Abstract: This thesis investigates whether stock returns can help forecast macroeconomic activity. Future earnings and dividends and current stock prices should contain information about the future state of firms and the consumption possibilities of consumers. These activities are linked to aggregate economic development and, hence, the stock markets should improve economic forecasting. We review the theoretical points that justify the importance of stock markets in economic forecasting. Recent literature on the stochastic discount factor in asset pricing and the real business cycle models has approached this connection. We try to show that the direction between financial markets and macroeconomy could be from stock markets to real economy. We empirically test the forecasting ability of stock markets with respect to macroeconomy. The unexpected part of stock return can be revealed with economic tracking portfolios (ETP), which are constructed so that the unexpected portion of the portfolio return has the maximum correlation with revisions to expectations of the target variable. ETP’s track how investors revise their expectations about relevant macroeconomic variables. The results show that specific stock portfolios track future changes in macroeconomic variables well. In the previous literature, stock returns have been connected to the business cycle. This connection is analysed by explaining stock returns with total factor productivity (TFP) as a factor. TFP is measured by corporate innovation variable, i.e. the change in a firm’s gross profit margin unexplained by changes in firm’s capital and labour. The TFP variable performs quite nicely in explaining stock returns and it can be related to stock market momentum. Next, the aim is to investigate the forecasting power of stock returns together with the TFP factor. Even though in our results the TFP contains no information relevant for economic forecasting, the stock returns continue to perform well.
Processing Request
No Comments.