Monetary Policy, Time Variation of Equity Returns, and Regime Switching

Abstract

The thesis studies time variation of the cross-sectional stock returns. The aim of the study is to identify and quantify economic factors behind the common variation of the cross-sectional returns both theoretically and empirically. Identifying and quantifying economic factors behind the common variation of stock prices is important. Common variation in stock prices affects the measure of systematic risk that rational investors use to evaluate stocks and financial wealth in the real economy. The study examines especially monetary policy’s heterogeneous effects on the firm level and further on the portfolio level.

The earlier studies have mainly focused on the unconditional linear econometric frameworks. However, the presence of regimes in the macroeconomy infers a regime switching structure in equity returns’ means, volatilities, autocorrelations, and cross-covariances. Thus, globally linear models may lead to an efficiency loss in the estimation process. This study thereby studies portfolios’ returns behavior by using both the conditional linear econometric framework and the nonlinear econometric framework. The nonlinear regime switching framework is constructed by using the Hamilton’s (1989) Markov regime switching framework and by calibrating the model to allow two regimes. The research data consists of U.S. stock market stocks from January 1990 to December 2020 sorted on book-to-market equity portfolios. The independent variables consist of monetary, business cycle, credit market, and investor sentiment variables. The key contribution of this thesis is to measure the impact of unconventional monetary policy actions by using Wu’s and Xia’s (2016) shadow rate. The shadow rate measures the monetary policy actions in the zero lower bound environment when the nominal interest rates are restricted by the zero lower bound.

The empirical results show that the research portfolios’ returns are time varying and regime dependent. Market volatility can be considered to be one factor affecting the regime switches. The portfolios’ beta-parameters behave asymmetrically between the portfolios and the states. The finding is coherent with the economic theories behind the behavior of portfolios’ returns. The nonlinear Markov switching econometric framework offers findings of the research portfolios’ parameter estimates that the simple linear econometric framework is unable to detect. The nonlinear econometric framework’s results suggest that monetary policy regimes have statistically and economically significant effect on the portfolios’ returns and the value premium (difference between high and low book-to-market ratio portfolios’ returns) in the research sample. The result is coherent with the economic role of the monetary policy and with the earlier research findings. Monetary policy regimes can be connected asymmetrically to expectations of firms’ cash flows and/or discount rates applied by investors. The finding of returns regime dependency provides support to the nonlinear econometric framework and highlights the importance of conditioning information in evaluation of expected returns.