Credit risk daily report · 2026-11-14

Beijing time: 2026-11-14 11:29 CST. Listed-security data run through the completed November 13 U.S. session. FRED series use their own stated observation dates. Public overall level: Yellow. Cash-bullet deploy trigger: watch. This is a risk monitor, not investment advice, and it does not execute trades.

1. Five-indicator dashboard

HY OAS is the highest-weight signal and the key one-to-three-month credit-event lead indicator. Its raw ICE/FRED level is not redistributed publicly. The internal threshold check did not activate the HY-yellow override, but a withheld highest-weight value prevents a public all-green conclusion, so the published framework remains Yellow/watch.

2. What changed / what matters

• HY OAS first: the observation advanced to November 12; the raw level stays withheld, while the internal threshold check remains below the >5% yellow boundary.
• KRE rose 0.30% and is up 0.75% over five sessions, far from both its price and momentum warnings; volume was only 0.52× the prior 30-day average.
• VIX closed at 19.30 on November 13, up 3.21% for the day but still comfortably below 25.
• T10YIE rose to 2.24%, moving above the 2.20% lower edge of the green band.
• None of the three large-bank proxy signals is active. Bank equities were firm over five sessions, KBE rose 0.82%, and the LQD-versus-SPY five-day spread was -4.24 percentage points.

3. Market detail

All listed-security returns are computed locally from one consistent Yahoo adjusted-close/volume series. Every required symbol has 128 usable observations, nonzero price denominators and 30 prior volume observations. No split event was reported. Yahoo query1 succeeded for every symbol, so query2 and other history fallbacks were not used. An independent callable finance lookup matched the November 13 closes for KRE, JPM, BAC, WFC, C, GS, MS, KBE, LQD and SPY within 1%. The preferred FMP connector was searched by the exact mcp__financial_modeling_prep__ namespace in three fresh bounded checks, but no FMP tool was exposed in this session; this is a discovery limitation, not a claim that the server is down.

Signal 1 is inactive because no monitored bank fell more than 5% in one day with volume above 2×. Signal 2 is inactive because KBE gained 0.82% over five sessions rather than falling more than 10%. Signal 3 is inactive because LQD -1.65% minus SPY +2.59% equals -4.24 percentage points, not strictly above +3.

4. Credit, volatility and inflation boundary

Direct fredgraph.csv reads were isolated by series and used a 20-second hard timeout. The first grouped Python read did not complete within its bounded window; the single permitted retry used separate curl reads and succeeded. BAMLH0A0HYM2 is dated November 12, T10YIE is 2.24% on November 13, and VIXCLS is 18.70 on November 12. These dates remain within the series' normal publication cadence and are not marked STALE_CRITICAL. VIXCLS is an official dated cross-check, not a simultaneous observation; the fresher November 13 Yahoo session value of 19.30 controls the VIX row. The HY OAS level is checked against the rules internally but withheld from public redistribution.

5. Overall algorithm and deploy implication

Any deep red, or at least three red indicators, makes overall deep red. One or two reds, or at least three yellows with no red, makes overall red. One or two yellows makes overall yellow. HY OAS yellow alone upgrades overall to red. All five indicators must be publicly observable and green before an all-green public claim.

Deploy implication: Yellow means prepare and review the watchlist only. The current status is watch, not an instruction to deploy. Red would activate a human review of the pre-planned staged cash-bullet framework; deep red would trigger crisis-playbook review. No trade should be executed automatically from this report.

6. Data health and boundaries

Data health: FMP was searched through the exact namespace, but no task-relevant tool was exposed after three fresh checks within the discovery budget. Yahoo query1 therefore supplied the consistent history, the callable finance lookup supplied listed-quote cross-checks, and direct FRED supplied all macro series after one retry. No required market item is missing or stale. Public confidence remains deliberately capped because HY OAS is DATA_WITHHELD; the raw value is neither published nor used to claim an all-green state.