Hillary Stein

Abstract: We leverage supervisory micro data to uncover the role of global banks' risk limits in driving exchange rate dynamics. Consistent with a model of currency intermediation under risk constraints, shocks to dealers' risk limits lead to price and quantity adjustment in the foreign exchange market. We show that dealers adjust their net exposure and increase spread in response to granularly identified limit shocks, leading to lower turnover and an adjustment in currency returns. These shocks exacerbate effects of net currency demand on exchange rate movements, as predicted by theory, and trigger deviations from covered and uncovered interest parity.

Abstract: I examine the effect of exogenous terms of trade shocks on an exchange rate by turning to New Zealand’s dairy auctions. Dairy is New Zealand's largest export category, making up almost 20 percent of exports. Specifically, whole milk powder accounts for 6 to 11 percent of total exports, and its price is determined in twice-monthly auctions. I use event studies to quantify the impact of surprise auction results on the New Zealand dollar on a high-frequency basis. I find that a 1 percent increase in whole milk powder prices has a modest, but nevertheless significant, effect on the nominal exchange rate that does not seem to be explained by interest rate movements. Rather, the effect seems to be driven by a combination of two channels: a financial flows channel and a direct terms of trade channel. The methodology developed here can potentially be applied to other commodity exporters.

Abstract: Even though financial risk management can generate value, the use of financial derivatives among nonfinancial corporations remains limited. We identify a channel that contributes to this limited use:  the decoupling of derivatives losses and operational gains. We explore this phenomenon using a unique dataset built from the universe of US dollar-Mexican peso currency derivatives transactions in Mexico along with customs data. Using a regression kink design, we find that when losses from previous transactions increase 1 percentage point, firms become 4.24 percentage points less likely to take out a new derivatives position within 90 days.