AXIS Desk·The free series

The Weekly Drop

One tested NQ stat a week, with its sample, its method and its misses. Kills get equal billing with keepers — the caveat is the credibility. Newest first.

Nº 06STUDY

Big orders move price. They don't predict it.

Every aggressive sweep on the NQ tape since late 2020 — 3.68 million of them, 56,401 of 100 lots or more. A big sweep moves the mid 10 to 33 ticks and keeps it; after that the average forward move is zero at every horizon to 30 minutes. What comes back first, what doesn't, and the three things that actually change the aftermath.

Nº 05KEEPER

Morning holds, afternoon pays.

If NQ hasn't traded 0.3% below its 9:00 open by 11:00, adding long has paid about +0.30R a trade over 603 sessions — positive every year since 2020. The method, the caveats, and the mirror that fails.

Nº 04STUDY

The Three Opens

The 8:30 level, the 9:30 cash open and the 10:00 price — every ordering across 1,401 NQ sessions. We went looking for a directional signal and found a map: mixed states revisit their anchors ~80% of the time, and only the clean bull stack tilts direction. No entry. Not yet through the null.

Nº 03KEEPER

One number at 10:30 tells you what kind of day it is

Travel from the RTH open, divided by ATR, read once at 10:30: it calls balance vs directional with +7.0% genuine skill after the shuffled-returns null — and the quiet end is the strong end. Modest, real, useful by lunch.

Nº 02KILL

An 81% win rate that pays $0.00

Early high plus a 70% give-back by 10:20 and the session makes a new high only 29.9% of the time (n=284, base 63.8%). Traded the obvious way it wins 81% of the time and makes −0.6 points per trade. Hit rate is not edge.

Nº 01KILL

Is that the high? We measured it. Then we killed it.

A calibrated, walk-forward-validated curve for whether the current extreme holds — 1,403 NQ sessions, 100,990 observations. Then the structureless null: 97% of the curve is geometry. The kill is worth more than the curve.

Want the next one when it lands? Join the free list. No signals — research and context only.