Managing the procyclical effects of margin requirements at HKEX
Our principles and industry best practices, protect the integrity of the Hong Kong financial markets, and promote the efficiency of our markets in terms of funding, capital and liquidity efforts were founded on three principles: ensure consistency with international risk management.
Hong Kong Exchanges and Clearing will soon launch a new margin methodology as part of our Next Generation Post Trade Program. The methodology will be more sophisticated than the current methodology and is designed specifically to provide more efficient and accurate representation of portfolio risk exposure.
While there are many benefits to a more risk based methodology, it is increasingly important, if not fundamental, to assess broader systemic risk considerations that have been set out in regulatory policy initiatives–notably, for example, measures to mitigate procyclicality. We also recognize and appreciate that the industry is increasingly sensitive to margin funding costs and capital requirements to support clearing. Both of these themes motivated very insightful and engaging discussions with our participants as we progressed through the design phase of our new margin methodology.
The Next Gen risk management team at HKEX, led by Ketan Patel, the deputy group risk officer, played a central role in the development of the new margin methodology. Our efforts were founded on three principles: ensure consistency with international risk management principles and industry best practices, protect the integrity of the Hong Kong financial markets, and promote the efficiency of our markets in terms of funding, capital and liquidity.
The initial design phase took place over six months, focusing on how to best calibrate a value-at-risk (“VaR”) based model for the products we clear as well as the characteristics of our participants and the Hong Kong financial market. The next phase was focused primarily on engagement with the various types of participants that clear at HKEX, including large international clearing firms, market makers and retail investors. Further to engaging bilaterally with our participants, we also hosted town hall events to provide broader market awareness and education as well as working group deep dive sessions.
The range of feedback that we received through the engagement has been invaluable and has allowed us to incorporate the risk and market expertise of our participants into the methodology.
Regulatory policy objectives – practicalities of procyclicality
During the course of this process, we frequently discussed what measures should be employed to manage and mitigate procyclical tendencies that would otherwise amplify margin movements in times of market volatility. As the new methodology is intended to more precisely forecast market volatility in terms of timing and magnitude, the tradeoff for higher precision in margin requirements is that it could exhibit procyclical tendencies. Therefore, it has been important for us to consider measures to appropriately assess sudden changes in market volatility to mitigate and, to the extent practical, throttle potentially destabilizing procyclical changes in margin requirements.
Recognizing the importance of this risk, international standard-setting bodies such as CPMIIOSCO as well as domestic authorities such as the European Securities and Markets Authority have published guidance on measures and techniques to mitigate procyclicality. We carefully analyzed this guidance and incorporated anti-procyclical measures as appropriate for the Hong Kong markets.
These measures include techniques such as volatility smoothing and incorporating stress scenarios and margin floors, which have been proven to be effective in managing procyclical tendencies. While volatility smoothing can be implemented in different ways, this technique generally involves taking averages of historic data and recent data, thereby reducing fluctuations in margin requirements. Margin floors are typically set as a percentage or absolute dollar for minimum margin levels (i.e., this prevents a model from having margin values that are too low). The floor can be set based on a variety of parameters depending on the market and model. For example, it can be set at a portfolio’s gross notional, the higher of long or short positions within a portfolio, or even at an underlying reference or instrument level.
Market characteristics also affect how measures to assess procyclicality are incorporated. For example, the Hong Kong market is equity-centric and is comprised of many retail investors in addition to large brokers and international clearing firms (e.g., both large and small with varying funding and capital profiles). It is rather common for companies listed on the Hong Kong market to be new issuers (e.g., there were 161 new listings this past year) that have limited or no price history. It is also common for companies to initiate corporate actions, such as bonus issues, stock splits, mergers and dividends. Both of these are examples of market characteristics where the methodology needed to be further assessed and incorporated to our design.
The associated chart demonstrates how measures to address procyclicality would impact margin rates for a single stock listed in Hong Kong during a period of materially increasing volatility.
The dates noted in the table were among those analyzed because at that time the Hong Kong market was experiencing a material volatility regime change, where one could expect corresponding margin requirements as calculated by models to react similarly. Without applying any measures to assess procyclicality, the margin rates produced by a pure, filtered, historical simulation VaR model would have increased by 475% and 599% for a long and short position, respectively, during these two weeks.
In designing the Next Gen Model, we employed extensive historical analysis to inform how to best calibrate anti-procyclical measures and assess their effectiveness in reacting to numerous types of stress events. While it is fundamental that a methodology reacts to market volatility, it must also assess the materiality and suddenness of corresponding margin requirements to ensure that the methodology does not unnecessarily overreact. The Next Gen Model seeks to find this optimal balance to best manage the procyclicality of margin increases and decreases in reaction to market volatility when compared to a pure FHS VaR model, as illustrated in the above table.
Techniques to address procyclicality will continue to be an evolving science that requires routine assessment and careful calibration as markets continue to develop and expand. Guidance provided by regulatory authorities is very much welcomed and will continue to guide and inform further enhancements to our risk framework.
Further information related to the Next Generation Post Trade Program can be found online here.
Vincent Cheung is the group head of quantitative risk model development at HKEX. He leads the development and system implementation of the risk models including initial margin and stress testing for the HKEX clearinghouses.
Ryan Ingram leads clearing risk policy and FMI strategy efforts for group risk management at HKEX, including the analysis of global clearing regulatory initiatives.
Ketan Patel is a managing director in group risk management and serves as the deputy group risk officer. He heads up the group credit and quantitative analysis function and oversees risk policy and risk system development. He is chief risk officer for OTC Clear, the group's clearing service for interest rate swaps and other OTC derivatives.
