On a test set, the bagged model was significantly more accurate than the single model with four covariates. It is also more stable. This is one possible fix. The multiplicity problem and its effect on conclusions drawn from models needs serious attention.
Leo Breiman, “Statistical Modeling: The Two Cultures”