forex-scalping.com

After a data spike, the chart looks tradable before the market is.

I wait for spread snap-back on EURUSD: live spread back inside my normal filter for three consecutive seconds, not a single flicker. That is the first gate. Second gate is a level I already marked. Third is time - I want the initial expansion printed so I am not first-cross noise.

This is not order-flow romance. It is a cost filter wearing a tape-read name. On the last two CPI prints, price left a clean pullback on the M1 while spread was still double my limit. Skipping those kept me out of fills that would have needed nearly the whole target to break even.

When snap-back never comes within my window, I cancel the idea for that release. The go signal is not the candle. It is the candle plus a market that is willing to show me a normal spread again. I log the wait time. If I keep overriding it, the filter is theatre and I cut permission for post-news entries for a week.

LondonNewsTrader, the three consecutive seconds rule is what separates this from a plain spread filter. With EURUSD on my feed, the first drop back under the limit after a big print is often a single tick, and it widens again straight after as the next wave of orders hits.

I logged the wait times after the last eight US releases I traded. Snap-back came between about 25 and 90 seconds on most, but on two of them it never arrived inside my four minute window, and both times the pullback that looked clean failed.
This illustration is a typical CPI print, with the early flicker at around 15 seconds that I ignore and the real snap-back starting near 50 seconds. My limit there is 1.2 pips, and the spread sat above 3 for most of the first half minute.

One edge case I watch is a spread that looks normal while the quote has stopped updating. If price has not ticked for two or three seconds, I treat the normal spread as stale and restart the count.

Do you scale the window by release? I give NFP longer than CPI, since the second leg on payrolls usually comes later.