The First 5-Minute Candle: Does the Opening Range Predict Where QQQ and SPY Close?

Curated by FolioViz · Published 2026-10-01 · Updated 2026-10-01 · 9 min read

The verdict · Suggestive, not conclusive

After a strong start, QQQ and SPY closed higher 72% of the time, versus 47% on any day.

After a strong start (the next three 5-minute candles all close above the first one), the day closed above the first candle 72% of the time across 144 strong starts in 496 trading days, versus 47% on any day. It held in QQQ and SPY separately, but they move together, so it is one result, not two, and it is not a forecast.

After a strong start, against any day

Based on 496 trading days, two ETFs, 2024-10-08 to 2026-09-30. Green closed above the first candle, grey inside, red below. The dashed line marks the any-day rate.

Data table

Signal	Days	Close below	Close inside	Close above
Strong start up	144	19.4%	8.3%	72.2%

28% of strong starts still did not close above the first candle.

How the day closed after each kind of start, against any day

After a strong start, the day closed above the first candle 72% of the time, against 47% on any day. Green closed above the first candle, grey inside, red below. The dashed line marks the any-day rate.

Data table

Signal	Days	Close below	Close inside	Close above
Strong start up	144	19.4%	8.3%	72.2%
Leaning up	182	24.7%	15.4%	59.9%
Neutral	387	42.9%	14.0%	43.2%
Leaning down	157	51.6%	11.5%	36.9%
Strong start down	122	63.9%	14.8%	21.3%

On Sep 25, 2026, QQQ had a strong start and closed above the first candle's range at $744.45.

One real QQQ session, from the open to the close

This is one example, not proof. The first candle ran from $741.66 to $743.69, and the day ended at $744.45.

Data table

Item	Value
Session	QQQ 2026-09-25
First candle high	743.69
First candle low	741.66
Day close	744.45
Close vs first candle	above

Closes after strong and weak starts overlap heavily, so the first candle is only a small clue.

Where the day closed after each kind of start, as overlapping bands

Strong starts lean right, but every band is wide: the first candle is a small yardstick for a full day.

Data table

Signal	Days	Mean close position
Strong start up	144	2.77
Leaning up	182	1.85
Neutral	387	-0.44
Leaning down	157	-1.88
Strong start down	122	-4.52

Following the signal every day earned 0.07% per trade; simply buying every day earned 0.05%.

Growth of $1 following the signal, against buying every day

605 trades. The win rate was 52.2% following the signal and 57.5% when always buying. The worst fall from a peak was 21.7% against 21.3%.

Data table

Strategy	Trades	Win rate	Avg return/trade	Profit factor	Max drawdown
Signal (long on bull days, short on bear days)	605	52.2%	0.069%	1.26	21.7%
Always long, same days	605	57.5%	0.047%	1.17	21.3%
Random direction (mean of 1000 shuffles)	605		-0.000%

Taken alone, QQQ closed above the first candle 78% of the time after a strong start and SPY 67%.

QQQ and SPY separately, strong start against any day

QQQ alone: 496 trading days.

Data table

Signal	Days	Close below	Close inside	Close above
Strong start up	72	13.9%	8.3%	77.8%
Leaning up	92	22.8%	18.5%	58.7%
Neutral	193	38.3%	17.1%	44.6%
Leaning down	76	56.6%	10.5%	32.9%
Strong start down	63	68.3%	14.3%	17.5%

SPY, the same comparison

Any-day rates were 47% for QQQ and 47% for SPY. The two trade the same sessions, so this is one result seen twice, not two confirmations.

Data table

Signal	Days	Close below	Close inside	Close above
Strong start up	72	25.0%	8.3%	66.7%
Leaning up	90	26.7%	12.2%	61.1%
Neutral	194	47.4%	10.8%	41.8%
Leaning down	81	46.9%	12.3%	40.7%
Strong start down	59	59.3%	15.3%	25.4%

The first candle does not tell you the day. It leans; it does not call.

The first candle does not tell you the day. A strong open is a lean, not a call: plenty of strong starts still fade.

Go deeper

The supporting charts, method and plain-English terms. Each opens on click.

The first candle sets a price range, and the next three candles are labelled by where they close.

How a strong start is defined

The first candle sets H1 and L1. Each of bars 2-4 is labelled by where it closes: above (A), inside (I) or below (B).

Data table

Term	Meaning
H1 / L1	High and low of the 09:30-09:35 ET candle
A	Bar closes above H1
I	Bar closes between L1 and H1 (including exactly on either)
B	Bar closes below L1
Close position	Day close in first-candle ranges, 0 = midpoint, +/-0.5 = H1 / L1

Most days end far outside the first candle's range, because that candle is small next to a full day.

Where the day closes, measured in first-candle ranges

Across 496 trading days, two ETFs, the close above the first candle happens 47% of the time. The dashed line is a normal curve with the same mean and spread.

Data table

Bin from	Bin to	Days
-56.00	-6.00	47
-6.00	-5.75	4
-5.75	-5.50	0
-5.50	-5.25	8
-5.25	-5.00	7
-5.00	-4.75	8
-4.75	-4.50	6
-4.50	-4.25	8
-4.25	-4.00	14
-4.00	-3.75	14
-3.75	-3.50	7
-3.50	-3.25	13
-3.25	-3.00	17
-3.00	-2.75	13
-2.75	-2.50	18
-2.50	-2.25	21
-2.25	-2.00	17
-2.00	-1.75	28
-1.75	-1.50	25
-1.50	-1.25	31
-1.25	-1.00	21
-1.00	-0.75	34
-0.75	-0.50	37
-0.50	-0.25	27
-0.25	0.00	40
0.00	0.25	30
0.25	0.50	30
0.50	0.75	44
0.75	1.00	36
1.00	1.25	22
1.25	1.50	39
1.50	1.75	24
1.75	2.00	31
2.00	2.25	29
2.25	2.50	33
2.50	2.75	26
2.75	3.00	24
3.00	3.25	12
3.25	3.50	21
3.50	3.75	12
3.75	4.00	20
4.00	4.25	14
4.25	4.50	13
4.50	4.75	12
4.75	5.00	7
5.00	5.25	7
5.25	5.50	4
5.50	5.75	2
5.75	6.00	4
6.00	56.00	31

There are 24 three-candle patterns, and many are too rare to trust.

Share of days closing above the first candle, by three-candle pattern

Green cells close above the first candle more often than the baseline, red cells less often. Hatched cells have under 20 days and show no rate.

Data table

Pattern (bars 2,3,4)	Days	Close above
AAA	144	72.2%
AAI	33	48.5%
AAB	2	n too small
AIA	21	57.1%
AII	27	44.4%
AIB	8	n too small
ABI	2	n too small
ABB	2	n too small
IAA	67	68.7%
IAI	27	66.7%
IAB	2	n too small
IIA	83	55.4%
III	196	42.3%
IIB	57	50.9%
IBA	4	n too small
IBI	34	32.4%
IBB	65	24.6%
BAA	2	n too small
BIA	3	n too small
BII	42	40.5%
BIB	21	38.1%
BBA	2	n too small
BBI	26	34.6%
BBB	122	21.3%

Most days follow a few common paths from the first three candles to the close.

Flow of days from candles 2-4 to the day close

Read left to right: candles 2, 3 and 4, then the day close. Green is above the first candle, grey inside, red below.

Data table

Step	From	To	Days
Bar 2 to bar 3	above	above	179
Bar 2 to bar 3	above	inside	56
Bar 2 to bar 3	above	below	4
Bar 2 to bar 3	inside	above	96
Bar 2 to bar 3	inside	inside	336
Bar 2 to bar 3	inside	below	103
Bar 2 to bar 3	below	above	2
Bar 2 to bar 3	below	inside	66
Bar 2 to bar 3	below	below	150
Bar 3 to bar 4	above	above	213
Bar 3 to bar 4	above	inside	60
Bar 3 to bar 4	above	below	4
Bar 3 to bar 4	inside	above	107
Bar 3 to bar 4	inside	inside	265
Bar 3 to bar 4	inside	below	86
Bar 3 to bar 4	below	above	6
Bar 3 to bar 4	below	inside	62
Bar 3 to bar 4	below	below	189
Bar 4 to day close	above	above	213
Bar 4 to day close	above	inside	40
Bar 4 to day close	above	below	73
Bar 4 to day close	inside	above	167
Bar 4 to day close	inside	inside	54
Bar 4 to day close	inside	below	166
Bar 4 to day close	below	above	84
Bar 4 to day close	below	inside	36
Bar 4 to day close	below	below	159

Returns from the fourth candle to the close overlap heavily across signals.

Return from the candle 4 close to the day close, by signal

Medians sit close to zero and the boxes overlap heavily across signals.

Data table

Signal	p10	p25	Median	p75	p90
Strong start up	-0.74%	-0.20%	0.23%	0.60%	0.87%
Leaning up	-0.78%	-0.31%	0.11%	0.39%	0.75%
Neutral	-0.86%	-0.38%	0.01%	0.39%	0.75%
Leaning down	-0.78%	-0.28%	0.11%	0.45%	0.85%
Strong start down	-1.09%	-0.43%	0.02%	0.30%	0.90%

Does the size of the first candle change how reliable the signal is?

Share of signal days closing in the signalled direction, by first-candle size

Smallest (Q1) to largest (Q5) first candle. The dashed line marks 50%.

Data table

Quintile	First-candle range	Days	Hit rate	95% interval
1	0.04-0.14%	198	57.0%	48.4%-65.3%
2	0.14-0.20%	198	63.0%	54.1%-71.2%
3	0.20-0.26%	199	60.0%	51.2%-68.2%
4	0.26-0.36%	198	60.5%	51.7%-68.6%
5	0.36-0.98%	199	67.9%	58.6%-75.9%

Is the signal stable over time?

Rolling 60-session success rate of strong signals, against the any-day rate

A stable edge would stay clear of the dashed baseline.

Data table

Date	Hit rate	Signal days
2024-10-10	0.0%	1
2024-11-08	69.2%	13
2024-12-09	70.6%	17
2025-01-10	70.0%	10
2025-02-10	56.3%	16
2025-03-12	77.8%	18
2025-04-09	84.2%	19
2025-05-09	75.0%	16
2025-06-09	46.2%	13
2025-07-10	75.0%	12
2025-08-07	91.7%	12
2025-09-08	92.9%	14
2025-10-06	66.7%	15
2025-11-04	55.0%	20
2025-12-03	75.0%	20
2026-01-05	73.3%	15
2026-02-03	81.3%	16
2026-03-05	81.8%	22
2026-04-02	61.5%	13
2026-05-04	70.0%	10
2026-06-02	56.3%	16
2026-07-02	46.7%	15
2026-07-31	53.3%	15
2026-08-31	50.0%	16
2026-09-29	63.6%	22

When does price first leave the first candle's range?

Number of 5-minute candles until the first close outside the range

Counts of days by the candle on which the first close outside the range happened, split by direction.

Data table

Bar	Up breaks	Down breaks
2-3	239	218
3-4	96	103
4-5	83	57
5-6	37	31
6-7	22	19
7-8	17	15
8-9	6	12
9-10	3	4
10-11	7	6
11-12	3	2
12-13	1	1
13-14	2	0
14-15	1	1
15-16	0	1
16-17	1	0
17-18	0	0
18-19	0	1
19-20	0	0
20-21	0	1
21-22	0	0
22-23	0	0
23-24	0	0
24-25	1	0
25-26	0	0
26-27	0	0
27-28	0	0
28-29	1	0

Plain-English terms

first candle: The first five minutes of trading (09:30 to 09:35 ET), drawn as one price bar with a high and a low.
strong start: The next three 5-minute candles (09:35 to 09:50 ET) all close above the high of the first candle.
A, I, B: A candle that closes above (A), inside (I) or below (B) the first candle's range. A three-letter pattern such as AAA is candles 2, 3 and 4 in order.

Method and limits

Regular-session five-minute bars only. A close exactly on the first candle's high or low counts as inside. Half-days use their own final bar.
No commissions or slippage are modelled.
The sample covers about two years and two highly correlated ETFs, so the pooled days overlap and the effective sample is roughly half the pooled count.
24 patterns were examined, so some will look strong by chance.
Candles 2-4 are part of the same session as the close, so some persistence is mechanical.
No commissions, slippage or borrow costs modeled.
Entry at bar-4 close, exit at the close of the final regular-session bar.
Equity compounds one full-size trade per day; flat on neutral days.
Everything here is descriptive; none of it is a forecast.

FAQ

What is the first 5-minute candle?

The price bar for the first five minutes of the US session (09:30-09:35 ET). Its high and low form the range the rest of the day is compared with.

Does the first 5-minute candle predict the close?

After a strong start, QQQ and SPY closed higher 72% of the time, versus 47% on any day. After a strong start (the next three 5-minute candles all close above the first one), the day closed above the first candle 72% of the time across 144 strong starts in 496 trading days, versus 47% on any day. It held in QQQ and SPY separately, but they move together, so it is one result, not two, and it is not a forecast.

Does this work for SPY and QQQ?

The study pools both ETFs and also reports each alone. They are highly correlated, so treat them as one market view, not two independent samples.

Not financial advice.