Arcane’s base case is contained stress, not a broad market break: confidence is 73%, breadth is 75% normal, and 2 of 8 assets are stressed, with NVDA the only high-risk vector.
Confidence 73% [Review] | Breadth 75% normal [Narrowing] | Stressed assets 2/8 | High Risk 1 | Coverage 8/8 | Risk vector NVDA
NVDA is the only High Risk asset. SPY is elevated but not High Risk. That keeps the stress pocket localized until breadth or core proxies confirm it.
Breadth is still 75% normal and only 2 of 8 monitored assets are stressed. That keeps the regime call in contained-stress territory rather than a broad break.
The evidence table shows the stress cluster directly: SPY and NVDA. QQQ, AAPL, and BTC remain normal, which is why the base case has not moved to broad regime review.
| Symbol | State read | As of | Vector | Score | Confirmed |
|---|---|---|---|---|---|
| NVDA | High Risk | 20:00 EDT | Growth leadership | score 0.59 | 85% confidence |
| SPY | Elevated | 20:00 EDT | Core equity proxy | score 0.50 | 50% confidence |
| BTC | Normal | 20:00 EDT | Crypto risk appetite | 45% conf. | 45% confidence |
| QQQ | Normal | 20:00 EDT | Core equity proxy | 100% conf. | 100% confidence |
| TSLA | Normal | 20:00 EDT | Growth leadership | 75% conf. | 75% confidence |
| VIX | Normal | 20:00 EDT | Volatility complex | 60% conf. | 60% confidence |
Confidence and breadth are answering different questions. Confidence at 73% measures signal agreement across the stack; breadth at 75% normal measures how much of the monitored universe is still unstressed.
The chart plots those two objects together: the confidence line shows system agreement, while the bars show how many assets are in Elevated or High Risk. A wider stress bar set with falling confidence would be the warning configuration.
The read deteriorates if NVDA weakness spreads into SPY, QQQ, or VIX or breadth falls below 73% with additional High Risk assets.
The warning fades if NVDA stabilizes while breadth holds above 73%, VIX stays orderly, and no additional asset enters High Risk.
Confirmation requires distribution, not just persistence. The base case only changes if the stress vector starts pulling core proxies or breadth into the same direction. If the lead vector stabilizes and the stressed count stops expanding, the desk can treat the move as localized rather than self-reinforcing.
Base case: keep the broader regime call in place, but shorten review cadence around NVDA. If the confirm trigger fires, escalate from localized stress to broader regime review. If the invalidate trigger holds, step the warning back to localized noise.
That argues for tighter surveillance around NVDA, not a wholesale regime rewrite. The broader stack should only change if the confirmation path is met.