Essays on the demand for money by firms

This thesis consists of two essays on money demand by firms and the roles trade credit and financial innovation play in its determination.The aim of the first essay is to provide theoretically sound foundations for analyzing firms' demand for liquid assets. To meet that end, I develop an endogenous liquidity-in-advance firm model that formally captures the role bank-like financial institutions play in reducing the costs associated with capital market imperfections. The model provides a general theoretical underpinning for the role of what I refer to as distribution services (or financial innovation) whose role is to reduce the costs incurred by firms when buying and selling liquid assets to produce some desired level of liquidity. I show that endogenizing the liquidity in advance constraint by allowing trade credit to serve as another means of payment has critical theoretical implications. Ignoring trade credit, in fact, renders the cost of liquidity---or financial cost---irrelevant to the choice of production inputs mix. Therefore, the importance of financial cost is significantly reduced in determining the demand for liquid assets and incorrect theoretical predictions can obtain. The developed model also provides a framework for analyzing the demand for trade credit. The ambiguous relationship between the rate of interest charged on trade credit and its use that results from the model provides theoretical support for the corresponding ambiguous empirical results.The second essay examines the empirical implications of accounting for the use of trade credit in the analysis of money demand by firms and the effect of financial innovation. I derive two testable implications relating to the importance of the transaction role of trade debt in the long run. Since, in addition to real balances, trade credit constitutes a means of payment, it follows that a more stable relationship with output should obtain when trade debt is added to real balances than when it is not. Along the same line of reasoning it is clear that trade debt and real balances are substitutes. Therefore, a long run relationship with output and interest rate should reflect this substitution condition. Using Johansen's maximum likelihood estimation method for detecting the presence of cointegration I find substantial support for these two implications. Previously elusive long run relationships relating money to its traditional determinants are found when trade debt is included in the cointegrating vector autoregression. I use the second cointegrating relationship (relating real balances, trade debt, output and interest rate) in an error correction money demand equation that also accounts for financial innovation. Impulse response functions to shocks in interest rate and output are derived and analyzed.