the afternoon IB is an initial balance setup that runs on a custom session from 2:00 to 4:00PM ET instead of the regular 9:30AM New York open. we started testing the afternoon IB because one of the most popular setups in our community, the morning IB by rejection, stopped giving traders a usable bias over the last 3 months.

the market's been awful. tons of news, whippy price action, and a morning session that chops through levels instead of respecting them.

so instead of forcing the same setup in bad conditions, we pulled the data on a different part of the day. and the 2:00 to 4:00PM ET window in the NY session came back with numbers the morning session hasn't given us in months.

this post breaks down the full afternoon IB trading strategy: what it is, why the morning version struggled, how to build the custom session on edgeful, the exact stats on ES, NQ, YM, and RTY, and a real example from Thursday, July 16 on NQ.

what is the afternoon IB

the afternoon IB is the initial balance of a custom session that runs from 2:00 to 4:00PM ET. the IB forms during the first hour of that window, 2:00 to 3:00PM ET, and then you trade the final hour, 3:00 to 4:00PM ET, using the range that first hour gives you.

if you've traded the regular initial balance, everything here will feel familiar. same concept, same reports, same rules. the only thing that changes is the window you run it on.

and that one change matters, because the two sessions have been behaving completely differently. the morning session has been choppy and news-driven. the afternoon session has been giving traders a clean, two-sided bias.

quick background: the initial balance and IB by rejection

the "initial balance" is the range price builds during the first hour of a trading session. on the standard New York session, that's 9:30 to 10:30AM ET. the high and low of that first hour become the IB levels, and a lot of intraday setups are built around whether price breaks those levels, holds them, or trades back inside the range.

if the concept is new to you, we cover the classic version in our full initial balance breakout strategy guide.

the subreport this whole post revolves around is IB by rejection. here's what it tracks: during the IB window, which side of the range forms first, the high or the low?

that matters because of what typically happens next. once the range is set, price tends to move in the opposite direction of whichever side formed first, based on the data.

the low forms first, your bias is long and you're watching for a break of the IB high. the high forms first, your bias is short and you're watching for a break of the IB low.

that's the entire report: one question that hands you a bias. it's one of the most popular setups in our community, and a lot of traders build their morning session around it.

which is exactly why the last few months have been so frustrating.

why the morning IB stopped working

according to edgeful data, the morning IB by rejection has been one-sided for a while, and lately the weaker side fell apart completely.

start with the bigger sample. over the last 6 months on the regular 9:30AM to 4:00PM New York session:

• NQ, low formed first: price broke the high first 76.81% of the time
• NQ, high formed first: price broke the low first just 59.32% of the time
• ES, low formed first: price broke the high first 67.57% of the time
• ES, high formed first: price broke the low first just 59.26% of the time

so even over 6 months, the long side of the bias was strong and the short side was mediocre. usable, but lopsided.

now narrow it to the last 3 months, and the short side disappears entirely.

on ES over the last 3 months, when the high formed first in the morning, price broke its own high 50% of the time and broke the low first just 41.67% of the time. that bucket is only 24 sessions, so I wouldn't lean on the exact split, but the direction is obvious: there's no edge there. NQ wasn't much better, with price breaking the low first only 48.39% of the time after the high formed first.

basically 50/50. when the data looks like that, the honest move is to stop trading the setup.

but the IB concept itself didn't break. the session did. so we ran the same reports on a different window.

how to build the 2:00 to 4:00PM ET custom session

everything below comes from a custom session you can build on edgeful in under a minute:

1. create a custom session on the ticker you trade with a 2:00PM ET start and a 4:00PM ET end
2. set the IB period to 60 minutes, so the initial balance forms from 2:00 to 3:00PM ET
3. run the standard IB report and the by rejection subreport on that session

that's the whole setup. the reports do the same thing they do on the morning session... they're just measuring a different two-hour window.

quick note: the comparison dashboards in this post took a few minutes to build using the edgeful API. you don't need any of that to trade this. the standard report view on the platform shows you everything below.

and if session mechanics are new to you, our ETH vs RTH guide breaks down how futures sessions are defined in the first place.

the afternoon IB stats on ES and NQ

two things stood out when we ran the afternoon IB over the last 3 months (62 sessions on each ticker).

first: double breaks dropped

a "double break" is when price breaks both sides of the IB range in the same session. it's the outcome that punishes directional traders the most, because whichever side you leaned, at some point you were wrong.

over the last 3 months on the 2:00 to 4:00PM ET session:

• ES double breaks
  • afternoon session: 14.5% of sessions
  • morning session: 25%
• NQ double breaks
  • afternoon session: 12.9%
  • morning session: 17.2%

I'll be straight with you on this one, though. part of that drop is mechanical. the afternoon IB only leaves price about an hour of runway after the range is set, while the morning IB leaves about five and a half hours. less time means fewer chances to break both sides.

so the lower double-break rate is good to see, but it's not the actionable takeaway on its own.

second: the IB by rejection bias is clean in both directions

this is the real finding. on the afternoon session over the last 3 months, the by rejection numbers work on both sides:

• ES afternoon IB by rejection
  • high formed first: price broke the low first 65.7% of the time
  • low formed first: price broke the high first 74.1% of the time
• NQ afternoon IB by rejection
  • high formed first: price broke the low first 64.9% of the time
  • low formed first: price broke the high first 76% of the time

price moves opposite the side that formed first about 2 out of 3 times, and closer to 3 out of 4 when the low forms first. it holds on both ES and NQ.

compare that to the morning session, where the same short-side number has been sitting near 50/50 on ES. the afternoon session has been giving you something the morning simply hasn't: a bias you can act on in either direction.

does the afternoon IB hold up on other tickers?

we ran the same custom session on the other two major index futures over the same 3-month window (63 sessions each):

• YM afternoon IB by rejection
  • high formed first: price broke the low first 69.7% of the time
  • low formed first: price broke the high first 66.7% of the time
• RTY afternoon IB by rejection
  • high formed first: price broke the low first 77.1% of the time
  • low formed first: price broke the high first 60.7% of the time

so the opposite-direction bias holds on all 4 index futures. all 8 directional numbers lean the same way, and 6 of the 8 sit at 65% or better.

the two softer cells are worth naming. NQ's short side at 64.9% is essentially at that threshold, so I'd still treat it as tradeable context. the RTY long side at 60.7% is the genuinely soft one: it leans the right way, but I'd treat it as context rather than an edge worth trading on its own. if RTY is your ticker, that's worth knowing before you size up on the long side.

and that's the broader point: run the numbers on the specific ticker you trade before you touch this setup. this is a pattern that currently shows up in the data on these 4 tickers, nothing more. your instrument might look different.

a real example: NQ on Thursday, July 16

here's what the afternoon IB looked like in practice on Thursday, July 16 on NQ.

it's just past 3:00PM ET. the 2:00 to 3:00PM ET range on NQ is set, and the high formed first.

that gives you a short bias. for the last hour of the session, you're watching for price to break the IB low and keep moving down. and that's exactly what happened: the low broke first.

if the low had formed first instead, you'd flip it. long bias, watching for the break of the IB high.

you're answering one question after 3:00PM ET: which side of the range formed first. then you target the opposite.

how to trade the afternoon IB step by step

here's the full process:

1. check the data first. pull up the by rejection numbers on the ticker you trade and confirm the bias actually exists there right now. don't take my word for it.
2. on edgeful, build a custom session from 2:00 to 4:00PM ET. the initial balance forms over the first hour, 2:00 to 3:00PM ET.
3. once the range is set at 3:00PM ET, check which side formed first: the high or the low.
4. let that set your bias. the low formed first, you're leaning long and looking for a break of the IB high. the high formed first, you're leaning short and looking for a break of the IB low.
5. if price never breaks either side of the IB, you sit on your hands. no setup, no trade.

if you want the IB levels drawn on your charts automatically, our initial balance indicator for TradingView plots them for any session you define, including this one.

and if you'd rather see how far the IB concept can go when you automate it, the GC initial balance algo is built on the same underlying report.

honest caveats before you trade the afternoon IB

a few things I want on the table before you run this in a live session:

• the double-break improvement is partly mechanical. the afternoon window is shorter, so price has less time to break both sides. the by rejection bias is the robust finding here, because it's about which side breaks first, not how much total movement the session allows.
• some of the morning buckets are thin. the ES "high formed first" sample over the last 3 months is only 24 sessions. the conclusion (no edge in the morning short side right now) is solid, but don't quote the exact split as gospel.
• the RTY long side is soft at 60.7%, so don't build a trade around that cell on its own.
• these numbers are from a specific 3-month window and they will change. market conditions rotate, and the morning IB may well come back. re-run the reports before you trade this, not once a quarter.
• the afternoon session overlaps with the most volatile stretch of the trading day. new highs and lows into the close have their own patterns, which we broke down in our power hour breakout trading strategy post. know how your ticker behaves into 4:00PM ET before you hold through it.

none of this is plug-and-play, and I'm not going to pretend it is. you have to build the custom session, verify the numbers on your ticker, and put in the screen time to learn how the last two hours actually trade. the data does the heavy lifting, but only after you do the setup work.

key takeaways

• the afternoon IB runs on a custom session from 2:00 to 4:00PM ET, with the initial balance forming from 2:00 to 3:00PM ET
• the morning IB by rejection has been near 50/50 on the short side over the last 3 months on ES, with price breaking the low first just 41.67% of the time after the high formed first
• the afternoon IB by rejection has been clean both ways: ES 65.7% short side and 74.1% long side, NQ 64.9% and 76%, over the last 3 months
• the bias holds across all 4 index futures, with YM at 69.7% / 66.7% and RTY at 77.1% / 60.7% over the same window
• the rule is simple: low forms first, lean long toward the IB high. high forms first, lean short toward the IB low. no break, no trade.
• the lower afternoon double-break rate (ES 14.5% vs 25% in the morning) is partly explained by the shorter window, so treat the by rejection bias as the real edge
• verify the afternoon IB numbers on your own ticker before trading it, because these stats come from a specific 3-month window and will change