Momentum Cycle Sentry

Lagging Indicator

The multi-layer EMA smoothing and momentum lookback introduce lag relative to the actual price action.

Event	Lag	Detail
Raw momentum calculation	= length bars	Close − Close[length] (default length=20)
Fast layer (p1)	≈ smoothing bars	EMA with period = smoothing (default 5)
Slowest layer (p5)	≈ smoothing×5 bars	EMA with period = smoothing × 5 (default 25)
OB/OS corridor shift	≈ ob_lookback/2 bars	Rolling stdev window (default 50)
Retracement detection	= retrace_len bars	Requires consecutive falling/rising bars (default 2)

Formula for custom params: layer lag ≈ smoothing × layer_number, OB/OS lag ≈ ob_lookback / 2

Momentum Cycle Sentry is a multi-layer momentum oscillator ported from the LuxAlgo PineScript indicator. It computes raw momentum, applies five levels of EMA smoothing to create a layered "glow" visualisation, and derives dynamic overbought/oversold corridors from the standard deviation of the fast line.

How it works:

1. Compute raw momentum: Close − Close[length]
2. Apply EMA smoothing at 5 increasing periods (smoothing×1 through smoothing×5) and multiply by magnitude
3. Calculate rolling standard deviation of the fast line over ob_lookback bars
4. Build inner and outer OB/OS corridors using the stdev multiplied by ob_mult_inner / ob_mult_outer
5. Determine trend direction: bullish when fast line > 0, bearish when < 0
6. Detect retracements: when p1 is pulling back (falling in uptrend or rising in downtrend) for retrace_len consecutive bars

def momentum_cycle_sentry(
    data: Union[PdDataFrame, PlDataFrame],
    length: int = 20,
    smoothing: int = 5,
    magnitude: float = 1.0,
    retrace_len: int = 2,
    ob_lookback: int = 50,
    ob_mult_inner: float = 2.0,
    ob_mult_outer: float = 3.0,
    close_column: str = "Close",
) -> Union[PdDataFrame, PlDataFrame]:

Example

from pyindicators import (
    momentum_cycle_sentry,
    momentum_cycle_sentry_signal,
    get_momentum_cycle_sentry_stats,
)

# Calculate Momentum Cycle Sentry
df = momentum_cycle_sentry(df, length=20, smoothing=5)

# Generate trading signals
df = momentum_cycle_sentry_signal(df)

# Get statistics
stats = get_momentum_cycle_sentry_stats(df)
print(f"Bullish bars: {stats['bullish_bars']} ({stats['bullish_pct']}%)")
print(f"Bearish bars: {stats['bearish_bars']} ({stats['bearish_pct']}%)")
print(f"Retracing: {stats['retracing_bars']} ({stats['retracing_pct']}%)")
print(f"Zero crossings: {stats['zero_crossings']}")

Parameters

Parameter	Type	Default	Description
length	int	20	Lookback period for raw momentum
smoothing	int	5	Base EMA period — layers use smoothing×1 through smoothing×5
magnitude	float	1.0	Scalar multiplier for all layers
retrace_len	int	2	Consecutive falling/rising bars required for retracement
ob_lookback	int	50	Rolling window for standard deviation (OB/OS bands)
ob_mult_inner	float	2.0	Inner corridor multiplier
ob_mult_outer	float	3.0	Outer corridor multiplier

Output Columns

Column	Description
mcs_p1	Fast momentum layer (EMA period = smoothing)
mcs_p2	Layer 2 (EMA period = smoothing × 2)
mcs_p3	Layer 3 (EMA period = smoothing × 3)
mcs_p4	Layer 4 (EMA period = smoothing × 4)
mcs_p5	Slowest momentum layer (EMA period = smoothing × 5)
mcs_ob_inner	Overbought inner corridor (stdev × ob_mult_inner)
mcs_ob_outer	Overbought outer corridor (stdev × ob_mult_outer)
mcs_os_inner	Oversold inner corridor (-stdev × ob_mult_inner)
mcs_os_outer	Oversold outer corridor (-stdev × ob_mult_outer)
mcs_trend	Trend direction: 1 bullish, -1 bearish, 0 neutral
mcs_retracing	1 when momentum is pulling back within trend, 0 otherwise

Trend & Retracement

• Trend is determined by the fast line (p1): positive = bullish, negative = bearish
• Retracement is flagged when p1 has been falling for retrace_len bars in an uptrend, or rising for retrace_len bars in a downtrend
• The signal function combines both: +1 (bullish, not retracing), -1 (bearish, not retracing), 0 (neutral or retracing)

Signal Values (from momentum_cycle_sentry_signal)

Signal	Meaning
1	Bullish momentum — fast line > 0 and not retracing
-1	Bearish momentum — fast line < 0 and not retracing
0	Neutral or momentum pulling back (retracing)

Stats (from get_momentum_cycle_sentry_stats)

Key	Description
bullish_bars	Number of bars with trend == 1
bearish_bars	Number of bars with trend == -1
bullish_pct	Percentage of bullish bars
bearish_pct	Percentage of bearish bars
retracing_bars	Number of bars flagged as retracing
retracing_pct	Percentage of retracing bars
overbought_bars	Bars where p1 > inner OB corridor
oversold_bars	Bars where p1 < inner OS corridor
max_momentum	Maximum p1 value
min_momentum	Minimum p1 value
avg_momentum	Mean of absolute p1 values
zero_crossings	Number of times p1 crossed the zero line

Chart