Indicators
EMA Crossover Strategy for Crypto: The 9/20/50 Setup Explained
How the 9/20/50 EMA crossover works, why the 50-period line matters more than most traders think, and where this strategy tends to fail.
By CruxleeUpdated 3 min read
If you've spent any time looking at crypto charts, you've seen them: two or three smooth, wavy lines tracking price, crossing over each other every so often. Those are exponential moving averages (EMAs), and the crossover between them is one of the oldest, simplest trend-following signals in trading. Here's how the specific 9/20/50 setup works, and — just as important — where it tends to go wrong.
What an EMA actually is
An EMA is a rolling average of recent closing prices, weighted so that more recent candles count more than older ones. That weighting makes it react a bit faster to new price action than a plain simple moving average (SMA). In practice, the difference is often smaller than people assume — but EMA is the convention most traders default to, so it's worth using if only to speak the same language as everyone else looking at the same chart.
The three lines: 9, 20, 50
A common setup uses three EMAs together, each doing a different job:
- EMA 9 — the fast line, hugs price closely, reacts quickly
- EMA 20 — the medium line, smooths out some noise
- EMA 50 — the slow line, used less as a signal and more as a filter
The crossover signal
The core signal is simple:
- EMA 9 crosses above EMA 20, both sloping upward → potential uptrend starting
- EMA 9 crosses below EMA 20, both sloping downward → potential downtrend starting
That's it, at its simplest. But traded alone, this setup gets chopped up badly in sideways markets — you'll see the lines cross back and forth repeatedly with no real follow-through, racking up small losses each time.
Why EMA 50 matters more than people think
This is where the setup gets more useful: EMA 50 isn't there to generate its own crossover signal — it's there to filter which crossovers are worth acting on.
- If price, EMA 9, and EMA 20 are all above EMA 50, a bullish crossover is considered a stronger, more reliable signal — the longer-term trend agrees with the short-term signal.
- If the same bullish crossover happens while price is still below EMA 50, the move is fighting the dynamic resistance that EMA 50 represents — worth waiting for a confirmed break above it before trusting the signal.
The mirror is true for downtrends: a bearish crossover below EMA 50 is a stronger sell signal than one happening above it.
Where this strategy fails
Two situations consistently trip up traders using this setup:
- Flat, ranging markets. When EMA 9 and EMA 20 are both roughly flat and tangled together, crossovers fire constantly and mean almost nothing — there's no real momentum behind them. A trend-strength filter such as ADX is the usual fix: no crossover trades while it says the market is chopping.
- Treating it as a standalone signal. EMA crossovers work best as one input among several — paired with support/resistance levels, volume, or a confirming oscillator — not as the only reason to enter a trade. That stacking of independent signals is called confluence.
A variation: trade the pullback, not the cross
Crossovers are late by design: by the time EMA 9 crosses EMA 20, a good part of the move has often already happened, and your stop has to sit far below. Many trend traders flip the idea around. They let the slower EMAs define the trend, then wait for price to dip back to a faster line and enter when it turns up again. That's the EMA pullback strategy, and it's what Cruxlee's signals use, with EMA 200 as the big-trend filter and EMA 50 for its direction. The two entries side by side, with our backtest numbers, are in EMA crossover vs EMA pullback.
Here's that trend check on Bitcoin, updated every 4 hours:
The takeaway
The 9/20/50 EMA crossover is a genuinely useful trend tool, but its real value is in the filter, not just the crossover itself. Before trusting a signal, check: does the broader trend (EMA 50, and ideally the higher timeframe) actually agree with what the fast lines are telling you? If yes, you've got a real signal. If not, you're probably looking at noise.
Cruxlee's signal engine is built on the same filter idea — a slow EMA (200) for the big trend, EMA 50 for its direction, and pullbacks to EMA 20 for entries — checked automatically on every candle close for the coins you pick. If you'd rather have that done for you than eyeball it on every chart, see how Cruxlee works.