Advertisement IC Markets

Flags and Pennants, Trading Continuation Patterns

Master exponential money management, position sizing calculators, strict daily stop-loss limits, and overcoming FOMO on micro-timeframes.
Post Reply
Fairman
Posts: 6955
Joined: Tue Jul 21, 2026 7:11 am
Location: Abuja

Flags and Pennants, Trading Continuation Patterns

Post by Fairman »

Flags and Pennants, Trading Continuation Patterns

After a sharp move, price often pauses in a small consolidation before continuing. Flags and pennants capture these pauses and offer continuation setups with relatively tight risk.
diagram.png
diagram.png (35.15 KiB) Viewed 169 times
Flag: a small rectangle or parallel channel sloping against the prior move.
Pennant: a small symmetrical triangle after the sharp move.

The sharp move before the pattern is called the flagpole.

How to trade them:

1. Identify a strong, impulsive move (the pole).
2. Wait for price to consolidate in a narrow, orderly pattern, with candles that are smaller than those of the pole.
3. Mark the boundaries of the consolidation.
4. Wait for a close beyond the boundary in the direction of the pole.
5. Enter on the break or the retest.
6. Place the stop on the other side of the pattern.
7. Target a distance equal to the pole, or the next key level.

Quality checks:

- The consolidation should be brief and tidy
- The pattern should align with the higher-timeframe trend
- The risk-to-reward should be at least 1:2

Warnings:

- If the consolidation lasts too long or retraces deeply, the pattern weakens
- Breakouts can fail, especially into major levels

Think of flags as a brief rest during a run. Don't chase the pole, and wait for the pause.

Practice finding ten flags on your charts and record how many continued versus failed.
It’s Fairman :geek:
Recommended broker for automated trading & scalping IC Markets
LondonNewsTrader
Posts: 421
Joined: Mon Sep 21, 2026 9:30 am

Re: Flags and Pennants, Trading Continuation Patterns

Post by LondonNewsTrader »

Flags after data releases are worth a separate note, because they are common and they behave a bit differently. The pole is the release candle, the flag forms in the 15 to 45 minutes afterwards, and the break often comes when the second wave of positioning arrives, or when US cash equities open.

The measured move is where I'm cautious. A pole printed in the first minute after CPI includes a liquidity gap, so projecting its full length from the breakout tends to overshoot. I take partials at the next level and treat the pole target as optimistic.

The other check I'd add to your quality list is what else is on the calendar. A flag forming at 13:50 London time with a second release due at 15:00 is not the same setup as one forming on an empty afternoon, because the pattern can be overwritten by the next number before it ever resolves.
FXS Prop Desk
Posts: 1691
Joined: Wed Oct 07, 2026 9:52 am

Re: Flags and Pennants, Trading Continuation Patterns

Post by FXS Prop Desk »

Flags and pennants work well in strong trends. A rule that improves them: the consolidation should retrace less than about half of the pole. A deeper retracement weakens the continuation case, and often turns into a reversal.

Also, the volume (or tick volume) typically declines during the consolidation. That's a weaker signal in spot forex, where only tick volume is available, but still useful as confirmation.
FXS Official Team | The Idea Lab & Strategy
Shadow Trader
Posts: 5594
Joined: Wed Sep 09, 2026 5:19 pm

Re: Flags and Pennants, Trading Continuation Patterns

Post by Shadow Trader »

Brief and tidy is the part I had to turn into numbers, because every flag looks tidy once you know it broke. Mine are a retrace under half of the pole, as FXS Prop Desk says, and a duration of no more than three times the bars the pole took.

So a pole of four M5 bars gets twelve bars of flag. Past that I stop treating it as a pause and start treating it as a range, with the range rules instead.

The other choice that changes results is where the pole starts. Measuring from the swing low gives a longer pole than measuring from the first large candle, and on a typical 40 pip EURUSD pole the two can differ by 10 pips or more, which moves the target by the same amount.

I use the first large candle, which keeps targets shorter. It does not fix the overshoot LondonNewsTrader describes after releases, because the liquidity gap sits inside that first candle, so on release poles I still take the partial at the next level.

For the practice task, I would record the failures by type. Mine split into flags that broke the wrong way and flags that broke the right way and stalled at the next level. The fixes for the two are different.
Shadow Trader
Posts: 5594
Joined: Wed Sep 09, 2026 5:19 pm

Re: Flags and Pennants, Trading Continuation Patterns

Post by Shadow Trader »

To follow up my own practice task, I went back through 50 EURUSD M5 flags that passed both rules, the half pole retrace and the three times duration limit. Each one was sorted by how it ended.
img_12722_r2.png
img_12722_r2.png (68.52 KiB) Viewed 107 times
The chart shows the split. 22 reached the full pole target, 13 broke the right way but stalled before it, 10 broke the wrong way, and 5 ran out of time and were dropped as ranges.

The stalls were the group with an obvious fix. Nine of the 13 had a previous day high or low or a round number sitting inside the projected move, and price stopped within a few pips of it.

So when a level sits inside the projection, my target now goes just in front of that level instead of the full pole. That trims some of the 22 winners, but it turns most of the stalls into small winners rather than scratches.

The wrong way breaks had no clean pattern I could find. Most were simply the higher timeframe trend not cooperating, which supports the quality check in the opening post.
Shadow Trader
Posts: 5594
Joined: Wed Sep 09, 2026 5:19 pm

Re: Flags and Pennants, Trading Continuation Patterns

Post by Shadow Trader »

Candlestick patterns, such as engulfing bars or pin bars, are popular on all timeframes. On a one-minute chart, a pin bar might consist of a few seconds of unusual quotes.

The reliability of such patterns drops as the time scale shrinks. On short timeframes, noise dominates. A long wick might just be a spread spike. An engulfing candle might reflect a single large order, with no follow-through.

Testing candlestick patterns on tick-level data across thousands of occurrences shows weak predictive power on very short horizons. The patterns that work on daily charts, where each candle summarizes thousands of participants' decisions, don't carry the same meaning when a candle sums up a minute.

There is still a use for them as a description of what happened. A long upper wick means price rose and was rejected. That helps in understanding a market state, even if it doesn't predict.

The question for any pattern is: what does it do to the probability of the next move, after costs? Without a sample showing it, the pattern is a story.

One honest test is to code a pattern definition once, run it across a year of one-minute data, and measure the average price change over the next five bars, minus the spread. If the number is close to zero, the pattern has no value on that timeframe.

Which candle patterns have held up in testing on short timeframes, and which fell apart?
Post Reply