All trades discussed here are paper positions for analytical purposes only.
FDIC noncurrent loan rates held at 0.98% through Q1 2026, while net charge-offs reached 0.577% — both readings remain contained relative to prior credit cycles, but the directional trend warrants attention. The gap between performing and non-performing credit is compressing slowly, and BDC portfolios with floating-rate exposure are beginning to feel the asymmetry: coupons reset lower as base rates ease, but credit losses, when they arrive, don't reset with them. This is the quiet part of the cycle — not crisis, but the phase where the scorecard separates disciplined underwriters from yield-chasers.
The `exit_vs_peer_contradiction` signal is carrying the most regime-relevant information right now, with a 70% hit rate across 20 events and a +3.3% average abnormal return over 60 days — the highest conviction signal in the system this week. By contrast, `v3_borrower_stress` is flashing caution: 65 events, a 46.2% hit rate, and a deeply negative average abnormal return of -13.05%. That divergence tells you the market is not indiscriminately risk-off, but it is punishing specific credit deterioration hard when it surfaces. Spreads on middle-market senior secured are not yet dislocating, but the borrower stress signal suggests select names inside BDC portfolios are moving toward restructuring before any public announcement.
Ares Capital remains the benchmark against which middle-market credit quality is measured, and the `exit_vs_peer_contradiction` signal — 70% hit rate, +3.3% abnormal return — favors names where peers are marking credits down while the manager holds. ARCC's scale ($22B+ portfolio) gives it workout resources that smaller BDCs simply don't have. A sustained rise in FDIC noncurrent rates above 1.2% or deterioration in its first-lien coverage ratios would be the thesis-breakers.
Golub Capital BDC has concentrated its book in senior secured floating-rate loans to sponsor-backed borrowers — exactly the structure that benefits from the `exit_vs_peer_contradiction` dynamic when peers are mismarking. The signal's 20-event sample is small, but 70% hit rate at 60 days is material. Watch for any uptick in PIK income as a warning: PIK creep in a softening rate environment is often the first visible symptom of borrower stress before it reaches the formal watchlist.
The `v3_borrower_stress` signal — 65 events, -13.05% average abnormal return, 46% hit rate — is weighted toward venture lending and growth-stage credit profiles, where cash burn and refinancing risk are acute in a higher-for-longer environment. TriplePoint Venture Growth sits squarely in that exposure profile. A 46% hit rate means the signal is wrong more than half the time, which is precisely why this is sized as an underweight rather than a conviction short: the downside when it hits is severe (-13%), but you need confirmation from NAV marks or portfolio company disclosures before pressing.
PennantPark Floating Rate Capital's first-lien focus and its smaller average position size relative to portfolio count give it natural diversification against single-name blowups. The `exit_at_loss` signal showed a 49.4% hit rate across 168 events — essentially coin-flip — which means the market is not efficiently pricing exits at loss for this cohort. PFLT's manager has historically recycled exits into higher-yielding reinvestments; if that continues, NAV erosion risk is lower than peers. The thesis fails if charge-off rates in its healthcare and services book accelerate beyond Q2 trends.
The `debt_to_equity_conversion` signal produced only 7 events but the numbers are stark: 28.6% hit rate and -2.28% average abnormal return. SLR Investment Corp's portfolio construction includes equipment finance and asset-based lending — sectors where D/E conversions, when they occur, tend to lock up liquidity and impair realized NAV. At this sample size the signal is not actionable alone, but SLRC warrants close monitoring of its Q2 filing for any new equity positions taken via restructuring.
The system generated no formal trade decisions this week, and that restraint is the correct output. With `nim_leadlag` returning no data and `decisions` and `alerts` both empty, the platform is telling us the highest-confidence setups are not yet structured — they're still forming in the borrower stress data. We passed on three names where the stress signal fired but the hit rate was below 45%; below that threshold, the signal is noise until corroborated by mark changes in the 10-Q.
The forward question for the week of 28 August: whether the FDIC noncurrent rate's slow climb toward 1.0% begins to show up in BDC Q2 non-accrual disclosures — and which managers chose to mark early versus wait.
Screening research, not investment advice. All data from public SEC filings.