As regulators continue their push/pull with moving collateralized trading business onto CCPs (they’re good for risk management, but wait, they’re also the next too-big-to-fail institutions!), the Bank of International Settlements has released a working paper looking at how CCP margin policies impact the repo markets. We expect this will be an ongoing and interesting form of research going forward in securities finance when and if more transactional flow heads in the CCP direction.
The BIS paper, “The impact of CCPs’ margin policies on repo markets,” by Arianna Miglietta, Cristina Picillo and Mario Pietrunti, “quantifies the impact on the cost of funding in repo markets of the initial margins applied by central clearing counterparties (CCPs).” Margin, along with the clearing fees that CCPs charge, has been argued as a reason for using or not using CCPs over the years. The latest ICMA European repo market survey (data from June 2015) showed CCP cleared-trades in Europe at about 27%, barely changed from December 2014’s figures.
Generally speaking, CCPs are not always viewed as positive by the repo market but regulatory pressures may win out in increasing their use. According to the just-released ICMA’s analysis, “The potentially negative impact of the shift in business on CCP-cleared business (which is largely also electronic and is limited to government bond collateral) may have been mitigated by banks’ appreciation of the regulatory benefits of multilateral netting and a lower regulatory counterparty risk weight, as well as by the use of post-trade reporting of transactions executed directly or via voice-brokers.”
The BIS paper authors looked at Italy’s repo market and found increasing CCP volumes. “While unsecured trades have dropped off over time, stabilising at very low levels, the collateralised ones have steadily increased, now accounting for the largest share of money market transactions.” Average daily volumes on MTS were about EUR 84 billion in 2014, up from EU 82 billion in 2013. This is 40 times the volume of e-MID, a non-CCP electronic market. There is no mention of directly negotiated transactional volume. The paper concludes that funding costs for participants in CCP-cleared repo varies by 3-4 bps for every 100 bps change in CCP margin. Its a small amount but it is quantifiable.
This paper is not perfect. We’re not sure that the argument for the robustness of the Italian CCP-cleared repo market fully holds relative to bilateral volume. There is also the tricky issue of in case of a crisis, that LCH.Clearnet would move to cash settlement for Italian repo, ie, no CCP guarantee. This complicates an analysis purely on Italian repo as the market may skew towards factoring in the loss of the guarantee. The authors use the Eurepo rate published by the European Money Markets Institute as a benchmark for spread evaluation; we’re not sure that this is robust enough.
All that said, this paper is an important first step in quantitatively evaluating the financial impacts of CCPs in the funding markets. We expect more to come.