Decomposition techniques and disjunctive linear programming for fixed-income portfolio selection.
Item
-
Title
-
Decomposition techniques and disjunctive linear programming for fixed-income portfolio selection.
-
Identifier
-
AAI9732986
-
identifier
-
9732986
-
Creator
-
Wyatt, Katherine Grace.
-
Contributor
-
Adviser: Ken McAloon
-
Date
-
1997
-
Language
-
English
-
Publisher
-
City University of New York.
-
Subject
-
Mathematics | Operations Research | Computer Science
-
Abstract
-
Drawing on Sharpe's work in the '60s and '70s, researchers have developed a family of portfolio selection models that use absolute deviation, instead of variance, as a measure of dispersion of returns. We define a class of linear programs, called step-shaped programs, and show that programs for these absolute deviation models are step-shaped. The addition of logical requirements to programs in this class leads to the definition of disjunctive step-shaped programs. Variable decomposition for linear; step-shaped programs is discussed, and our results for variable decomposition of disjunctive step-shaped programs are presented. Algorithms for the solution of disjunctive step-shaped programs are outlined and verified. We describe a hybrid method of decomposition introduced by Van Roy, called cross decomposition, and show the effects of applying this method to step-shaped programs. A new model for fixed-income portfolio selection, the absolute deviation trade-off model, is introduced and the linear and disjunctive step-shaped programs that describe this model are detailed and analyzed. We include an overview of the financial ideas behind our model.
-
Type
-
dissertation
-
Source
-
PQT Legacy CUNY.xlsx
-
degree
-
Ph.D.
