The Essential Components of The Risk Management Framework for CCPs

By Dale Michaels, EVP, Financial Risk Management, OCC

Dale Michaels, EVP, Financial Risk Management, OCC

Central counterparties (CCPs) like OCC have performed extraordinarily well during many stressful periods due to innovations such as mark-to-market settlements, initial margin models, and default management processes. CCPs add a critical risk management function to the financial system as was evident during and after the 2008 global financial crisis.

As the only CCP for all U.S. exchange-listed options trades, OCC manages risk for those involved in options trading by being the buyer to every seller and the seller to every buyer. This instills confidence in the financial markets and the broader economy.

CCP processes are very transparent with the adherence to the Principles for Financial Market Infrastructures (PFMIs) and the distribution of qualitative and quantitative information, so market users can better understand the overall risk management of CCPs, which includes membership standards, initial margin requirements, a clearing fund, and sound default management processes.

Clearing Members

OCC strives for a broad, diversified clearing membership while maintaining rigorous membership standards. In the U.S., clearing members are either a registered broker-dealer or a futures commission merchant, and a corporate entity.

Each CCP monitors the credit risk of its clearing members by reviewing financial statements and market metrics. OCC also performs a risk review of our members at initial admission to membership and periodically after that to ensure they meet acceptable risk management standards.

Initial Margin

Initial margin covers the potential future exposure between the last margin collection and the closeout of a portfolio in the case of a defaulting member. CCPs' initial margin models are distinct from one another as they reflect differences in the products cleared and their inherent risks. Much of the work of CCPs is to calibrate and review initial margin models as conditions and products evolve. OCC developed a 10,000 scenario Monte-Carlo initial margin calculation methodology called STANS (System for Theoretical Analysis and Numerical Simulations). The STANS model is based on expected shortfall at a 99% confidence level, which means it includes all observations, including worst-case scenarios, and averages those amounts from 99 to 100 percent.

CCP processes are very transparent and the overall risk management of CCPs includes membership standards, initial margin requirements, a clearing fund, and sound default management processes

OCC also employs a two-day margin period of risk for its initial margin model, which exceeds the regulatory standards for exchange-traded derivatives of a one-day margin period of risk.

Margin models are not calibrated to cover 100 percent of the risks. CCPs utilize robust back-testing to assess the adequacy of its margin calibrations.

The margin period of risk (i.e., the estimated time needed to close out a defaulting counterparty's accounts), is another critical determination in calibrating appropriate margins. We believe the margin period of risk should be tied to the default management process. While the U.S. regulatory minimum margin period of risk for exchange-traded derivatives is one day, OCC believes that two days is more reflective of the practical timeframe needed to close out a defaulting counterparty's accounts, given the default management experience. OCC, therefore, sets the margin period of risk at the more conservative two days for its exchange-traded derivative products.

Many CCPs offer margin offsets for products that are economically and intuitively linked. These correlations must also be persistent and strong, and margin offsets should not be allowed for products that are tangentially correlated or in different asset classes. OCC runs de-correlation scenarios where there is an additional charge at a higher confidence level based on the greater of historical, zero, or perfect correlations of products to cover the risk of markets moving in a much different way than in the past.

Other considerations when reviewing margin models include daily calibrations, the length of lookback periods, liquidity add-ons, and concentration add-ons, wrong-way risk, and intraday margining capabilities, to name a few other important items.

Clearing Fund

Many of the considerations CCPs consider in establishing initial margin also apply to stress testing, as CCPs size their clearing funds at appropriate levels. This includes using several test scenarios, de-correlation scenarios, and long lookback periods. The regulatory minimum in the U.S. is "cover 1", meaning the CCP can cover the exposure of its largest clearing firm. In September 2018, OCC implemented a "cover 2" standard covering our two largest clearing firm exposures.

CCPs look to bring many similar products into clearing fund and default waterfall to allow them to be risk-managed together. A default waterfall refers to the financial safeguards available to a CCP to cover losses arising from a clearing member default and the order in which they would be expended. While a clearing member may not clear every product cleared by the CCP, it is important from a risk management perspective to have broad participation in a CCP clearing fund rather than small siloed funds for single products. If there is no recourse to access to other financial resources beyond the small siloed funds, a CCP default is more likely, which would be much more damaging to the customer, clearing firms, and the financial system.

Default Management

CCPs work closely with clearing member firms and other CCPs to design and test robust default management processes. This coordination includes vigorous testing of the auction process that includes participation from clearing members and clients. Of importance in the default management process and any stressed market environment is the CCPs’ ability to retain the flexibility to react to the facts and circumstances at the time, as each stress event typically looks very different from past events.