ES and NQ Divergence: How to Test Index Confirmation as a Futures Trade Filter

When ES breaks a prior high and NQ doesn't, is the breakout weaker? Many traders think so, and it's a reasonable question. The honest way to answer it is to log your own trades with and without the check, because I don't have public evidence that the filter improves results.

Why the two overlap

ES tracks the S&P 500 and NQ tracks the Nasdaq-100. They aren't the same set of stocks. The S&P 500 includes companies listed on both major US exchanges, while the Nasdaq-100 holds 100 large Nasdaq-listed non-financial companies. They share many of the same mega-cap names, which is why they usually move in the same direction. The weights differ and change over time, so check the index factsheets for the current picture.

Contract details matter for comparisons. ES is $50 per point with a 0.25 tick worth $12.50. NQ is $20 per point with a 0.25 tick worth $5. Micros are MES at $5 per point and MNQ at $2 per point. RTY is $50 per point with a 0.10 tick worth $5.

What divergence looks like

Divergence means one index makes a new session or prior-day extreme and the other does not confirm it. The four common cases traders describe:

1. ES makes a new high, NQ doesn't. The rally may lack tech participation.
2. NQ makes a new high, ES doesn't. Leadership is narrow.
3. ES makes a new low, NQ holds above its low. Weakness may be concentrated outside the largest tech names.
4. NQ makes a new low, ES holds. Tech is weaker than the broad index.

These are interpretations, not outcomes. A lagging index can catch up in the next minutes, or the leader can fade. Treat each case as a question to answer with price, not a verdict.

Charting it

The simplest method is two charts side by side, same timeframe, with the prior session high and low drawn on both. Watch whether both indexes break the level together. A second method is a ratio line, ES price divided by NQ price. At a hypothetical ES 6,000 and NQ 21,000, the ratio is 0.2857. A rising ratio means ES is outperforming NQ in percentage terms, and a falling ratio means the reverse. Comparing the ratio to its own moving average is something you'd have to test, since I'm not aware of a standard setting that works.

A dollar-neutral spread, if you try one

Equal contract counts do not give equal exposure. At the same hypothetical levels, one ES has about $300,000 of notional (6,000 x $50) and one NQ about $420,000 (21,000 x $20). Matching them takes something like 7 ES against 5 NQ, since both come to about $2.1 million. That is already a lot of margin and risk, and a spread can still lose on both legs. Margin is collateral, not a cap on losses, and each leg has its own stop slippage and costs. Micros give finer control: for the same idea you could use MES and MNQ in 10x smaller increments.

A confirmation filter, tested properly

Say you trade breakouts of a daily reference level on ES. The filter: only take the trade if NQ is also breaking its own corresponding level within a few bars. To find out whether this helps you, run an experiment:

1. Write the exact definition of confirmed (same direction, within N bars, beyond which level).
2. For every breakout setup, record both the trade outcome and whether the filter would have passed. Record the skipped ones too.
3. Compare expectancy in R multiples for confirmed versus unconfirmed setups, after costs.
4. Keep collecting. A few dozen trades won't prove anything, and a filter that cuts your trade count in half also halves your data.

If confirmed trades aren't better after costs, drop the filter. It will take you a minute per trade otherwise.

Adding RTY

RTY tracks small-cap stocks, so some traders check it as a third reference. The same test applies: record whether RTY agreed, and see if the results differ. Don't increase size just because three charts agree. Size comes from your dollar risk per trade.

Time of day

I pay less attention to divergence in the first few minutes after 9:30 am ET (when opening imbalances and gap moves distort everything) and during the thin midday hours. Those are my process preferences, not findings. Close-time order flow can also produce short dislocations. If time of day matters to you, tag it in your trade log and let your data decide.

Fixed daily reference levels are covered in KPL trading. The standard ES/NQ map has four levels anchored to the regular-session opening print, so it isn't complete before 9:30 am ET, and it doesn't move intraday. Run your expectancy numbers in the edge calculator.