With the purpose of stabilizing profits while maintaining predicable sales prices this study concerns itself with the properties of different cost-based internal pricing methods. It uses a performance attribution framework in order to study simulated data made to look like real production data that can be extracted from an ERP software in a multinational industrial company. 

The pricing methods differ in being either stochastic, such as the average overhead cost method, or deterministic, such as first-in-first-out and the modelling takes account of this main difference. Besides this the methods are studied under different interest and exchange rate circumstances. 

The results show clear similarities under one currency circumstances but differ more when more currencies are introduced. The method showing lowest profit variance was the three month updated price list closely followed by the last purchasing price and the average overhead cost methods.