Glossary — trading terms used across Reversal Labs

Quick definitions for the concepts you will see on the dashboard, in the TradingView indicator, and in the daily digest. Each entry is anchored so you can deep-link.

Mean reversion

The thesis that asset prices tend to snap back toward a short-term average after stretching unusually far in one direction. Mean-reversion setups enter positions when the stretch is extreme and bet on the snap-back. It is a contrarian style by construction. Background: Mean reversion (finance) on Wikipedia.

Trend

A sustained directional move in price. A trend signal fires when the engine classifies momentum as one-sided for long enough to rule out noise. Trend signals are lagging by design — the wait is the filter. Background: Market trend on Wikipedia.

Oscillator

A rules-based momentum reading in the RSI family. The Reversal Engine oscillator combines momentum with adaptive bands so that extreme readings only mark genuine stretch, not routine volatility. Background: Relative strength index and technical analysis on Wikipedia.

Support

A price level where buying has historically overwhelmed selling and halted a decline. The Reversal Engine derives support deterministically from recent swing lows. Adaptive support shifts as new swings print. Background: Support and resistance on Wikipedia.

Resistance

The mirror of support. A price level where selling has historically overwhelmed buying and capped a rally. Derived from recent swing highs; adaptive to new prints. Background: Support and resistance on Wikipedia.

Divergence

A disagreement between price action and momentum. Classic bullish divergence: price makes a lower low but the oscillator makes a higher low — momentum has stopped confirming the downtrend. Divergence rules on Reversal Labs fire only when the disagreement is structural, not noise-driven. Background: Relative strength index on Wikipedia.

Confluence

The count of independent rules firing on the same ticker on the same bar. Higher confluence means more lenses agree. Reversal Labs ranks signals by confluence first. Background: technical analysis on Wikipedia.

Win rate

The share of prior instances of a rule on a specific ticker that closed above the entry bar at the defined forward horizon (typically 5 bars). Computed per (signal, ticker) — not universally.

Average return

The mean forward return across prior instances of a rule on a specific ticker, measured at the same horizon as win rate. Expressed as a percentage.

Net P&L

The cumulative hypothetical return from taking every prior instance of a signal at equal sizing, measured over the full historical window. Used to sanity-check a signal against buy-and-hold on the same ticker over the same span. Excludes slippage and fees.

Falling knife

A bullish signal that fires inside an unresolved decline. The name comes from "do not catch a falling knife" — the signal looks like a reversal but the driver pushing price down is still active, so the bounce fails and the decline resumes. See the falling-knife check for how to screen these out.

Signal hit

A single firing of a rule on a ticker on a bar close. Stored as one row in the signal-hits table. The Market Radar is a filtered view of recent signal hits ranked by confluence and track record.

Triggered value

The numeric reading that caused the rule to fire. For an oscillator rule it is the oscillator value; for a trend rule it is a percentage or z-score. Useful for ranking within a single signal type — a deeper oversold reading on the same rule is a stronger trigger.

Rule type

Two kinds: strategy rules are complete entry definitions with a tracked historical record; signal rules are building blocks that may or may not appear together. Strategy rules are the ones to act on directly; signal rules are the raw inputs that confluence counts.

Direction

Bullish or bearish. Every signal carries a sign. The dashboard defaults to bullish because that is what most users trade, but bearish fires are written to the same table and can be filtered in.

Bucket

A grouping of signals by asset class: stocks, crypto, funds, other. Buckets are used on the Today page to let you see what fired in each slice of the market without scrolling a single long list.

Timeframe

Which bar the rule runs on — daily (1D) or weekly (1W). Daily rules fire on each trading-day close. Weekly rules fire on Friday close. Weekly is lower noise and better suited to funds and indices; daily is faster and suits stocks and crypto. See daily vs weekly.

Confluence score

Same as confluence. Sometimes referred to as confluence count or confluence value in older material — the three terms are interchangeable.