Research

This paper explores how high inflation affects household savings, aggregate capital, and informality. Using a large panel of countries, I find that high inflation correlates with high levels of informality and low levels of financial development, even when accounting for GDP per capita. To explain this, I extend the Aiyagari model with an informal sector subject to a cash-in-advance constraint. In a tractable deterministic version, I show that the total effect of inflation on informality decomposes into a direct headwind, as inflation taxes cash and discourages informal consumption, and a general equilibrium tailwind, as inflation compresses savings, raises the interest rate, and through a Stolper-Samuelson mechanism increases the relative cost of capital-intensive formal production. The net effect is governed by a single sufficient statistic, the capital gap between formal and informal sectors, and a theorem identifies the exact conditions under which the tailwind dominates: the formal sector must be sufficiently more capital-intensive, but not so much that savings capacity is exhausted. A fully-fledged heterogeneous-agent version with idiosyncratic uncertainty confirms the mechanism quantitatively. Calibrated to an average of Latin American economies, the model shows that increasing steady-state annual inflation from 10% to 90% is linked to a decline in the steady-state aggregate capital by 8% and a growth in informality as a share of GDP, from 40% to 54%.

This paper focuses on a Central Bank that seeks to profit from trading against irrational noisy traders but cannot directly observe their positions as they blend in with rational traders that are going against the noise themselves. The optimal strategy for the Central Bank is to trade against the remaining noise partially, but this results in an inefficiently small position size. With complete information, targeting the actual noise is optimal for higher expected profits. If the Central Bank had a prior on the net open position limit of the rational traders, it would target more than the remaining observable noise. I prove that the Central Bank can learn to differentiate actual noise from rational traders by fixing the prior for many periods, computing the sample variance of the observable noise, and comparing it with the actual noise variance. Finally, if rational traders have better information than the Central Bank about the future real exchange rate, intervening in the market may result in negative expected profits, so total inaction may be recommendable.

I document that the increase in prices and not in quantities are driving the rise in the value of exports of goods in Argentina in this century. I also show the country’s poor performance in capital formation, the low value of its market capitalization, and the tendency of consumers to save and firms to borrow in foreign currency. These three forces could help explain why Argentina is not exporting more. To understand the mechanisms, I build a model in which financial constraints and external debt will keep the economy closer relative to the same economy unconstrained. Moreover, this constrained economy allows for multiple equilibria, which could open room for optimal policy intervention. The critical channel driving the results is the low capital creation  arising from high real exchange rates and the balance-sheet effect, affecting future consumers’ wages through a decrease in the installed capital and its price, which forces the economy to appreciate in the future to compensate for that drop in income and keep the balance of payments in equilibrium. An economy “trapped” in a bad pace will initially export more due to high real exchange rates. Still, it will eventually be forced out of the international markets due to insufficient capital stock.