All positions below are paper trades in the Crypsos cross-sectional autopilot book. The strategy is UNVALIDATED — caps and a kill switch gate any live order. This is review, not execution.
FDIC noncurrent loans clocked 0.98% of total loans as of Q1 2026, the highest read in this cycle, while net charge-offs hit 0.577% — still below the 2010 peak but moving in one direction. Against that backdrop, the BDC sector's dividend coverage picture is fragmenting sharply: a handful of names are earning their distributions comfortably in cash, while a growing tail is papering over gaps with PIK income or simply not covering at all. The divergence between those two cohorts is the trade.
BDC spreads remain technically supported — the quality end of the market continues to price off a strong NAV-to-market floor — but the coverage alert queue this week is the longest it has been in several months. Seven names triggered a cash-coverage break signal, and the common thread is PIK creep: ARCC's PIK share reached 33.6% on 1.17x total coverage, leaving cash coverage at just 0.78x; NMFC sits at 16.5% PIK with cash coverage of 0.87x. Neither is in crisis, but both illustrate the regime dynamic — headline coverage ratios are increasingly optical. The stress end of the universe (TPVG cash coverage: 0.001x; ICMB: 0.243x) belongs in a different conversation entirely.
OBDC — HOLD (rank #1 of 46, composite +0.79)
Already in the paper book and staying there. OBDC leads the 46-name universe on two of the metrics we weight most: mark trend (z-score +2.8) and dividend coverage (z +1.0), meaning portfolio marks are improving sequentially and the distribution is well-supported in cash. A deterioration in either driver — a sequential NAV decline or coverage slipping below 1.0x — would trigger a re-rank review. OBDC
RWAY — BUY (rank #2 of 46, composite +0.71)
Runway Growth Finance is the strongest mark-trend signal in the book this week: z-score of +3.3 on portfolio appreciation, against a non-accrual share that is modestly below the peer median (z -0.3). That combination — marks moving up while credit quality holds — is precisely the quality signature the autopilot is designed to find. The thesis breaks if the mark trend is lag-driven by a single large name being carried above recoverable value; that requires position-level verification. RWAY
MAIN — BUY (rank #3 of 46, composite +0.62)
Main Street Capital carries the highest book-mark z-score in this week's long book at +2.5, reflecting a portfolio marked at a consistent premium to cost — a structural characteristic of its equity co-investment model. Mark trend is also positive (z +0.3). The risk is that the NAV premium to peers is already priced into the stock, meaning the informational edge from this signal may be weaker than in names where the market hasn't yet recognized the quality. Valuation compression, not credit deterioration, is the more likely invalidator here. MAIN
CION — BUY (rank #4 of 46, composite +0.42)
CION scores on two dimensions the system penalizes elsewhere: its PIK share is well below the peer median (z -1.1) and mark trend is positive (z +0.9). In a week when PIK inflation is the dominant noise in the coverage data, a BDC earning predominantly in cash with improving marks is a relative-value statement. The invalidating condition is straightforward: if PIK share rises materially in the next quarter's filing, the primary thesis leg breaks. CION
MSIF — BUY (rank #5 of 46, composite +0.41)
Morgan Stanley Income rounds out the long five with book mark at z +1.4 and a positive mark trend (z +0.4). It is the weakest composite in the long book, which matters — at a +0.41 composite, it is a marginal inclusion. The thesis is that book quality and sequential improvement are persistent signals, but the margin of safety is thin relative to OBDC or RWAY. Any single-quarter mark reversal would push MSIF out of the top quintile and trigger a review. MSIF
The autopilot flagged RAND (rank #46, composite -1.08) as the clearest short in the universe — non-accrual share at z +4.7 is not a rounding error, it is a structural stress signal — but the paper order returned a 422 error and was not filled; we are logging the signal, not forcing a workaround. ICMB, TPVG, OXSQ, and CGBD were also scored as shorts and are not shortable on the paper broker; TPVG's cash coverage of 0.001x and 34.3% PIK share make it the most acute stress read in the universe, and the system's inability to express that view is a known constraint we are documenting honestly rather than substituting a long-side proxy.
The week ahead will test whether the coverage divergence visible in Q2 filings begins to reprice the stress tail — or whether carry-hungry flows keep the bid under names the fundamentals no longer support.
Screening research, not investment advice. All data from public SEC filings.