|
Article Excerpt
Abstract
This paper presents business cycle stylized facts for the Greek economy and extends the relevant Greek literature in the following directions. First, the index of industrial production (IOP) is used to represent real economic activity and business cycle conditions. Second, the behavior of certain financial variables throughout the various phases of the business cycle is analyzed in order to assess their leading indicator properties. Third, possible non-linearities in these variables are investigated and tested for their relation to the business cycle states. The results imply that the most reliable leading indicators are real Treasury bill rates. Volatilities of real short-term interest rates may also contain useful predictive information for IOP volatility. Finally, mean non-linearities seem to be associated with business cycle asymmetries in the mean. (JEL E32, E44)
Introduction
In recent years, the role of financial variables as indicators of future real economic activity has attracted the attention of many researchers. The debate has been particularly lively in the U.S., where financial deregulation and innovation have led to the search of financial indicator variables that are predictably linked to the final targets of output growth. In choosing an appropriate financial indicator variable, two aspects are important. First, the variable should signal the strength and direction of monetary policy actions and, second, it should act as an early indicator of future developments in the target variable (for example, GDP growth).
Numerous empirical studies for the U. S. demonstrate that interest rate spreads (measured as the difference between long- and short-term nominal interest rates) are good predictors of future economic activity [Estrella and Hardouvelis, 1991; Estrella and Mishkin, 1996]. However, the impact of term structure innovations on economic activity is less documented (or even negligible) for European countries [Canova and De Nicolo, 1997; Saner and Scheide, 1995]. Similarly, the literature finds that interest rates and money supply are considered leading business cycle indicators for the U.S., whereas the evidence for European countries is less decisive [Fiorito and Kollintzas, 1994; Andreou, Osborn, and Sensier 2000]. Finally, studies of the dynamic relationship between economic growth and stock market returns indicate that the latter are related to turning points in the U.S. business cycle [Fama and French, 1989; Malliaris and Urrutia, 1991]. As with the previous financial variables, the magnitude of the stock market leading indicator is generally stronger for the U.S. than for European countries [Andreou et al., 2000].
The aim of this paper is to make an empirical contribution to the literature on the relationships between real, monetary, and financial dimensions of the economy, both in terms of fluctuations in rates of change and in terms of volatilities. The underlying motivation is to investigate whether information contained in monetary and financial variables can be used to forecast future real economic activity. While most studies have focused on G-7 countries, very little attention has been devoted to medium-sized economies like Greece, Portugal, Ireland, and Norway.
This paper attempts to remedy this deficiency by providing a comprehensive empirical study of the behavior of certain monetary and financial time series over different phases of the business cycle in Greece. The Greek literature mainly focuses on monetary aggregates [Varelas, 1997; Dimeli et al., 1997; Kaskarelis 1993], and finds that the money supply is weakly related to GDP fluctuations. Apergis and Eleftheriou [2002] report a significant positive coefficient when regressing stock prices on the industrial production index and other variables, whereas Papapetrou [2001] shows that stock returns do not lead to economic activity. Another set of studies deals with the impact of monetary variables on consumption and concludes that real money supply [Apergis et al., 2000] and non-housing loans to the private sector [Apergis and Velentzas, 2003] significantly affect consumption behavior in the post-financial deregulation period.
The analysis in this paper extends the relevant Greek literature in the following directions. First, the index of industrial production (1) (IOP) is used to represent real economic activity and business cycle conditions. The adoption of this measure is in preference to a broader output measure, such as GDP, because the higher frequency of IOP allows one to accurately define the turning points of real economic activity. It should be noted that Greece is a semi-industrialized country and the largest proportion of its output is related to trade, tourism, transportation, and communications. (2) The structure of Greek production does not invalidate the selection of IOP as a proper output measure given that the industrial product as a percentage of GDP is both synchronous and pro-cyclical with GDP [Dimeli et al., 1997]. Focusing on the examination of IOP's cyclical movements around its underlying trend and setting a number of criteria related to the duration and amplitude of the phase, a recession chronology for the Greek economy for the period 1962-2001 is proposed.
Second, the behavior of a wide variety...
|
