“…More importantly, the specific model is now found to contain significant predictive ability both in and out-of-sample. 8 Given that stock prices serve as a leading indicator and, hence, carries useful information for policy makers as to where the economy might be heading, future research would aim to investigate not only in-sample, but also out-of-sample predictability of real stock returns 9 based on a wider set of financial and macroeconomic variables (Choudhry, 2004;Chancharoenchai et al, 2005;Rapach et al, , 2010aRapach and Wohar, 2006;Ng, 2007, 2009, forthcoming;Carvalhal and de Melo Mendes, 2008;Goyal andWelch, 2008, Cakmakli andvan Dijk, 2010) by extracting factors to serve as explanatory variables in predictive regression models or even based on Bayesian vector autoregressive models, with both these approaches capable of handling huge data sets involving hundreds of variables. In addition, one might also want to delve into multifractal (Balcilar, 2003), long memory models (Franses and van Dijk, 2000;Balcilar, 2004) and even non-linear models 10 (Qi, 1999;McMillan, 2001) to capture stock return movements.…”