RollingCalmar

Description

Calmar ratio: annualised return divided by the worst rolling drawdown:

\[ \text{Calmar}[t] = \frac{\text{ppy}\ \cdot\ \text{mean}(r)}{\big|\,\text{RollingMaxDrawdown}(\text{implied price})\,\big|} \]

Takes a returns series; internally reconstructs the implied price path as a cumulative product price *= (1 + r) starting from 1.0, so the drawdown calculation is well-defined. Returns NaN when the path is monotonic up (no drawdown in window).

If you already have a price series, compose by hand:

NaN handling

Policy: ignore. A NaN in any input at index t causes the function to skip that step: output at t is NaN and internal state is unchanged. Subsequent finite samples are processed as if step t had not occurred.

Examples

Usage example

import numpy as np
import plotly.graph_objects as go
from plotly.subplots import make_subplots
from screamer import RollingCalmar

np.random.seed(0)
ret = np.random.normal(0.0005, 0.01, size=300)
calmar = RollingCalmar(window_size=63, periods_per_year=252)(ret)   # annualised return over worst rolling drawdown

fig = make_subplots(rows=2, cols=1, shared_xaxes=True,
                    row_heights=[0.55, 0.45], vertical_spacing=0.08)
fig.add_trace(go.Scatter(y=ret, mode="lines", name="returns"), row=1, col=1)
fig.add_trace(go.Scatter(y=calmar, mode="lines", name="Calmar",
                         line=dict(color="red")), row=2, col=1)
fig.update_layout(title="Annualised Calmar over 63 bars (RollingCalmar)",
                  margin=dict(l=20, r=20, t=60, b=20),
                  legend=dict(orientation="h", yanchor="bottom", y=1.02, xanchor="right", x=1))
fig.update_yaxes(title_text="returns", row=1, col=1)
fig.update_yaxes(title_text="Calmar", row=2, col=1)
fig.show()