Mean squared error of training dataset is higher than test data set

Question

mean_squared_error(y_train,y_train_pred)==468.2240 mean_squared_error(y_test,y_test_pred)==390.2098

Is this possible that model performance on testing set is batter t