sf_quant.performance.get_turnover_stats#

sf_quant.performance.get_turnover_stats(weights: DataFrame) → DataFrame#

Compute summary statistics for two-sided portfolio turnover.

Calculates the rolling 252-day mean two-sided turnover and summarises it with mean, min, and max statistics.

Parameters#

weightspl.DataFrame

Portfolio weights containing:

  • date (date): The observation date.

  • barrid (str): Security identifier.

  • weight (float): Portfolio weight.

Returns#

pl.DataFrame

Summary statistics with columns:

  • Mean Turnover (float): Mean of the rolling two-sided turnover.

  • Min Turnover (float): Minimum of the rolling two-sided turnover.

  • Max Turnover (float): Maximum of the rolling two-sided turnover.

Notes#

  • Two-sided turnover is the sum of absolute weight changes per date.

  • Rolling window is 252 trading days.

  • Null values (warm-up period) are excluded from summary statistics.

Examples#

>>> import polars as pl
>>> import sf_quant.performance as sfp
>>> import datetime as dt
>>> weights = pl.DataFrame({
...     'date': [dt.date(2024, 1, i) for i in range(2, 6)],
...     'barrid': ['A'] * 4,
...     'weight': [0.5, 0.6, 0.4, 0.5],
... })
>>> sfp.get_turnover_stats(weights)
shape: (1, 3)
┌───────────────┬───────────────┬───────────────┐
│ Mean Turnover ┆ Min Turnover  ┆ Max Turnover  │
│ ---           ┆ ---           ┆ ---           │
│ f64           ┆ f64           ┆ f64           │
╞═══════════════╪═══════════════╪═══════════════╡
│ 0.15          ┆ 0.1           ┆ 0.2           │
└───────────────┴───────────────┴───────────────┘
On this page