Banks’exposure to large-scale asset purchases, as measured by the relative prevalence of mortgage-backed securities on their books, affects lending following unconventional monetary policy shocks. Using a difference-in-differences identification strategy, this paper finds strong effects of the first and third round of quantitative easing (QE1 and QE3) on credit. Highly affected commercial banks increase lending by 2% to 3% relative to their counterparts. QE2 had no significant impact, consistent with its exclusive focus on Treasuries sparsely held by banks. Overall, banks respond heterogeneously, and the type of asset being targeted is central to QE.

Banks’ exposure to large-scale asset purchases, as measured by the relative prevalence of mortgage-backed securities on their books, affects lending following unconventional monetary policy shocks. Using a difference-in-differences identification strategy, this paper finds strong effects of the first and third round of quantitative easing (QE1 and QE3) on credit. Highly affected commercial banks increase lending by 2% to 3% relative to their counterparts. QE2 had no significant impact, consistent with its exclusive focus on Treasuries sparsely held by banks. Overall, banks respond heterogeneously, and the type of asset being targeted is central to QE. Received January 13, 2016; editorial decision January 18, 2017 by Editor Philip Strahan.

Abstract Do firms respond to cost shocks by reducing the quality of their products? Using microdata from a large Russian retailer that refreshes its product line twice-yearly, we document that higher quality products are more profitable than lower quality ones, but that the number of high-quality products experiences a relative decrease after a large ruble devaluation in 2014. We show that rising firm costs—and not shrinking consumer incomes—explains the reallocation, and rationalize the data with a model that features consumer expenditure switching between high and low qualities. The reallocation to lower quality products reduces average pass-through by 26%.