quantlab.portfolio.config.LedoitWolfEstimatorConfig#

class quantlab.portfolio.config.LedoitWolfEstimatorConfig(lookback_bars, max_stale_bars=5)[source]#

Config of LedoitWolfEstimator: a shrunk sample covariance of trailing returns.

Examples

>>> cfg = LedoitWolfEstimatorConfig(lookback_bars=252)
>>> cfg.lookback_bars, cfg.max_stale_bars
(252, 5)

lookback_bars#: Bars of trailing one-bar returns the covariance is estimated from; at least 2. A symbol needs a finite return on every one of them.

max_stale_bars = 5#: Largest staleness (bars since the symbol’s last real price) a symbol may have at the bar and still be covered; a halt no longer than this stays in the estimate, flat returns and then its gap.