Multivendor Portfolio Strategies In Cloud Computing

Cloud computing emerges as a powerful driver of the information technology industry and many companies are willing to exploit the advantages this development bears. However, services provided by the cloud are subject to default which can result in major economic damage for the client. Moreover, different cloud service providers may also bear a conjoint risk and may therefore not default independently. Hence, to implement effective cloud sourcing strategies, this paper postulates requirements for evaluating multivendor sourcing decisions to select cloud service providers, considering cost, cloud computing specific risk, and interdependencies. We develop an analytic model that meets these requirements and quantitatively expresses the specific cost and risk structure of cloud computing sourcing decisions. Our approach is based on Portfolio Theory with regard to the specifics of fungible cloud services by using exponential loss distributions and one-sided risk measures. Thereby, an evaluation and optimization of a client’s cloud service provider portfolio is possible. To determine the value added we use a simulation for the evaluation of our approach.