Brian
Amorim Cabaço

I study the macroeconomic effects of merger waves in the United States since 1980. Firm-level event studies show that merger announcements are associated with positive abnormal returns for acquirers and their competitors, but not for targets' competitors. This pattern is difficult to reconcile with industry-wide signaling alone and is consistent with anticompetitive effects. I aggregate the estimated abnormal returns into a proxy for merger activity and find that increases in the proxy are followed by persistent declines in GDP and labor productivity and a modest rise in inflation. To interpret these patterns, I develop a firm-dynamics model with an auction-based market for mergers and variable markups. The model matches merger-pairing patterns, firm sales concentration, and the markup distribution. Counterfactual simulations imply sizable welfare losses from mergers through higher markups, lower allocative efficiency, and lower firm entry.

We assemble a narrative record of all federal antitrust case filings by the DoJ and FTC since the mid-1950s. We document a shift from civil non-merger and criminal enforcement toward civil merger cases in the early 1980s. Non-merger and criminal filings reduce defendants' markups while increasing sales and input expenditure. We construct sector-level enforcement proxies from publicly listed defendants' stock-market reactions to filings. Increased enforcement strengthens competition, raises output, R&D capital and patenting, and improves allocative efficiency. For civil non-merger and criminal cases, output gains are accompanied by greater input use. For civil merger cases, output rises without sustained increases in inputs, consistent with a role for productivity and resource allocation.

We identify four dimensions of monetary policy in the euro area by applying narrative restrictions to high-frequency data. Well-documented episodes—including Mario Draghi's "whatever it takes" speech—allow us to distinguish conventional-policy, forward-guidance, quantitative-easing, and asymmetric-country-risk-premia shocks using one narrative dominance restriction per shock. After accounting for predictable high-frequency asset movements and state-dependent variances in a Bayesian factor model, we find limited evidence of economically meaningful information shocks. We then use the identified shocks to estimate the aggregate effects of different monetary policy instruments. A Bayesian VAR with distributed lags and stochastic volatility permits overidentifying restrictions on the impulse responses. Contractionary conventional-policy, forward-guidance, and quantitative-easing shocks reduce activity and inflation. Adverse asymmetric-country-risk-premia shocks reduce activity while raising inflation modestly, albeit imprecisely. Forward guidance has small effects on economic activity; asset purchases have larger effects, but both are less potent than conventional policy.

I provide the first semi-structural estimates of exchange rate effects on macroeconomic aggregates in an emerging market using a granular instrumental variable (Gabaix and Koijen, 2024) constructed from financial fund equity flows into South Africa. I find that appreciations against the USD are inflationary and cause expansions in output and credit—responses resembling credit shocks rather than traditional trade effects. I document a powerful exchange rate credit channel: in countries with substantial foreign-denominated debt, local currency appreciations reduce debt burdens, spurring higher credit. These findings reveal that the USD matters both as the trade invoicing currency and the funding currency for debt. USD-denominated liabilities constrain domestic monetary policy autonomy in emerging markets while extending the reach of the US Federal Reserve, underscoring the need for international central bank cooperation.