FinautorFinautor
Signals

Reading & using Signals

How to read a regime score and put it to work in your own research.

Finautor Signals turn a complex market question — "what regime are we in?" — into a single, systematic reading you can check at a glance. This guide covers how to read one and how to put it to work.

Educational market research, not investment advice. Signals are generic and identical for every subscriber; the methodology is proprietary.

What a signal is

A signal is a daily, machine-learning regime score: a number from 0 to 100 with a regime label. Two are live today — a composite market-regime score for broad U.S. equities, and a credit-stress regime for broad credit markets. Each reading also shows a 20-day direction (▲ / ▼) and an as-of date with a staleness flag.

How to read the score

  • The score (0–100) places today's environment on a scale — the higher end is the more constructive one for that signal, the lower end the more defensive.
  • The regime label names the band in plain language (risk-on / risk-off for equities; Tight / Normal / Widening / Stress for credit).
  • The direction tells you which way the last 20 days have leaned — momentum, not a forecast.
  • The as-of date is the reading's timestamp; a staleness flag appears if the latest data is behind.

Putting it to work

A regime score is context for your own research, not a trade and not personalised advice. Use it to frame the environment — for example, to sanity-check whether your own read of the market matches a systematic one. The score is generic: every subscriber sees the same number.

The current reading on every signal is open to read. A subscription adds the full published history (up to five years), so you can see how a regime has moved over time, plus API and MCP access to pull the reading into your own tools.

What's under the hood

Deliberately, the score is all you get: the inputs, weights, and formula are proprietary and never returned. For the approach — what the model reads and how it classifies regimes, without the secret sauce — see the Signals methodology.

Next