How to read coverage in an Oracle signal

Coverage describes available model inputs

Coverage reflects the availability of the indicators that apply to the selected model, using their stated weights. An operating-company stock can use earnings, valuation and analyst inputs alongside technical indicators. An ETF uses its applicable price-action model. Crypto also faces identity, history and market-quality checks.

Coverage is not the percentage of the internet searched or the probability that a trade will make money. A result with complete model-input coverage can still carry substantial market risk.

The methodology explains which indicators and weights apply to each asset class.

Score and coverage answer different questions

The score summarizes how the available inputs map to the model’s point rules. Coverage describes the inputs supporting that calculation. A score of 90 out of 100 does not mean a 90% expected return or a 90% chance of success.

Consider a hypothetical screen with a score of 82 and 80% coverage. Those numbers describe two different properties. They do not establish a return target or replace the actual classification and its eligibility checks. Read the model and supporting indicators before comparing the result with another asset.

Missing inputs remain part of the result

When an analyst target, earnings input or technical observation is unavailable, the model uses its defined missing-input rules and reduces applicable coverage. A fallback point value is a scoring rule; it is not a verified analyst target, earnings estimate or market observation.

Unavailable also differs from zero. An unverified maximum token supply does not mean the supply is zero or unlimited. Check what each field represents rather than treating every blank as the same kind of information gap.

See stock indicators and crypto data checks for asset-specific context.

Risk and confidence need their own reading

A BUY classification can coexist with high observed risk. A favorable score does not eliminate volatility or an instrument’s structural risks. Confidence describes aspects of indicator coverage and the score’s distance from a classification boundary; it is not a calibrated win probability.

For an ETF, confirm the exact fund and its exposure. A leveraged or inverse product can behave differently from an ordinary index fund even when their model scores look similar.

Read the ETF signal guide before comparing funds with different objectives.

Why a screen can be UNRATED

UNRATED means the available indicators cannot support a directional classification under the applicable requirements. It does not predict that the asset will trade sideways. Insufficient coverage, unusable history, unresolved identity or a blocking quality check can prevent a directional result.

Being discoverable through search does not guarantee that an asset qualifies for analysis. Read the explanation shown with the result to understand which requirement could not be satisfied.

Put the measurements together

Confirm the asset first, then read the classification, score, coverage, risk and supporting indicators. Check the quote time and calculation time before comparing observations. A saved search requests another screen; it does not freeze the earlier result.

Continue with the time behind an Oracle signal or the daily-list reading guide.