Intraday Market Regime Classification: How to Know What Kind of Day You Are Trading

A mean-reversion entry on a trend day and a breakout entry on a range day lose money for the same reason. The tactic didn't match the session. Where you enter matters, but deciding what kind of day you're in decides which entries deserve consideration at all. Below are the day types I use for ES and NQ, the inputs worth checking before and after the open, and a way to log your classifications so you find out whether your reads hold up.

The Five Day Types

Trend day. Price keeps making higher highs and higher lows (or the reverse) through the regular session. It stays mostly on one side of VWAP, pullbacks are shallow relative to the prior leg, and the opening range gets extended rather than revisited. Volume is usually heavier than normal, and the daily range often runs well past the average.

Range day. Price rotates between a defined high and low and fails each time it tries to break out. It crosses VWAP several times, order flow is mixed, and volume tends to be lighter. The opening range often contains most of the day's action.

Neutral or overlap day. Price accepts the middle of the recent range, the day's value area overlaps the prior day's, and neither side takes control. Volatility is usually low and neither trend tactics nor fades have much to work with. Sometimes the right trade is none.

News-driven day. A scheduled release (FOMC, CPI, jobs data) sets the session's direction through repricing instead of routine intraday flow. Gaps, wide opening ranges and fast one-way moves are common, and ordinary technical reference points can fail more often than usual.

Expansion day. A quiet range breaks a level where stops have stacked up, such as a multi-day boundary, and the session turns into a trend. It starts looking like a range day, which is why it catches fade traders.

What to Check Before the Open

These are context inputs, not predictions. Treat each as a reason to adjust your plan, then confirm after the open.

• Gap location. An open outside the prior day's range puts you in a different situation than an open back inside it. Check whether the overnight session held the gap or filled it.
• Economic calendar. Note every scheduled release and its time in ET. A day with an 8:30 AM ET release isn't a normal day.
• Recent range versus ATR. After several unusually quiet sessions, a catalyst can produce a larger move. That's a reason to respect breakouts, not a forecast.
• Overnight behavior. A one-directional overnight session tells you where the market already moved before the cash open. It doesn't tell you whether it continues.

Early-Session Reads (9:30 to 10:15 AM ET)

The regular US cash open is 9:30 AM ET, except on holidays and early-close days. By about 10:15 you have 45 minutes of data. Four things to look at:

• Opening range versus 14-day ATR. Many traders use roughly half of ATR and one full ATR as dividing lines. A small range leans toward a quiet day, a range near or beyond ATR points to an unusual one. Those cutoffs are starting parameters you should test on your own data, not established facts.
• VWAP behavior. Holding one side of VWAP leans trend. Crossing it repeatedly leans range. Whipping across it on big candles usually means a news reaction.
• Cumulative delta. A steadily rising or falling reading means aggressive orders are consistently on one side. A flat, oscillating line means they aren't. Delta depends on your data feed's bid/ask classification, so see the data feed note below.
• NYSE TICK. Readings pinned at one extreme for long stretches suggest broad one-sided activity. Readings that bounce around the middle suggest balance. Which numbers count as extreme changes with market conditions, so calibrate against your own charts.

KPL are fixed reference prices anchored to the regular-session opening print, so they can't be finished before 9:30 AM ET and they don't move afterward. Once they exist, they give you fixed places to judge whether a move through a level is happening. See how KPL trading works for how the four standard ES and NQ levels are built.

A Checklist at 10:15 AM ET

1. Is there a tier 1 release today, and has it already happened? If yes, classify as news-driven and treat usual setups cautiously until the reaction settles.
2. Has price left the prior day's RTH range? If yes, favor expansion or trend as your working assumption.
3. Has price crossed VWAP at least once? If yes, favor range as your working assumption.
4. Is the opening range large relative to ATR? If yes, assume a larger-than-normal day and size accordingly.
5. Is delta directional or mixed?

When these disagree, you've found a day you can't classify. That's useful information. Smaller size or no trade is a legitimate answer.

Matching Tactics to the Day

On a suspected trend day, the usual candidates are opening range breaks in the trend direction, pullbacks toward VWAP, and break-and-retest entries, often with a runner. On a suspected range day, they are VWAP deviation band fades and failed-breakout fades with targets back near VWAP. Whether a given automated approach fits a given regime is specified in the bot documentation (see the Marty bot specs), not something to assume from the chart. On news days, many traders stand aside through the initial reaction, cut size, and widen stops to match the abnormal volatility, which also means each contract risks more dollars.

When the Day Changes Underneath You

The costly mistake is holding a range-day mindset after the day has become a trend. Watch for a break of the prior day's high or low or of a KPL on expanding volume, a delta line that suddenly drifts one way after hours of oscillation, a TICK reading that stays one-sided when it hadn't before, and a range boundary that has been tested repeatedly and finally gives way. When that happens, re-check every open position against the new assumption and stop adding to the old one.

Test Your Classifications

Don't trust your read until you've checked it. For each session, write down your 10:15 classification and the evidence behind it, then record what the day actually looked like at the close. Track how often your early call matched. Thirty days of notes is a start for spotting habits, but it isn't proof that your rules work, and a sample that small can't separate skill from luck. Keep logging, and change one threshold at a time so you can tell what helped.