Advanced

Prognostisering av aktieavkastningar med hjälp av makroekonomiska variabler - en svensk studie

Simonsson, Louise and Hääger, Therése (2006)
Department of Economics
Abstract
Forecasting stock market returns is an interesting topic since more and more Swedes enter and invest in this market. Theory implies, however, that such exercises should be impossible. The aim of this thesis is to investigate the possibility to forecast future stock returns by looking at macroeconomic variables’ history. The study is limited to the Swedish market as it is based on the OMXS30-index which represents the 30 most exchanged stocks on Stockholms¬börsen, the Swedish stock exchange. This study uses four macroeconomic variables: Swedish unemployment, interest rate, industrial production and inflation, all available on a monthly basis. We evaluate forecasts over 2002-2005, although our study makes use of all historical observations... (More)
Forecasting stock market returns is an interesting topic since more and more Swedes enter and invest in this market. Theory implies, however, that such exercises should be impossible. The aim of this thesis is to investigate the possibility to forecast future stock returns by looking at macroeconomic variables’ history. The study is limited to the Swedish market as it is based on the OMXS30-index which represents the 30 most exchanged stocks on Stockholms¬börsen, the Swedish stock exchange. This study uses four macroeconomic variables: Swedish unemployment, interest rate, industrial production and inflation, all available on a monthly basis. We evaluate forecasts over 2002-2005, although our study makes use of all historical observations back to 1995. Forecast models containing macroeconomic variables are compared and contrasted to two common benchmark models for forecasting returns. The thesis concludes, as theory suggests, that macroeconomic variables are not useful for forecasting the evolution of stock prices. The best performing model is a random walk model which predicts that the return in each period is zero. The second best forecast model is an AR(1)-specification for returns that excludes macroeconomic variables altogether. The result could be explained by the stable stock market in Sweden during the investigated period of time. (Less)
Please use this url to cite or link to this publication:
@misc{1335629,
  abstract     = {Forecasting stock market returns is an interesting topic since more and more Swedes enter and invest in this market. Theory implies, however, that such exercises should be impossible. The aim of this thesis is to investigate the possibility to forecast future stock returns by looking at macroeconomic variables’ history. The study is limited to the Swedish market as it is based on the OMXS30-index which represents the 30 most exchanged stocks on Stockholms¬börsen, the Swedish stock exchange. This study uses four macroeconomic variables: Swedish unemployment, interest rate, industrial production and inflation, all available on a monthly basis. We evaluate forecasts over 2002-2005, although our study makes use of all historical observations back to 1995. Forecast models containing macroeconomic variables are compared and contrasted to two common benchmark models for forecasting returns. The thesis concludes, as theory suggests, that macroeconomic variables are not useful for forecasting the evolution of stock prices. The best performing model is a random walk model which predicts that the return in each period is zero. The second best forecast model is an AR(1)-specification for returns that excludes macroeconomic variables altogether. The result could be explained by the stable stock market in Sweden during the investigated period of time.},
  author       = {Simonsson, Louise and Hääger, Therése},
  keyword      = {aktieavkastning,prognostisering,Random Walk,makroekonomisk variabel,Economics, econometrics, economic theory, economic systems, economic policy,Nationalekonomi, ekonometri, ekonomisk teori, ekonomiska system, ekonomisk politik},
  language     = {swe},
  note         = {Student Paper},
  title        = {Prognostisering av aktieavkastningar med hjälp av makroekonomiska variabler - en svensk studie},
  year         = {2006},
}