B is correct. Ideally, the analyst will use both the actual and hypothetical returns for backtesting since both yield informative comparisons. If the model passes backtesting with hypothetical but not actual returns, the problem in the model lies with intraday trading. If the model does not pass backtesting with hypothetical returns, the modelling methodology should be re-examined.

A is incorrect. The descriptions of the cleaned returns are not correct, and these are not the correct two sets of returns that the analyst should use. Cleaned returns will also subtract any fees, commissions, net interest margin from actual returns. The correct definition of cleaned returns can sometimes be used as an approximation of hypothetical returns, but not the incorrectly defined returns stated here.

C is incorrect. These are not the correct two sets of returns that the analyst should use, and the description of cleaned returns is incorrect.

D is incorrect. There is not a return set that is called trading returns.