Show HN: STIX – Institutional positioning in US equities revealed as a score
What it is: STIX produces a per-ticker score for US equities and ETFs, plus sector and market aggregates, updated daily. Higher and rising scores suggest stronger structural support; lower and falling scores suggest potential distribution or fragility. It’s not a price predictor—think of it as a distilled view of where larger players appear to be accumulating or distributing risk.
Why I built it: Most trading tools surface signals that are incredibly complex to understand and require a lot of subjective interpretation. I wanted a positioning-centric tool that simply and concretely expressed to users whether a stock had a higher probability of appreciating or deteriorating.
How it works (high level): Each day STIX ingests trade/quote data from all US exchanges and engineers features around size/lot anomalies, time-of-day behavior, persistence, breadth, clustering and more. A rules-plus-ML pipeline weights these features and normalizes them from which a 0–100 score is derived per symbol. Zones where larger players campaign tend to mark areas of interest; holding above those zones can indicate supportive positioning, while trading below can indicate distribution or short pressure.
Outputs: Current score, short-term change, sector rollups, and simple screens.
Try it: https://stix-v1.onrender.com — use promo code STIXTRIAL to waive the $2.50 day pass and get 24 hours free.
Feedback I’d love: Does the signal explanation make sense? What, if anything, is confusing in onboarding or UX? Did you hit any latency or performance bottlenecks? Which programmatic outputs (API or CSV) would be most useful?
Notes: This is research/analytics, not financial advice. No trackers or ads. Emails (if used) are for key generation only. I’m iterating actively—candid criticism is welcome, and I’ll be in the thread.
No comments yet