Essays on Trade under Uncertainty

Abstract

This dissertation consists of three chapters focused on the role of uncertainty in trade. The first chapter studies trade policy uncertainty in an environment with multiple stages of production. The recent increase in trade policy uncertainty affects a variety of industries, and in particular, uncertainty is important for industries whose final good is produced in multiple stages that are located across different countries. These industries are the most concerned about trade policy and reduce investment during uncertainty periods. This paper analyzes trade policy uncertainty in a two-country dynamic, stochastic, general equilibrium model with multistage production where a firm?s decisions today depends on the future tariff path. Studies with one stage of production that measure the effect of trade policy find that uncertainty, as a second moment shock, does not play a big role in explaining the changes observed in the economy. Introducing multistage production, which generates a magnified response of trade to tariff changes, provides a better mechanism to analyze the role uncertainty in future tariffs plays in the economy. In the second chapter, I revisit the trend of U.S. manufacturing inventories and the role of delivery times. U.S. manufacturing inventories have been increasing since 2005, reversing a declining trend that lasted for decades. The rise is observed across U.S. manufacturing industries and types of inventories. While the long term decline is well-understood as a consequence of improvements in transportation and information technology, the reversal of the trend has not yet been studied. This paper explores the role of increasing delivery times due to the creation of global supply chains. As foreign inputs become cheaper, firms choose to source more inputs from abroad, and in particular inputs from China which face long delivery times and frequent delays. This increases the firms' exposure to volatility in demand leading to a greater incentive to hold inventories. I build a dynamic trade model that features stochastic delivery times for different inputs in the presence of idiosyncratic demand risk. In this framework, firms face a tradeoff when sourcing inputs from different locations between their relative price and delivery times. I find that the initial decrease in delivery times explains $61\%$ of the decline in inventories from 1992 to 2004, and the increase in reliance on inputs from China, which face longer and more volatile delivery times, explains $34\%$ of the increase in inventories from 2005 to 2018. In the third chapter, I explore the cost of supply chain disruptions and examine the role of inventories as insurance. Supply chain disruptions have become increasingly common in today's integrated world. As the risk of disruptions increases, so does the value of inventories, as firms rely on their inventories to smooth and insure their production process. In this paper, I build a model to study supply chain disruption in an environment with demand risk and inventory storage. Then, I use the model to study the recent supply chain disruptions cause by COVID-19. I allow for specially long delivery delays of inputs, and sudden rise of sales, and track a firm's choices from 2020 onwards. With this model, I am able to quantify the effect on prices, analyze the use of inventories to insure and smooth out production during the period, and study its effect on sourcing choices and imported inputs substitutability during this period of long delivery times. Finally, I study the effects of the increasing and more volatile delivery times in firms sourcing choices. The rise in delivery times increases the cost of sourcing from farther countries, and could force firms to change or diversify their current supply chain networks.