What is Confluence?
Reviewed by Karthikeyan Anbarasan (Founder, TradeLyser) · Updated 19 Oct 2026 · Editorial policy
Every setup rests on some reason to expect price to react at a place. Confluence is the practice of waiting until more than one reason agrees. The logic is that different groups of traders watch different things: some watch prior-day levels, some watch VWAP, some trade off moving averages. When those groups all have orders at the same price, a reaction there is more plausible than at a level only one group cares about.
Building a confluence checklist
| Factor family | Examples | Count as |
|---|---|---|
| Price level | Prior day high or low, swing high or low, round number, pivot | One factor |
| Trend | Direction of the 20 or 50 EMA, higher highs and higher lows, higher timeframe trend | One factor |
| Volume-weighted price | VWAP, value area, point of control | One factor |
| Trigger | Rejection candle, engulfing bar, break of a small range | One factor |
| Time and context | Session window you trade well, no major event in the next hour | One factor |
The checklist works only if it is written before the trade and each factor comes from a different family. Three momentum oscillators all showing oversold are one opinion repeated, not three. Write the list into your trading plan, define each item so it can be ticked yes or no, and set a minimum score for entry.
Worked example on Nifty futures
Illustrative levels. Nifty futures pull back in the morning towards 24,880. The previous day’s high was 24,875, the rising 20 EMA on the 15-minute chart is at 24,890 and VWAP is at 24,885. A bullish engulfing candle closes on the 5-minute chart at 24,880. That is level, trend, VWAP and trigger: four of five factors, with session time the only one missing because it is just before lunch.
Confluence does not change the stop logic. The stop goes where the shared level is clearly broken, here 40 points below entry, and the size is set from that distance. Note that the EMA and VWAP are both averages of recent prices, so on a quiet morning they can sit close together for the same reason; a strict checklist might count them as half each. Note also what the confluence did not do: it did not make the trade safe. If Nifty futures fall through 24,840, all four reasons failed together, and the loss is the same ₹2,600 per lot that a one-factor trade with the same stop would have cost. What confluence can change is how often the trade works over many attempts, and only your logged results can show whether it does.
Testing whether confluence pays
| Score at entry | Trades | Win rate | Expectancy |
|---|---|---|---|
| 1 to 2 | 46 | 41% | +0.02R |
| 3 | 38 | 47% | +0.18R |
| 4 to 5 | 12 | 58% | +0.35R |
The table is illustrative, 96 trades split by score. The 4-to-5 bucket looks best, but with 12 trades the margin of error on its win rate is close to 30 points either way, as Figure 1 shows. Splitting trades into buckets shrinks every sample. Keep logging the score with each trade, and only raise your minimum score once the higher bucket has at least 30 trades and still leads.
- Score before you click buy. A score written after the result tends to drift towards whatever explains the outcome.
- Keep the checklist short. Five well-defined items are easier to apply the same way every day than twelve loosely defined ones.
- Count missed trades too. If high-score setups you skipped would have paid, the issue is execution, not the checklist.
- Review the factors one at a time every few months. A factor that never changes the result can be dropped to simplify the list.
Reference guide
| Context | Value | Reading |
|---|---|---|
| Factors from different families: level, trend, volume and a trigger | Independent | Each one adds information the others do not |
| RSI, stochastic and CCI all oversold counted as three factors | Double counting | They measure the same thing from the same prices; count them as one |
| Factors added after the trade to explain it | Hindsight | Score the checklist before entry and log the score with the order |
Questions
What is confluence in trading?
Several independent reasons, such as a key level, the trend, VWAP and a trigger candle, pointing to the same trade at the same price and time.
How many confluence factors are enough?
There is no fixed number. Many traders use a five-item checklist and a minimum such as three. Test the minimum on your own logged trades before raising it.
Can too much confluence be a problem?
Yes. Each extra filter cuts the number of trades, and a very strict checklist fitted to past trades may not hold in future ones. Counting similar indicators twice also inflates the score.
How do I track confluence in a trading journal?
Record the checklist score and which factors were present before entering. Review expectancy by score bucket monthly once each bucket has a reasonable number of trades.