Skip to content

Trend day or failed breakout: why the breakout candle cannot tell you

By Author of ORDER FLOW — Market Mechanics

Two rules about breakouts circulate side by side. One says a break of the Initial Balance on strong delta is real. The other says the candle with the largest volume and delta of a move is its climax. On the day it matters, both rules describe the same candle.

The breakout candle cannot settle which one applies; the next two or three candles can, and even “the largest-delta candle” depends on a setting the reader chooses. The answer decides more than one trade: it decides whether the session is being read as a trend day or as a rotation that just tested its edge.

What the breakout candle records

The Initial Balance is conventionally the range of the first hour of the regular session, 9:30 to 10:30 New York time on Nasdaq futures. A break of its high puts price somewhere the first hour did not trade. The candle that does it records two things: how much traded, and how much of it was aggressive buying net of aggressive selling. That net figure is the delta, and how reliable it is depends on whether the data attributes each trade from the real bid and ask or infers it.

What the candle cannot record is what happens next. Heavy aggressive buying through a level is exactly what a move looks like when it starts, because buyers have to pay up to get filled. It is also exactly what a move looks like when it ends: stops above the level trigger, late buyers chase, and passive sellers who had been waiting for that liquidity fill them. The print is the same in both cases, and the difference is whether anyone keeps buying.

A break that starts a move Breakout candle: delta +540, 3,600 contracts IB high Closes CVD from the prior candle peak priorbreak1234 CVD peak: +1,490, last candle Price holds outside the IB A break that exhausts it Breakout candle: delta +540, 3,600 contracts IB high peak priorbreak1234 CVD peak: +600, breakout candle Closes back inside the IB 2 candles later The breakout candle is identical in both cases. What changes is where the CVD peaks.
Schematic of two breaks of the IB high with an identical breakout candle. 5-minute candles; CVD accumulated from the candle before the break. Values are illustrative.

Both breakout candles in the schematic carry a delta of +540 on 3,600 contracts. On the left, CVD keeps building after the break and peaks on the last candle at +1,490. On the right, CVD peaks on the breakout candle itself at +600, gives it all back to 0, and price closes back inside the range.

Start or climax: what the next candles show

The book reads the distinction from the sequence of candles. When a break starts a move, the two or three candles after it hold price outside the range, volume holds up or grows, and CVD keeps accumulating rather than topping out on the break. The footprint of those candles shows no fresh aggressive selling coming back in.

When the break is the climax, the sequence runs the other way. The breakout candle is the highest-volume, highest-delta candle of the move. The candles after it lose mass straight away, CVD flattens or reverses, and bid-side imbalances appear that were not there before. Whoever was defending the edge still had size: they had stopped defending it passively and waited for the break to sell into it. A stacked imbalance on the breakout candle does not settle the question either: the same stack can mark an initiation or a climax, depending on where it sits in the move.

A break that starts a move 1,400 prior 3,600 break 2,900 1 3,100 2 2,850 3 3,000 4 4 of 4 later candles above the mean A break that exhausts it 1,400 prior 3,600 break 1,900 1 2,100 2 1,500 3 1,300 4 0 of 4 later candles above the mean first-hour mean: 2,800 Both breaks clear the mean on the candle that breaks. Only one keeps the volume on the candles that follow.
Schematic of volume per 5-minute candle in the same two breaks, against the mean of the first-hour candles. Values are illustrative.

In the schematic, the first-hour candles average 2,800 contracts. Both breakout candles clear that comfortably. After the break, 4 of the 4 following candles stay above it in one case and 0 of 4 in the other. Neither reading is available until those candles exist.

“The largest-delta candle” is a measurement

The climax rule has a weak point: whether the breakout candle is the largest-delta candle of the move depends on how long a candle is. Group the same trades into different intervals and the peak can move.

1 minute crosses the level +260 3-min candles +75 +330 largest delta, contains the break +300 +270 6-min candles +405 contains the break +570 largest delta Same minutes, same total delta, +975. The largest-delta candle is the breakout candle at 3 minutes and the next one at 6.
Schematic of the same twelve minutes of delta grouped into candles of two different lengths. Values are illustrative.

With 3-minute candles, the minutes in the schematic read +75, +330, +300, +270, and the breakout candle is the largest. With 6-minute candles they read +405 and +570, and the largest candle is the one after the break. The total is +975 either way. This is the same kind of choice as the swing threshold behind a delta divergence: a setting that decides what the data appears to say. The cumulative line is less sensitive to it than any single candle, which is one reason to read whether CVD keeps building, and to fix the interval before the session starts.

What it says about the day

A break that is accepted turns the session into a candidate trend day, and the cleaner setup is the first pullback to the broken edge, where the footprint has to show the level being defended. A break that fails is information too: the market tested new prices, found no interest and came back. The day reads as rotational, the edge that seemed broken is reinforced, and the setups that make sense are rejections at the extremes rather than continuations.

An accepted break also leaves the prior session's point of control behind. If price never comes back, it becomes a naked POC, and whether it is still naked a week later depends on rules the reader sets as well as on the market.

This is also why a narrow first hour cannot classify the day on its own. In the book's glossary, a narrow IB is where most trend days come from, and also where the no-opportunity days come from: sessions in which every extension fails and the flow never builds a story. The behaviour at the edges is what tells them apart.

Limits and false positives

The answer arrives late. Waiting two or three candles means price has already moved. The cost of reading it correctly is a worse entry or no entry.

Trend days pause. A real trend can stall for a candle or two after the break, with CVD flattening briefly. One quiet candle is normal; aggressive selling reappearing at the broken edge is a different matter.

Major data releases. Around a scheduled release, volume and delta spike for reasons unrelated to the level, and both breaks and failures stop meaning what they mean on an ordinary session.

The reference is a choice. Comparing volume against the first-hour mean is one reference among several. Whatever reference is used has to be fixed in advance, or it will be chosen to fit the outcome.

A held break can still fail. Order flow records what happened at the level. A break that holds for three candles can still fail later in the session, and the reading has to be updated when it does.