# Strat Stats: ES and NQ futures, 2018-2026

Descriptive statistics for TheStrat candlestick setups on CME E-mini S&P 500 (ES) futures, built from 3,070,709 one-minute bars (2018-01-01 to 2026-09-22), with a replication on Nasdaq-100 (NQ) futures. Published by 345 Strategies (https://345strategies.com). Format after The Strat Stats by @notprofgreen (https://x.com/notprofgreen). Research-only: not a strategy, not a forecast, not financial advice.

Interactive page: https://345strategies.com/research/strat-stats
Machine-readable results (JSON): https://345strategies.com/research/strat-stats.json

## Key findings

- The setup label alone barely beats a random break. Of 290 setup cells with at least 100 trades, 6 differ from their baseline at q < 0.05 (Benjamini-Hochberg).
- Across every first break, +1R is reached before the stop about 49-54% of the time, depending on timeframe and direction.
- The one cell flagged on both ES and NQ is a negative: 15m 2D then 2D, traded long (-4.8 points vs baseline on ES; -2.5 on NQ in our run, -2.8 on @notprofgreen's page).
- Timeframe continuity (TFC) adds nothing measurable to these first breaks.
- Replication: run on NQ over @notprofgreen's window (2019-06-02 to 2025-03-14) and timeframes, the same code reproduces his page: bar counts within 0.1%, 186 setup cells with lift correlation 0.99, same sign 95% of the time, mean difference 0.38 points. His only significant cell, 15m 2D-2D long, flags too (-2.5 vs his -2.8).
- What it means: the edge, if any, is not in the bar label. It has to come from context, targets, or management, which this study does not test.

## Definitions

- Bar types compare each bar with the one before, highs and lows only. 1 = inside (broke neither side). 2U / 2D = broke the prior high only / low only. 3U / 3D = broke both; 3U if it closed above its own open, otherwise 3D. Equal is not a break.
- Failed 2 (F2U / F2D): a 2 that closes back inside the prior bar's range, edges included. Only 2s can fail; there is no F3.
- Combo: two closed bars, C2 then C1. C1 is the setup bar.
- Entry bar: the bar after C1 closes. Long = first trade through C1's high, short = first trade through C1's low (a stop order; if the minute gaps through, the fill is that minute's open). No break, no trade. A minute that breaks both sides is dropped.
- Stop: the other side of C1 plus one tick. 1R = entry to stop.
- Outcome: +1R (and +2R) reached before the stop within the entry bar plus two more bars, checked on 1-minute bars. The entry minute is skipped; a minute touching stop and target counts as the stop.
- Baseline: all first breaks on the same timeframe and direction. Wilson 95% intervals, z-tests, Benjamini-Hochberg q-values.
- Sessions: 18:00 ET (15:00 PT) start; 4h bars at 18, 22, 02, 06, 10, 14 ET. Prices back-adjusted at each contract roll from raw prices.

## Baselines (ES, all first breaks)

| Timeframe | Direction | Trades | +1R before stop | +2R before stop | Trigger bar closes beyond entry |
|---|---|---|---|---|---|
| 15m | long | 85,564 | 53.3% | 23.6% | 50.3% |
| 15m | short | 80,272 | 53.9% | 26.7% | 47.8% |
| 30m | long | 42,135 | 52.9% | 23.8% | 51.1% |
| 30m | short | 38,738 | 52.7% | 27.6% | 47.2% |
| 1h | long | 22,675 | 52.0% | 24.4% | 51.6% |
| 1h | short | 19,986 | 52.3% | 28.4% | 46.9% |
| 4h | long | 6,008 | 52.3% | 26.5% | 54.1% |
| 4h | short | 4,990 | 50.9% | 29.7% | 47.3% |
| D | long | 1,124 | 48.6% | 18.6% | 54.3% |
| D | short | 835 | 50.2% | 27.1% | 45.6% |

## Setup cells that differ from baseline (ES, q < 0.05)

| Timeframe | C2 | C1 (setup) | Direction | Trades | +1R before stop | vs baseline (points) | q |
|---|---|---|---|---|---|---|---|
| 15m | 2D | F2D | long | 6,280 | 49.2% | -4.1 | 0.001 |
| 15m | 1 | 1 | long | 2,080 | 57.6% | +4.3 | 0.016 |
| 15m | 2D | 2D | long | 1,880 | 48.5% | -4.8 | 0.032 |
| 30m | 1 | F2D | long | 1,473 | 58.1% | +5.2 | 0.033 |
| 30m | 2D | F2D | long | 3,024 | 49.2% | -3.7 | 0.033 |
| 15m | 2U | F2U | short | 6,375 | 51.6% | -2.3 | 0.045 |

Trades that follow one another share price paths, so q-values are optimistic. Treat these as reasons to look again, not as proof.

## Limitations

- One trade definition: a symmetric 1R bracket over three bars. Exits, targets (magnitude, exhaustion) and management are not tested.
- No commissions or slippage. In-sample description; a dev (to 2022) and holdout (2023 on) split is on the interactive page.
- Unfillable data gaps: 2019-02-26 19:40-22:45 ET and 2020-06-30 18:00-19:59 ET. Nothing is interpolated.

## Reproduce

Everything on the page is rebuilt from 1-minute bars by open code: https://github.com/345-strategies/strat-enrichment (bar types in strat_enrichment/, the statistics in stats/). Market data is licensed and not redistributed, so bring your own.

## Credit

Format and trade definition after The Strat Stats (NQ) by @notprofgreen: https://x.com/notprofgreen. TheStrat methodology by Rob Smith. Bar-type colours and definitions follow TheStrat Suite: https://thestratsuite.com