Extending an EconoPhysics Value Model for a Pre-contract Award DOD Acquisition Investment Decision

Abstract

The Department of Defense acquisition professionals have a fundamental problem: there is no quantitative, accepted, comparable measure of value. Without this critical element, acquisition professionals cannot assess acquisition investment portfolios on the basis of the value that investment options bring to the DoD enterprise. This limits program managers to using historical cost estimates to predict the future performance of their programs. While historical measures can be useful, particularly on relatively mature programs, these methods are lacking with regard to the reliability and perhaps even the validity of trying to determine future program outcomes and subsequent program performance. Program managers need a more reliable predictive method that provides real insight into program performance from a return on investment strategy, not simply a cost method based upon comparing actual to estimated cost relative to work performed. Using the nexus between financial investment theory and physics, we hope to show that future performance of a an information technology (IT) can be predicted more a accurately than using historical cost data alone.

When there is no unique quantitative value metric with which to take advantage of commonly used financial performance ratios, the acquisitions program manager (PM) is forced to use metrics that do not have the predictive power resident in metrics that incorporate quantitative value estimates. Examples of rigorous financial metrics that can be used when there is a quantitative (i.e., common units) estimate of value include productivity based performance ratios, such as return on investment (ROI) and benefits/cost ratios. The current research study focused on developing an extension of our basic econophysics model (Housel, Baer, and Mun, 2015; Baer and Housel, 2017; Baer, Bounfour, and Housel, 2018) to generate a quantitative value metric, i.e., protovalue , that can be used to optimize acquisition portfolios on the basis of the relative returns on, and potential adoption rate of, DoD technology acquisitions. This extension includes parameters for cognitive biases that influence acquisition professionals and vendors expectations about the risk to successful performance of IT acquisitions. Further, this study will provide a potential extension of Earned Value Management (EVM) using the physics of the thermodynamics of a turbulent flow model.

Armed with a better understanding of the risks inherent in cognitive biases, a defensible econophysics based quantitative value metric (i.e., protovalue) and the application of a turbulent flow model to the current EVM framework, the acquisition professional will be better prepared to develop more precise approaches to predicting the performance of IT acquisitions. In what follows, we review the basic econophysics model in a simplified form and in a more detailed form. Acquisitions decision makers can use the simplified form to make rapid, rough-cut estimates of the future value of an IT application acquisition. A review of the more detailed basic econophysics model, in the context of an IT acquisition, provides the scaffolding for inclusion of the cognitive bias risk parameters in the proposed extended model. The report includes a review of Prospect Theory which was used to develop the new cognitive bias parameters in the extended econophysics model.

The proposed extended econophysics model should lead to better prediction of the value and potential adoption rate of future IT applications. Understanding and developing uses for the econophysics model will require a major learning curve that may cause acquisitions professionals to eventually abandon the current EVM approach. However, the proposed extensions of the current EVM approach, to include the phenomena of turbulent flow, will help acquisition professionals re-conceptualize EVM with the purpose of detecting cost, schedule, value problems earlier than the current model permits. This extension of the current EVM approach may not require a major learning curve.