Date of Graduation

5-2016

Document Type

Thesis

Degree Name

Bachelor of Science in Industrial Engineering

Degree Level

Undergraduate

Department

Industrial Engineering

Advisor/Mentor

Pohl, Ed

Committee Member/Reader

Rainwater, Chase

Abstract

Portfolio optimization techniques are methods used to determine the best set of stocks in which to invest. Mean-variance optimization, one method of portfolio optimization, attempts to find the set of portfolios that have the maximum expected return at each level of risk (Jorion, 1992). Another technique, Monte Carlo simulation, uses random number generation to create a probability distribution of potential returns (Kwak & Ingall, 2007). This can be used to determine the risk of potential investments not returning a certain desired amount (Thompson & McLeod, 2009). Though traditionally used in the world of finance, these tools can also be utilized by professional sports teams, such as those in Major League Baseball, to make more efficient investments in personnel and increase their likelihood of reaching the postseason.

This research effort explores strategies to optimize the allocation of a baseball team’s resources in the free agent market. In this effort, we use a portfolio optimization approach and explore a variety of baseball performance metrics. A prototype optimization model is created and evaluated. This model is designed to assemble the team with the highest likelihood of making the playoffs while accounting for various budget and roster constraints faced by Major League Baseball teams. The prototype is utilized to create an optimized 2015 roster for three teams: the Boston Red Sox, Kansas City Royals, and San Diego Padres. These optimized rosters are then compared to each team’s actual 2015 opening day roster. Several iterations of this model are discussed in an attempt to find the option that returns the most value. After multiple alternatives are analyzed, three different options are identified that compare favorably to the teams’ actual opening day rosters with regards to 2015 performance of the players selected. Weaknesses of the model are then discussed, as well as ways in which it can be improved.

Keywords: portfolio optimization, mean-variance optimization, Monte Carlo simulation, expected returns, risk, performance metrics, stochastic optimization, linear optimization

Download

Share

COinS