About CFI
A research framework for earnings-driven investors who analyze narrative risk.
We don't predict what will happen.
We predict when everyone else is wrong.
Most consensus tools tell you what analysts expect. CFI tells you how structurally vulnerable that consensus is—and what would break it.
How it works
We track 50 S&P 500 companies through their earnings cycles. Our system extracts thesis pillars—the core assumptions underpinning analyst narratives—and monitors for contradictions in news and data.
When contradictions accumulate against weighted pillars, fragility rises. The output is a score (0-100) quantifying how exposed the consensus is to breaking at the next catalyst.
We don't forecast direction. We measure thesis stability.
Calibration
CFI was calibrated on 152 earnings events spanning 7+ years of market history. Break rates by fragility bucket:
| Fragility | Break Rate | Base Rate |
|---|---|---|
| Low (0-35) | 0% | 24% |
| Medium (35-50) | 31% | 24% |
| High (50+) | 89% | 24% |
Low-fragility consensus held in every case. High-fragility consensus broke nearly 9 in 10 times.
Current coverage skews toward high-beta tech. We're expanding to defensive sectors and publishing forward results quarterly to validate out-of-sample performance.
Early access
CFI is free during validation. We're publishing predictions and results through Q2 2026 earnings. If you want to evaluate the framework against live data, now's the time.
Questions? Get in touch
Important: CFI is a research tool, not investment advice. We provide fragility scores for informational purposes only. Always conduct your own research and consult qualified advisors before making investment decisions.