Confluence is several independent reasons pointing to the same trade at the same place. Across timeframes it means the higher chart's structure, the trading chart's level and the entry chart's trigger all agree — timeframe alignment. Its value comes from independence: a daily level and an hourly trigger are two pieces of evidence, but a moving average and a MACD on the same chart are one, because both are computed from the same prices. Count reasons, not indicators, and set the minimum before the chart is opened.
Confluence is several independent reasons pointing to the same trade at the same place. The word matters less than the adjective: reasons that come from the same source are one reason, however many names they have.
Part 9 of Foundations introduced confluence as a way of choosing between retracement levels. This part makes it the way trades are ranked — across timeframes, across kinds of evidence — and puts a rule on it so that it cannot be gamed.
What counts as a reason
The test is independence. Two reasons are independent when one could be true while the other is false. A daily level and a 1-hour trigger pass: the level can be there and the trigger never come. A rising 50-period average and a MACD above zero do not: both are computed from the same recent closes, so if one is true the other nearly always is.
This is the single most common way confluence is miscounted, and Foundations warned about it from the other side — stochastic and KDJ ended with the advice to stop adding indicators. The moving averages, RSI and MACD parts each describe a useful tool. Stacked, they describe one tool three times.
Independent sources look like this:
- Structure on the higher timeframe — the direction
- A level with a reason: a swing point, a flip, a range edge (Part 2)
- A timeframe below, showing its own structure change — the trigger
- The session, when a move happens in hours that carry participation (Part 8)
- A related instrument confirming or failing to (Part 11)
- One indicator, if the plan uses one — momentum, or volatility — counted once
A trade can have a reason from each. It cannot have three from the third bullet and call it five.
Alignment across timeframes
The strongest form of confluence is the one Part 1 built: three charts, three different questions, three answers that agree. That is timeframe alignment, and it is strong precisely because the three sources cannot be the same thing — a daily structure read and an hourly trigger are not computed from each other.
When the three do not agree, the disagreement has a name and a procedure of its own — Part 4. For now the point is narrower: alignment is what a full count looks like, and anything less is a smaller number.
Confluence at a price
Reasons can also stack at a single level rather than across charts. A daily flipped level, a prior 4-hour swing low, a 61.8% retracement and a round number all landing within a few pips of each other is a place worth waiting at. The same four reasons spread across a hundred pips are four places, and none of them is special.
The zone that results is drawn as in Part 2 — as wide as the reasons that make it, no wider. Stacking does not make a level exact. It makes it populated: several groups of participants have a reason to act there, and that is what a level is.
What confluence changes — and what it does not
More reasons make a trade better. They do not make it different. This is the part most often got wrong in practice, because a high count feels like permission.
The stop stays where the idea is wrong. Stop placement put it beyond the swing that invalidates the read, and five reasons for the read do not move that swing. A trader who widens the stop “because the setup is so good” has converted confluence into a larger loss.
Size is where confluence is allowed to act — and cautiously. A common structure: below the plan’s minimum, no trade; at the minimum, standard risk; well above it, standard risk and a note in the journal. Size up on high-confluence setups only after the journal shows they actually did better. Until then, the extra reasons buy patience, not lots.
Key takeaway Confluence answers “is this worth taking, and with how much of the budget?” It never answers “where is the stop?” The stop was answered by structure, and structure does not count reasons.
Counted before, not found after
There is a version of confluence that is confirmation bias with a checklist. It goes: want the trade, then look for reasons, then count the reasons found. With three charts and a handful of tools, three reasons can always be found — that was the bottom-up problem of Part 1, wearing a number.
Two rules close the gap.
The minimum is written in the plan. “Three independent reasons, from the list” — decided before any chart is open, as part of writing a trading plan. The count is then a comparison against a fixed number, not a judgement made while wanting something.
The list is closed. The reasons that count are the ones the plan names. A reason invented on the day — “and it is Tuesday, which has been strong lately” — does not count, however plausible, because a list that can grow on demand is not a list.
Confluence and confirmation
One more distinction, because the two words are used interchangeably and mean different things.
Confluence is counted before the setup completes: the reasons that make a place worth watching. Confirmation is waited for after: the reaction at that place — the rejection wick, the higher low, the break of the last minor high from Part 2 — that shows the level is doing what the reasons said it would.
Confluence without confirmation is entering because a place should hold. Confirmation without confluence is entering because something did happen, somewhere that had no reason to matter. The setup needs both, in that order, and the entry-trigger part of this level (Part 16) is about pricing the second one — how much of the move waiting for confirmation costs, and what it buys.
Next: When timeframes disagree — the lower chart is falling while the higher one rises. Pullback, or the start of a reversal, and how to tell.
- Confluence is several independent reasons for the same trade at the same place.
- Timeframe alignment is confluence across charts: direction, level and trigger agreeing.
- Two indicators derived from the same price series are one reason, not two.
- The minimum number of reasons is set in the plan, before the chart is opened.
- More confluence can justify a larger share of the risk budget; it never justifies a wider stop.