Skip to content

Navigating the Minefield of EMIR Segregation

Introduction

Here is a link to an interesting article in Risk Magazine dealing with the issue of segregation under EMIR.

Article 39 of EMIR (“Segregation and Porting”) requires a central counterparty (“CCP”) to keep separate records and accounts that will enable it to distinguish the assets and positions of:

  • one clearing member (“CM”) from those of any other CM;
  • a CM from those of its clients (“omnibus client segregation”); and
  • a client of a CM from any other client of that CM (“individual client segregation”).

The Problem of Choice

EMIR is clear that every CM (and anyone offering indirect clearing services[1]) must offer its clients at least a choice between omnibus client segregation and individual client segregation, but that this list is non-exhaustive.  In reality, this has given rise to a number of different approaches to segregation among CCPs which it seems has done as much to muddy the water as provide clarity in the area of central clearing.  The article notes that Swapclear currently offers both individual and omnibus segregation, and plans to roll out two further segregation models in the second half of this year.  Eurex offers full segregation, net-margined omnibus segregation, and a futures-style elementary clearing model, with a fourth US-style “legally segregated, operationally commingled” (“LSOC”) model planned for later this year.  CME Clearing Europe provides a physically segregated account, an unsegregated omnibus account as well as an LSOC-model, with a hybrid ‘LSOC-with-excess-posted-to-CCP’ model planned.  ICE also plans to offer the options of full or omnibus segregation as well as an LSOC-based solution.

Further Hurdles to a Solution

As the article notes, in its basic form, full segregation involves the transfer of margin between three parties – the client, CM and CCP.  In order to mitigate the obvious counterparty risk, the viability of quad-party segregation is also being assessed.  Under this structure, a custodian or central securities depository would maintain legal title to margin assets at all times, with beneficial entitlement being tracked by way entries on an updated register.  However, whilst it has obvious advantages, there are also a number of challenges which this model must overcome.  Of most concern at this point appears to be the requirement under EMIR that margin and default fund contributions must “where available, be deposited with operators of securities settlement systems that ensure the full protection of those financial instruments.”[2]  Unfortunately, recent ESMA guidance on EMIR[3] states that the depositing of financial instruments with an operator of a securities settlement system (“SSS”) via a custodian does not constitute a deposit with an operator of a SSS for the purposes of EMIR.  Instead, such a structure would instead amount to a deposit with an authorised financial institution, an approach which is permitted under EMIR, but which ESMA has stated is only permissible if the relevant CCP is able to demonstrate that it cannot access an SSS.  Accordingly, concerns remain that the quad-segregation model may simply be non-compliant from an EMIR perspective.

The Pros and Cons of Segregation Models

Broadly, whilst benefitting from a higher degree of protection, individual client segregation will typically also involve higher cost.  This is mainly driven by the fact that the full segregation of a client’s assets will require that client to post margin in support of cleared trades on a gross basis, without the benefit of netting.  In addition, CMs may be required to pre-fund this gross margin requirement at a CCP before actually receiving margin from the client.  In contrast, whilst the omnibus client segregation model is theoretically less protective of underlying assets, collateral can be commingled with that of a clearing member’s other clients.  As a result, a CM will be able to post margin required in relation to an account on a net basis, resulting in cost savings which can be passed on to clients.

EMIR also requires CCPs to commit to attempt to port transactions of underlying clients affected by the default of a CM.[4]  However, there are practical problems associated with porting in times of market stress.  Two obvious problems include the fact that replacement CMs are not obliged to accept positions (and are less likely to do so in relation to omnibus portfolios of clients than in relation to an individual client), as well as the fact that the consent of all clients within an omnibus account is required in order to port.  As the article notes, less obvious but equally important is the fact that, under LSOC, CMs cannot guarantee to return the same assets as were originally posted by a client due to the fact that assets are commingled and posted on a net basis to a CCP.  Accordingly, clients should get back the value of the underlying assets, but not necessarily the assets themselves.  This may not be acceptable to some clients which would face the prospect of rebuilding underlying portfolios.

The pros and cons of various types of segregation model can be summarised in the table below:

Segregation Model 

Pros

Cons

Individual   Client Segregation
  • Highest   level of protection
  • Facilitates   Porting
  • Return of assets guaranteed
  • Expensive
  • Loss   of netting benefit
LSOC
  • High   level of protection
  • Easier   to administer
  • Return of assets not guaranteed
  • Porting more difficult
Omnibus   Client Segregation
  • Cheaper alternative
  • Easier to administer
  • Lower level of protection
  • Porting more difficult

 

 

 


[1] EMIR, Article 4(3)

[2] EMIR, Article 47(3)

[3] ESMA/2013/324

[4] EMIR, Article 48(5)

Contact Us