Skip to content

Expected Shortfall

Intuição

O Expected Shortfall (ES) (também conhecido como CVaR) mede a perda média condicional nas piores realizações além do VaR.

Definição

Para retornos r, janela n e confiança c:

  • alpha = 1 - c
  • q = quantile(r, alpha)
  • ES = -mean(r | r <= q)

Uso

from quantmaster.features.risk import expected_shortfall

df["es"] = expected_shortfall(df, window=252, confidence=0.95)

API

Source code in src/quantmaster/features/risk.py 
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
def expected_shortfall(
    data: pd.DataFrame | pd.Series,
    *,
    window: int = 252,
    confidence: float = 0.95,
    price_col: str = "close",
    log_returns: bool = True,
) -> pd.Series:
    window = validate_positive_int(window, name="window")
    try:
        confidence = float(confidence)
    except (TypeError, ValueError) as exc:
        raise TypeError(f"confidence must be float, got {type(confidence).__name__}") from exc
    if not (0.0 < confidence < 1.0):
        raise ValueError(f"confidence must be between 0 and 1, got {confidence}")

    price = get_price_series(data, price_col=price_col).astype(float)
    price = price.where(price > 0)

    if log_returns:
        rets = np.log(price).diff()
    else:
        rets = price.pct_change()

    alpha = 1.0 - confidence
    out = rets.rolling(window).apply(lambda x: _expected_shortfall_window(x, alpha=alpha), raw=True)
    out = out.clip(lower=0.0)
    out.name = f"expected_shortfall_{window}_{confidence:g}"
    return out