The dynamics of stock market returns and macroeconomic indicators: An ARDL approach with cointegration

Abstract: Macroeconomic indicators are amongst the most important and used tools for investors as they provide an outlook for the economy and thus improve the assessment of investments e.g. for asset allocation. The purpose of this thesis is to investigate the short- and long-run relationship between the US stock market index S&P500 and six selected macroeconomic indicators during different time regimes during 2000-2016. The chosen indicators are Personal spending, Initial jobless claims, M1 Money supply, Building permits, Michigan Consumers Sentiment index and the ISM Manufacturing index as they measure different parts of the economy and are commonly used by investors. We achieve the purpose by using the Autoregressive Distributed Lags model (ARDL) as it has several advantages in relation to comparable time series models. The results show that all indicators except Personal spending are significant in the long-run on the 1-percent level, in at least one time-regime. All indicators have significant results also in the short-run except the Money Supply (M1), depending on which time period that is under investigation. Our conclusion is that our chosen indicators have different characteristics depending on the current dynamics of the stock market, economic state and other related markets. The practical implication for investors is that different indicators are of limited use depending on the current market dynamics and investors must evaluate the underlying premises of the development of the indicator rather than interpreting a specific datapoint.