Best Time of Day to Day Trade: What 723 Sessions of SPY Show

Michael Whitman · Trading since 2017, full-time since 2020 · Published 2026-09-25

“When should I trade?” has a measurable answer for any instrument, because the busy and quiet parts of the session repeat. This page measures them for SPY and turns the result into a trading window you can write down.

Before you start

Your round-trip cost as a percentage of price. Take it from a filled order: commission both ways plus the spread you crossed. On a $500 share, a total of $0.10 is 0.02%.

A chart that shows each bar’s time in New York time. Every figure below is Eastern time, and a chart set to another zone will put the busy hours in the wrong place.

A written rule for when your window opens and closes. Two clock times on paper, before the session.

How it works

As of September 2026, the regular US stock session runs from 9:30am to 4:00pm Eastern, and activity inside those hours is uneven. Overnight news and orders arrive at the open, the middle hours thin out, and index funds and other benchmark holders trade into the close. The New York session page describes that shape, and the Nasdaq overnight session planned from 6 December 2026; this page counts the regular hours.

Two things matter to a day trader, and both change by the hour. Volume sets how easily you get filled near the price you see. Range, the distance from an hour’s high to its low, sets how much movement there is to capture. A fixed cost takes a bigger share of a small range.

The steps

1. Pull intraday bars for the instrument you trade

Use 5-minute or 60-minute bars covering at least a few months. SPY is used here; a single stock or a future has its own profile.

2. Compute each time slot’s share of daily volume

For every day, divide each slot’s volume by that day’s total, then average across days. Averaging raw volume would let a few huge days dominate.

3. Compute each slot’s average high-to-low range as a percentage

Divide each slot’s high minus low by its opening price. Percentages let a $50 stock and a $500 one be compared.

4. Divide your round-trip cost by each slot’s range

This is the share of an average move your costs eat. Mark the slots where it is smallest.

5. Set your window where volume and range are both high

Write the start and end times down. For SPY on this data that is 9:30 to 10:30am Eastern.

6. Stop trading when the window closes

Close or leave any open position under its plan and take no new trades until the next window.

A worked example

Use the 5-minute figures for SPY over 60 sessions, 2 July to 25 September 2026, and a hypothetical round-trip cost of 0.02% of price.

The first half hour, 9:30 to 10:00, averaged a range of 0.322% of price. The cost is 0.02 divided by 0.322, which is 6.2% of the average move.

The quietest half hour, 1:30 to 2:00pm, averaged 0.135%. The same cost is 0.02 divided by 0.135, or 14.8% of the move, more than twice the share.

The last half hour, 3:30 to 4:00, averaged 0.204%, for 9.8%. It carried the most volume of any half hour, 21.0% of the day, but much of that sits in the final five-minute bar, which alone averaged 9.5% of daily volume. That bar is where closing orders execute, and it moves far less than its volume suggests.

So on this data a trader paying 0.02% keeps the most of each move by trading the first hour, and pays the most, relative to movement, just after lunch.

How to tell it worked

After 20 sessions, compare your trades inside the window with any outside it. Record each trade’s cost as a percentage of the range captured.

The check is concrete: the average cost share inside the window should be lower than outside it, as 6.2% was against 14.8% in the example. If more than 2 trades in 10 fall outside the window, the rule is not being followed.

The original data

60-minute bars for SPY across 723 full sessions, 27 October 2023 to 25 September 2026, from Yahoo Finance. The first hour, 9:30 to 10:30am, carried 22.4% of the day’s volume on average and had the widest range, 0.52% of price. The hours from 11:30 to 2:30 carried 9.7% to 10.7% each, with ranges of 0.32% to 0.34%.

The final bar, 3:30 to 4:00, is half an hour long and carried 19.9% of volume with a 0.29% range. Rounded to two decimals, the average hourly ranges were 9:30am 0.52%, 10:30am 0.39%, 11:30am 0.34%, 12:30pm 0.33%, 1:30pm 0.32%, 2:30pm 0.31% and 3:30pm 0.29%. Each volume share here is measured against the total of Yahoo’s 9:30am-to-4:00pm bars; on 25 September those bars summed to about 30.3 million shares, against the 35.2 million Yahoo reported for the whole day, so the split is approximate.

By half hour over the 60 sessions, the average share of daily volume was 9:30am 14.3%, 10:00am 8.4%, 10:30am 7.1%, 11:00am 6.7%, 11:30am 5.4%, 12:00pm 5.3%, 12:30pm 4.7%, 1:00pm 4.8%, 1:30pm 4.0%, 2:00pm 5.1%, 2:30pm 6.1%, 3:00pm 7.1% and 3:30pm 21.0%.

The day’s high or low was set in the first hour on 512 of 723 sessions, 70.8%: the high on 240 and the low on 286. The last half hour held the high or low on 250, or 34.6%. Over the shorter 60-session window, the first half hour alone held one of them on 31 of 60 days.

What that suggests: the extremes a day trader is trying to catch are mostly made early. That is also why the open is the hardest period to trade well, as the trade the open page explains.

In this site’s study of 24,971 trading and investing videos, deduplicated by video id, 52 titles are about the time of day to trade. 47 of them, from 16 channels, put the “power hour,” the last hour, in the title, many as recurring live streams. The median across the 50 that show a view count is 1,720.

On this data the last hour is heavy in volume but narrower than the first, so the popular window and the widest one are not the same. Every figure is in the intraday CSV.

Why the close looks busier than it trades

Much of the late volume is orders meant to fill at or near the official closing price. Index funds and benchmarked accounts send them to the closing auction. In the 5-minute data the final bar alone averaged 9.5% of the session’s bar volume, yet the last half hour’s range was well under the first half hour’s. That is real volume, but it does not create a wide range to trade.

The last hour is still useful for positions that should not be held overnight, and for traders whose day trading plan includes being flat by 4:00pm. It is a place to finish, more than a place to start, and the trade the close page covers how to handle it.

When it fails

The first failure is copying SPY’s profile onto another instrument. A small stock on news day, a future that trades nearly 24 hours, or a currency pair each has its own shape. Run the steps on what you trade.

The second is treating the busiest window as the easiest. The first hour has the widest ranges and the most extremes, and it also has the fastest reversals and the widest spreads of the day, especially in the first few minutes.

The third is trading the quiet hours out of boredom. Midday hourly ranges, 0.32% to 0.34%, were 35% to 38% smaller than the first hour’s 0.52% on this data, so the same cost takes a bigger bite, and small moves invite overtrading.

A fourth is ignoring scheduled news. Data releases at 8:30am and 10:00am Eastern, and central bank statements in the afternoon, can make a quiet hour violent. An economic calendar check belongs in the morning routine.

And a fifth is treating 60 or 723 sessions as permanent. The numbers here describe one fund over one stretch, and a new market regime can reshape it. Re-run the measurement every few months.

Related

The New York session page describes the shape of the US trading day and the overlap with London. The trading sessions page compares activity across the global sessions for markets that trade around the clock. And the day trading page covers the cost base that makes the choice of window matter in the first place.

This page is educational, not financial advice. Test every idea on your own charts before risking money.