Lab · B · Cross-sectional ranking
The internal results of the model that answers the question "which stock will do better than the others".
Feature importance
What it is: The share of each feature in the model's decisions: momentum, reversal, volatility, volume, residual return sums, momentum relative to the sector, and the graph leader score.
What data feeds it: The engineB result of the selected run; the variance reduction of the gradient boosted tree model in the last walk-forward fold. The ridge coefficient is the sign of the same feature in the linear model.
What it shows: The relative importance of the features at the selected horizon.
What can be concluded from it: At the 21-day horizon, a single feature (12-month momentum) dominating with a negative coefficient indicates that the model has learned the mechanical reversal built into the construction of the residual return. Verify this with the "residual / raw return" table on the Performance screen.
Information coefficient
What it is: For each day, the Spearman correlation between the model's predicted rank and the realised residual return rank; the dashed line is the mean.
What data feeds it: The engineB result of the selected run; only the predictions in the walk-forward test windows.
What it shows: Bars of the daily information coefficient, its mean, the t statistic and the number of days.
What can be concluded from it: A small positive mean (0.02–0.05) is realistic. A very large value (such as 0.2) says that leakage or an artefact should be looked for first; in this application, the source of such a value is the definition of the residual return.
The investment information, comments and recommendations given here are not within the scope of investment advisory services. Investment advisory services are provided under an investment advisory agreement to be signed between a client and brokerage houses, portfolio management companies, or banks that do not accept deposits. The signals here are produced from historical data with statistical models, are shown the same to everyone and are not personalised; they may not suit your financial situation or your risk and return preferences. Therefore, making investment decisions based solely on the information given here may not produce results that meet your expectations.