Portfolio Optimization for Multiple Group Credit Unions

摘要

Recently passed legislation (H.R. 1151 Credit Union Membership Access Act) permits credit unions to add new select employee groups (SEGs) to their fields of membership. The field of membership defines the members served by a credit union. Given this new ability, along with guidance from the National Credit Union Administration (NCUA) to diversify, credit unions now have the opportunity to market their services to specific employee groups or industries which can reduce the overall risk to the credit unions' health or solvency. This thesis explores the use of Modern Portfolio Theory applied to a credit union's portfolio of assets (in this case SEGs or industries) using a mean-variance optimization technique (quadratic programming). Altman's Z- Score, a thoroughly tested and broadly accepted predictor of corporate failure, is used as an asset's 'return' or 'health' measure in the mean-variance algorithm and the variance of the Z-Score over a time interval is used as the 'risk' measure. The algorithm can be constrained as necessary and used to develop an efficient frontier showing optimal portfolios for each viable level of health and risk. Credit union management can use the results of the optimization as a tool to assist in their decision making process regarding the addition of new SEGs. The thesis examines two credit union case studies in which the technique is applied.