Abstract
Modern open economy macroeconomics is to a large extent characterized by dynamic models with explicit micro-foundations. The models are often highly non-linear, and hard or even impossible to solve explicitly. Rather than explicit solutions, analyses are often limited to determine signs of derivatives.
In order to find out whether a model corresponds to data, or in order to estimate the quantitative effect of a shock or of different kinds of fiscal and monetary policy, one needs a much wider analysis. It is not satisfactory only to be able to state the direction in which the variables of the model move, and not even know whether changes are significant or not.
This paper has two main objectives:
1. First it attempts to discuss methods to estimate how much macroeconomic variables such as production, consumption or the real interest rate change in the short- and the long run as a result of an exogenous shock. The methods presented here are based on methods in Uhlig (1997).
2. Secondly, an example of its application is shown, to discuss the
effect of sticky prices and monopolistic competition in an open-economy framework. This is done by extending a model in Obstfeld and Rogoff (1995a and 1996) and fitting it into Uhlig's framework. It is shown what powerful results that can be found in the field of dynamic shock analysis. The main focus is on monetary shocks, and it is shown how shocks in the nominal money stock can lead to permanent real effects. Solving and simulating the model is done using MatLabTM source code.
In addition this paper provides long-run simulations and compares the
results with historical data. Finally, the paper takes up important issues such as the quality of the linear approximations, and the sensitivity of the final model for mistakes in the calibration. Also, attempts are made to provide implications for fiscal and monetary policy of the impulse responses that are drawn.
The author has already, together with Houeland in Houeland and Lien (2003), shown how the baseline model in Obstfeld and Rogoff (1995a) can be analysed quantitatively using the framework in Uhlig (1997) - and thereby contributed somewhat to the second objective mentioned above. This paper builds to some extent on their work in reaching the second objective, but corrects some mistakes in addition to extending and generalizing the model.
Chapter 2 presents the general method of linearizing a dynamic stochastic model, to write it as a system of difference equation and how to solve this system. The chapter is not complete, as potential problems and more details follow in the subsequent chapters as an example of application is shown.
Chapters 3 and 4 present an extended version of the open-economy model with sticky prices in Obstfeld and Rogoff (1995a) and show how to apply in practice the methods from chapter 2. Chapter 5 continues this by calibrating the model, and chapter 6 analyses impulse responses of shocks. Chapter 7 simulates first- and second-order moments of the model variables and compares with real data. Chapter 8 discusses possible weaknesses of the calibration, of the model setup and of the linear approximation methods. Chapter 9 concludes.