Trading the Close: How to Trade the Last Hour of ES and NQ Futures

The last hour of the US cash session, 3:00 to 4:00 PM ET, does not behave like the 10:00 AM hour, and trading it with morning rules is a good way to hand back a profitable day. Different orders are working, the clock is short, and your own P&L starts to push on your decisions. This article walks through what changes, three setups worth testing, and the mistakes that show up most in the final stretch.

What Changes Between 3:00 and 4:00 PM ET

Futures keep trading after the cash close. Regular US equity cash trading ends at 4:00 PM ET, but CME equity futures like ES and NQ run on Globex until the daily break, so the close is a marker for the stock market, not the end of your instrument. Holiday and early-close days have different schedules, so check the CME holiday calendar rather than assuming 4:00 PM.

Closing auctions matter. Index funds and other participants send market-on-close orders that execute at the official closing price. Exchanges publish order imbalance information shortly before the close, and the exact publication times and order cutoffs are set in each exchange's rules, so read them instead of trusting a number from a blog post. One caveat: that data covers individual stocks, not ES or NQ directly. A big buy imbalance in the cash market does not mean futures must rise. Treat it as context, not a trigger.

Options expiration can add movement. Around monthly expiration, dealers hedging large near-expiry positions may push price in the final hours. That is a mechanism, not a guarantee, and it varies by day. Mark expiration days in your journal and see what your own trades show.

Liquidity and fills can differ. Spreads and slippage may change late in the day compared with the busiest morning minutes. Don't assume either way. Compare your actual fills from 3:00 to 4:00 PM against your morning fills and see whether the gap is large enough to change your size or order type.

Three Setups to Test, Not Rules to Trust

I don't have a sourced hit rate for any of these, so treat each one as a hypothesis you test in a journal or in replay before risking real size.

1. Trend extension into the close

If the session has trended since the morning and the afternoon did not reverse it, you can look for continuation. A simple filter is that VWAP is still sloping in the trend direction at 3:00 PM. For entries, watch a KPL level on a pullback. KPL levels are fixed reference prices anchored to the regular-session opening print, so by 3:00 PM they are already set and do not move. The plan: enter on a confirmed reaction at the level, put the stop beyond the next level, and target a new session high. The strategy overview is at /strategies/kpl-trading.

2. Fading an unconfirmed 3:00 PM spike

Sometimes price makes a sharp move around the 3:00 PM transition without broad participation, and it can retrace. Define the spike in numbers before you trade it. A hypothetical definition: at least 10 ES points in 5 minutes, with TICK not confirming and the move against the VWAP trend. Entry is after the spike candle closes and the next candle closes the other way. Stop goes beyond the spike extreme, target is VWAP.

Hypothetical example: you enter 4 points off the extreme with a 6 point stop and an 8 point VWAP target. That is $300 of risk on one ES ($30 on MES) for $400 of potential reward on ES, about 1.3 to 1. Add commissions and slippage before deciding that is worth it.

3. Imbalance data as context only

An imbalance reading can add or remove conviction on a trade you already have a reason to take. A large buy imbalance on a strong up day is consistent with holding a long. The same reading on a down day may do nothing. Don't build an entry rule from it without your own logged results.

How to Test These Before Risking Size

Tag every trade in your journal with entry time and setup name, then review the 3:00 to 4:00 PM bucket separately from the morning. Record spread paid, slippage versus intended price, and the reason you exited. Thirty trades is not a statistically proven edge, and neither is fifty simulated ones. A simulated fill also tells you nothing about queue position or real slippage. Use the sample to find process errors, and keep size small until live fills confirm what the journal suggests.

Timing Mistakes in the Last Hour

New complex entries late. A setup that needs 45 minutes of follow-through is a poor fit at 3:35 PM. Late entries should be quick mean-reversion ideas with small targets or continuation of an existing position, and you should decide in advance whether you hold into the evening session.

The final minutes. Price can move erratically around the close and fills can differ from what the candle suggests. Many traders set a personal flat time a few minutes before 4:00 PM ET. Pick yours, write it down, and test whether it helps.

Trading your P&L instead of the chart. A trader who is up wants to add to a good day, and a trader who is down wants to recover. Both are about money, not setup quality. Ask the same question you ask at 10:00 AM: is this on my plan, and is the risk defined?

Automated Strategies and the Close

Mean-reversion logic and directional logic can both be affected by end-of-day flow, which is one reason session windows and cutoffs matter. For what the current bot variants do around the close, read /docs/marty-bot-specs and /docs/kpl-bot-specs instead of relying on a number from an article. If you run any automation alongside manual trades, set your own cutoff for new entries, decide who manages open positions after it, and label every time in your plan with its timezone.