عدم قطعیت سیاست اقتصادی و نرخ بازده بازار سهام در کشورهای حاشیه اقیانوس آرام: شواهد مبتنی بر مدل خود رگرسیون برداری پانل بیزین Economic policy uncertainty and stock market returns in pacific-rim countries: Evidence based on a Bayesian Panel VAR Model

۱٫ Introduction In the wake of the financial crisis, economic policy uncertainty has raised a lot of interest due to its potential negative effects on economic activity (Bloom et al, 2007; Bloom, 2009; Antonakakis et al., 2013; Pastor and Veronesi, 2012, 2013; Aasveit et al., 2013; Shoag and Veuger, 2013; Baker et al., 2015; Brogaard and Detzel, 2015; Gulen and Ion, 2015). For example,the Federal Open Market Committee (2009) and the International Monetary Fund (2012, 2013) have suggested that uncertainty about US and Europeanfiscal, regulatory andmonetary policies have contributed to a steep decline in 2008–۲۰۰۹٫ Furthermore, many authors, such as Baker et al. (2015) have also suggested that the high levels of policy uncertainty are behind the weak recoveries after the 2007 financial crisis. The economic literature points to different channels through which uncertainty might negatively affect economic growth. Considering the demand side of the economy, in a highly uncertain environment, firms will reduce investment demand and delay projects (Bernanke, 1983; McDonald and Siegel, 1986; Dixit and Pindyck, 1994), while households will reduce their consumption of durable goods (Carroll, 1996). On the other hand, on the supply side, firms’ hiring plans will be also negatively affected by high uncertainty levels (Bloom, 2009). Policy uncertainty is also believed to have these potential effects on different macroeconomic variables (Friedman, 1968; Pastor and Veronesi, 2012, 2013; Fernández-Villaverde et al., 2015). Among the different measures of policy uncertainty,the economic policy uncertainty index based on newspaper coverage frequency proposed by Baker et al.(2015) has become a benchmark1 for measuring economic policy uncertainty (Sum, 2012a, 2012b; Antonakakis et al., 2013, 2016; Gulen and Ion, 2015).2 Fig. A2 in the Appendix A plots the EPU indices of the countries considered in the paper. The historical evolution of the US EPU index, for example, shows that policy uncertainty sharply increased after several events, such as Black Monday’s stock market fall in 1987, the 9/11 attack and the 2nd Gulf War. According to this index, the highest policy uncertainty levels correspond to the 2011 debt-ceiling dispute. It is impossible to deny the international influence of the U.S. economy as an exporter of international uncertainty spillover effects (Klößner and Sekkel, 2014; Yin and Han, 2014), which in turn, justifies the analysis of the impact of this variable on international stock market returns. When the EPU indexes for Australia, Canada, China, Japan and Korea are considered, the data reveal that these indexes reached their peaks in 2011, coinciding with different national events, coupled with high international political uncertainty due to the Eurozone fears. An exception is the Japanese case where the index reached its highest level in 2010, with the Bank of Japan’s monetary easing. In Canada, although the index spikes in 1995, with the Quebec Referendum and in 2008 with the collapse of Lehman Brothers, it also reaches its highest level in 2011, as it does in Australia (Baker et al., 2015). In Korea, the main spikes in the indexes correspond to the enforcement of the real-name financial transactions law’ in August 1993, under Kim regime and the death of Il-Sung Kim in July 1994. Other episodes with high EPU indexes coincide with the bankruptcy of Daewoo Motors in 2000, the beginning of Roh regime and the disaster at a subway station in Daeggo in 2002, and the global financial crisis initiated by the collapse of Lehman Brothers. Again, the index reached its peak in 2011 with the serial bankruptcy of savings bank and the death of Jung-II Kim (Choi and Shim, 2016). In China, the index spikes with the township and village enterprises bankruptcy in 1995–۱۹۹۶, privatization and restructuring in 1997–۲۰۰۰, accession to World Trade Organization in 2001, global financial crisis in 2008–۲۰۰۹, and euro crisis in 2010. The Chinese index also reaches its peak when Xi-Li Administration began with legislation aimed at corruption and poverty in 2011 (Kang and Ratti, 2015). These high levels of policy uncertainty are considered as one of the key differences of the on-going recovery from previous recoveries, by some authors such as Baker et al. (2015) and attests to the severity of the recent crisis (Yin and Han, 2014). The impact of policy uncertainty, on stock market returns, has been already studied in the literature, primarily based on time-series approaches which involve single-country vector autoregressive (VAR) models. Even if multiple countries are analysed, individual VAR models and at times quantile regressions are used in the process, ignoring the role of dependence between markets in a globalized world. However, although the results seem to suggest that uncertainty negatively impact stock returns (Pastor and Veronesi, 2012; Antonakakis et al., 2013; Kang and Ratti, 2013, 2015; Chuliá et al., 2015; Chen et al., 2016), the results are far from conclusive. Given the dominance of the U.S. economy, Sum (2012a,b) examines the existence of international uncertainty spillovers and finds that US EPU shocks are non-significant in the stock returns in China, Brazil and India, while Momin and Masih (2015) find the same results when analysing their impact on the BRICS countries. Li et al. (2015) examine the causal link between US economic policy uncertainty and stock returns in India and China and they do not find evidence of causality between the two variables. Furthermore, Mensi et al. (2014, 2016) and Balcilar et al. (forthcoming) examine how an increase in US policy uncertainty could positively affect international stock markets, since it could lead to an improvement in foreign stock markets through the diversification channel of investor portfolios. In this context, the objective of this paper is to analyze the effect of economic policy uncertainty on stock market returns in a sample of six Pacific-rim countries, which include Australia, Canada, China, Japan, South Korea and the US, using monthly data from January 1998 to December 2014, by means of estimating a restricted PVAR, estimated using the Stochastic Search Specification Selection (SSSS) prior (PVAR-SSSS) of Koop and Korobilis (2016). We use monthly data as the EPU indices, of the countries in our sample, are only available at monthly frequency. While the choice of the US economy is natural, given its global influence on other financial markets, the decision to look at the Pacific-rim countries is driven by the increased transmission of stock market return among these markets over recent periods (Balcilar et al., 2015). The main contributions of the paper are as follows: First,the PVARmethodology applied in this paper rather than the pure time-series based approaches, is an excellent way to examine internationaltransmission of different shocks, allowing for cross-sectional dependence, given strong evidence of stock market return spillovers, interdependence and financial contagion, already found in the literature (Bekaert et al., 2009, 2014; Arouri et al., 2011; Diebold and Yilmaz, 2015), which as we show exists statistically as well. So, by using a panel approach, since we combine both time series and cross-sectional elements of the data set, we gain in efficiency over time series models (Rapach and Zhou, 2013), and by using non-homogenous coefficients, it allows us to obtain impulse responses for each of the six countries separately, rather than an average impulse response obtained under standard panel data approaches. The cost of overparameterization due to the usage of heterogeneous coefficients in the PVAR, is in turn solved using the Bayesian methods proposed by Koop and Korobilis (2016). Second, and in order to account for international uncertainty spillovers, we not only analyze the impact of the own country’s EPU shocks, but also the U.S. EPU shocks on the various stock markets. Furthermore, the sign and persistence of these spillovers based on impulse response functions will help us understand the mechanism through which international uncertainty spillovers affect stock market returns and for how long. To the best of our knowledge, this is the first paper to use impulse responses in a heterogeneous coefficient Bayesian PVAR model to analyze the impact of own and US EPU shocks on stock returns of Pacific-rim countries.3 Hence, we extend this literature on stock market and EPU, primarily based on time-series approaches, which fail to account for cross-sectional dependence in international stock markets and thus, could be leading to inaccurate inferences. In addition, our study also deviates from the existing time series work, which primarily looks at G7 or BRICS countries. The remainder of the paper is structured as follows. Section 2 discusses the methodology used in the paper. Section 3 describes the data and shows the empirical analysis. Section 4 summarizes the main findings.