Economic uncertainty, asset returns and the real economy

Numerous studies have estimated the effect of economic uncertainty on investment, but found no conclusive evidence. Using a novel measure of economic uncertainty in a theoretically motivated macroeconometric model could guide the debate on the effect of economic uncertainty on investment. The strands of literature analysing the effect of economic uncertainty on asset prices and international transmission of global economic uncertainty shocks are still in infancy. This thesis contributes to the literature by examining the effects of economic uncertainty on investment and asset returns for individual countries. In addition, it investigates macroeconomic linkages among countries with specific attention to the transmission of economic uncertainty shocks across countries. The thesis comprises three interrelated empirical studies. The first study examines the effect of economic uncertainty on fixed investment for 26 countries for the 1990-Q1 to 2014-Q3 period. A neoclassical investment model augmented with economic uncertainty is estimated. It is argued that economic uncertainty increase leads to delayed or postponed fixed investment because of its irreversible nature, which is also known as the wait-and-see channel. Using the autoregressive distributed lag model for estimation, the study finds that in the long run, economic uncertainty exerts a negative effect on investment for developed countries (the United States, South Korea, Japan, New Zealand, Malaysia, Hong Kong Special Administrative Region and Spain), whereas it exerts a positive effect for some developing countries (Brazil, Chile, Italy, Colombia and Indonesia). A negative effect in developed countries may be due to policy uncertainty, while a positive effect in developing countries is explained by buffer stock creation by working more, anticipating high economic uncertainty or long investment lags. Output, measured in terms of real gross domestic product (GDP), persistently has a positive effect on fixed investment across the countries and models. Effects of real interest rate and financial friction on fixed investment are mixed. The key policy recommendation is that developing an active second-hand market for fixed assets can lower the degree of irreversibility for fixed investment, and consequently reduce effects of economic uncertainty on investment. The second study examines the differential effects of common uncertainty and idiosyncratic uncertainty on returns on three asset classes: stocks, bonds and real estate. Idiosyncratic uncertainty refers to disagreement among economic agents, while common uncertainty refers to perceived variability of aggregate macroeconomic shocks. Using the international capital asset pricing model augmented by economic uncertainty, the study demonstrates that a one standard deviation increase in economic uncertainty lowers stock returns, bond returns and real estate returns by 3.2%, 3.56% and 0.27%, respectively. Idiosyncratic uncertainty has a more pronounced effect on asset returns than does common uncertainty. This finding is consistent with the notion that under idiosyncratic uncertainty, which reflects policy uncertainty, forward-looking investors require an extra risk premium. The significance of idiosyncratic uncertainty calls for ensuring transparent, well-communicated policies to reduce asymmetric information, and prescribing rule-based fiscal and monetary policies. In the third empirical study, international transmission of economic uncertainty shocks across countries is examined. The study adopts the global vector autoregressive (GVAR) modelling framework of M.H. Pesaran, Schuermann and Weiner (2004) and Dees, di Mauro, Pesaran and Smith (2007). To the best of author’s knowledge, this study is the first to examine global uncertainty shocks in the GVAR framework. The model includes five domestic variables (real GDP, inflation, real equity price, real exchange rate and short interest rate), five foreign counterparts of domestic variables (trade-weighted domestic variables) and two global variables (oil price and global uncertainty). Global uncertainty shocks affect real GDP, inflation, real exchange rate, real equity price and short-term interest rate; however, the effects are not statistically significant. Evidence of synchronised business cycles and policy synchronisation is found through contemporaneous effects of foreign variables on domestic variables. The US financial market explains a large amount of global uncertainty shocks but China is insulated from global uncertainty shocks, consistent with the decoupling hypothesis. A key policy implication is that economies should coordinate policies with their trading partners and diversify trade from the United States and European Union to China or any other country that is less affected by global shocks.