Credit Crunch: Where AI Fits With Cognitive Credit’s Rob Slater

Rob Slater, CEO and founder of Cognitive Credit, tells Bloomberg Intelligence’s Credit Crunch podcast the firm is automating parts of the credit analytic workflow beginning with reliable financial data extraction and progressing toward scalable valuation. He frames AI pilots as low-risk, attractive risk/reward experiments and emphasizes the growing imperative for credit investors to adopt a formal credit data strategy as technology permeates the credit-investing landscape.

Analysis

Market structure: Automation of credit analytics disproportionately benefits cloud providers (MSFT/AMZN/GOOGL), data vendors (SPGI, MCO, ICE) and quant-driven asset managers that scale models cheaply; manual, high-cost credit shops and some BDCs risk margin erosion as pricing for alpha compresses by an estimated 100–300bps over 1–3 years. Faster, cheaper credit valuation increases tradeable supply and shortens holding-periods; expect secondary market bid/offer compression in IG corporates and tighter spreads in liquid names within 6–18 months.

Risk assessment: Key tail-risks are model failure or systemic bias (1–5% chance annually with >10% mark-to-market shock for levered books), regulatory intervention (EU AI Act/SEC guidance within 6–18 months) and cloud concentration outages (single-event liquidity shocks). Hidden dependencies include data-licensing cost inflation (licenses could rise 10–30% annually) and vendor lock-in; catalyst set: a high-profile quant fund outperformance or a cloud outage will accelerate re-rating.

Trade implications: Favor long positions in data/infra winners and short operationally vulnerable credit managers. Expect a 6–12 month window to harvest edge: bid for SPGI/MCO and cloud names; hedge with protection on cyclical credit exposures. Use options to buy convexity into winners and defined-risk bearish structures on BDCs/active credit managers.

Contrarian angles: Consensus understates the cost of reliable labeled data — data licensing, not models, will be the recurring monopoly rent; market may underprice that revenue stream today, creating 20–40% upside potential for dominant data owners. Conversely, automation could produce correlated mispricings in stress (flash de-risking), so pure long-credit plays without liquidity hedges are underpriced in risk.

Key Decisions for Investors

• Establish a 2–3% net long position split: SPGI (S&P Global) 1.5% and MCO (Moody's) 1.5% over 6–12 months; add 12-month ATM call options (delta ~0.5) if share price rallies >10% to lever upside while capping cash outlay.
• Rotate 3–4% of active credit allocations into cloud/infra: MSFT 2%, AMZN 1%, GOOGL 1% over the next 3–9 months to capture a projected 10–20% incremental cloud spend by credit managers; dollar-cost average on any pullback >5%.
• Initiate a 1–2% short or put-spread position against PSEC (Prospect Capital) or AINV (Apollo Investment) sized to limit max loss to 2% portfolio equity (e.g., 3-month 10% OTM put spread) — thesis: NAV compression and execution risk may shave 10–20% off BDCs in 6–12 months.
• Reallocate 50% of fees currently paid to high-cost active credit managers into passive IG exposure (iShares LQD) plus 1–2% allocation to quant/tech-enabled credit managers within 90 days to lock in cost savings; use 3–6 month LQD put protection (5–7% OTM) if credit volatility rises above historical 12-month average by >30%.