lange earman on the projectibility of grue

#LANGE EARMAN ON THE PROJECTIBILITY OF GRUE SERIES#

His idea was that a spectral-density approach would give a richer and more comprehensive picture than a one-shot GC measure that is supposed to apply across all periodicities (e.g., in the short run, over the business-cycle frequencies, and in the long run).

#LANGE EARMAN ON THE PROJECTIBILITY OF GRUE SERIES#

The idea of further disentangling the Granger causality relationship between two time series was first suggested by Granger (1969).

While traditional approaches to GC have yielded many interesting insights, they generally tacitly ignore the possibility that the strength and/or direction of the GC – if any – could vary over different frequencies. While not exhaustive, the above enumeration clearly demonstrates the widespread use of the GC concept in economics and business. In addition, GC tests have also been used extensively for model selection (Teräsvirta, 1986). In marketing, GC testing has been used predominantly to discern competitive reaction patterns (see Hanssens, Parsons, & Schultz, 2001, p.314, for a review). In the financial literature, GC testing has been applied to identify price-leadership patterns among national stock prices (Peiers, 1997), to study the stock price–volume relationship (Hemstra & Jones, 1994), and to get insight into the dynamic behavior of bonds and stocks (Park & Shenoy, 2002) or the causal relationship between equity and real estate returns (Okunev, Wilson, & Zurbruegg, 2002). In addition, the predictive content of survey data is often translated into a GC framework (Hanssens & Vanden Abeele, 1987). For instance, the well-known “export-led growth” hypothesis has been studied repeatedly in a GC framework (Abual-Foul, 2004), as has the relationship between economic growth and various other variables, such as business-cycle volatility (Dopke, 2004), the degree of openness (Bahmani-Oskooee & Niroomand, 1999), defense spending (Al-Yousif, 2002), and electricity consumption (Thoma, 2004).

Specifically, a process X t is said to Granger cause another process Y t when the variance of the error in forecasting future values of Y t, using an (optimal) forecast based on the observed values of both X t and Y t, is strictly smaller than the variance of the prediction error, using an (optimal) forecast based only on the observed values of Y t.Ī wide range of bivariate GC tests exist (e.g., the Granger–Sargent test, the Granger–Wald test, the Haugh–Pierce test, and the Sims test see Gouriéroux & Monfort, 1997, for a general overview), which have been used extensively to study a wide range of substantive economic issues. GC reflects a restricted sense of causality, i.e., the extent to which a process X t is leading another process Y t, and builds upon the notion of incremental predictability. To assess the causality between two processes in a common and well-defined (non-experimental) framework, one usually refers to the concept of Granger causality (hereafter GC), introduced in 1969 by the 2003 Nobel prize winner in Economics. Investigating causality is a topic of great interest in scientific research. This predictive content is found to vary widely with the frequency considered, illustrating the usefulness of not restricting oneself to a single overall test statistic. In addition, we apply the methodology in the context of the predictive value of the European production expectation surveys. To provide insights into the relative performance of this test, we study its power properties by means of Monte Carlo simulations. In this paper, we compare two existing approaches in the frequency domain, proposed originally by Pierce and Geweke, and introduce a new testing procedure for the Pierce spectral Granger causality measure. This may yield new and complementary insights compared to traditional versions of Granger causality. Decomposing Granger causality over the spectrum allows us to disentangle potentially different Granger causality relationships over different frequencies.

Lange earman on the projectibility of grue