SPY and QQQ Daily Read: 10-Minute Workflow

A working morning routine on SPY and QQQ takes about ten minutes and has a fixed shape: write down five premarket levels, capture four charts, ask one structured question, and end with two scenarios per instrument each carrying an invalidation price. An AI copilot compresses the writing-up part of that, which is where most of the ten minutes used to go. It does not replace any of the data gathering or the decision.

Why SPY and QQQ are the right pair to standardize on

Most US-focused traders touch these two charts every session whether or not they trade them, because they are the de facto read on the market. SPY tracks the S&P 500 and QQQ tracks the Nasdaq-100, and their top holdings overlap heavily.

That overlap is the point. Two charts with shared constituents and different concentrations give you a cheap breadth signal. When both are trending and agreeing, the tape is broad. When QQQ is making highs and SPY is stalling, the advance is narrow and concentrated in a handful of large-cap technology names. When SPY holds and QQQ breaks down, the rotation is out of that same group. You get that read from two pictures, in seconds, without a single screener.

Standardizing on a pair also makes reads comparable over time, which a rotating watchlist never will. Fifty reads of the same two charts, produced by the same question set, is a dataset about your own judgment. Fifty reads of fifty different tickers is a scrapbook.

Minutes 0 to 2: write the five levels

Before any chart, before any AI, write these down. They are data, not analysis, and the model cannot get them for you.

Two minutes, five numbers, per instrument. This is the least glamorous part of the routine and the part that changes outcomes most, because a level you wrote down is a level you will not discover at the worst possible moment.

Minutes 2 to 4: capture four charts

One chart per image, price axis and time axis visible, volume on, at most two studies.

• SPY daily, about six months. Enough to carry the swing structure and any multi-month shelf.
• SPY 15-minute, last three sessions. Enough to see yesterday's shape and this morning's premarket.
• QQQ daily, about six months.
• QQQ 15-minute, last three sessions.

On the intraday panes, VWAP earns its space. On the dailies, a moving average pair is enough for slope. Nothing else. Every additional subgraph shrinks the candles, and candle geometry is what a structural read is made of. The capture mechanics for specific platforms are in our thinkorswim and TradingView screenshot guides.

Minutes 4 to 7: one structured question

Use the same prompt every session. The consistency is what makes reads diffable.

Four images: SPY daily (6 months), SPY 15m (3 sessions),
QQQ daily (6 months), QQQ 15m (3 sessions).
My levels, already measured:
  SPY prior high X, prior low Y, prior close Z,
  overnight high/low A/B, premarket high/low C/D
  QQQ (same five)
For each instrument:
1. Daily trend from the swing structure, not the averages.
2. Which of my levels price is currently between.
3. Any level on the daily I have not listed that price
   has clearly reacted to, marked as read-from-axis or
   estimated.
4. A continuation scenario and a reversion scenario,
   each with an invalidation PRICE.
5. One sentence on whether SPY and QQQ structures agree,
   and what the disagreement implies if they do not.
No recommendation, no sizing, no prediction.

The two instructions doing the heavy lifting are "marked as read-from-axis or estimated" and "invalidation PRICE". The first is a hallucination detector: a tool that cannot distinguish a measured level from a guessed one will fail it visibly. The second is the field that makes a scenario something you can risk-manage rather than a mood. Our AI chart analysis page shows the output shape, and Bullynx pricing covers what reads cost.

Minutes 7 to 10: gap scenarios and the decision gate

By now you have levels and structure. The remaining question on most mornings is what the open is doing relative to yesterday, and that is worth classifying explicitly rather than reacting to.

Gap inside the prior range. Price opens between yesterday's high and low. This is the common case and the least informative. The prior high and low remain the day's likely boundaries, and the routine ends with "no edge until one of those is tested".

Gap above the prior high. Two clean paths. Continuation holds above the prior high and uses it as the invalidation reference. Fill rejects from the open and targets the prior close, invalidated by a reclaim of the opening high. Write both, with prices.

Gap below the prior low. Mirror image. The distinction that matters is whether the gap has a cause. A gap following an overnight macro print behaves differently from a drift-down on no news, because the first has a reason to extend. Gap trading strategies covers the classifications in more depth.

Gap with a known event today. CPI, FOMC, a major earnings report from a top holding. The honest output here is often that the structural read is real and not actionable until after the print. Writing that down is a result.

Then the gate: every number the read produced goes back onto your live chart and gets confirmed against the actual axis before it touches sizing. Only after that does it reach a position size calculator and your own risk per trade rule.

The three ways this routine goes wrong

Worth naming, because each one is common and each one is cheap to avoid.

Reading the 15-minute first. If you open the intraday chart before the daily, you form a view from noise and then look at the daily for confirmation of it. The order in the sequence above is deliberate: structure, then position inside structure. Reversing it is how a two-week swing read becomes a scalp you did not plan.

Letting the read supply the levels. If you skip the two minutes of measuring and ask the model to find the levels instead, you get estimates presented as facts. The prompt above hands it your measured numbers on purpose, and asks it to flag anything it adds as read-from-axis or estimated. A tool that cannot make that distinction is telling you something about its other numbers too.

Running it on a day you were never going to trade. Half days, the session before a holiday, the hour before an FOMC statement: the routine still produces a clean read, and acting on it is the mistake. Writing "no action, event risk" is a complete and correct output, and a journal full of those entries is a sign the process is working rather than failing.

What the AI does not do

It does not have your premarket levels. It has an image. The five numbers in the table above are yours to measure, which is why they go into the prompt rather than coming out of it.

It does not see live price. The read describes a screenshot taken at a moment. Ten minutes into the session on an index ETF, that moment has passed. This is a pre-open and between-setups workflow.

It does not know the calendar. Earnings from a top-five holding, a CPI release, an FOMC statement, a half day before a holiday: none of it is in the picture, and all of it can make a structurally clean read irrelevant.

It does not see breadth, options positioning, or flows. The SPY-versus-QQQ comparison above is a crude breadth proxy built from two images. It is not advance-decline data, it is not gamma exposure, and it should not be treated as either.

It cannot substantiate an accuracy claim. No vendor in the category, ourselves included, publishes a benchmark that would let you verify one. Our evaluation methodology publishes the tests rather than a score.

It does not decide. You get a structured description and two conditional paths per instrument. Which one you act on, at what size, with what stop, is yours alone.

Making the ten minutes stick

The routine works because it is closed-ended. Five levels, four captures, one prompt, two scenarios each, then stop. Open-ended chart browsing before the open feels like preparation and mostly produces a view you then look for reasons to justify.

Log the output. A trading journal holding the read, the level you traded off, and what happened is what turns a daily habit into feedback. After thirty sessions you will know whether your continuation scenarios or your reversion scenarios are the ones that actually pay, which no amount of reading about SPY will tell you.

For a deeper single-instrument treatment see our SPY chart analysis guide, for the futures version of this routine see prop firm challenge chart review with AI, and for the alert-driven variant see TradingView alerts to AI analysis.