In today's turbulent stock market conditions, the ability to leverage powerful computational resources to provide timely and accurate risk analysis has become critical to the successful operation of stock trading firms and hedge funds. One such risk measure, Value-at-Risk (VaR), is a market standard used by senior management and regulators to quantify the risk level of a firm's holdings. Applications for performing VaR calculations often have very dynamic computational workloads and are time critical. The ever-changing computing environment of such VaR applications makes it necessary for a firm’s cyber-infrastructure to be able to handle bursts in computing and storage needs. A traditional approach to address this problem involves investing in high-end hardware that can accommodate spikes in demand. However, this approach is rather costly, and leaves these resources underutilized during off peak times. Cloud computing has emerged as a powerful paradigm that provides computational resources, software, and middleware as services that can be ‘rented’ on demand. However, for the most part, cloud computing companies provide inexpensive low-end resources that do not meet the intensive computational requirements of VaR applications. In this work we utilize cloud abstractions to provision and provide high-end resources as a utility to end-users. We provide on-demand scalability, dynamic provisioning, and the integration of distributed high performance computing (HPC) resources. The proposed ‘HPC as a service’ has the potential to cost effectively support spikes in resource requirements needed by VaR applications, integrating clouds with computing platforms and data centers, as well as developing and managing applications that can utilize this platform.