Harvard University Job Market Candidate Papers

• Demand Shocks, Transportation Mode and the Distance Elasticity of Trade Flows, this version: May 2008 (first version: November 2007).
  Abstract: There is considerable variation in the distance elasticity of trade flows across sectors. This paper argues that the relative prevalence of air transportation vs. slower transportation, and the premium paid for it, also vary across sectors and can account for a large share of this variation in distance elasticities.
  Our model integrates demand uncertainty and choice of transportation mode to explain varying prevalence rates based on the distribution of demand uncertainty in each sector. The basic firmlevel trade-off in choosing transportation mode consists of weighing the extra cost of fast transportation against the gained mitigation of uncertainty risk: shorter lags between production decisions and sales allow for more precise forecasts and better matches between quantities produced and sold. So firms facing higher demand uncertainty have higher incentives to choose air transportation. And firms beyond a certain level of uncertainty (the exporting threshold) do not export at all.
  This sorting pattern and its drivers have three main implications for trade patterns at the sector level. First, in sectors where transportation costs are more sensitive to distance, trade flows are also more sensitive to distance. Second, in sectors with more firms choosing slow transportation, trade flows are less elastic to distance. Third, in sectors with less dispersion in the distribution of uncertainty, trade flows respond more to distance because the share of exporters responds more to induced movements in the exporting threshold. We also derive implications for the determinants of the prevalence of each transportation mode. We find strong support for our predictions: overall the factors we highlight explain over 60% of the variation in distance elasticities across sectors.



• Corruption, Delays and the Pattern of Trade, with Quoc-Anh Do, May 2008.
  Abstract: We argue that corruption deters international trade by causing delays in exporting and importing, both at customs and in other required administrative procedures. We study three manifestations of corruption as a barrier to trade. The corruption effect is both significant and economically sizeable. We first show the negative relationship between the exporter’s and importer’s levels of corruption and trade volumes at the country level in a gravity framework. This country-level effect implies that a standard deviation increase in the exporter’s corruption level causes a 27% drop in exports. We then show that corruption indeed operates through delays: we establish that this effect stronger in sectors in which goods are more time-sensitive. The magnitude of this interaction effect is large: a standard deviation increase in the exporter’s corruption level causes a decrease in exports ranging from 7% in the least time-sensitive sector to 42% in the most time-sensitive sector. Finally, we find that corruption also decreases more the probability of positive trade in sectors in which goods are more time-sensitive. We use unpredictability of sales as our measure of time-sensitivity. Our results are robust both to controlling for a variety of alternative explanations and to instrumenting corruption to alleviate concerns of endogeneity.



• Global Currency Hedging, with John Y. Campbell and Luis Viceira, May 2007.
  Abstract: This paper considers the risk management problem of an investor who holds a diversified portfolio of global equities or bonds and chooses long or short positions in currencies to manage the risk of the total portfolio. Over the period 1975-2005, we find that a risk-minimizing global equity investor should short the Australian dollar, Canadian dollar, Japanese yen, and British pound but should hold long positions in the US dollar, the euro, and the Swiss franc. The resulting currency position tends to rise in value when equity markets fall. This strategy works well for investment horizons of one month to one year. In the past 15 years the risk-minimizing demand for the dollar appears to have weakened slightly, while demands for the euro and Swiss franc have strengthened. These changes may reflect the growing role for the euro as a reserve currency in the international financial system. The risk-minimizing currency strategy for a global bond investor is close to a full currency hedge, with a modest long position in the US dollar. Risk-reducing currencies have had lower average returns during our sample period, but the difference in average returns is smaller than would be implied by the global CAPM given the historical equity premium.