QIS-A-300 QISAGENT.COM THE INTELLIGENCE LAYER REV 2026-07-28 · BUILD 10.0
The Intelligence Layer · Utility

The QIS tape.

A rules-based read of the instruments that actually determine systematic strategy behaviour — trend, carry, volatility and dispersion proxies — refreshed each US trading day. Deterministic, published, and reproducible: the code that generates it is on this page, so you can run it yourself and get the same numbers.

QIS-A-300·Rules-based·No discretion, no forecasts

Tape QIS regime proxies Loading…
Hover to pause Delayed, end-of-day. Not a quote feed. snapshot.json

Levels are end-of-day and delayed. This is a regime read for research context, not an execution feed, and nothing here is a recommendation or an offer.

The Briefing

The tape, in your inbox, every morning.

The same numbers plus the one paragraph that explains what moved and why it matters to a systematic book. Free, daily, and you can leave in one click.

Daily tape brief on trading days; the weekly Signals briefing on Sundays. One email address, both publications, no third-party sharing ever.

The Wire

Quant finance, as it happens.

Headlines from the desks, journals and feeds that matter to systematic strategy. Sources are configured in feed-sources.js and rendered client-side; nothing is stored and nothing is tracked.

The wireNo sources configured

An empty wire is the honest default. Sources are added deliberately, not scraped indiscriminately — see the workflows for how we treat provenance.

Reproducibility

Run it yourself.

The generator is a single Python file with two dependencies, run on a schedule by GitHub Actions. It takes no discretionary input, so the same date produces the same output on your machine as on ours. That is the whole point.

git clone <this repository>
pip install yfinance pandas
python daily_snapshot.py        # writes snapshot.json

# The tape reads snapshot.json. Same file, same numbers,
# no hidden state and no forecast anywhere in the pipeline.

The methodology — instrument selection, the lookbacks, and why each proxy was chosen — is documented in the research workflows. If you find an error in it, we would rather hear about it than not.