bayesline.api.equity.MaximizeAlphaObjective#

model MaximizeAlphaObjective#

Maximize expected return net of a risk penalty (mean-variance).

The penalty is risk_aversion times the holdings’ total variance from the risk model (factor and idiosyncratic), so the objective trades expected return against risk in annualized return units.

Fields#
Field	Type	Required	Default	Constraints
`alpha_source`	`IdioVolProportionalAlpha`	No	factory
`enabled`	`bool`	No	`True`
`kind`	`Literal`[‘maximize_alpha’]	No	`'maximize_alpha'`
`risk_aversion`	`float`	No	`1.0`	gt=0
`weight`	`float`	No	`1.0`	gt=0

field alpha_source: IdioVolProportionalAlpha [Optional]#: Where the expected returns come from.

field enabled: bool [Optional]#: Whether this objective contributes to the optimization problem.

field kind: Literal['maximize_alpha'] [Optional]#

field risk_aversion: float [Optional]#

Weight of the variance penalty against expected return.

Constraints:

gt = 0

field weight: float [Optional]#

Relative weight of this objective in the combined problem. Only the ratios between objective weights matter; they need not be normalized to sum to one.

Constraints:

gt = 0