TonalityIQ
Executive Conviction Analysis. Not Sentiment. Not NLP.
4,664 companies. 169,000 transcripts. 11 conviction types. Calibrated on real forward returns.
This is not NLP.
NLP asks “is this sentence positive or negative?” Every quant fund has tried it on earnings calls. The results are marginal. Sentiment analysis gives you “positive” or “negative” and that's about as useful as reading the headline.
TonalityIQ asks a different question: is this CEO restructuring with conviction or restructuring with fear? That's a management psychology question, not a text mining question.
The methodology comes from upper echelons theory, 40 years of academic research on how executive cognition shows up in language. When a CEO shifts from “we will” to “we believe,” that's not a sentiment change. That's a confidence change. When restructuring language spikes alongside commitment language, that's not positive or negative. That's a turnaround signal. NLP can't see that because it doesn't have the management theory underneath.
We score each CEO against their own historical baseline. A CEO who always says “challenging” isn't stressed. That's their pattern. A CEO who never says it and suddenly starts? That's a 3-sigma event.
73.3%
Crack pattern accuracy (constraint + hedge spike + ratio collapse predicts next-quarter deterioration). n=1,508.
-55%
Median 12-month return when NARRATIVE_HYPE detected (buzzwords without fundamentals). n=65. The rarest and most devastating signal.
11
Conviction types. From GROWTH_CONVICTION to NARRATIVE_HYPE. Each calibrated against real forward returns across 169,000 transcripts.
We detected Apple's AI collapse one quarter early
TIQ Score: 71 — Highest structural deterioration in SaaS
ENERGIZED/ENERGIZED CEO-CFO configuration detected
Pattern 10 identified before the 17% move
No tickers with ALL signal
TonalityIQ analyzes earnings call transcripts for structural shifts in executive language.
Click any ticker to see the full constraint/conviction arc.