NQ, 5-minute bars, New York regular session. Data: Yahoo Finance (unofficial, may be delayed).
Not investment advice. The rules below are published and tested on history; the forward journal shows what they do going forward. About two trades in three lose, and most of the profit came from a few strong years.
Loading…
Green and red lines are the noise-area limits at each check (the stepped lines change every 30 minutes). Arrows are entries by the rules, circles are exits.
Today's levels
All twelve checks of the day. The lines depend only on the open, the previous close and past sessions, so they are known right after 09:30 and do not need a live feed: put them on your own chart.
Loading…
How to use this page
After 09:30 New York read the table above: for every check it gives the upper and the lower line.
On your own live 5-minute chart watch the bar that closes at each check time (10:00, 10:30 … 15:30). If it closes above the upper line, the rules say long; below the lower line, short.
Enter at the open of the next bar. The stop is the stop column of that check (and at most 58.5 points from your entry); the target is 1.5 σ from your entry; close everything at 16:00.
The data on this page comes from a free source and runs about 15 minutes behind the market. Use the page for the levels and the history, and your broker or TradingView chart for the moment of entry.
The rules in six lines
Twelve checks a day: the closes of the 5-minute bars ending 10:00, 10:30 … 15:30 New York time.
Noise σ for a check is the average move from the day open to that same time over the last 14 sessions.
Upper line = max(open, previous close) × (1 + σ); lower line = min(open, previous close) × (1 − σ).
A check that closes above the upper line is a long signal, below the lower line a short one; entry at the next bar's open.
Stop 0.24 σ beyond the line (it moves with the lines), plus a hard limit of 58.5 index points from the entry; target 1.5 σ from the entry.
Everything closes at 16:00; no new entries after four losing trades in a day; after a stop, flat until the next check.
Tested history
Real NQ futures, January 2015 – July 2025, 5-minute bars, costs and one tick of slippage included. * 2025 is partial.
Loading…
Position size is chosen so that the stop costs exactly this share of the account, so the account size does not matter, only the risk per trade does: more risk gives a higher return and a deeper drawdown (compounded, estimated between tested values 0.25%, 0.5%, 1%, 1.5%, 2%, 3%). Years of weak trends (2016, 2017, 2019) were flat or negative. We reproduced the published rules in two separate implementations (Python and the TypeScript behind this page); their daily results agree on all 2,611 sessions.
Calculated now from the source history. This is not the forward journal. Click a row to see that day on the chart.
Loading…
NQ noise bands: an intraday momentum rule for Nasdaq-100 futures, tested and published
This page applies a published intraday momentum rule to Nasdaq-100 futures (NQ, with the micro contract MNQ for small accounts). For each half hour of the New York session it measures how far price usually travels from the day open, builds an upper and a lower line from that average, and treats a 5-minute close beyond a line as a signal in that direction. The stop trails the line, the target is a fixed multiple of the noise, and everything is closed at 16:00. The rule comes from the research of Zarattini, Aziz and Barbon on intraday momentum and from an open-source implementation that documents every parameter; we re-implemented it ourselves and checked that our code reproduces our independent Python replication day by day.
Tested on about ten years of real NQ futures (2015 to July 2025) with commissions and one tick of slippage, it earned about 28 percent a year when risking 1 percent of the account on each trade, with a worst drawdown near 46 percent; risking half as much roughly halves both (about 14 percent a year, 26 percent drawdown). Roughly two trades in three lose; the profit comes from fewer, larger winners in trending days. About two thirds of the total came from three years (2018, 2022, 2023), and 2016, 2017 and 2019 were flat or negative.
How to use the page and what it does not promise
The status card tells you whether the rules currently ask for a long, a short or nothing, with the stop and target levels. The chart shows the two lines and every entry and exit. Decisions stay yours: the page is a calculation of published rules, not advice, and past results do not guarantee future ones. Data comes from Yahoo Finance, an unofficial source that can lag, so compare the price with your broker before acting.
To keep the claim honest, every complete session from 30 October 2026 is stored automatically in the forward journal, with the result the frozen rules produced, and is never edited. Our own research on this site found no reliable edge in simple level or pattern rules, which is why this rule was accepted only after independent replication on a long history. The forward journal is the real test: judge the rule by it after at least several months.
Questions and answers
What is the noise area?
For each check time it is the typical distance price has moved from the day open over the last 14 sessions. A close beyond it means the move is larger than ordinary noise.
Is this a signal service?
No. It shows the output of published rules and a public forward journal. You decide whether to trade, and you should check price and size with your broker.
Why does the profit look uneven?
Momentum rules earn in strong trending periods and lose small amounts in quiet ones. Most of the historical profit came from three years.
Does it work on crypto?
We tested the same logic on six large coins with hourly bars and found no edge after costs, so it is offered for Nasdaq futures only.
