| This dissertation investigates the cross-market impact of U.S. equity order flow on exchange rates. Order flow is defined as the difference between buyer-initiated and seller-initiated trades. To compute it, I use a unique sample of intraday trading data for over fifteen years. My main findings can be summarized as follows: (i) there is a highly significant correlation between U.S. stock order flow and exchange rate changes; (ii) for most currencies the correlation is positive, i.e. an increase in order flow leads to an appreciation of the U.S. dollar; (iii) for commodity countries", namely Australia, Canada, New Zealand and South Africa, the correlation is negative; (iv) the link is stronger at higher frequencies and during more volatile periods. As robustness checks, I also look at the correlation for shorter subsamples and by industry. Results are not affected, suggesting that the correlation is overall robust across sectors and time periods.;I argue that, for non-commodity countries, the effect of order flow is consistent with the portfolio rebalancing channel emphasized by Hau and Rey (2004). An increase in order flow is associated with an increase in purchases of U.S. stocks by foreigners. The rise in foreign purchases raises the demand for U.S. dollars. As a result the U.S. dollar appreciates.;Empirically, an increase in order flow is not associated with an increase in the net purchase of U.S. stocks by agents from commodity countries. Moreover, an increase in order flow is associated with an increase in commodity prices. I interpret this positive correlation as reflecting the rise in the demand for commodities on the part of U.S. agents. The rise in demand leads to an appreciation of the commodity currencies. |
