Borders and Nominal Exchange Rates in Risk-Sharing

Models of risk-sharing predict that relative consumption growth rates across locations should be positively related to real exchange rate growth rates across the same areas. We investigate this hypothesis using a new multi-country and multi-regional data set. Within countries, we \u85nd evidence for risk-sharing: episodes of high relative regional consumption growth are associated with regional real exchange rate depreciation. Across countries however, the association is reversed: relative consumption and real exchange rates are negatively correlated. We de\u85ne this reversal as a bordere¤ect and show that it accounts for 53 percent of the deviations from full risk-sharing. Since cross-border real exchange rates involve di¤erent currencies, it is natural to ask how much of the border e¤ect is accounted for by movements in exchange rates? We \u85nd that over one-third of the border e¤ect is due to nominal exchange rate uctuations. We develop a simple open economy model that is consistent with the importance of nominal exchange rate variability in accounting for deviations from cross-country risk-sharing