Fair Value Gap vs Support and Resistance
Michael Whitman
·
Trading since 2017, full-time since 2020
·
YouTube creator, 137K subscribers
·
Published 2026-09-08
·
Updated
·
Reviewed
Fair value gaps are strips of untraded price between the outer wicks of three fast candles. Support and resistance mark prices where the market turned before, such as a swing low. On the S&P 500 both held about equally often on the first touch; only the gap's width changed how often it broke.
A swing low is a price where buyers showed up. A fair value gap is a range where nobody traded much. Built from opposite evidence, both are used the same way: as a place to watch when price comes back. On the S&P 500 index the first return to each ended about the same way.
What each one is
A fair value gap leaves a band, not a line. In a bullish one, price skipped the stretch between candle one’s high and candle three’s low. The full rule is on the fair value gap page.
Support and resistance are prices the market stopped at before. Here they are five-bar swing lows (support) and swing highs (resistance), a low or high that beats the five sessions on each side. Support and resistance explains why they are bands in practice.
A level is a line; a gap is a band. That difference in shape drives most of what follows.
Where they differ
What made them. A swing low exists because price reversed there. A gap exists because price moved through quickly. One records a reaction and the other the absence of trading.
When they can be marked. A five-bar swing low is known five sessions after it prints. A gap can be drawn once candle three closes, two sessions after candle one.
How often price comes back. Gaps sit close to where price is, so most were touched within 60 sessions; swing lows are often left behind by a rising market.
What “broken” means. A level breaks on any close past it. A gap has an edge to touch and a far edge to close past, so a close inside the band is neither held nor broken.
Which one to use: Fair Value Gap or Support and Resistance?
Build your map from support and resistance levels, since they come from prices where the market already turned and others watch the same swings.
Use a gap when you want an area with room inside it, and treat its far edge, not its near edge, as the line that decides whether it failed.
When a gap and a swing level coincide, judge them on one record, not two. On this index neither held more often than the other on its own.
The first touch, compared
The test. Take each bullish gap and each swing low followed by at least 60 sessions, and note the first session within those 60 that reaches it. A gap held if that session closed above the band; a level held if it closed at or above the level. Bearish gaps and swing highs are scored the same way, upside down. Swing levels count from the day they become known.
Bullish gaps: 1,791 of 2,058 touched (87.03%). On the first touch 847 held (47.29%), 480 closed inside the band (26.80%) and 464 closed below it (25.91%). Swing lows: 364 of 616 touched (59.09%), of which 161 held (44.23%) and 203 closed below (55.77%).
Held against held, the gap and the level were within chance (Fisher exact test, p 0.30). Closed below against closed below, they were far apart (p below 0.001), and that difference is the band’s width: a gap has a whole band to fall through before it counts as broken, a line does not.
Two September 2026 gaps
One held. The index’s high on 17 September 2026 was 7,646.86 and its low on 21 September 7,691.19, leaving a bullish gap between them. On 24 September the low of 7,662.57 reached into it, and the session closed back above the band: a hold on the first touch.
One broke. The 2 September high of 7,681.19 and 4 September low of 7,706.12 left another bullish gap. On 8 September the low of 7,666.99 went through the band and the close of 7,673.52 sat below it.
Same index, same month, opposite outcomes, which is what a hold rate near one in two looks like.
The original data
The sample. 10,519 daily bars of the S&P 500 index (^GSPC) from Yahoo Finance, 2 January 1985 to 2 October 2026, highs, lows and closes. The start is 1985 because earlier years in this download include runs of bars with no high-low range. 3,426 gaps; 616 swing lows and 569 swing highs had 60 sessions after them.
The bearish side. 95.61% of bearish gaps were touched; 456 held (35.46%), 397 closed inside (30.87%) and 433 closed above (33.67%). 483 of 569 swing highs were touched (84.89%); 187 held (38.72%) and 296 closed above (61.28%). Held against held: p 0.222.
The file. Every gap and swing level with its first-touch result is in the S&P 500 gap and level file.
What is left out. Later touches, the size of any bounce and what followed are not measured. A level here is one exact price; traders who draw support as a band would score it differently.
When it fails
The failure is crediting the gap with a strength that is only its width. A gap closes “below” less often than a line because price has to cross the whole band first. Count a close inside the band as a hold and the gap looks far better than support; count it honestly and the two hold about equally. The width is real, and it costs a wider stop.
The second failure is putting the stop at the gap’s near edge. More than half of first touches closed beyond it.
A third is choosing levels after the reaction. Then every level held.
A fourth is reading index results as single-stock results. A single stock can behave differently.
And a fifth is counting every touch as a test. The first touch is the one with a clean record.
Related
Fair value gap has the band rule. Support and resistance covers levels and why they are bands. And imbalance is this site’s other name for the same three-candle gap.
Judge a gap and a level on the same first touch before you decide one is better. On the S&P 500 they held about equally often. The gap’s advantage was only its width, so if you use one, put the stop below the whole band, not at the edge.
— Michael Whitman
This page is educational, not financial advice. Test every idea on your own charts before risking money.