Identifying key macroeconomic shocks to Canadian GDP

This paper seeks to identify the largest two shocks that can explain the movement in Canadian GDP for the period 1981Q1 to 2011Q4. I employ a very flexible identification method proposed by Uhlig (2003) that allows us to identify the key shocks from the time series data without imposing any strict identification assumption. The largest two shocks are extracted by maximizing the forecast error variance of GDP for a ten years horizon. Two shocks are sufficient to explain most of the variation in the GDP in Canada. My findings suggest that TFP news shock is the key driver of GDP in the medium run and it creates significant positive co-movements among the aggregate variables at business cycle frequencies. Demand shock dominates in the short run, however, its hard to pin down the exact source of the shock. The findings are robust to alternative SVAR identification strategy and variable specification.