Central Clearing of Interest Rate Swaps: A Comparison of Offerings
Imperial College London; CNRS - Universite de Paris VI
Radu Paul Mondescu
DRW Trading Group
DRW Trading Group
March 11, 2011
Regulatory changes have motivated the development of a variety of solutions for the clearing of interest rate swaps. Margin payments associated with clearing lead to modifications in cash flows which result in differences in the valuation between cleared and non-cleared swaps. We propose a framework for computing these differences and show that they lead to two types of modifications in contract value: a convexity effect and a “Net Present Value” (NPV) effect, which can be significant for long-dated swaps. As a result, modifications in contract design are required in order for a centrally cleared interest rate swap to be economically equivalent to its uncleared counterpart. Among the currently available offerings for cleared interest rate swaps, three offerings are shown to be economically equivalent to their uncleared counterparts – the “Price Alignment Interest” method used by LCH. Clearnet and CME, as well as a new adjustment method used by the Eris Exchange – while a fourth method, used in the IDCG swap futures contract, is shown to lead to substantial deviations in valuation with respect to a non-cleared interest rate swap. Using a Hull-White model calibrated to the market data as of December 2010, we find the difference between the IDCG futures swap rate and the corresponding uncleared swap rate to be around 18 basis points for a 10 year contract and about 60 basis points for a 30 year contract. An interest rate environment with higher volatility will result in larger differences.
Number of Pages in PDF File: 25
Keywords: interest rate swap, central clearing, convexity, swap futures, price alignment interest, net present value effect
JEL Classification: G13, G20working papers series
Date posted: March 14, 2011
© 2015 Social Science Electronic Publishing, Inc. All Rights Reserved.
This page was processed by apollo6 in 0.328 seconds