3.10
Mean-Reversion Weighted Regression
PYTHON PLUGIN
Stocks
Mean Reversion
Description
Strategy 3.10: Weighted regression residuals for mean reversion
Strategy Logic
Strategy 3.10: Mean Reversion with Weighted Regression.
Use weighted (exponentially decaying) regression residuals for
mean-reversion signals. Fit: log(P_t) = a + b*t + epsilon_t
with exponential weights. Trade on the sign and magnitude of
the most recent residual.
Parameters
| Parameter | Default Value | Type |
|---|---|---|
| lookback | 60 | int |
| halflife | 21 | int |
| entry_z | 2.0 | float |
| exit_z | 0.5 | float |
Risk Configuration
| Risk Parameter | Value |
|---|---|
| Max Position Pct | 5.0% |
| Stop Loss Pct | 3.0% |
| Take Profit Pct | 6.0% |
| Max Drawdown Pct | 10.0% |