Do Inside Bars Work?

An inside bar is one whose entire range fits within the previous bar: a lower high and a higher low, the market quietly pinching in. Chartists treat it as a pause before a move and trade the break of its range in either direction. We wanted to know two things. Does an inside bar predict which way price goes next, and does it actually precede a bigger move the way the setup promises?

We ran it on SPX, the S&P 500 index; QQQ, the Nasdaq 100; and SMH, the semiconductors, on both daily and weekly bars across their full history. For every inside bar we measured two things: the close-to-close return over the next bar, to test direction, and the size of the very next move against an average bar, to test whether the range really does expand.

A single inside bar predicts nothing

Take them one at a time first. Inside days and inside weeks are common, hundreds of each across the decades, and they are followed by returns that sit right on top of each instrument’s baseline.

Next-bar close-to-close return and the size of that next move versus an average bar. For reference, the all-bar green rate runs 52–54% daily and 55–56% weekly, so the inside-bar returns land on baseline. Full listed history through mid-2026.

Pattern (count)	Next bar (avg / % up)	Next move vs a normal bar
SPX inside day (1,426)	+0.02% / 52%	same (0%)
QQQ inside day (687)	+0.03% / 54%	+3%
SMH inside day (658)	+0.06% / 50%	+2%
SPX inside week (265)	+0.15% / 59%	+1%
QQQ inside week (104)	−0.01% / 52%	+2%
SMH inside week (143)	−0.11% / 52%	−3%

The direction is a wash: 52% green after an inside day on the S&P against a 53% baseline, and the other names land just as close to their own averages. The more telling column is the last one. The whole appeal of an inside bar is that a tight range stores up energy for a bigger move. It does not. The bar that follows a single inside bar is the same size as an ordinary bar, within a couple of percent everywhere, and on the semis’ weekly chart it is actually a touch smaller. A quiet bar is simply followed by another ordinary bar.

Two in a row is where a bigger move shows up

Stacking two inside bars back to back changes the second half of the story, but not the first. The move that follows does grow.

Average size of the next move (up or down) after two consecutive inside bars, versus an average bar. Occurrence counts in parentheses; the weekly buckets are small.

Two in a row (count)	Normal move	After 2 inside bars	Bigger by
SPX daily (53)	0.73%	0.91%	+24%
QQQ daily (17)	1.14%	1.19%	+4%
SMH daily (23)	1.55%	1.46%	−5%
SPX weekly (15)	1.67%	2.08%	+25%
QQQ weekly (4)	2.45%	2.97%	+21%
SMH weekly (4)	3.29%	3.90%	+19%

On the weekly chart the effect is consistent: after two inside weeks, the next week travels about a fifth to a quarter farther than a normal week, and it holds on all three (SPX +25%, QQQ +21%, SMH +19%). On daily bars it shows up clearly only on the S&P, at +24%; QQQ is flat and the semis are slightly quieter. So a double contraction does compress the range enough to matter. What it does not do is tell you which way the release goes. The direction after two in a row is a coin flip, and on the tiny weekly samples it even skews slightly red by chance.

You cannot stack them past two

The obvious next question is whether three or four in a row compress even harder. They do not happen. Three inside bars in a row occurred exactly once in the whole record on the S&P, on the daily chart and on the weekly, and never on the Nasdaq or the semis. Four in a row has never happened on any of them, on either timeframe. The reason is mechanical: each inside bar demands a lower high and a higher low, so a third one running would need the range to shrink to almost nothing, and at that point it nearly always breaks out instead of pinching in again. The pattern runs out of room. Two is the practical ceiling, and two is where the expansion tops out.