Momentum Scalping: Entering With the Move, Not After It

[FXS Prop Desk]

Momentum scalping tries to join a move early enough that the remaining part still pays for the spread and the stop. The hard part is telling a real move from a spike that will retrace.

Signs that a move has substance:

• Several consecutive M1 closes near the candle extremes.
• Tick volume rising, not just one large bar.
• A break of a level that others watch: session high, previous day high, round number.
• Correlated pairs moving the same way, e.g. EURUSD and GBPUSD both breaking higher.

Entry options that avoid buying the top of a single spike:

• The first small pullback that holds above the broken level.
• A break of the first tight consolidation after the impulse.

Exits are usually quick: a fixed target, or as soon as momentum candles shrink and overlap.

Momentum setups are also where slippage is largest, so testing them with realistic fills matters more than for slower setups.

How do you decide in real time whether a move is worth joining?

[Shadow Trader]

My filter for joining is distance from the level, not candle count. If price is already more than one M5 ATR beyond the break, I let it go, because what remains has to pay for a stop back under the level.

A typical London open case looks like this. EURUSD clears the Asian high at 1.1712 with four straight M1 closes in the top quarter of their bars, then pulls back to 1.1715 and holds three pips above the broken level.

img_13753_r1.png (63.74 KiB) Viewed 4 times

The chart marks the buy stop at 1.1721 above the pullback high, with the stop loss at 1.1709 below the old high, so 12 pips of risk. The exit came at 1.1735 once the M1 bodies started overlapping, which is the fading momentum the post describes.

GBPUSD is my confirmation, but I only need it to hold its own Asian high, not break it in the same minute. Requiring both to break together mostly made me late.

On tick volume I am a bit sceptical. A retail feed counts quote updates, so it rises with volatility in general and tells me little that bar size does not.

How do you handle a pullback that dips back below the level by a pip? I skip those, though some run anyway.