In this paper, we study a two-product, multi-period stationary newsvendor problem with budget constraint. In the problem, the inventory of each product is not allowed to be carried between one period and another. At the beginning of each period, the newsvendor has to determine the order quantities of the two products with budget constraint. We apply the weak aggregating algorithm (WAA) to study this problem. WAA, developed in the field of computer science, is a competitive online prediction method of combining expert opinions. By regarding each fixed stock level as an expert opinion and updating its weight according to the cumulative loss it achieves, we first provide an online ordering strategy to decide each period's order quantity. Furthermore, we theoretically prove that this online ordering strategy possess competitive property, i.e., the cumulative loss it achieves is as small as that achieved by the optimal expert. The salvage value and stock-out cost are further considered to obtain extended results. Finally, the numerical analysis illustrates that the online ordering strategies proposed in this paper are competitive whether with or without salvage value and stock-out cost.