武大讲义-DynamicOptimizationIII..doc

III Dynamic Optimization A person who insists on understanding every tiny step before going to the next is liable to concentrate so much on looking at his feet that he fails to realize he is walking in the wrong direction. － I. Stewart (1975) 3.1 Introduction to dynamic optimization This section is a very informal attempt to motivate the exposition of the dynamic optimization methods. We use a simple dynamic macroeconomic model to introduce several important concepts through numerical examples. Let , , and be aggregate income, consumption, investment and capital, respectively; then a simple macroeconomic model of income determination would be To obtain a growth model we formally equate the net rate of change in capital stock to investment and . Solution: Optimal borrowing We consider the possibility of borrowing to augment the initial stock of capital. Let the size of the loan be, with continuously compounded interest at the rate ; the amount to be repaid at the expiry date of the loan, say , is . We use the production function , and the initial condition ; letting the price of capital be 1 at any time, it is now possible to begin with units of capital; is to be optimally chosen to maximize the amount of capital available at time , after the loan has been repaid. Solution: with initial condition (1) (2) At time the loan is repaid and what remains is FOC Substituting into yields the value of the remaining stock after repayment: In order to obtain the solution for , we must not simply deduct from (2). Instead, we must use as the new boundary condition and determine anew the constant in (1). This implies that we jumt onto a new trajectory at time . Fiscal policy Here we add a government to the preceding model and give it the power to tax; the government’s tax revenue is invested, so that it is a case of forced savings. The model is