Trend & directionJurik Moving Average · JMA
Mark Jurik's adaptive filter, prized for being smooth and low-lag at the same time.
Works best in trending marketsEngine-computed on a fixed sample series
What it is
The Jurik Moving Average, or JMA, is an adaptive smoothing filter created by Mark Jurik that tries to achieve the holy grail of moving averages: a line that is both very smooth and very fast. Every ordinary moving average forces a trade-off, because reducing the jitter of the line means adding lag, and cutting lag means letting more noise through. JMA was engineered specifically to soften that trade-off, producing a curve that stays close to price and turns quickly at real reversals while still filtering out the small random wiggles that trigger false signals. For a beginner, think of it as a premium trend line that gives you an EMA's responsiveness with far less of the choppiness that makes a fast EMA hard to trust. It is widely admired precisely because it feels like it cheats the usual lag-versus-smoothness rule, even though no filter escapes that rule entirely.
How it is calculated
JMA is a proprietary algorithm, so its exact internal formula is not fully published, but its behavior and its controls are well understood. It combines an adaptive filter that responds to volatility with a phase-adjusted smoothing stage, so that when the market moves decisively the filter speeds up and when it drifts the filter calms down. Three parameters shape the output: a length that sets the base period like any moving average, a phase control that shifts the balance between overshoot and lag, and a power or speed setting that dials smoothness against responsiveness. Higher phase values let the line lead more aggressively at the risk of overshoot, while lower values keep it conservative. Because implementations across platforms differ, the same length can produce slightly different curves depending on the coder's approximation of Jurik's method.
Reading it, step by step
Read JMA the way you read any trend line, but with more confidence in its small movements because it filters noise so well. When the line slopes upward and price holds above it, the trend is up; when it slopes down with price beneath, the trend is down. Because JMA lags so little, a change in its slope is an earlier and more trustworthy signal than the equivalent turn in a standard EMA, which typically arrives several bars later. A flattening JMA warns that momentum is stalling before price has fully reversed. Crossovers of price through the JMA, or of a fast JMA through a slow one, generate cleaner signals than the same crossovers on ordinary averages precisely because the JMA does not chatter around price during consolidations.
Best timeframes and settings
Because JMA's whole purpose is to let you run a short, responsive average without the noise penalty, it is popular among intraday and swing traders who want early signals. A common approach is to set a length in the range of seven to fourteen for active trading, where a comparable EMA would be too noisy to use. The phase parameter defaults near zero, and most traders leave it there unless they specifically want more lead, in which case they raise it cautiously. It performs across all timeframes, from five-minute scalping charts to daily swing charts, because its adaptive core adjusts to whatever volatility it is fed. The core trade-off still exists: pushing the length shorter and the phase higher makes JMA faster but eventually reintroduces the overshoot and whipsaws it was designed to suppress.
When and where to use it
JMA is a trend-following backbone and is at its best in markets that trend, where its early slope changes let you enter sooner than a lagging average would allow. It suits liquid instruments across stocks, futures, forex, and crypto, since its adaptive filter handles varying volatility gracefully. In a genuine range it will still outperform a noisy EMA, but no moving average thrives when price has no direction, so a flat JMA is a signal to stand aside. Traders reach for it specifically when they need speed and cannot tolerate the false turns that a short EMA would generate. Avoid over-relying on its precision, since its proprietary nature means you cannot fully audit why it turned when it did.
Strategies that use it
The most direct strategy trades JMA's slope: go long when the line turns up and price is above it, exit or reverse when the slope rolls over, using the low lag to capture more of each move than a slower average would. A crossover approach pairs a fast JMA with a slow JMA and takes the direction of their cross, benefiting from the smoothness that keeps the two lines from tangling in noise. A price-cross method enters when price closes decisively through the JMA and treats the average itself as a trailing stop, exiting on a close back across it. Because the JMA hugs price closely, some traders also use it as a dynamic support and resistance rail, buying pullbacks that hold the rising line in an uptrend.
Combining it with other indicators
Since JMA supplies a clean trend read, it pairs naturally with a momentum oscillator such as the RSI, the MACD, or the stochastic to confirm that a slope change has force behind it. An ADX reading helps confirm that the market is trending strongly enough to trust JMA's crossovers rather than fade them. Volatility tools like the ATR or Keltner Channels give context for stop placement around the JMA line. Some traders combine a fast JMA for entries with a much slower JMA or a two-hundred period average as a higher-timeframe bias filter, only taking JMA signals aligned with the larger trend. The aim is to add confirmation of momentum and regime, because JMA already handles the smoothing and timing better than most tools.
Where it fails
JMA's proprietary nature is its biggest practical drawback, because different platforms implement it differently and results are not perfectly reproducible, which complicates backtesting and comparison. More fundamentally, no filter truly escapes the lag-versus-smoothness trade-off, so when traders push the length very short and the phase very high in search of maximum speed, JMA overshoots and whipsaws just as any fast average would. In a dead, rangebound market it will still generate losing crossovers because there is no trend to follow. Traders sometimes develop a false sense of security from its smoothness and size positions too aggressively on its signals. The safeguards are to keep the settings within reason, confirm signals with an independent momentum or regime filter, and remember that a beautiful line is not the same as a correct forecast.
A worked example
Suppose you trade a currency pair on the one-hour chart and set a JMA with a length of ten and a phase of zero. Price has been chopping sideways around one-point-one-zero-five-zero, and a standard ten-period EMA has whipsawed you three times as it crossed back and forth through price. The JMA, by contrast, stayed nearly flat through the chop without issuing false crosses. When a genuine breakout begins, price pushes to one-point-one-zero-eight-zero and the JMA slope tips clearly upward one bar after the move starts, well before the EMA confirms. You enter long at one-point-one-zero-eight-five, place a stop at one-point-one-zero-five-zero beneath the consolidation, and ride the trend to one-point-one-one-five-zero, trailing your stop along the rising JMA and exiting only when the line finally rolls over.