Tracking down the business cycle: A dynamic factor model for Germany 1820-1913

We use a Bayesian dynamic factor model in order to calculate an economic activity index for Germany prior to World War I. The procedure allows us to incorporate information from a vast number of time series, which are underutilized by historical national accounts. Therefore, our indicator provides an alternative measure for economic activity, based on a broader database. To investigate industrialization, we compare our aggregate measure of economic activity with sectoral activity indices. We find that the industrial transition was completed earlier than agricultural output and employment shares suggest, since the indicator for agriculture had already decoupled from the aggregate business cycle measure during the 1860s. Moreover, we find that stock prices are strongly correlated with our indicator, and lead it by 1-2 years.