Modern Portfolio Theory (MPT) is a framework for building a portfolio of risky assets such that the ratio of risk to return is minimized. While this theory has been used in the field of financial economics for over sixty years, the method has not yet been applied to compensatory personnel selection. A common method for personnel selection is multiple regression to maximize the predicted performance of the selected group given a cut-off score on the predictor(s). Recognizing that maximizing the performance of the selected group is not the only consideration, and that, for many jobs and organizations, the outcomes of false positives and false negatives can be drastically different in terms of costs, is central to this study. MPT is offered as an additional method for generating weights that produce fewer false positives than multiple regression.
MPT generates a set of all possible combinations of predictors within the plane of risk and return and finds an optimal set of weights on the efficient frontier, the hyperbola that represents the best possible set of trade-offs between risk and return. This study uses Monte Carlo simulations to estimate boundary conditions where MPT can outperform multiple regression. Comparisons are drawn between MPT, multiple regression, and unit weighting, applying weights uniformly across all predictors. Comparisons between the methods are drawn consistent with Signal Detection Theory, categorizing prediction-criterion pairs in terms of "correct selections," "false positives," "correct rejections," and "false negatives." Boundaries for suitability for initial sample size, applicant pool size, and cutoff score of the performance measure are explored. Finally, an application of MPT for reducing adverse impact and promoting diversity by choosing combinations of variables that reduce the weight given to cognitive ability is explored.
Date of Degree
Master of Arts (MA)
Social and Behavioral Sciences
Leingang, Eric, "Selection Portfolio: Applying Modern Portfolio Theory to Personnel Selection" (2017). All Theses, Dissertations, and Other Capstone Projects. 718.
Creative Commons License
This work is licensed under a Creative Commons Attribution-Noncommercial 4.0 License