Alloyed steel raw material is very cost intensive due to the necessary addition of high price alloy elements, e.g. molybdenum, chrome or nickel. By this it is a considerable cost factor for the metalworking industry and a fraction amounting about 60 percent of the total costs is nothing unusual. Using an optimal purchasing policy that defines which quantities to order at which points in time the acquisition costs in the operational-tactical area can be reduced significantly while improving the logistical efficiency simultaneously. Still no adequate procedure dealing with the purchasing object group “alloyed steel raw material” that allows to derive an optimal purchasing policy in practice in a feasible way has been published so far. This dissertation closes the mentioned gap in research work and supports the competitiveness of the metalworking industry. The process development was focused on a low level of complexity in exercise aiming at an improved acceptance in business environment and a facilitated consideration of the implicitly existing know-how of the persons in charge of disposition. The described procedure intends the installment of an initial solution strategy which shall be followed in case of decreasing or stable prices of alloy elements where the minimum order quantity shall be equivalent to the technically and economically determined lot size of the steel producer. In case of an expected price increase the initial solution strategy shall be reviewed at least partly: the procedure shall also check the reasonability of purchasing more than one lot size. Hereby the speculative gain induced by a larger order quantity is compared with additionally generated storage costs. After the development of the strategy it was subject to validation. The result showed that the described procedure is an appropriate strategy for the area of the investigated purchasing object group and superior to other yet published solutions. According to Popper’s philosophy of science the rational verification of this strategy was furthermore supported by subsequently adding empirical data by using a comparison of the described procedure with 642 alternative variants of disposition.