Theoretically, an argument could have no premises and still be valid. However, most such arguments are not interesting because the conclusions are essentially tautologies.
For example:
Assume P ->Q
Assume not Q
Assume for contradiction: P
Therefore Q, but we already said not Q, so P must be false.
Therefore not P.
Therefore not Q -> not P
So (P->Q) -> (not Q->not P)
That proves one direction of the equivalence of a conditional and it's contrapositive. No premise is necessary to prove the conclusion because the conclusion simply says that IF you have the statement P->Q, THEN you have it's contrapositive also. It doesn't assume that you actually have the original statement or any of it's components. It has no premises, but the conclusion is still valid.
However, the conclusion is just a tautology, which is usually the case in these kind of arguments.