Using Stock Returns to Identify Government Spending Shocks : New Insights

The existing instruments of government spending using accumulated stock returns of military contractors generate vastly different consumption and investment impulse responses when compared to the narratively identified war news shocks as per Ramey (2011). We show that a reason for this difference is because of the persistence in the accumulated stock returns. Instead, a return spread between diversified portfolios of defense firms minus private consumption and investment-good firms (DMP) renders persistence and generates responses akin to war news shocks. DMP return spread is a relevant instrument for post-1963 period and the spread Granger-causes shocks identified using standard VAR approach