Previous Day High and Low: 8,471 SPY Sessions Counted
The previous day high and low are the highest and lowest prices of the prior regular session, and they are among the first reference levels many traders mark on a new day's chart. What matters is less whether price trades through them than whether it closes beyond them.
Many day traders draw two lines before anything else: yesterday’s high and yesterday’s low. They are objective, everyone can see them, and they frame the new session before a single bar prints. This page defines them, follows SPY across one real week, and counts how often those levels were broken, held and broken back since 1993.
How it forms
Each session leaves two extremes behind. The high is the most anyone paid during the regular session; the low is the least anyone accepted. Overnight, those two prices become the edges of the map for the next day.
Orders gather around them. Some traders who missed the move place buy-stop orders just above the prior high to join a breakout, and some holders protect gains with sell stops just under the prior low. Where such orders cluster, a trade through the level can fill many of them at once, and a quick reversal afterward leaves those buyers near the top of the move.
The open sets the first question. If the market opens between the two levels, the day starts inside yesterday’s range and the levels act as boundaries. If it opens above the high or below the low, the gap has already crossed one, and the question becomes whether the gap holds.
Regular-session and extended-hours charts disagree. A chart that includes premarket and after-hours trading will often show a different high and low from the 9:30 a.m. to 4:00 p.m. session. This page uses regular-session figures throughout, because those are the prices in the official daily bars.
The definition used here
“Took the high” means the session’s high was above the prior session’s high by any amount. “Took the low” is the mirror. A day can do both, which is an outside day, or neither, which is an inside day in the loose sense (a strict inside bar also needs its high and low to differ from the prior day’s).
“Held” means the close finished beyond the level. A day that trades above the prior high and closes above it held the break. A day that trades above it and closes back below it did not.
Two samples. Daily bars cover SPY from 29 January 1993 to 25 September 2026, 8,471 sessions that have a prior session. One-hour bars, which the Yahoo chart API serves for the most recent 730 days, cover 27 October 2023 to 25 September 2026: 730 sessions and 729 day-to-day comparisons, with each session’s high and low rebuilt from its own hourly bars. On the seven half days the source adds a 1:00 p.m. bar with no volume that holds prints from outside the session, so that bar is left out.
A worked example
Start on Friday 18 September 2026. SPY’s regular session ran from a low of $757.97 to a high of $762.00. Those were the levels for Monday.
Monday 21 September opened at $766.25, $4.25 above Friday’s high, so the gap had already cleared it. The session reached $774.89 and closed at $773.50, well above $762.00. That break held.
Tuesday 22 September shows the other outcome. The new prior high was $774.89. SPY reached $775.14, just $0.25 through it, in the first hour of trading, then closed at $773.38, back under the level. A buy-stop placed a few cents above $774.89 was filled near the top of the day.
Wednesday 23 September went the other way. Tuesday’s low was $772.57. SPY fell to $766.50 and closed at $767.81, $4.76 below that low, so the downside break held. Thursday opened at $764.07, under Wednesday’s $766.50 low, and closed at $767.18, back inside. On Friday 25 September SPY traded up to $772.28, through Thursday’s $768.95 high, and closed at $771.35 above it.
One week, four kinds of day. A gap that held, a poke through that failed, a break down that held, and a gap down that was bought back. Every one of them “broke” a prior-day level.
The original data
Across 8,471 SPY sessions, trading through a prior-day extreme was the normal state of affairs. The high was taken out on 4,560 days (53.8%) and the low on 3,799 (44.8%). Exactly one side went on 6,645 days (78.4%). Both went on 857 (10.1%), and neither on 969 (11.4%).
Closing beyond the level was much less common. Of the 4,560 days that took the prior high, 2,669 closed above it (58.5%) and 1,891 closed back below it (41.5%). Of the 3,799 days that took the prior low, 1,967 closed below it (51.8%).
Where the day opened made a large difference. SPY opened above the prior high on 1,895 sessions (22.4%), and 1,280 of those closed above it (67.5%). On the 2,665 days that opened at or under the prior high and then traded through it during the session, 1,389 closed above it, or 52.1%: close to an even split.
The hourly data shows when it happened. In the 729 comparisons, SPY traded above the prior high on 434 sessions. On 328 of them (75.6%), the first trade above it came in the opening hour, 9:30 to 10:30 a.m. Eastern, and 356 (82.0%) had happened by 11:30. The six later hourly bars took 28, 19, 25, 12, 8 and 14, the last covering only the half hour to the 4:00 p.m. close. The low was taken on 301 sessions, 198 of them (65.8%) in the first hour.
Gaps explain part of that, but not all. 190 of the 434 sessions opened above the prior high, which puts them in the first hour automatically. Of the 244 that opened below it and crossed later, 138 (56.6%) still crossed in the first hour. The session file behind these counts is published as SPY prior-day breaks by hour.
How the mix has shifted by decade
Days that took neither side have become rarer on SPY. They were 13.3% of sessions from 1993 to 1999, 11.7% in the 2000s, 11.4% in the 2010s and 9.3% from 2020 to September 2026. Taking the prior high rose from 51.0% in the 2000s to 56.5% since 2020.
The upward tilt shows up clearly. In every decade the prior high was taken more often than the prior low. That reflects an index that rose over the period, and it is a reason not to treat a break of the high and a break of the low as mirror images on SPY.
When it fails
The poke and reverse is the familiar failure. On 41.5% of the days that traded above the prior high, SPY closed back below it, as it did on 22 September 2026. A buy-stop just over the level can turn into a fill near the day’s high.
Outside days take out both sides. On one session in ten (857 of 8,471), both prior-day levels traded before the close, so orders resting at both were both filled. A false breakout at the first level can set up exactly that.
Gaps can jump the level entirely. When the open is already beyond yesterday’s extreme, the level was never traded at, and orders parked there fill at the open price rather than at the line.
A different chart gives a different level. Extended-hours highs, futures versus the ETF, and dividend-adjusted versus unadjusted data all move the numbers slightly. A level a few cents off can decide whether a stop is hit.
And the level ages quickly. In the hourly counts above, only 78 of the 434 first breaks of the prior high came after 11:30 a.m. When the prior high fell, it had usually fallen by late morning.
Related
Support and resistance covers the broader idea these two levels belong to, and why old extremes attract orders. The opening range breakout builds a second pair of levels from the first minutes of the new session. And pivot points combine the prior high, low and close into a formula that projects further levels for the day.
Mark the prior high and low before the bell, then decide in advance what a close beyond them would change. A trade through the level in the first hour is common enough that I wait to see where the session settles before I lean on it.
— Michael Whitman
This page is educational, not financial advice. Test every idea on your own charts before risking money.