Post 2008 financial crisis BCBS and IOSCO began developing a framework for minimum Initial and Variation Margin requirements (IM & VM respectively) for uncleared OTC derivatives. It was a basis for Margin Rules adopted by various jurisdictional regulating agencies.
Sections 731 and 764 of the Dodd-Frank Wall Street Reform and Consumer Protection Act (Dodd-Frank Act) require the Federal Reserve, FDIC, OCC, DTCC, Farm Credit Administration, Federal Housing Finance Agency, CFTC and SEC (collectively, the “agencies”) to adopt rules jointly for swap dealers, major swap participants, security-based swap dealers and major security-based swap participants (collectively, “covered swap entities” or “CSE”) that establish initial and variation margin requirements on all swaps that are not cleared by a registered derivatives clearing organization or a registered clearing agency.
Understanding Margin (Initial & Variation) Regulations
Initial Margin requirements were phased-in starting on September 1, 2016 based on the size of the aggregate average notional amount (“AANA”) of non-centrally cleared derivatives of the corporate family to which a covered swap entity belongs, with the final phase going live on Sep. 1, 2020. At the same time, Variation Margin requirements were put in effect on March 1, 2017 for all CSEs registered with CFTC or SEC regardless of the size of uncleared derivatives portfolio.
Both IM and VM are calculated for a bilateral portfolio of uncleared OTC derivatives traded between CSE and financial end user counterparty (that also might be a CSE). While a collateral transfer related to VM is unidirectional (from one counterparty to another, based on the VM calculations results), the IM is posted by both sides to custodian account(s).
Minimum amount of IM may be determined by one of two methods:
- Based on standardized look-up table (quite punitive, as IM calculated this way is on average several times larger than if using a permitted model);
- A model approved by a relevant regulating agency.
Failure to get an IM Model approved by regulators might be very costly to a CSE: not only does the table-based IM become significantly larger in most cases, it may even result in losing a sizable portion of business, since some counter-parties might refuse to trade with the firm whose model was not approved.
ISDA developed a methodology for IM calculation (ISDA SIMMTM) to provide OTC derivatives market participants with a uniform method that meets requirements established by regulators in line with BCBS – IOSCO recommendations. Its use greatly simplifies the reconciliation process and is supposed to reduce the number of IM disputes between counterparties if the same model is used by both sides. So far, all CSEs required to post / collect IM (Phases 1 through 3) adopted ISDA SIMMTM as a standard risk sensitivity based tool to calculate IM.
IM related Matters Requiring Attention
The number of financial institutions that are in-scope for IM implementation in phases 4 and 5 (years 2019-2020) is considerably larger than for phases 1 through 3 (several thousand vs. 80+ respectively), while their resources will be much more limited.
Developing internal IM model or incorporating ISDA SIMMTM presents many challenges. Despite relative simplicity, SIMM methodology contains many details requiring accurate implementation.
An IM calculator should include both an IM model (possibly ISDA SIMMTM) and the table-based methodology – as required by regulations – if risk based methodology cannot be used. It should be properly integrated with existing production systems and flexible enough to allow fast implementation of updates in regulations and/or new versions of methodology produced by ISDA. This calculator should be available not only for end-of-day calculation of IM obligations, but also for pre-trade “what-if” analysis of margin related funding costs that Front Office desks might incur. This leads to another, more complex issue of calculating Margin Valuation Adjustment (“MVA”), and it needs to be decided whether to apply this adjustment on an individual trade basis, book basis, portfolio basis, or on a Legal Entity basis.
Recent involvement in approval and ongoing performance monitoring of ISDA SIMMTM during Phases 1 and 2 of Margin Rules implementation allowed the author to observe numerous parts of the process that might cause troubles if not addressed properly. Smaller CSEs should be aware of these issues and be prepared to avoid them.
Those concerns include:
- Proper understanding of ISDA SIMMTM assumptions and limitations as related to the firm’s business strategy and OTC Derivatives portfolio composition;
- Accurate calculation of risk sensitivities and their transformation from a production system utilized for Books & Record into the standard set of sensitivities prescribed by SIMM methodology;
- Efficient quality control of SIMM data inputs;
- Comprehensive identification of risk factors present in the OTC derivatives portfolio, but missing from SIMM (“Risks-not-in-SIMM”) and ability to correctly quantify them;
- Proper back-testing of SIMM outputs and benchmarking against comparable models and IM methodologies utilized by clearing organizations (as required by Margin Regulations);
- Understanding of statistical meaning of performance monitoring metrics and proper governance around underperformance of the model.
Below are major observations from Phases 1 and 2:
- It is important to carefully plan the whole end-to-end process starting at downstream feed of data from pricing models to Margin calculator and finishing at upstream feed of its results to Collateral Management system. This planning process must include the inputs from various firm’s units: Front Office, Legal & Compliance, Risk, Model Developers and Validators, Collateral Management and Treasury, etc.
- These are some essential questions to be answered before the development and implementation process starts:
- Should the firm stop trading OTC derivatives to avoid high IM burden or try to stay under allowed IM thresholds with in-scope counterparties?
- How the cost of IM funding, if any, will be allocated?
- How IM might be optimized?
- How the hedging of OTC derivatives should be done considering prudent IM management?
- The review of Phase 1 and 2 approvals demonstrated that most firms did not fully follow the industry best practices for model risk management outlined in FRB SR 11-7. Some of the major deficiencies noted were lack of proper separation between model development and model validation, inadequate quality level of model development and model validation documentation, often insufficient level of model testing, among others.
- The firms often relied too heavily on the analysis and testing performed centrally by ISDA and ignored the fact that idiosyncratic properties of their uncleared OTC Derivatives portfolios require additional review and analysis of the model’s conceptual soundness, its assumptions and limitations, as well as the model’s suitability to their business strategy.
- The same is valid for SIMM testing – back-testing and benchmarking of the model performed centrally by ISDA cannot be considered fully sufficient without analysis of a firm’s specific exposures. SIMM is sensitivity based methodology. It approximates 99th percentile of a bilateral portfolio’s P&L over 10 days (Margin Period of Risk defined by Margin regulations) by using the Taylor formula with first and second order derivatives (whereas, Gamma terms are replaced by Vega based terms, although this approach works well only for vanilla options). This approximation does not always work well for potentially large moves of risk factors over 10-day periods, especially for such volatile asset classes as commodities. This is one of the reasons it is necessary to perform exhaustive benchmarking and back testing during the implementation stage, as well as a part of ongoing model performance monitoring.
Necessary steps to be taken:
The Top Tier OTC Derivatives markets participants (those in scope for Phases 1 and 2) spent up to two years implementing Margin related models and processes, and still some of them were not ready to be approved by regulators until very close to their deadlines. For Phase 4 and especially for Phase 5 eligible firms, the resources both financial and intellectual are much more limited, so it is crucial to start early planning, development and implementation processes.
Irrespective of whether the firm decides to build an IM calculator internally or to buy it from a vendor, it must be incorporated into the production system and thoroughly tested using the most complex portfolios that adequately represent future business strategy and live OTC derivatives exposures post compliance date.
A Model Validation team should not rely on the test results and documents produced by either ISDA or the vendors and instead conduct full validation of IM model including analysis of conceptual soundness, assumptions and limitations of the model, perform full set of tests required by the firm’s Model Risk Policies and Procedures, including correctness of implementation, exhaustive backtesting and benchmarking of the model against comparable CCP IM methodologies and internal fully validated and approved VaR-type risk measures, sensitivity and stress testing, etc. Special attention should be paid to analysis of risks not covered by IM model and design of compensating measures to be prepared to control IM deficiencies.
A Model Risk team should formulate IM related Model Risk Governance program that covers all stages of IM model use, outlines model performance monitoring process and lists compensating controls and escalating actions required to mitigate issues discovered during ongoing model performance monitoring.
It is very important prior to going live to take part in a centralized reconciliation process driven every year by ISDA, to reconcile with as many counterparties as possible and to use the most complex bilateral portfolios for this purpose in order to avoid surprises after compliance date.
With the exponential growth of financial institutions in scope for Margin regulations there is potential for severe constraints on the talent pool, so it is important to start assembling participating teams early enough to properly prepare the firm for compliance with regulations.
At Aleph Analytics our knowledge of potential issues and experience in resolving them in collaboration with financial institutions during the early stages of development and implementation will help a CSE save time, resources and efforts during development, implementation, approval and go-live stages.
To request additional information, send email to info@alephanalytics.org or call us at +1 914 320 0013.