Did the Federal Reserves' Quantitative Easing (QE) in the aftermath of the financial crisis have macroeconomic effects? To answer this question, we estimate a large-scale DSGE model over the sample from 1998 until 2020, including data of the Fed's balance sheet. We allow for QE to affect the economy via multiple channels that arise from several financial frictions. Our nonlinear Bayesian likelihood approach fully accounts for the zero lower bound on nominal interest rates. We find that QE increased output by about 1.2 percent, reflecting a net increase in investment of nearly 9 percent accompanied by a 0.7 percent drop in aggregate consumption. Both government bond and capital asset purchases effectively improved financing conditions. Especially capital asset purchases significantly facilitated new investment and increased the production capacity. Against the backdrop of a fall in consumption, supply side effects dominated, leading to a disinflationary effect of about 0.25 percent annually.
Keywords: Quantitative Easing, Liquidity Facilities, Zero Lower Bound, Nonlinear Bayesian Estimation. JEL: E63, C63, E58, E32, C62Working paper no. 691