Model Shows Network Density Affects Derivatives Trade Costs

Views and opinions expressed are those of the authors and do not necessarily represent official positions or policies of the OFR or Treasury.

In the updated OFR Working Paper, Intermediation Networks and Derivative Market Liquidity: Evidence from CDS Markets, OFR’s Associate Director for Financial Institutions Mark Paddrik and OFR Research Principal Stathis Tompaidis examine how intermediation networks and dealer exposures affect the liquidity of the over-the-counter (OTC) derivative markets. Using a model and empirical tests, the authors determine that the density of the intermediation network (1) has a significant influence on the liquidity provided by dealers and (2) affects the cost of trade differentially in the dealer-to-client and interdealer segments. The results highlight the need for policymakers to consider how regulations lead to changes in counterparty relationships.

OTC markets are decentralized. Buying and selling parties in an OTC market trade securities, commodities, currencies, and other derivative products directly with each other, without using a central exchange. Dealers in an OTC market act as market makers by quoting prices at which they will buy and sell a security. Thus, dealers intermediate risk and provide liquidity through holding and managing exposures.

The authors examine how intermediation networks and dealer exposures affect the liquidity of one of the OTC derivative markets—specifically, the market for single-name credit default swaps (CDS). The authors introduce a model that links OTC intermediary relationships and derivative market liquidity provision. They empirically examine the U.S. single-name CDS market by using supervisory data on two market segments: interdealer and dealer-to-client.

The authors obtain the following notable findings:

1. A dealer’s willingness to provide liquidity, in terms of taking increased inventory onto its balance sheet, is positively associated with how well connected the dealer is to its clients and other dealers.
2. A dealer’s execution costs are primarily driven by its transactions with clients, while the dealer’s bid-ask spreads are primarily driven by its ability to intermediate trade with other dealers, rather than with clients.
3. A dealer’s interdealer execution cost declines as the proportion of relationships it maintains with clients increases—however, perhaps surprisingly, this execution cost is not related to the dealer’s relationships with other dealers.
4. The bid-ask spread a dealer receives with its clients declines as the completeness of its interdealer network increases, while the dealer’s interdealer bid-ask spread is not related to its interdealer network.

The paper’s model predictions highlight that intermediation network density significantly influences the liquidity provided by dealers, both on an individual and a collective basis, as seen through trade volumes, inventory management, and transaction costs. These results help shed light on the importance of trading relationships in maintaining market liquidity. The authors empirically validate these predictions and highlight several shifts in dealer behavior during the periods of study in which several regulations took effect, as interdealer trade and dealer participation declined and inventory management tightened. All these shifts are consistent with a decline in market liquidity.