DP8765 Internalization, Clearing and Settlement, and Liquidity
|Author(s):||Hans Degryse, Mark Van Achter, Gunther Wuyts|
|Publication Date:||January 2012|
|Keyword(s):||anonymity, clearing and settlement, internalization, liquidity, transaction fees|
|Programme Areas:||Financial Economics|
|Link to this Page:||cepr.org/active/publications/discussion_papers/dp.php?dpno=8765|
We study the relation between liquidity in financial markets and post-trading fees (i.e. clearing and settlement fees). The clearing and settlement agent (CSD) faces different marginal costs for different types of transactions. Costs are lower for an internalized transaction, i.e. when buyer and seller originate from the same broker. We study two fee structures that the CSD applies to cover its costs. The first is a uniform fee on all trades (internalized and non-internalized) such that the CSD breaks even on average. Traders then maximize trading rates and higher post-trading fees increase observed liquidity in the market. The second fee structure features a CSD breaking even by charging the internalized and non-internalized trades their respective marginal cost. In this case, traders face the following trade-off: address all possible counterparties at the expense of considerable post-trading fees, or enjoy lower post-trading fees by targeting own-broker counterparties only. This difference in post-trading fees drives traders strategies and thus liquidity. Furthermore, across the two fee structures, we find that observed liquidity may differ from cum-fee liquidity (which encompasses the post-trading fees). With trade-specific fees, the cum-fee spread depends on the interacting counterparties. Next, regulators can improve welfare by imposing a particular fee structure. The optimal fee structure hinges on the magnitude of the post-trading costs. Noteworthy, a fee structure yielding higher social welfare may in fact reduce observed liquidity. Finally, we consider a number of extensions including market power for the CSD, anonymous trading and differences in broker size.