Momentum

systematica.api.models.momentum

Momentum factors capture price trends across multiple time horizons, incorporating volatility adjustments and weighted scoring mechanisms to generate normalized signals between -1 and 1.

Momentum Calculation Components

The momentum model computes scores based on the following key components:

Component	Description	Function
Short-term Momentum	Recent price trends	get_momentum_components_nb()
Medium-term Momentum	Intermediate price trends	get_momentum_components_nb()
Long-term Momentum	Full dataset price trends	get_momentum_components_nb()
Volatility Adjustment	Risk normalization	get_volatility_adjustments_nb()
Weighted Scoring	Combined signal generation	weighted_average_scores_nb()

API Classes for Momentum

The momentum API provides four main classes:

Class	Purpose	Core Function
RollingMomentumFactor	Per-asset rolling	rolling_momentum_nb()
RollingAverageMomentumFactor	Averaged rolling	rolling_momentum_average_nb()
MomentumFactorCV	Per-asset CV	momentum_cv(combined=False)
AverageMomentumFactorCV	Averaged CV	momentum_cv(combined=True)

RollingAverageMomentumFactor

RollingAverageMomentumFactor(
    window: int = 365,
    minp: int = None,
    short_term_mom: int = 30,
    medium_term_mom: int = 50,
    short_term_vol: int = 30,
    short_term_weight: float = 0.5,
    medium_term_weight: float = 0.3,
    long_term_weight: float = 0.2,
    score_scale: float = 2.0,
)

Momentum variation factor for calculating and scoring momentum between two symbols:

• Multiple Periods: The Momentum indicator is calculated through different time horizons, giving more weight to recent data for faster response.
• Volatility Adjustments : It incorporates both momentum and volatility adjustments to generate a normalized score.

Workflow:

1. Trend Components Calculation: For each symbol, calculate the short, medium and long-term momentum using the average price returns over specified periods.

2. Volatility Adjustment: Calculate the short-term and long-term volatility using the standard deviation of returns over the respective periods.

3. Standardization: Normalize the momentum components by dividing them by the corresponding volatility values to account for risk.

4. Scoring Mechanism: Combine the normalized scores from the short, medium and long-term periods, applying different weights to each based on their importance (e.g., more weight for short-term momentum).

5. Signal Calculation: Amplify the combined score by a scaling factor and clip the final signal to a range between -1 and 1 to represent the direction and strength of the momentum.

6. Handling Multiple Symbols: The process is repeated for each pair, with the final signal being the equal-weighted average, offering a broader market view.

Method generated by attrs for class RollingAverageMomentumFactor.

Ancestors

• systematica.models.base.BaseStatArb
• abc.ABC

Descendants

• systematica.api.models.momentum.RollingMomentumFactor

Instance variables

• long_term_weight: float: Long term weight coefficient. Defaults to 0.2.

• medium_term_mom: int: Medium term momentum window. Defaults to 50.

• medium_term_weight: float: Medium term weight coefficient. Defaults to 0.3.

• minp: int: Mininim period. Defaults to None, which means no minimum period is applied.

• score_scale: float: Scale final score. Defaults to 2.0. This factor amplifies the final score, enhancing the signal strength.

• short_term_mom: int: Short term momentum window. Defaults to 30.

• short_term_vol: int: Short term volatility window. Defaults to 30.

• short_term_weight: float: Short term weight coeff. Defaults to 0.5.

• window: int: The size of the rolling window. Defaults to 365.

RollingMomentumFactor

RollingMomentumFactor(
    window: int = 365,
    minp: int = None,
    short_term_mom: int = 30,
    medium_term_mom: int = 50,
    short_term_vol: int = 30,
    short_term_weight: float = 0.5,
    medium_term_weight: float = 0.3,
    long_term_weight: float = 0.2,
    score_scale: float = 2.0,
)

Rolling momentum variation factor for calculating and scoring momentum on mulitple symbols. See RollingAverageMomentumFactor.

Method generated by attrs for class RollingMomentumFactor.

Ancestors

• systematica.api.models.momentum.RollingAverageMomentumFactor
• systematica.models.base.BaseStatArb
• abc.ABC

MomentumFactorCV

MomentumFactorCV(
    training_window: int = 365,
    testing_window: int = 60,
    splitter: str = 'from_custom_rolling',
    custom_splitter: str | None = None,
    custom_splitter_kwargs: dict | None = None,
    short_term_mom: int = 30,
    medium_term_mom: int = 50,
    short_term_vol: int = 30,
    short_term_weight: float = 0.5,
    medium_term_weight: float = 0.3,
    long_term_weight: float = 0.2,
    score_scale: float = 2.0,
)

Momentum cross-validation variation factor for calculating and scoring momentum between two symbols. More information, see RollingMomentumAverageFactor.

Method generated by attrs for class MomentumFactorCV.

Ancestors

• systematica.api.models.momentum.AverageMomentumFactorCV
• systematica.models.base.BaseStatArb
• abc.ABC

AverageMomentumFactorCV

AverageMomentumFactorCV(
    training_window: int = 365,
    testing_window: int = 60,
    splitter: str = 'from_custom_rolling',
    custom_splitter: str | None = None,
    custom_splitter_kwargs: dict | None = None,
    short_term_mom: int = 30,
    medium_term_mom: int = 50,
    short_term_vol: int = 30,
    short_term_weight: float = 0.5,
    medium_term_weight: float = 0.3,
    long_term_weight: float = 0.2,
    score_scale: float = 2.0,
)

Momentum cross-validation variation factor for calculating and scoring momentum between two symbols. More information, see RollingMomentumAverageFactor.

Method generated by attrs for class AverageMomentumFactorCV.

Ancestors

• systematica.models.base.BaseStatArb
• abc.ABC

Descendants

• systematica.api.models.momentum.MomentumFactorCV

Instance variables

• custom_splitter: str | None: Custom splitter to use for data partitioning. Defaults to None. If set, it should be a string that matches a custom splitter function.

• custom_splitter_kwargs: dict | None: Additional keyword arguments for the custom splitter. Defaults to None. If custom_splitter is set, this should contain any necessary parameters for the custom splitter function.

• long_term_weight: float: Long term weight coeff. Defaults to 0.2. This is the weight applied to the long-term momentum in the final score calculation.

• medium_term_mom: int: Medium term momentum window. Defaults to 50. This is the window size used for medium-term momentum calculations.

• medium_term_weight: float: Medium term weight coeff. Defaults to 0.3. This is the weight applied to the medium-term momentum in the final score calculation.

• score_scale: float: Scale final score. Defaults to 2.0. This factor amplifies the final score, enhancing the signal strength.

• short_term_mom: int: Short term momentum window. Defaults to 30. This is the window size used for short-term momentum calculations.

• short_term_vol: int: Short term volatility window. Defaults to 30. This is the window size used for short-term volatility calculations.

• short_term_weight: float: Short term weight coeff. Defaults to 0.5. This is the weight applied to the short-term momentum in the final score calculation.

• splitter: str: Default splitter to be used if custom_splitter is not passed. Choices are from_rolling, from_custom_rolling, from_expanding, from_custom_expanding. Defaults to from_custom_rolling.

• testing_window: int: The size of the testing window. Defaults to 60. This is the period used for testing the model after training on the training window.

• training_window: int: The size of the training window. Defaults to 365.