Can the stock markets predict changes in macroeconomic variables?
|Author:||Bc. Marek Vařeka|
|Year:||2017 - summer|
|Leaders:|| prof. PhDr. Ladislav Krištoufek Ph.D.
|Work type:|| Bachelors
|Awards and prizes:|
|Abstract:||There is a consensus in the literature, that the stock market can predict
the Gross domestic product on quarterly base or the industrial production,
which is good proxy for GDP, on monthly basis and that the causal relationship
between stock market and GDP should work both ways. However,
using Vector autoregression model on US data since 1950, model shows that
the stock market can not only predict the Industrial production on monthly
basis, but also ISM non-manufacturing index, which is a good proxy for
services in the economy. Furthermore I have managed to prove, that the
unemployment can be predicted by past realizations of the stock market
and managed to explain almost one third of all variations in change in unemployment
using S&P500 and oil prices during last 20 years. The Granger
causality test concluded that stock market does cause the unemployment
but not vice versa, at least during last 20 years.