confluence in trading is one of those terms everybody uses and almost nobody defines.

ask 10 traders what confluence means and you'll get some version of "when multiple things line up"... and then nothing about what those things should be, or how to measure whether they actually work.

that's a problem, because "things lining up" is a feeling.

and you can't backtest a feeling.

in this post I'm going to give you a definition of confluence in trading you can actually test: multiple reports pointing in the same direction at the same time. then I'll walk through a real example on NQ using the initial balance, where stacking 2 checks takes you from a decent bias to a true A+ setup. every number in this post comes straight from edgeful's reports, and I'll show you exactly which report each one comes from.

table of contents

• what is confluence in trading
• why confluence in trading beats a single data point
• the foundation: the initial balance almost always breaks
• check #1: which side of the IB formed first
• check #2: where price ends the first hour
• stacking the checks: what confluence does to the numbers
• how to trade the A+ setup: entries, stops, and targets
• when to sit on your hands
• common mistakes with confluence in trading
• key takeaways

what is confluence in trading

confluence in trading means multiple independent data points agreeing on the same direction at the same time. the more independent checks that agree, and the stronger each one is on its own, the better the setup.

the traditional version of this idea is technical: a trendline, a moving average, and a fibonacci level all sitting near the same price. traders call that confluence, and it's better than nothing. but none of those pieces carries a number. you can't say how often that exact combination has actually played out in your favor, so you're still trading on how good the chart looks.

here's the definition I want you to use instead: an A+ setup is when multiple reports all point in the same direction at the same time. each report has a win rate you can check. one strong number gives you a bias. multiple strong numbers agreeing gives you an A+ setup.

that small change matters, because it turns "this looks good" into "the last 6 months of data say this works X% of the time." you can measure it and decide in advance what's good enough to trade.

why confluence in trading beats a single data point

here's the whole argument in 2 numbers, according to edgeful data on NQ over the last 6 months in the NY session:

• when the IB low formed first, price broke the IB high first 75% of the time (51 out of 68 days)
• when the IB low formed first AND price finished the first hour pressing right up against the IB high, that jumped to 92.3% (36 out of 39 days)

the rest of this post walks through exactly how that example works, step by step, so you can run the same process yourself.

the foundation: the initial balance almost always breaks

quick definition in case you're new to the IB: the initial balance is the range set by the high and low of the first hour of the NY session, 9:30AM to 10:30AM ET. we covered the full report in our initial balance breakout strategy post, but here's what you need for this setup.

once that first hour range is set, there are only 3 things price can do. here's how they've played out on NQ over the last 6 months in the NY session:

• single break: 80% of days (price breaks one side of the IB and never breaks the other)
• double break: 16.9% of days (price breaks both sides)
• no break: 3.1% of days (price stays inside the range all day)

add the first two together and price has broken at least one side of the IB on 96.9% of days (126 out of 130).

so the break itself is close to a given. but that stat alone is NOT an A+ setup, because it doesn't tell you which side breaks. a break happens almost every day whether you're positioned right or wrong.

that's where the 2 checks come in.

check #1: which side of the IB formed first

the IB by rejection subreport tracks which side of the range forms first during the opening hour, the high or the low.

the pattern is simple: whichever side forms first, price tends to break the opposite side after the range is set. if the low of the first hour printed early and price spent the rest of the hour moving up, the sellers had their chance and couldn't hold it.

on NQ over the last 6 months in the NY session, when the IB low formed first, price broke the IB high first 75% of the time (51 out of 68 days).

check #2: where price ends the first hour

this is the check that surprises people, and it's a customization inside the by rejection subreport called the closing zone.

the closing zone answers one thing: when the IB finishes forming at 10:30AM ET, which quadrant of the range is price sitting in?

think about what it means when the low forms first AND price spends the rest of the hour climbing, ending right up against the IB high. buyers took control early and kept control the entire hour.

here's what the data shows on NQ over the last 6 months in the NY session:

• low formed first + price ends the hour in the top quadrant of the range, pressing the IB high: the high breaks first 92.3% of the time (36 out of 39 days)
• high formed first + price ends the hour in the bottom quadrant, pressing the IB low: the low breaks first 86.7% of the time (26 out of 30 days)

notice the short side. on its own, "high formed first" was a weak 58.1% lean. add the closing zone check and it becomes 86.7%. that second check turned a weak lean into a tradeable setup.

stacking the checks: what confluence does to the numbers

here's the full picture on NQ, last 6 months, NY session, so you can see exactly what each layer of confluence in trading adds:

• IB low formed first, no other check
  • high breaks first: 75% (51 of 68 days)
• IB low formed first + hour ends pressing the IB high
  • high breaks first: 92.3% (36 of 39 days)
• IB high formed first, no other check
  • low breaks first: 58.1% (36 of 62 days)
• IB high formed first + hour ends pressing the IB low
  • low breaks first: 86.7% (26 of 30 days)
• either side formed first + hour ends mid-range
  • break toward the expected side: 40% to 57.9%, depending on direction

that last line is just as important as the big numbers. when price ends the first hour in the middle of the range, the edge is gone. those days have broken toward the "expected" side just 57.9% of the time when the low formed first, and 40% when the high formed first. barely better than 50/50, and not worth trading. half the value of a confluence trading process is the days it tells you to skip.

this is also the exact idea behind edgeful's what's in play feature, which shows you where price sits relative to each report's key levels in real time. instead of checking each report by hand, you can see which ones agree on one screen.

how to trade the A+ setup: entries, stops, and targets

by 10:30AM ET you've run both checks above.

say the low formed first and price is pressing the IB high: the data leans long. here's how to structure the trade.

entries first. there are 2 common ways to do it, both using the retracement levels of the IB range:

• tighter: enter when price retraces to the 25% level of the IB range, stop at the 50% level. you won't always get filled, but the risk is small relative to the target.
• wider: enter at the 50% level, stop at the IB low (on a long setup). you still may not always get filled, but you give the trade more room.

your first target is the opposite side of the range: the IB high on a long, the IB low on a short.

after the break, the IB by levels subreport tells you how far price has historically extended past the range, measured in multiples of the IB size. on NQ over the last 6 months in the NY session, looking only at breakout days:

• price touched the 0.1 extension 98.4% of the time (60 of 61 days)
• the 0.2 extension 75.4% of the time
• the 0.5 extension 45.9% of the time

so instead of targeting a random 2:1 because someone said that's what your risk-to-reward should be, you can take the majority of the trade off around the 0.2 extension and leave runners toward the 0.5.

if you want the IB levels plotted on your charts automatically, our initial balance indicator for TradingView draws the range and the extension levels for you.

one thing I need you to sit with before you trade this: A+ setups still fail. Steph Curry makes over 90% of his free throws, and you still wouldn't bet your net worth on a single shot. a 92.3% setup loses roughly 1 morning out of every 13. size your trades so that morning doesn't matter, always with a risk-first mindset.

when to sit on your hands

confluence in trading cuts both ways. the same process that hands you an A+ morning will also tell you, clearly, that today has nothing for you.

skip the trade when:

• the checks disagree. the low formed first but price faded back to the middle of the range by 10:30AM ET. that's the 40% to 57.9% bucket, and there's no edge in it.
• price rotated all the way back to the side that formed first. the sample is small on those days, so I won't quote a number, but the original bias is off. stand down.
• your numbers are different. these stats are NQ, NY session, last 6 months. before you trade this on ES, GC, or anything else, run the reports on YOUR ticker and session. the numbers will be different, and they change as the market environment changes.

that last point is the effort part nobody likes to hear. this setup took me 2 checks to explain, but it only works if you actually pull the data for your market and keep it current. edgeful does the counting for you, but the process is still yours to run.

common mistakes with confluence in trading

mistake 1: stacking data points that measure the same thing

2 moving averages agreeing is not confluence in trading, because they're built from the same prices. the checks in this post work because they answer different questions: one measures which side of the range formed first, the other measures where the hour closed. independence is what makes the win rate jump.

mistake 2: calling it confluence in trading without a number

if you can't say how often the combination has played out historically, you have a nice-looking chart, and that's all you have. every layer you add should come with a win rate attached, like the 75% to 92.3% jump we walked through above.

mistake 3: taking the trade when the checks disagree

the whole value of a confluence trading process is that it filters your mornings. if you override the mid-range days because you "feel like" the break is coming anyway, you've paid for a filter and thrown it away.

mistake 4: treating an A+ setup like a sure thing

92.3% is a great number and it still fails. size every trade like it could be the loser, because eventually one is.

key takeaways

• confluence in trading means multiple independent, data-backed checks agreeing on direction, each with its own win rate you can verify
• on NQ over the last 6 months in the NY session, the IB broke at least one side on 96.9% of days, so the real question is direction, and that's what the confluence checks answer
• check #1: when the NQ IB low formed first, the high broke first 75% of the time (51 of 68 days)
• check #2: add a close pressing the IB high and that jumps to 92.3% (36 of 39 days). the short side goes from a weak 58.1% to 86.7%
• a mid-range close at 10:30AM ET removes the edge (40% to 57.9%), and sitting out those days is part of the setup
• target the break using the extension data: the 0.2 extension was touched on 75.4% of NQ breakout days, the 0.5 on 45.9%
• run the same reports on your own ticker and session before trading this. the numbers change by market and over time, and keeping them current is your job

edgeful provides historical performance data to help traders make informed decisions. this does not constitute financial advice. past performance is not indicative of future results. all trading involves risk — always do your own analysis and manage your risk accordingly.