Oracle's 5Y CDS is at 198bps. The credit market is asking a question equity has not.
The first Public Equities Ledger. The Credit Monitor goes live, with one indicator already on watch. What it means, and what to track next.
number: 1 title: "Oracle's 5Y CDS is at 198bps. The credit market is asking a question equity has not." subhead: "The first Public Equities Ledger. The Credit Monitor goes live, with one indicator already on watch. What it means, and what to track next." date: "2026-05-11" preview: "ORCL CDS at 198bps vs a 60bps baseline. BBB OAS calm. HY OAS tightening. Composite score: 1 (BENIGN). And what equity is missing." nextWeek: "A deeper look at the Mag 7 capex-to-FCF ratio and what it implies for credit headroom into 2027." readthrough: "The dashboard says BENIGN. The score is 1. But that headline number hides the most important signal: a single name is on watch, and it is the most leveraged hyperscaler. Equity is still treating this as idiosyncratic. Credit is treating it as a question worth asking. When ORCL's CDS first broke 100bps, it was a curiosity. At 198bps, it is a measurable shift in implied default probability. If the move spreads to META or AMZN, the regime changes."
Most leveraged hyperscaler; cleanest AI-stress proxy. A CDS at this level implies a measurable annualized default probability, materially above its own three-year average.
Just started trading November 2025. Useful as a baseline for high-capex, internal-payback AI builders.
Fortress balance sheet. Any widening is a systemic worry, not a name-specific one.
Same logic as MSFT. A tight cross-check on systemic AI-credit stress.
FCF is cracking on capex. The first place a hyperscaler capex hangover would show up in credit.
Broad investment-grade BBB option-adjusted spread. The non-AI credit context.
High-yield broad spread. A leading indicator of risk-off across the entire credit complex.
Investment-grade corporates relative to 7-10y Treasuries. Falling = IG underperforming Treasuries.
High-yield ETFs relative to 7-10y Treasuries. Falling = junk under stress.
Equity-implied 30-day volatility. The cross-asset confirmation when credit moves.
Treasury-implied volatility. Rate vol hits long-duration tech hardest.
§01 What the dashboard is telling you
The Credit Monitor is the data spine of this publication. Eleven indicators, color-coded by status, refreshed weekly. The composite score is a heuristic: alarms count double, watches count single, the sum tells you where on the playbook to be.
This week's score is 1. One indicator on watch, ten on green. The regime is BENIGN. The action playbook says: no defensive overlay required.
That headline number understates the most important signal. The one indicator on watch is Oracle, the most leveraged hyperscaler. At 198 basis points, ORCL's 5-year CDS is implying a roughly 3.3% annualized risk-neutral default probability. Three years ago, that number was below 1%. The market is pricing real capital structure stress at a name with $80B+ of capex committed against an AI-services revenue ramp that has not yet arrived.
The cross-asset signals are calm. BBB OAS at 115bps is two basis points off baseline. HY OAS at 295bps has tightened 25bps off its baseline (a constructive signal, the high-yield market is not flagging broad risk-off). LQD/IEF and HYG/IEF ratios are at baseline. VIX and MOVE are slightly elevated but inside their watch bands.
The reading: this is a single-name story, not a systemic one. Yet.
§02 The Playbook
| Score | Regime | Action |
|---|---|---|
| 0 to 1 | BENIGN | Credit markets calm. Stay positioned per rebalance plan. No defensive overlay required. |
| 2 to 3 | NORMAL | Some indicators flagging. Maintain rebalance. Watch ORCL trend. |
| 4 to 7 | ELEVATED | Credit stress building. Add 3-month 10% OTM SPY puts on $200K notional. Trim AI-semis 20%. |
| 8 to 11 | STRESSED | Multiple alarms. Cut semis (AMD, MRVL, NVDA) by another 30-50%. Raise puts to 15% notional. |
| 12+ | CRISIS | Systemic credit stress. Cut equity exposure to 50%, raise cash to 30%, max put protection. |
The playbook is a decision tree. At score 1, you do nothing. At score 4 to 7, you start hedging. At score 12+, you cut equity exposure to 50%.
The discipline is the playbook itself. Sophisticated investors do not improvise during stress events. They define the response in advance and execute it when the dashboard moves.
§03 What equity is missing
Three observations on what the credit market is signaling that the equity market has not yet priced in.
Oracle's CDS leads its bond spread, which leads its equity volatility. The CDS market is the most informationally efficient corner of corporate credit. When a single-name CDS widens 138 basis points off baseline, the equity options market typically follows within four to six weeks. ORCL's 30-day implied vol is currently at 28%, against a 1Y realized of 31%. The vol surface has not yet repriced the credit signal.
The capex divide is sharpening. Microsoft, Alphabet, and Amazon have fortress balance sheets relative to their AI capex commitments. Their CDS spreads are inside 40bps and barely moving. Oracle has a different profile: it took on debt to fund its capex pivot and the leverage shows. Meta sits in between (the META CDS only began trading in November 2025; baseline data is thin). The market is not pricing AI capex as a single risk factor. It is pricing balance sheet capacity to absorb it.
Cross-asset confirmation is the tell. A single-name CDS move can be noise. The signal becomes credible when (a) it widens to multiple names, (b) the broad credit indices follow, or (c) cross-asset vol picks up. None of those three has happened yet. When they do, the score moves to 4+ and the playbook activates.
§04 Three names to watch this week
AMZN. The model flags Amazon as the place "FCF cracking on capex would first show up in credit." Q1 capex of $26B against operating cash flow of $25B means the FCF cushion is gone. CDS at 38bps is OK; baseline is 30bps. A move above 55bps is the watch threshold.
META. Capex pivot is the most aggressive in absolute dollars. CDS at 55bps is OK but already above its short baseline of 35bps. A move above 70bps changes the cross-asset picture, because Meta is the cleanest test of "AI capex with no current AI revenue."
MSFT. The cross-check on systemic stress. At 30bps, any widening to 50bps+ is the signal that this is no longer a name-specific story. Microsoft is the largest cloud business in the world; its CDS does not widen for idiosyncratic reasons.
§05 The frame for sophisticated readers
For investors at the Druckenmiller and Buffett level, this publication exists to surface the cross-asset signals that public-equity research desks consistently miss. Equity research is built on income statements and earnings calls. Credit research is built on balance sheets and refinancing calendars. The dislocations between the two views are where most of the alpha lives in this cycle.
For investors earlier in their journey, every indicator on this dashboard has a plain-English explanation in the Public Equities Foundations. Start with Reading a CDS spread, then come back here.
Both audiences should track the same thing: when the composite score moves above 3, this publication's frame is different.
The dashboard says BENIGN. The score is 1. But that headline number hides the most important signal: a single name is on watch, and it is the most leveraged hyperscaler. Equity is still treating this as idiosyncratic. Credit is treating it as a question worth asking. When ORCL's CDS first broke 100bps, it was a curiosity. At 198bps, it is a measurable shift in implied default probability. If the move spreads to META or AMZN, the regime changes.
A deeper look at the Mag 7 capex-to-FCF ratio and what it implies for credit headroom into 2027.