LEVERAGED BOND PORTFOLIO OPTIMIZATION UNDER UNCERTAINTY

Autor: Dan R. Pieptea
Rok vydání: 1987
Předmět:
Zdroj: The Financial Review. 22:87-109
ISSN: 1540-6288
0732-8516
DOI: 10.1111/j.1540-6288.1987.tb00320.x
Popis: This paper develops a multiperiod management model for balance sheets of bond assets and liabilities. The decision variables are the amounts to buy, sell, or hold for each bond maturity class and the amount to borrow or debt to be repurchased for each liability class. Decisions are made at the beginning of each period conditional on the revealed interest rates scenario. The investor's utility is maximized in a state-preference approach. The maturity structure of the initial holdings is an important component of the input parameters. The possible states of the world are described in terms of interest rate scenarios. The solution is subject to funds flow, inventory, liquidity, leverage, policy, and expiration constraints. The model leads to a large-scale linear programming model, with multi-indexed variables, that is solved with the Dantzig-Wolfe decomposition algorithm. A numerical application that was motivated by discussions with top management at the RepublicBank Trust Department is presented. The experiments suggest that optimal portfolios frequently have a barbell structure consisting of short and long maturities with few intermediate maturities.
Databáze: OpenAIRE
Externí odkaz: