Methodology

Trifecta Signals is a rules-based stock screener. After every market close it scans US exchange-listed stocks for a rare oversold setup, waits for independent confirmation before a signal counts, and grades every confirmed signal against the S&P 500 at fixed horizons. This page documents the rules the public track record is built under — the universe, admission, the two signal gates, and the accounting standard — at the level the record itself is published. Last updated: November 2026.

The universe and admission

The screen covers US exchange-listed common stocks only — NYSE, Nasdaq, NYSE American, NYSE Arca, and Cboe. There are no OTC or penny stocks. A stock is admitted at its Early Signal only if it passes a number of proprietary filtering criteria, applied point-in-time from data available on the signal date.

Admission is final. A stock admitted at its Early Signal stays in the record even if it later falls outside those criteria — the signal log is append-only, and a name that stops trading stays in the record, marked DELISTED.

Gate one: the Early Signal

An Early Signal fires when technical indicators point to a stock sitting in oversold territory.

An Early Signal is a candidate, nothing more. It enters a monitoring queue; it does not enter the official track record, and no headline statistic is computed from it.

Gate two: confirmation

A candidate becomes a Confirmed signal only when additional technical indicators signal a bullish reversal. If that confirmation never comes, the signal is closed out as never confirmed. It leaves the current screen, its figures freeze, and it is listed separately on the Track Record, labelled CLOSED — never confirmed.

The gate exists because most oversold stocks keep falling. Of the Early Signals that never confirmed and had a full 20 sessions to run, 57.8% were trading lower 20 sessions after the Early Signal (record as of November 6, 2026). Requiring independent confirmation before a signal counts is what separates the record from a list of falling knives. As of November 6, 2026, 12,940 of the 20,199 Early Signals on record had confirmed.

Categories and groups

Every signal carries a category from its point-in-time data at the signal date. Quality signals passed the screener's quality checks; Low Quality signals did not, and were admitted only under stricter filtering criteria. Where the filings available by the signal date cannot establish the tests either way, the signal is left uncategorized rather than guessed at.

Quality signals are also assigned to exactly one group, ranking the strength of the bottom evidence present at the Early Signal. The groups prioritize investigation only — they are not an admission gate:

• Washout — a quality company sold far below its long-term trend on climax selling; the business intact, the price not.
• Capitulation — the washout climaxes: volume spikes as the last holders give up and forced selling burns itself out.
• Exhaustion — the decline runs out of sellers; the rarest pattern the screen produces.
• A ▼ Downtrend marker replaces the group for certain repeat Early Signals in established downtrends.

In an October 2026 study of the confirmed cohort, the share that finished positive measured from the Early Signal was 85% for Washout (mean +38%), 77% for Capitulation (mean +59%), and 75% for Exhaustion (mean +32% — a small sample, early evidence), against 63% for all confirmed signals.

How the record is graded

• Measurement origin. Headline statistics are measured from the Confirmation Close — the adjusted close on the day a signal confirms — never retroactively from the Early Signal price. The record also publishes since-Early-Signal figures, because the two decision points are graded separately.
• Horizons. Each confirmed signal is graded at 30, 60, and 90 calendar days after confirmation. A horizon counts only once it has fully elapsed; younger signals show a dash rather than a partial result.
• Benchmark. Forward returns are reported alongside the S&P 500 over the same window, so a signal's result can be read against what the index did.
• One issuer, one count. Confirmed signals from different share classes of the same issuer landing close together are counted once in the statistics; every class stays in the published log.
• Nothing is removed. Wins, scratches, losses, and delisted names remain in the record permanently. No record is retroactively adjusted or deleted.

Where the history comes from

The record reaches back to January 2022. Its earlier years are a point-in-time backfill: every historical signal was tested only against data available on its own signal date — SEC EDGAR facts as originally filed on or before that date, and shares outstanding from the closest filing on or before it — and the whole history is graded under the current rules, so the record reads as one standard throughout. Two caveats are part of the record, not footnotes to it: the universe is built from current listings, so companies delisted before the backfill do not appear; and the backfill is a rules-based reconstruction, not the performance of any actual investment account. A live daily scan now extends the record after every close. The full record, with every signal and outcome, is on the Track Record, and the guide to reading a track record shows how to interrogate it.

What Trifecta Signals is not

Trifecta Signals is an algorithmic research publication operated by Sappen LLC. It does not execute trades, manage funds, or provide individualized financial advice, and nothing on this site is a recommendation to buy, sell, or hold any security. Past performance does not guarantee future results.

Educational material only — not investment advice.