We evaluate the enforcement of information barriers—China Walls—within conglomerates. Our setting is the 25 million trades in 2018–2024 in the Israeli Shekel market, where the SEC tightened the China Walls around dealers from July 2018. Our difference-in-differences design compares the trade volumes and profits of funds that are affiliated with, clients of, or entirely unrelated to a dealer around the days when the dealer is likely to hold valuable information. Before July 2018, dealers extensively share information with both their clients and affiliate funds, and price discovery is rapid. After July 2018: (i) dealers cease sharing information with their affiliate funds; (ii) they continue sharing with their clients unabated; (iii) price discovery slows; (iv) the trading costs paid by affiliate funds decline; and (v) the trading costs paid by other funds rise. A back-of-the-envelope calculation shows that the China Walls eliminated 90.1% of profits by the affiliate funds on event days.

88% of foreign exchange transactions involve the US dollar. An important reason for this is its use as a vehicle currency: Traders often route trades through the dollar to access better liquidity. We develop a novel methodology to identify dollar cross-trades using high-frequency data from the two primary interdealer platforms. We show theoretically and empirically that cross-trades drive substantial price variation in dollar exchange rates. Using an instrumental variable approach, we find that increased cross-trading causes higher aggregate dollar volatility. This reveals a trade-off: While dollar dominance enhances liquidity, it also exposes the dollar to uncertainty from other currencies.

Using granular transaction-level data, this paper investigates the characteristics and implications of dealer-client trading relationships in the over-the-counter FX derivatives market. We first document that dealer-client trading relationships are persistent over time. Then, to shed light on the role of relationship strength for client access to these instruments during times of dealer stress, we examine the collapse of Credit Suisse in March 2023. Our analysis reveals that clients with greater exposure to Credit Suisse experienced a larger increase in spreads at the client level relative to unexposed clients by about 16 basis points per notional dollar traded on average across maturities, although their trading activity remained unchanged. The greater spread increases paid by clients who relied more heavily on Credit Suisse occurred through their trades with non-Credit Suisse dealers. While more exposed clients continued to trade with Credit Suisse in the post-period, less exposed clients reduced their trading activity with Credit Suisse, but increased their trading activity elsewhere, indicating an ability to substitute counterparties. These findings underscore the critical role of search and bargaining frictions in this market, particularly when a relationship dealer encounters adverse shocks.