Research

Job market paper

Does unconventional monetary policy affect the market power of firms? I address this question both empirically and theoretically. Comparing firms purchased under the European Central Bank’s Corporate Sector Purchase Programme with eligible but unpurchased ones in a staggered difference-in-differences design, I show that a purchase raises a firm’s markups by 7.1 percent. The transmission channel runs from issuing conditions to capital: purchased bonds are issued at a 22 basis point lower yield to maturity, and purchased firms invest 2.2 percentage points more. These effects materialise only under two conditions: purchases must be repeated over the years, and they must happen within a week of the bond issuance. A general equilibrium model rationalises these findings, combining endogenous markups and financial frictions. In a segmented bond market, a purchase raises the price of the very issue it absorbs, relaxes the issuer’s capital constraint, and thus raises its output, market share and markup. A quantitative version compares different purchase designs: several, including the ECB’s own with its deviations from market neutrality, deliver almost the same welfare gain, while concentrating purchases on a few large firms or directing them to the smallest eligible ones reduces it.

We develop a general-equilibrium model of loan insurance and bank capital, emphasizing the wedge between the capital insurance releases and the risk it transfers. Priced at the loss it transfers, protection is bought for the capital relief. Because bank lending creates deposits, the price level moves with credit, and the real value of the wages workers hold as money moves against it. The model explains why euro-area banks insure loans whose capital charge is high relative to their risk and become riskier relative to their capital. Quantitatively, insurance barely moves credit or prices but thins bank capital, and a tighter requirement leaks into the insurance market. After a shock the size of the global financial crisis, the price of protection rises as bank capital becomes scarce. When banks cannot raise all the capital they need, lending contracts and the burden shifts from bank owners to workers, whose wages it finances.

In a monetary union without a fiscal union, each government pays its own interest rate while its bonds are held across borders. Spread differentials that arise from financial frictions rather than default risk therefore redistribute resources between the households of different member states, a channel distinct from the level of interest rates. We document the interest flows that spread differentials generate across euro-area countries, using bilateral holdings imputed from the ECB’s Securities Holdings Statistics. We then build a two-country New Keynesian model in which financial frictions generate these differentials. After a common financial shock, the spread of the more indebted country rises more, and resources flow from its households to those of the other country. Quantitative easing allocated by the capital key barely narrows the differential and leaves most of this redistribution in place; purchases targeted at the high-spread country close the differential and largely offset it.