|
Article Excerpt
Lockheed Martin Space Systems Company spends millions of dollars on the maintenance and modernization of its infrastructure each year. Projects often involve investments that cannot be justified purely in terms of net present value or other classical investment-evaluation methods. The options are also restricted because funds that are not spent within a given time frame must be relinquished. Furthermore, some projects may be delayed and the unplanned carryover of their costs moved into the next fiscal year; this causes the postponement or cancellation of other unrelated projects because of in-budget transfers. In this paper, we used multiattribute utility theory and chance-constrained programming to optimize the selection of infrastructure projects. Our solution ensured the selection of high-value projects to maximize the company's performance. These selections were subject to the constraints that a portfolio did not exceed the available budget and that the carryover of the unspent funds to the next fiscal year did not exceed predetermined limits. We used Microsoft Excel to ensure broad accessibility, transparency, user interaction, improved data collection and asset management, and ease-of-use by managers.
Key words: capital rationing; chance-constrained programming; multiattribute utility theory; budgetary carryover; spreadsheet applications. History: This paper was refereed.
**********
Lockheed Martin Space Systems Company, based in Denver, Colorado, is located on a 5,500-acre facility with 37 major buildings--most constructed prior to 1970. Each major building contains a multitude of systems, all exhibiting various degrees of wear and tear. The Facility Operations and Services (FO&S) Department at Lockheed Martin is responsible for choosing a portfolio of projects to improve the condition of the site. Allocating capital to the maintenance and modernization of the facility infrastructure and office space is a nontrivial task given the size of the facility. Previously, FO&S selected projects based on the judgment of decision makers within the department and qualitative input from various stakeholders. This lengthy process often involved stressful negotiations, but produced suboptimal decisions. Lockheed Martin needed a robust analytical method for optimizing its project selection. Therefore, we developed a decision-support system based on multiattribute utility theory and chance-constrained programming.
We used multiattribute theory to clarify and measure the department's objective and quantify the potential contribution of each project to that objective. We addressed the risks and constraints inherent in capital-rationing problems explicitly via chance-constrained programming. We used a spreadsheet program to develop, solve, and present our model. This ensured broad accessibility and transparency, improved data collection and asset management, and ease-of-use by managers. Using this tool, FO&S achieved a substantial improvement in the value of Lockheed Martin's infrastructure.
Problem Description
During the fourth quarter of each year, Lockheed Martin allocates more than $10 million in capital funds to FO&S, which maintains a list of approximately 300 potential projects. Broadly speaking, these projects are in one of three categories: construction, installation, or purchase. Although FO&S's overall capital allocation is fixed and should not be overspent, some small overruns are usually tolerated. It selects a project portfolio in the third quarter of each year; during the fourth quarter of that year, it spends its available budget on those projects. The budget must be spent within the fiscal year in which it is allocated because unspent money cannot be carried over and added to the budgets of subsequent fiscal years.
Implementation of an analytical approach in project selection was long overdue. Historically, FO&S decision makers (10 to 15 managers, project managers, engineers, and planners) spent an inordinate amount of time, effort, and emotion each year on capital-project-portfolio selection and funding, making their selections based on their personal knowledge of the facility. There were myriad considerations, such as complying with local and federal law, modernizing office environments, minimizing the risk the facility might pose to aerospace products manufactured and tested on site, maintaining the infrastructure in working condition, and ensuring the facilities are properly configured to support the pursuit of new business. Given the diverse areas of expertise of the decision makers and their levels of familiarity with each project, they often expressed widely varying opinions, and spent substantial time and money negotiating. Because there was no quantitative and objective way to measure a project's value, each project necessitated a debate each year. However, because of the sheer numbers of available projects, the decision makers lacked the time to debate each project's value thoroughly. Therefore, they selected a small subset of seemingly important projects and debated their significance until they reached a consensus. This consensus might result in the portfolio that best benefited the company; however, it might just as easily be the one that allowed the debate to end.
Another drawback of the previous practice was that it considered only a particular project's importance to the company and its estimated cost. It did not consider the timing of cash flows or the uncertainty in timing and magnitude of cash flows. Unless a project was a multiyear endeavor, questions about its duration were usually not asked. High-valued projects were often selected because they were perceived as important; however, they typically introduced unexpected fiscal uncertainty. If these projects were completed later than planned, they could cause substantial disruption in financial-planning cycles. During the portfolio-selection process, schedule and budget uncertainty were not addressed; consequently, the predictive power of the FO&S's planning was limited.
Replacing the previous practice with an analytical, computer-based approach would allow the department to significantly reduce the time and effort invested in portfolio selection, improve the value of selected portfolios, and minimize the budget overruns.
Related Literature
The central issue in capital-rationing problems is the allocation of limited financial resources among alternative projects with the aim of achieving the maximum profit over time. Generally, the projects are weighted in comparison to each other using net present value (NPV), present value ratios, or incremental analysis of alternatives. The decision maker then chooses a project portfolio that maximizes expected return on investment while adhering to the capital-budget constraint. When the choice between...
|
