Template-Type: ReDIF-Paper 1.0
Author-Name: Kevin Hoover
Author-Name-First: Kevin
Author-Name-Last: Hoover
Author-Name: Selva Demiralp
Author-Name-First: Selva
Author-Name-Last: Demiralp
Author-Name: Stephen J. Perez
Author-Name-First: Stephen J.
Author-Name-Last: Perez
Author-Workplace-Name: Department of Economics, University of California Davis
Title: A Bootstrap Method for Identifying and Evaluating a Structural Vector Autoregression
Abstract: Graph-theoretic methods of causal search based in the ideas of Pearl (2000), Spirtes,Glymour, and Scheines (2000), and others have been applied by a number of researchersto economic data, particularly by Swanson and Granger (1997) to the problem of findinga data-based contemporaneous causal order for the structural autoregression (SVAR),rather than, as is typically done, assuming a weakly justified Choleski order. Demiralpand Hoover (2003) provided Monte Carlo evidence that such methods were effective,provided that signal strengths were sufficiently high. Unfortunately, in applications toactual data, such Monte Carlo simulations are of limited value, since the causal structureof the true data-generating process is necessarily unknown. In this paper, we present abootstrap procedure that can be applied to actual data (i.e., without knowledge of the truecausal structure). We show with an applied example and a simulation study that theprocedure is an effective tool for assessing our confidence in causal orders identified bygraph-theoretic search procedures.
Length: 38
File-URL: https://repec.dss.ucdavis.edu/files/TZJsa7SZh94XX2ukePgG2643/06-14.pdf
File-Format: application/pdf
Number: 233
Classification-JEL: C30, C32, C51
KeyWords: vector autoregression (VAR), structural vector autoregression (SVAR),causality, causal order, Choleski order, causal search algorithms, graph-theoretic methods
Creation-Date: 20060327
Handle: RePEc:cda:wpaper:233