with other BDCs, allowing Pillar 2 to be computed directly.
• Partial — Fewer than 10 shared loans. Composite is computed from Pillar 1 and Pillar 3 only, with a 0.75 confidence penalty applied.

59 of 118 ranked BDCs have Full coverage; 59 have Partial. The top of the table should be read with this badge in mind — Full rankings are the most defensible.

A Business Development Company is a specific SEC-regulated structure, not a listing type. All BDCs — publicly traded, non-traded, and private — file the same Schedule of Investments disclosures.

• Publicly traded — 34 funds. Listed on a US exchange. Examples: Ares Capital, Blackstone Secured Lending, Main Street Capital.
• Non-traded — 8 funds. Registered with the SEC but not listed. Includes the largest private credit vehicles in America, such as Blackstone Private Credit Fund ($89B), HPS Corporate Lending Fund ($27B), and Ares Strategic Income Fund ($22B).
• Private — 95 funds. Registered but not broadly offered. Often institutional co-investment vehicles.

All three are treated equally in the ranking.

• Non-BDC private credit vehicles that don’t file with the SEC
• Bank-originated private credit books
• CLO tranches
• Direct-lending vehicles that report only to institutional investors
• Sector-specific or concurrent distress flags that require narrative text analysis (planned for v2)
• Dividend coverage, NII coverage, or relative-value metrics (planned for v2 as a fourth pillar)

We cover the universe of SEC-disclosed BDC loan positions. That’s the largest transparent dataset of private credit marks in existence — but it isn’t the entire private credit market.

Every data point in Pillars 1 and 2 traces back to a specific SEC filing.

• Pillar 1 inputs — Check the Schedule of Investments in the BDC’s latest 10-Q. Count positions in each mark bucket.
• Pillar 2 inputs — Find a loan held by the BDC you’re checking and at least one other BDC. Compare the marks directly.
• Pillar 3 — The validation methodology is disclosed above. The specific patterns are proprietary. Subscribers receive full pattern detail under NDA.

We welcome analysts to challenge specific numbers. We’ll publish corrections publicly if we’re wrong.

Pillar 3 is the AI layer — a proprietary pattern engine trained and validated on historical BDC loan outcomes.

Pillars 1 and 2 are data engineering. Useful data engineering — nobody had assembled a cross-BDC mark dataset at this scale before — but the analysis itself is arithmetic, not machine learning.

We’re not claiming Bletchley is an autonomous intelligence that decides which BDCs are stressed. It’s a structured pattern engine running over the largest transparent cross-BDC dataset ever assembled. That’s the honest version.

The league table is free. Everything underneath sits behind a paywall.

Free — BDC League Table

All 137 BDCs ranked on P1, P2, and P3. Coverage flags (Full / Partial). Appendix of excluded BDCs with reasons. Updated quarterly.

$2,000 — 5 Borrower Reports

Pay-as-you-go credit pack. Use within 12 months. Each borrower report includes: all lenders holding the borrower with their marks, cross-BDC mark divergence analysis, which Bletchley patterns are firing, historical mark trajectory, Bletchley distress verdict, and peer comparison within the borrower’s sector.

$2,500/month — 50 Borrower Reports + Full Analytics

Subscription tier. Includes 50 borrower reports per month (additional at $40 each), full pattern detail under NDA including the 9 active pattern definitions and precision/lift data, quarterly methodology brief, priority email support, and early access to v2 features.

Best for LPs, family offices, institutional allocators, and analyst teams with active private credit exposure.