Article ID: | iaor19921360 |
Country: | United States |
Volume: | 22 |
Issue: | 1 |
Start Page Number: | 55 |
End Page Number: | 71 |
Publication Date: | Jan 1992 |
Journal: | Interfaces |
Authors: | Ben-Dov Yosi, Hayre Lakhbir, Pica Vincent |
Keywords: | financial, programming: mathematical |
During the last four years, Prudential Securities Inc has built a premier mortgate-backed securities (MBSs) department. Due to the complexity of the securities, standard fixed-income valuation tools are inadequate for MBSs. The development of sound and accurate models has been an integral part of the growth of the department. These models use a variety of management science techniques, including regression analysis, probabilistic modeling, Monte Carlo simulation, and linear, integer, and nonlinear programming. The models allow the firm to quickly and accurately value, and hence trade, complex MBSs, to properly hedge MBSs in inventory, and to structure clients’ portfolios to achieve given objectives while staying within specified constraints. The models are used hundreds of times each day by traders, salespeople, and clients to evaluate MBSs, and they form the basic tools necessary for the firm to participate successfully in the mortgage market.