Central Counterparts, as part of their Daily Margin Process with their clearing members, are enhancing their Initial Margin computation methodology. Historically based on a SPAN model or equivalent Risk aggregation models, the Clearing Houses are moving progressively, in addition to Liquidity and Wrong Way Risk add-ons, to Portfolio-based Initial Margin indicators (Value at Risk or Expected Shortfall); mostly privileging Historical Simulation. This global shift is accompanied by a de-standardization of computation methodologies, every CCP promoting its own HS margin model.
This coincides well with the other requirements of clearing financial institutions and has significant implications for them.
Need for anticipation: it’s been clearly identified by the ESMA, relayed by CCPS and expressed by the clearers themselves
Although the new margin frameworks, driven by regulatory constraints (EMIR Regulation, PFMI Principles, are supposed to ensure the limitation of their pro-cyclical effects, participants cannot afford to discard the implications of intraday variations of their positions and induced margins, especially in periods of high market volatility as this past spring, during the commencement of the COVID-19 crisis.
In order to ensure a close monitoring of volatile exposure, as well as reduce the cut-off pressure for financing the end-of-day margin calls, increasing intraday computation frequency provides risk departments with key-information.
Need for consolidation: Central counterparts will at best provide computation and simulations services limited to their own activity perimeters and cannot consolidate the exposures of Non – Clearing members for their cross-market business. On top of that, non-cleared OTC operations and re-delivery need to be considered as well.
Need for optimization: the fragmentation of sources of information, which can lead at some point to manual, semi-automated, desktop macro-based process, constitutes in turn a limit to both scalability and optimization of security collaterals. The latter, in a context of lastingly low interest rates, are to be favoured over cash, but, less fungible and less flexible, need more elaborated valuation / haircutting processes.
As a consequence, an efficient risk monitoring process needs to deal with a wide variety of challenges, of computation centralization, and high integration capabilities of exogenous data.
In addition, the more and more frequent use of Historical Simulation models implies elaborated qualification and cleansing procedures for time series.
Adding to this the increasing computation frequencies, the answer can only be a robust solution, with high processing capabilities (CPU/GPU, data storage).
SLIB has developed an expertise for more than 10 years in the following areas:
- Algorithm creation and use of external computation modules, including client’s proprietary algorithms.
- Data integration and cleansing, particularly events management on data series (Corporate Actions, missing prices).
- Ability to aggregate modular or multi-market portfolios.
- STP and centralized margin call management, leading to operational risk mitigation.
- A technology enabling to manage high data volumes, while ensuring high frequency risk computation, with real-time back up facilities (mirroring).
SLIB RMS, launched by SLIB in 2019, answers the needs of broker dealers, individual clearing members, and global clearers operating on various geographical areas.
SLIB RMS benefits are:
- A powerful and scalable solution, enabling clearers to manage a high volume of activity so as to face volatility peaks (example: COVID -19 crisis in March 2020) and numerous portfolio simulations.
- A proprietary algorithm implementing historical V@r methodology.
- The replication of the main European CCPs algorithms enabling contradictory, high frequency computation, and End-of Day anticipation.
- A reactive solution, allowing a real time monitoring (intraday margin up to 5-minutes frequency) and a real-time alert system.
- A modular and centralising solution to provide a unified vision of risks, by consolidating data from other back- office systems, private risk computation systems, and CCPs.
