Forecasting industrial production in Brazil: Evidence from a wavelet approach Benjamin Miranda Tabak a,b, * , Mateus Araujo Feitosa c a Banco Central do Brasil, Research Department, Brazil b Graduate Program in Economics, Universidade Catolica de Brasilia, Brazil c Department of Business Administration, Universidad Carlos III de Madrid, C/Madrid 126, 28903 Getafe, Spain article info Keywords: Multiresolution analysis Yield spread Term structure of interest rates abstract The present paper studies the informational content of the Brazilian term structure of interest rate. We apply multiresolution decomposition to the spread and to the industrial production, and investigate if the term spread presents information that can help predict the path followed by industrial production. We also investigate if all the predictive ability of the spread is due to the monetary policy. The results suggest that the yield spread contains relevant information, and this predictive power varies across time patterns. Inasmuch, the results indicate that the predictive power of the spread is not totally explained by mone- tary policy. Ó 2010 Elsevier Ltd. All rights reserved. 1. Introduction Forecasting output is crucial for policy makers and economic agents in general. Therefore, a large body of the literature has fo- cused on different forecasting models and comparing the predic- tive content of such models (Anderson, Hoffman, & Rasche, 2002; Banerjee & Marcellino, 2006; Black, Corrigan, & Dowd, 2000; Heravi, Osborn, & Birchenhall, 2004; Malik & Nasereddin, 2006; Nordhaus, 1987; Seip & McNown, 2007). One of the variables that may help predict inflation and output growth is the yield spread, that is, the difference between long and short-term interest rate. Theoretically, short-term rates do not concentrate all the information that can help predict future eco- nomic activity, since investment decisions are made considering the loan credits, which by its turn depends of different maturity rates. Black et al. (2000) compare the interest rate spread, mone- tary aggregates and the federal funds rate as predictors of indus- trial production, and empirical results indicate that the model that use lagged values of industrial production along with the interest rate spread as explanatory variables is the one with better predictions. 1 Despite the recent research effort on studying the informational content of interest rates spreads a few questions remained open to debate. Does the informational content of interest rates spread de- pend on the monetary policy in place? Does the informational con- tent depend on time patterns? Is it useful in emerging markets economies? In order to provide some answers to the above questions this paper focuses in the Brazilian economy, an emerging market, which has an Inflation Targeting framework for monetary policy. A multiresolution approach is employed in order to test the infor- mational content of the Brazilian yield spread. We decompose the series using wavelet methodology, and then study the relationship between the variables by means of univariate regressions. Our aim is to investigate if the yield spread can be used as a predictor of fu- ture economic activity. The main contribution of this paper is that it employs a recent methodology in order to test the ability of the yield spread to pre- dict economic growth in an emerging market, an issue that does not present a consensus in the research area. 2 Furthermore, it shows that the predictive content is not due solely to monetary policy framework. Wavelets methods have been shown to add sig- nificant insights for financial analysis in a variety of contexts (see Bayraktar, Poor, & Sircar, 2004; Fleming, Yu, & Harrison, 2000; Guo, Sun, Li, & Wang, 2008; Kim & Noh, 1997; Kyong & Han, 2000; Li & Kuo, 2008; Sharkasi, Ruskin, & Crane, 2005; Tabak & Fei- tosa, 2009; Tang, Tabg, & Sheng, 2009). The rest of the paper will be structured as follows: In Section 2 it is presented a brief literature review, and the wavelet methodol- ogy is presented in Section 3. Section 4 contents the results for the industrial production, the role of monetary policy is investigated in Section 5, and Section 6 concludes the paper. 0957-4174/$ - see front matter Ó 2010 Elsevier Ltd. All rights reserved. doi:10.1016/j.eswa.2010.02.086 * Corresponding author. Address: Banco Central, DEPEP, SBS Quadra 3, Bloco B., Ed. Sede, 13 Andar, Brasilia, DF 70074-900, Brazil. Tel.: +55 6134142045. E-mail address: Benjamin.tabak@bcb.gov.br (B.M. Tabak). 1 See also Seip and McNown (2007), which report that the yield spread is a leading indicator for the path followed by industrial production. 2 Stock and Watson (2003) concluded a survey on the subject arguing that the universality of the role of the yield curve as a predictor for growth is unresolved. The evidence for emerging markets has been virtually nil due to lack of data. Expert Systems with Applications 37 (2010) 6345–6351 Contents lists available at ScienceDirect Expert Systems with Applications journal homepage: www.elsevier.com/locate/eswa