In semiconductor testing companies, depreciation cost takes the largest part of the cost structure. Several factors result in the existing situation. Firstly, due to the uncertainty of demands, investment decisions need to be made in a short time. Secondly, some parts have relatively short period of demand. For instance, according to Moore's Law, the period of demand of an integrated circuit (IC) is about 18 to 24 months. In the past, investment plans were made subjectively; inaccurate decision might cause the increase of cost. The current study aims to formulate the best investment plan. To achieve this, the uncertainty of market demand and period of demand were taken into consideration. Alongside, key components for investment suggested by the decision-makers were collected and the weights of decision strategies were assessed using analytic hierarchy process (AHP). Then, the actual cost of a product was calculated by incorporating scenario planning and decision tree analysis (DTA).