We are releasing today a new target, target_nomi, that you can optionally use to train your models. You should still submit predictions in the same exact way as before but just using the outputs of the newly trained model. Fundamentally, this target represents the same thing as target_kazutsugi, but just a bit more faithfully. Consequently, it is compatible with target_kazutsugi and we will continue scoring using target_kazutsugi for the time being. We will give plenty of warning before we eventually switch to scoring with target_nomi (likely late October). In my experiments, training models using target_nomi and scoring against target_kazutsugi, gave about the same scores on average as training on target_kazutsugi, but with less variance era-to-era, and therefore higher Sharpe. On our side, this target allows us to construct portfolios with higher expected returns. We expect this to be win-win and want to make the transition as smooth as possible. Even after the transition to scoring on target_nomi, your old models trained on target_kazutsugi will still work, but will likely score slightly worse than if you had retrained on target_nomi. So let us know how it goes!
You can download it here as a parquet file:
Or here as csv:
You can watch the discussion of the new target in OHWA here:
