Skip to content

How the signal is computed — Cerebras Systems

Cerebras SystemsShort · ~2 weeksas of 2026-10-05Bearish

Structural led a bearish call; the other agents split, so conviction reflects the disagreement.

Coverage
3/7
Agreement
0.52
.33
Conviction
How the swarm reasoned
✓Question›✓Coverage›3Driver›4Consensus›5Gate›6Verdict
🧭
Structural Driver
Industry Index 13.4 (falling -9.4 over window) — structural headwind.
−0.334
Industry Index falling -9.4 [no link]−0.25442%
Industry Index level 13.4 [no link]−0.08013%
2 item(s) · sums to -0.334
Momentum (medium) [no link]−0.06911%
Moving-average trend (short) [no link]−0.0346%
MACD histogram (medium) [no link]−0.0315%
Momentum (short) [no link]−0.0203%
MACD histogram (short) [no link]−0.0163%
OBV (short) [no link]−0.0112%
Bollinger %B (short) [no link]−0.0102%
Stochastic (short) [no link]−0.0071%
RSI trend (short) [no link]−0.0041%
ADX / DMI (medium) [no link]−0.0020%
ADX 12 (trend strength 0.30) [no link]0.0000%
Bollinger %B 0.28 (extreme 0.00) [no link]0.0000%
RSI 45 (overextension penalty 0.00) [no link]0.0000%
13 item(s) · sums to -0.205
Unusual-activity net skew [src]+0.06611%
1 item(s) · sums to +0.066
📰
News Abstained
no recent news
—
📈
Growth Abstained
Fundamentals inform the ~3-month call, not a 2-week move.
—
🏦
Smart money Abstained
13F filings move too slowly for a 2-week call.
—
🏰
Moat Abstained
The moat prior is a ~6-month structural view, not a 2-week move.
—
Consensus · did they agree?

Most voters leaned one way; structural dissented. Directional agreement 0.52.

The gate · score → conviction → verdict
Raw score
−0.473
×
Agree·cover·strength
0.52
→
Conviction
0.33

A clear bearish score; conviction 0.33 reflects how strongly the agents agree. Conviction is confidence, not a probability of the move.

as of 2026-10-05 · config sig-v4-2026-07-26

Signal history

☆← Back to cerebras · Methodology