Strategy
What Is Confluence in Trading? (And Why One Indicator Isn't Enough)
Why experienced traders almost never act on a single indicator, and how combining signals from different tools actually improves your odds.
By CruxleeUpdated 4 min read
Ask five traders which indicator is "the best," and you'll get five different answers — and all five are probably half-wrong. The real answer most experienced traders land on eventually isn't a single indicator at all. It's confluence.
What confluence actually means
Confluence just means: multiple, independent signals pointing in the same direction at the same time. Not two variations of the same thing — a candlestick pattern and an EMA level is confluence; two slightly different moving average settings basically aren't, because they're measuring almost the same thing twice.
A single RSI reading of 72 tells you the market might be overbought. That same reading, at a level that was resistance three times before, with price forming a shooting-star candle — that's confluence. Each piece alone is weak. Stacked together, they're a genuinely stronger case.
Why no single indicator holds up alone
This isn't a hot take — it's close to a universal rule among the indicators traders actually use:
- MACD produces frequent false signals in flat markets and is explicitly meant to be used as an early-warning tool, confirmed by trend indicators.
- RSI and Stochastic overbought/oversold readings are framed as "potential turning points," not automatic triggers — price can stay overbought far longer than seems reasonable (more on that in RSI settings for swing trading).
- Parabolic SAR is explicitly unreliable in ranging markets and needs a trend filter to mean anything.
- Fibonacci retracements are widely watched, but even their strongest advocates will say they shouldn't be traded in isolation — the real edge comes from a fib level lining up with existing support/resistance or a candlestick pattern.
Notice the pattern: every single one of these tools comes with the same caveat. That's not a coincidence — it's the core limitation of technical analysis itself. Each indicator measures one narrow slice of market behavior (momentum, volatility, trend, volume), and markets are influenced by all of them at once. Picking one and ignoring the rest means trading with a fraction of the available information.
What confluence looks like in practice
A high-confluence setup might look like:
- Price breaking above a horizontal resistance level that's held three times before
- A bullish EMA 9/20 crossover, with both EMAs above EMA 50
- MACD crossing above its signal line, above the zero line
- Rising volume (or OBV) confirming real participation, not a thin move (see volume confirmation)
No single piece of that is remarkable on its own. Together, they describe a market where trend, momentum, structure, and participation all agree — which is about as strong a case as technical analysis can make.
A confluence checklist on real charts
Cruxlee's signals are built as exactly this kind of checklist: separate checks, each measuring something different — the big trend (EMA 200), Bitcoin's direction, trend strength (ADX), a pullback entry, momentum (RSI) and participation (volume). A signal needs all of them on the same closed 4h candle. Here's Ethereum on that checklist, updated every 4 hours:
Does confluence actually help? What our backtests say
We tested this on our own rules instead of taking it on faith. The 4h pullback checklist has four checks that can be switched off without changing what the strategy is: Bitcoin's direction, trend strength (ADX), volume and RSI. We ran every on/off combination of those four, 16 versions, on 20 liquid coins over three years (2023-10 to 2026-10, after 0.1% costs per trade):
- All four checks on was the best version, at about +0.09R per trade. Every version with one or more checks switched off landed between -0.01R and +0.04R.
- Bitcoin's direction mattered most. In an earlier run, removing only the BTC filter turned the result from +0.05R to -0.01R per trade. More on why in do altcoins follow Bitcoin?
- Volume helped too. Requiring an entry candle with at least 1.25x its usual volume improved results on the coins we tuned on and on 10 coins we hadn't.
- RSI added almost nothing. Switching it off barely changed the results.
So confluence did help, but by less than the word suggests, and not every check pulls its weight. It also didn't turn the strategy into a sure thing: on coins the rules had never seen, the full checklist was about break-even (-0.01R per trade over 317 trades). Confluence filters out bad trades; it can't create an edge the entry itself doesn't have.
The catch: confluence takes time to check manually
The honest downside of trading this way is that scoring five or six signals across every coin on your watchlist, every time you check the market, is tedious. Most traders either skip it (and trade off a single indicator anyway) or burn an hour a day manually cross-checking charts.
This is exactly the gap Cruxlee's signal engine is built to close — on every candle close it checks the coins you pick for trend, BTC's direction, trend strength, a pullback entry, momentum and volume, and only alerts you when all of them agree, with the stop-loss and target already worked out. See how it works.