How does corporate debt denominated in foreign currency impact international pricing decisions in response to exchange rate fluctuations, as measured by the exchange rate pass-through? We explore this question with a unique dataset that combines firm-level foreign currency debt with disaggregated international trade flows for Colombia. Our findings reveal that the exchange rate pass-through to import prices is greater than one for indebted industries. Conversely, the exchange rate pass-through to export prices is incomplete regardless of debt levels. To rationalize these findings, we propose a theoretical mechanism featuring pricing complementarities and foreign currency debt limited by occasionally binding borrowing constraints. When borrowing constraints bind following an exchange rate depreciation, domestic firms raise prices due to increased marginal costs. Pricing complementarities prompt foreign firms to pass on exchange rate movements to a greater extent, leading to higher pass-through to import prices. Pricing complementarities also induce domestic firms to absorb the increase in marginal costs and exchange rate fluctuations when setting export prices. Our findings suggest that pricing complementarities ultimately determine the exchange rate pass-through, even when firms set prices in a vehicle currency, such as the US dollar.

Why did the US dollar depreciate during the height of the COVID-19 pandemic despite its safe-haven role? We answer this question empirically and theoretically by uncovering a causal relationship between variation in COVID-19 cases and US dollar bilateral exchange rates. Utilizing rich spatial and temporal variation in a daily-frequency panel dataset, we find that a one standard deviation increase in US COVID-19 cases depreciated the US dollar by 1.23 percent on average after one month. To rationalize this finding, we introduce COVID-19 as a mixture of supply and demand shocks in a two-country, multi-sector open-economy model with incomplete markets, portfolio adjustment costs, and imperfect labor substitutability. The model quantitatively matches the empirical findings by calibrating the model with data on COVID-19 cases in the US and the rest of the world.

This paper sheds novel light on how government spending shocks affect firm investment. Using the narrative military spending news shock to identify exogenous variation in government spending (Ramey, 2011b; Ramey and Zubairy, 2018), we find that increases in government spending cause capital expenditures of publicly-listed firms to increase by up to one percentage point on average. The investment response of the average firm is not driven by the set of firms directly affected by the government spending news. Instead, we show empirically that government spending leads to a persistent decline in long-term real interest rates. Firms respond to falling costs of capital by issuing more debt and increasing corporate investment.

This paper uses firm-level data to test the so-called markup channel of uncertainty shocks. To that end, we measure markups based on a production function approach using quarterly data from US Compustat. We identify uncertainty shocks from macroeconomic data using a Bayesian VAR with Stochastic Volatility. Contrary to the predictions of DSGE models with nominal price rigidities, we find that markups are procyclical to uncertainty disturbances. Specifically, a one-standard-deviation shock to uncertainty decreases markups by approximately 0.30 percentage points, and the effect is long-lived, lasting up to three years for most uncertainty measures. We conducted several sensitivity checks and found that conditional procyclicality of uncertainty holds. Our results are consistent with previous findings in the literature that rely on aggregate measures of markups and alternative identification strategies of shocks to uncertainty.

We study how fiscal rules affect fiscal multipliers. While the empirical literature has focused solely on the effects that fiscal rules have exerted on local governments budgetary discipline, our results show that both the presence and the type of fiscal rule in place are important factors determining the effectiveness of government spending increases, as measured by the fiscal multiplier. Importantly, these empirical findings are not driven by correlated measures of fiscal fragility, such as the level of development or the level of debt across countries. Additionally, the paper also investigates whether the presence of fiscal rules affects the dynamics of long-term government bond yields (10-year maturity). The results indicate that the presence of fiscal rules attenuate bond-yields increase in response to the fiscal shocks. We interpret both findings as stemming from different expectations regarding fiscal discipline in the near future. Since any fiscal rule targeting directly the level of debt or the budget balance affects expectations regarding the discounted flow of future real primary results, the presence of fiscal rules determines the resulting state-dependent fiscal multipliers.

In this paper, we introduce in an otherwise standard neoclassical growth model the required time lag for public investment to be turned into public capital (time-to-build process) and distortionary taxes responding to the level of public debt. Our goal is to quantitatively isolate the macroeconomic effects of the increase in public investment observed in the Brazilian government investment package "Growth Acceleration Program" (PAC in the Portuguese acronym). Depending on the implementation delays associated with the time-to-build of public capital and the stringency of fiscal adjustment, the PAC could have led to a GDP decline between 0.2% and 0.4% for up to four years.

Applying the methodology proposed by Auerbach and Gorodnichenko (2012) to monthly GDP data and recurrent central government revenues and expenditures, fiscal multiplier estimates are virtually zero in situations of fiscal fragility, regardless of the indicator used to measure the soundness of public accounts. If we use the total general government deficit—our preferred measure, the cumulative multiplier for the one-year period is negative at -0.1. In periods of fiscal soundness, however, the output responses to government spending can be substantial, as the estimated fiscal multiplier reaches 2.5 for the same period, while, in the linear model, the estimates are 0.4 at impact and 0.7 at the peak of the trajectory.

In this paper, we document estimates of the Brazilian fiscal multiplier, finding hump-shaped responses and point estimates of 0.6 on impact with a peak of 0.9 in the quarter following the shock. On the other hand, the tax multiplier is negative and reaches -0.2 and -0.6 for the same periods. Moreover, while the transfer multiplier is 0.7 at its peak, the government investment multiplier is substantially higher, reaching 2.2 under the same metric. However, we take the latter estimate with caution due to anticipation effects associated with the time-to-build lags of public capital. Finally, in contrast to the conventional predictions in the neoclassical and new Keynesian literature, we also document that household consumption and private investment crowd-in in response to government spending shocks with corresponding fiscal multipliers of 0.6 and 1.8, respectively. Overall, the results indicate that the Brazilian fiscal multiplier is relatively large for an emerging economy but still lower than comparable estimates in advanced countries. One potential explanation for our results is Brazil's relatively closed economy.