I develop an open economy model with variable markups to analyze the effect of commodity booms on misallocation. Commodity booms create a real exchange rate appreciation that reallocates resources across firms within the tradable sector. This reallocation occurs for two reasons. First, there is tougher foreign competition on domestic producers (competition channel), which pushes large firms to reduce their markups and small firms to exit. Second, a reduction in the relative cost of imported materials (cost channel) induces large firms, who use them more intensively, to increase their markups. I calibrate the model to Chile and replicate the increase in the price of its main export product (copper) during the early 2000s. I find that there is a substantial reallocation within the tradable sector, which decreases misallocation. Both channels matter quantitatively. Without heterogeneity in the share of imported materials, the decrease in misallocation would be halved. Furthermore, markup dispersion falls more in industries that experience a higher increase in foreign competition. Finally, I estimate markups using Chilean firm-level data and show that the broad patterns are qualitatively consistent with the model predictions.

Commodity prices are widely viewed as a major source of business-cycle fluctuations in emerging-market economies (EMEs). This view is largely based on representative agent models, where agents behave according to the permanent income hypothesis (PIH). However, recent empirical evidence points to large deviations from PIH (Bracco et al., 2021) . Do these deviations from PIH behavior affect the importance of commodity price shocks for EME’s business cycles? To answer this question I first show analytically that deviations from PIH can matter via two channels. On the one hand, hand-to-mouth agents (i.e. those who violate PIH) amplify the income effect of commodity price shocks in the short-run and dampen it in the long-run. On the other hand, they dampen the indirect interest-rate effect of these shocks. I then estimate a structural model to quantitatively explore the effect of PIH deviations on the importance of commodity price shocks as business-cycle drivers. The model is a standard small open economy model with two agents, and is estimated to match data from Brazil, Chile, and Colombia. Using variance decompositions I show that hand-to-mouth agents increase the importance of commodity price shocks for output fluctuations, and can increase (decrease) the importance of commodity price shocks for consumption in the short-run (long-run). Finally, I quantify the importance of two mechanisms in generating these results. First, the indirect interest rate effect is a relevant transmission channel for commodity price shocks, but it does not appear to be dampened by hand-to-mouth agents. Second, wealth effects on labor supply account for most of the amplification of hand-to-mouth agents on income and consumption.

Recent literature has highlighted that differences across countries in the share of Hand-to-Mouth households (HtM) are important for the transmission of aggregate shocks as well as monetary and fiscal policy. How can we explain cross-country differences in the share of hand-to-mouth households? In this paper, we first document significant heterogeneity in the share of HtM households across European countries. This heterogeneity is driven by large differences in the share of wealthy HtM households, who hold illiquid but no liquid wealth. On the contrary, the share of poor HtM, who hold neither liquid or illiquid wealth, is similar across countries. Second, we develop a two-asset life-cycle model with incomplete markets and uninsurable income risk and calibrate it to Spain. Through the lens of the model we study the role country differences in income risk, the life-cycle profile of earnings and retirement benefits. Although we report substantial differences across countries in mean income, risk and retirement benefits, results suggest that they cannot explain cross-country differences in the share of HtM by themselves. We further gauge the potential of other drivers, such as differences in financial frictions, preferences, demographics and unexplained initial conditions, from the lens of our calibrated model. Among them, differences in financial frictions are the most promising avenue for future research.