Listings, narratives, and our AI news stack

Narratives move markets before price

In crypto, the story often leads the chart. Listing rumors, partnership tweets, regulatory headlines, and influencer cascades all shift liquidity within minutes. Quantive Invest ingests CryptoPanic, LunaCrush, exchange listing calendars, and additional news APIs into a unified scoring layer we call the narrative engine internally.

The engine does not “predict the future.” It ranks urgency: which symbols gained credible attention, which headlines duplicate known spam patterns, and which feeds disagree.

AI that ranks, humans that veto

Machine learning models consume live and historical datasets—millions of ticks in production storage—to estimate continuation versus mean reversion probabilities. Outputs are proposals: size, horizon, and asset. Risk officers can veto or cap any proposal before it hits treasury.

This split is deliberate. Fully autonomous trading sounds modern but fails compliance and sanity checks in real stress. We optimize for consistent process, not lottery tickets.

Exchange listing sleeve

Tokens with credible listing announcements receive a dedicated playbook: verify the announcement channel, measure liquidity depth, check wallet concentration, and only then scale entries. Fake listing screenshots are common; we treat unverified images as hostile input.

Data hygiene

News spam is a denial-of-service attack on traders. Filters include deduplication hashes, source reputation weighting, language detection for bot farms, and cross-checks against on-chain activity (is anything actually moving?). When feeds conflict, the default is reduce size, not increase it.

Relationship to investor plans

Published plan rates reflect long-run desk targets and treasury policy—not a guarantee that today’s headline trade wins. The narrative stack improves decision quality; it does not remove loss days.

Transparency

We publish high-level descriptions of this stack so investors understand we are not a static “4% button.” Operations remain proprietary—exact model weights and thresholds are trade secrets protected for the benefit of the pool.

How to read our journals

Articles like this one explain how we think. They are not invitations to mirror trades externally. Your interface remains the dashboard: deposit, plan selection, accrual history, withdrawal requests.