Inside a footprint chart, most price cells are roughly balanced — buyers and sellers trade in similar size. Now and then one side overwhelms the other, and that lopsidedness, an imbalance, is a fingerprint of concentrated aggression. When several imbalances stack in a row they mark the levels where one side truly pressed its advantage.
What a bid/ask imbalance is
A footprint prints two numbers at every price: the volume that traded at the bid (aggressive selling, shown red on the left in the diagram) and the volume that traded at the ask (aggressive buying, shown green on the right). An imbalance is when one of those numbers is far larger than the other it is compared against — one side of the auction dominated that level. A buy imbalance means aggressive buying overwhelmed the opposing selling; a sell imbalance means the reverse. Individually, a lone imbalance is common and means little, because order flow is naturally lumpy. Its significance comes from size relative to a threshold and, above all, from whether it repeats at consecutive prices.
Why the comparison is diagonal
Imbalances are measured diagonally, not straight across the same price level, and the reason is structural. At a single price, the bid volume and the ask volume are the two halves of the same set of trades, so comparing them like-for-like is misleading. The meaningful comparison pits the ask volume at one price against the bid volume at the price one tick below it, because those represent buyers reaching up versus sellers stepping down — genuinely opposing initiative across the level. So a buy imbalance compares the green ask at a price to the red bid diagonally beneath it; a sell imbalance compares the red bid at a price to the green ask diagonally above. This diagonal framing is the convention on virtually every footprint platform.
The imbalance ratio and threshold
An imbalance is declared when the dominant side exceeds the other by a chosen ratio, most commonly 3 to 1, though traders set 200%, 300% or 400% to taste. If 300 contracts trade at the ask against 75 at the diagonal bid, that is a 4-to-1 buy imbalance and clears any usual threshold. A higher threshold produces fewer but more meaningful flags; a lower one produces many and risks noise. Most platforms also ignore cells below a minimum volume so that a 3-to-0 print on a handful of contracts is not dignified as a signal. The exact number matters less than applying it consistently and understanding that it is simply filtering for lopsided, one-sided aggression.
Stacked imbalances
The signal traders actually act on is not one imbalance but several in an unbroken vertical run — stacked imbalances, the bold-outlined cells in the diagram. Three or more consecutive buy imbalances mean aggressive buyers pressed their advantage across a whole span of prices without sellers matching them, which is sustained, deliberate initiative rather than a single lumpy trade. Stacked sell imbalances are the mirror. Because they represent a zone where one side clearly and repeatedly won, the prices they cover tend to matter later: buyers who were that aggressive often defend the area on a pullback. Stacking is what separates a meaningful footprint read from random noise in the tape.
Where imbalances form tells you what they mean
The location of a stack within the move changes its message. Stacked buy imbalances forming as price advances, low in a bar or into support, are confirmation — aggression is fuelling the trend and often marks a level that will hold. The same stack forming right at the top of an extended rally is more suspect: it can be late, trapped buyers piling in at the highs just before absorption reverses them. Sell imbalances read the same way inverted. You therefore never read an imbalance in isolation from its context; a stack at the origin of a move and a stack at its exhaustion point can look identical on the cell but mean opposite things.
How to use them and what to avoid
Practically, stacked imbalances are used to confirm entries in the direction of aggression and to mark levels for later — the prices a stack covers become candidate support or resistance on a retest. They are strongest when they line up with independent structure such as a prior swing, a value-area edge or a high-volume node. The common errors are acting on a single imbalance, using footprints on thin instruments where the bid/ask split is unreliable, and forgetting that a large diagonal print can be one iceberg order rather than many participants. Treated as one input among several, with attention to threshold and context, imbalances are among the most concrete order-flow signals available.