summary: the ICT opening retracement report shows how often price retraces back to the midnight open during your session. this is the trade-it layer: the number to read, the one gotcha that trips everyone up, a worked example, setup, and reconciliation. for the full measurement detail, see the ICT opening retracement report: what it measures and how to read it.
what it measures
the reference level is the midnight open, the open of the 12:00 AM ET candle. the report answers: does price come back to that level during your session?
there are two setups, decided by where the session opens relative to the midnight open. "opened above" asks how often price retraces down to it. "opened below" asks how often it retraces up to it. only retracements inside the session window count (9:30 AM to 4:00 PM ET for NY, 3:00 to 8:00 AM ET for London).
the number that matters
start with standard for your ticker and setup, then always check by-weekday, because the standard number averages everything together and the real edge usually lives in specific days.
aim for 60% or higher, ideally 70%+. if standard is borderline (some tickers sit near a coin flip), drop into by-weekday and by-size to find the higher-probability slice.
what to do with it
the midnight open is your target. if the data shows a strong retracement rate for today's setup, plan to trade toward the midnight open: on an "opened above" day, look for a move down to it. on an "opened below" day, look for a move up to it. exit at the level.
for confluence, line it up with the gap fill report. gap fill targets the previous session close, ICT targets the midnight open. on days those two levels sit at the same price, you have two independent edges pointing at the same target, and a stronger reason to take the trade.
size your stop and time the trade
once you've got a setup, two more variants help you manage it:
by spike — on the days that retraced, how far price first ran away from the midnight open before reversing back. use the average spike as a minimum stop distance: if your stop is tighter than the typical spike, you'll get shaken out of trades that historically still retraced. and if price blows well past the average spike, the retracement is less likely — a cue to tighten up or step aside. you can also wait for price to spike toward that average before entering, for a tighter stop and a better level.
by fill time — of the days that retraced, how many filled before vs. after a time you pick. it tells you when retracements usually land in the session, so you know how long to give the trade instead of bailing early (or waiting on a fill that rarely comes that late).
the size-bucket gotcha
the by-size buckets are the single most misread part of this report. they are NOT the size of the retracement. they're the size of the opening displacement, the percentage move from the midnight open to the session open.
to find your day: calculate the percentage move from the midnight open to the session open, then read the bucket that contains that number. if NQ's midnight open was 21,000 and the NY session opened at 21,063, that's a +0.30% displacement, so you read the 0.20 to 0.39% row. reading a different row means you're looking at a different cohort of days.
a worked example
say it's an "opened below" day on NQ, and by-weekday shows today's weekday runs hotter than the standard rate.
the plan: look for a long back up to the midnight open, exit at the level, invalidate if price extends further away instead of retracing. check by-spike so your stop sits beyond the typical spike, and if the gap fill target sits at the same price, that's added confluence for the same trade.
the edge is in the data. the result still takes customizing it to your ticker and managing the trade.
set it up right (session and instrument)
use the NY or London session, the two that match the ICT framework. switching sessions changes both the open price that decides the setup and the window where retracements count, so a day can be "opened above" on NY and "opened below" on London. treat each session as its own analysis, don't compare across them.
this report is built for 24-hour markets (futures, forex, crypto). it doesn't really apply to stocks, which only trade 9:30 to 4:00 ET and have no meaningful midnight open.
why your numbers don't match
a few specific ones for this report:
the by-size buckets are opening displacement, not retracement size (see the gotcha above).
price hitting the midnight open in pre-market doesn't count, only inside the session window.
an "invalid configuration" error means your reference time is set after the session open. set it before the session start.
the numbers roll as new data comes in, so trust the live report over any example in an article or video.
for timezone and session matching across all reports, see why your edgeful data doesn't match TradingView.
the limits
check the sample size in the weekday or size bucket you're trading, and favor reads that hold across lookback windows. the report and the ICT opening retracement TradingView indicator are two different things: the report tells you whether there's an edge today, the indicator just plots the level on your chart.