Counterparty Risk Minimization by the Optimal Netting of OTC Derivative Trades

• Published in: The Journal of derivatives Vol. 24; no. 2; pp. 48 - 65
• Main Author: O’Kane, Dominic
• Format: Journal Article
• Language: English
• Published: New York Pageant Media 01.12.2016
• Subjects: Agreements; Algorithms; Collateral; Default; Derivatives; Financial institutions; Interest rate swaps; Linear programming; Optimization; Over the counter trading; Risk assessment; Securities markets; Simulation; Studies
• ISSN: 1074-1240; 2168-8524
• DOI: 10.3905/jod.2016.24.2.048

The OTC derivatives market has always suffered from the fact that there are generally many contracts among market intermediaries where neither counterparty is an end-user. Active OTC market makers may have thousands of very similar contracts with one another. To the extent that each one needs to be monitored and margined, substantial improvements in efficiency may be possible through "compression" trades in which sets of existing contracts are replaced by a smaller number of essentially equivalent new ones among the same counterparties. But how best to minimize counterparty risk exposures without mismatching terms? In this article, O'Kane explains how compression is currently done for interest rate swaps and reviews both linear and quadratic programming approaches. Using simulation, he explores how the possible degree of compression is affected by the number of trades and the number of counterparties involved. The linear compression algorithm appears to allow more full unwinds of existing contracts than the quadratic algorithm, which is a feature highly desired by the firms in the industry.