In the world of practical finance, the management of working capital—cash, marketable securities, receivables, and inventories is, perhaps, the most pressing and most frequently encountered problem for financial managers. Yet, in the realm of theoretical finance, and even at the level of “textbook” finance, working capital is given minimal attention.

Whenever the firm must borrow funds, it must also decide maturity of the new debt. Yet, the decision models which have dealt with the debt maturity decision have done so almost incidentally, as an extension of the decision to exercise the call provision on outstanding bonds ([6], [10], [23]). There has been little direct examination of the corporate debt maturity decision. In an attempt to fill this gap, this paper is an exploration of the debt maturity decision for a firm which is concerned with minimizing the present value of the expected costs of borrowing. This paper develops a discrete dynamic programming model of the debt maturity decision, in a world where interest rates follow a finite Markov process, and where the yield curve is formed from expectations regarding the future course of interest rates. With this optimization model, the influence on the debt maturity strategy of variables such as flotation costs and liquidity premiums will be explored. There will be no consideration of the risks associated with alternative borrowing strategies.

Linear programming models of specialized financial decision problems such as working capital management [21], short-term financing [22], or capital bug-geting [24] are deficient in that they may lead to decisions which are suboptimal with respect to the firm as a whole. Each model attacks a single decision problem and neglects its interaction with the other activities of the firm. On the other hand, a model which reflects these interdependences and interactions by including the various financing, investment, and operating decisions in a single model tends to become excessively large and inefficient to use. What is needed is a model that incorporates the efficiencies inherent in smaller, more specialized models which can be utilized on a decentralized basis and which can simultaneously lead to decisions that are optimal for the firm as a whole.