Coffee, the money market, the real exchange rate, and economic fluctuations in Colombia
This thesis analyses the effects of coffee booms on the money market, the real exchange rate, and the business cycle in Colombia. Chapter 2 presents an overview of the coffee sector in the country, including a brief description of its macroeconomic role, and unique institutional structure. Chapter 3 investigates, from a simulation perspective, two empirical difficulties that arise in econometric modelling when using quarterly data, as is done in chapters 4 and 5. The first practical concern is whether to conduct the econometric analysis on data that have been subjected to seasonal adjustment or in terms of unadjusted data. The simulation results provide a justification for using seasonally unadjusted data, as the use of filters reduces the power of the Dickey-Fuller and Phillips-Perron cointegration tests. The second difficulty concerns an empirical regularity encountered when analysing the Colombian quarterly series of money supply and GDP, both of which exhibit a structural break (or change) in the seasonal pattern. We find that these structural breaks bias both unit root and seasonal root tests, so that new critical values must be tabulated allowing for a change in either the level and/or the seasonal pattern of the underlying series. Chapter 4 examines the monetary consequences of coffee booms. The theoretical work on this subject shows that under a regime of fixed exchange rates, export booms affect both the demand and the supply for money. Within this theoretical framework, we assess whether the coffee booms of the second half of the seventies and mid eighties led to excess money supply in Colombia. We find a direct association between coffee export booms and excess money supply, implying that external disturbances jeopardise the ability of the economic authorities to carry out successful monetary policy. Chapter 5 uses the Johansen procedure to estimate a real exchange rate determination model for Colombia. We find one cointegrating vector, which can be thought of as a long-run real exchange rate equation. The deviations of the real exchange rate from its long-run equilibrium relationship, after correcting for the short-run dynamics, are interpreted as a measure of real exchange rate misalignment. The simulation performance of the model, during the period of estimation and three years into the future, is particularly good, with the simulated real exchange rate reproducing the general long-run behaviour of the actual series. Chapter 6 develops an intertemporal disequilibrium model in order to analyse the effects of temporary, anticipated, and permanent coffee price shocks on a small open economy under Keynesian unemployment. Our results indicate that a coffee price boom (whether temporary, anticipated or permanent) increases nontradable output in the short and long run (a similar result is obtained when we discuss other disequilibrium regimes). The basic model is then extended by including a government sector that administers a coffee price stabilisation fund, and by allowing capital market imperfections. Our results indicate that when the government is able to borrow on more favourable terms in international capital markets than households, the stabilisation fund neutralises part of the short-term effect of a temporary coffee price boom. On the other hand, when the government and the private sector borrow on the same terms, the stabilisation fund turns out to be redundant.