Services

Understanding and Addressing Model Risk

Our Model Risk team offers advisory based on many years of Ph.D.-level “quant” experience of developing and validating various types of valuation, market risk and economic capital models, including those required by Basel III and new regulations on Capital and Margin Requirements (SA-CCR, Initial Margin, etc.).

Proficiency across all stages of model risk governance (i.e. model development, validation, internal audit, ongoing performance monitoring) allows us to effectively support development and improvement of model risk governance process.

Our holistic analysis helps to control model risk, to prevent losses associated with it and to enhance key stakeholders’ understanding of models. We also help organizations manage the model risk of their portfolio by assessing and suggesting improvements to their model governance programs in line with regulators’ recommendations (e.g. FRB SR 11-7).

We will assist the firm in assessment of model development and implementation documentation to ensure that it is complete and consistent with industry standards and best practices (e.g. with FRB SR 11-7), covering following:

  • Regulatory expectations;
  • Model’s conceptual soundness;
  • Analysis of model’s assumptions and limitations;
  • Stress testing of model parameters and analysis of outputs’ sensitivities;
  • Model ongoing performance monitoring plan.

We will also assess completeness of model validation performed or advise on what need to be performed by answering following questions:

  1. Does validation address all model risks related to its development, implementation, use and performance monitoring?
  2. Is the thoroughness of validation testing consistent with model risk tier assigned and with industry best practice?
  3. Is validation documentation comprehensive?  Does it follow firm’s Model Risk Management policies and procedures and provide sufficient information about validation testing performed?
  4. Are conclusions and proposed remediation actions consistent with the results of validation testing and review?

We will be able to advise on development of customized quantitative models, refine and calibrate existing models, and design stress testing and scenario analysis programs to supplement existing analytics.

One key area of focus is model performance monitoring. This encompasses creation of suitable benchmarking and backtesting procedures and metrics for a wide variety of models, a framework for measuring model performance, and governance around treatment of underperforming models.

 

Derivatives Valuation and Independent Price Verification

Price is what you pay. Value is what you get.

                                           – Warren Buffett

Correct valuation of financial instruments became more and more important during post-crisis years.  For complex or illiquid derivatives valuation process employs quantitative models.  FRB/OCC Supervisory Guidance on Model Risk Management (OCC 2011-12/FRB SR 11-7) defines a model as “a quantitative method, system, or approach that applies statistical, economic, financial, or mathematical theories, techniques, and assumptions to process input data into quantitative estimates.”  Models are simplified and idealized representations of the real world and are prone to errors.  Their outputs heavily depend on model inputs that are usually marked by the traders.  Thus, prudent risk management must involve the control of this input marking process, which is done via Independent Price Verification (IPV).

IPV is the process of determining the fair value of financial instruments independently of risk taking units (i.e. front office).

It is defined by Basel II as “the process by which market prices or model inputs are regularly verified for accuracy” (BCBS “Supervisory guidance for assessing banks’ financial instrument fair value practices”, 2009).

IPV of all the fair valued positions should be performed on a regular basis (usually at each month end) and on “as needed” intra-month basis.  This process may uncover deliberate mis-marking by trading desk in order to inflate the value of trading portfolio or to conceal true trading performance.  Quite often it might indicate the weaknesses of the valuation models used for Books & Records. 

The results of IPV should be directly linked to the performance review of the models used by front office.  These links should be expressed in the terms of explicit quantitative thresholds.  Crossing of these thresholds should trigger the revalidation of those models.

Our team offers advisory based on years of experience of running IPV of Commodities vanilla and exotic derivatives portfolio and validating IPV methodology for FI/IR portfolio for top tier investment bank and examining Derivatives Valuation and IPV policies and procedures as part of regulator’s oversight of Swap Dealers adherence to Risk Management Regulations.

 

Valuation Adjustments (xVA)

In order to remain competitive market participants must be able to price in various risks and regulatory capital charges related to specifics of the parties to the trade.  This is usually done via utilization of various Valuation Adjustments commonly known as xVAs, such as CVA, FVA (including the asymmetric funding rates and possibility of rehypothecation of collateral), KVA, MVA, ColVA, etc.

Adoption of this approach requires deep understanding of the nature of these risks and their interaction with other existing or potential trades in the firm’s portfolio.

 

 

Margin related Matters Requiring Attention

The number of financial institutions that are in-scope for Initial Margin collateral management implementation in phases 4 and 5 (years 2019-2020) is significantly larger than for phases 1 through 3 (2,000+ vs. 80+), while their resources are much more restrained.

Our knowledge of potential issues and experience in resolving them in the early stages of development and implementation will help a Covered Swap Entity (“CSE”) save time, resources and efforts during development, implementation and approval stages.

Recent involvement in approval and ongoing performance monitoring of ISDA SIMMTM during Phases 1 and 2 of Margin Rules implementation demonstrated that despite relative simplicity of the model there are multiple hazards in the process that smaller CSEs should be aware of and ready to avoid.  Those include:

  • Quality control of SIMM data inputs;
  • Identification of risk factors present in the OTC derivatives portfolio, but missing from SIMM, and their quantification;
  • 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;
  • 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.