Real Economic Effects of the Federal Reserve's Quantitative Easing Program: A Structural VAR Approach

Abstract

As the Federal Funds rate approached the zero lower bound and macroeconomic indicators continued to show little signs of recovery, the United States Federal Reserve implemented an unprecedented policy of large-scale asset purchases known as Quantitative Easing. We explore the real economic effects of these policies using a novel approach to applying sign and zero restrictions in a Structural Vector Autoregression on post-2009 U.S. data, building on the framework of the SVAR model in Schenkelberg and Watzka (2013) for the Japanese economy. These identifying restrictions are drawn in general from the existing literature on monetary policy transmission and structural economic disturbances when benchmark interest rates are at near-zero levels. We find that a QE shock leads to statistically significant increases in excess reserves, industrial production (after a lag of 10-12 months) and the price level. The magnitude of these effects reaches similar proportional peak levels to comparable studies, though does not persist significantly for the full estimation period. Interestingly, we also find that the marginal real economic effects of QE2 and QE3 were comparable to the first program, despite QE1's proximity to the financial market rebound associated with the special liquidity facilities opened in 2007 and 2008. Analysis of variance and shock decompositions suggests that our identification approach is robust with regards to separating quantitative easing shocks from exogenous disturbances to aggregate supply and demand derived from business cycle theory.