Benchmarking Services: How It Helps a Client with Model Risk Management

Pain Points

The following outlines key challenges the client faces to improve the Model Risk Management (MRM) framework:

1. Resource-Intensive Validation Process: Independent model validation is a resource-intensive process, often resulting in substantial costs for the client.
2. Limited Availability of Skilled Validators: The scarcity of skilled quantitative validators and the time pressures driven by urgent business demands can conflict with the requirement for independent and objective model validation.
3. Lack of Suitable Benchmarks: the bank lacks suitable internal or external benchmarks against which to evaluate and/or challenge production pricing and risk models, especially for complex instruments.
4. Data Limitations: Data limitations pose a significant challenge, particularly for OTC derivatives models, where trade data is often sparse and difficult to source.

The Solution

To address the pain points, the client engaged with S&P Global to source an independent service for model validation/performance monitoring. The solution is benchmarking services which are designed based on the cloud-native solution; Risk-as-a-Service (RaaS). The client utilized Benchmarking Services to validate MtM and xVA for the portfolio of OTC Derivatives and MtM sensitivities (e.g. PV01, CS01, FX Delta, FX Vega, IR Vega, Commodity Delta) for the portfolio of Exchange Trade Derivatives and Securities. Client’s annual subscription was based on the portfolios with over 36,000 securities/ETD and OTC trades, with the option to add more positions/portfolios with additional cost. Benchmark results were provided monthly after receiving the client's position data.

As shown in the picture below, the client has the option to utilize the solution in two distinct schemes based on their need and data availability.

1. Model Benchmarking: The client provides in-house market data along with the position data.
2. Independent Risk Verification (IRV): The client provides the position data while using S&P’s market data.

In addition to the options mentioned, S&P Global converted the client’s portfolio data from the trade booking system to feed into the simulation/aggregation engine, fraimWRX. It covers multiple use cases over Market risk, Counterparty Credit Risk (CCR) and xVA with model library covering over 180 instruments across all asset classes.

Hosted on S&P Cloud (AWS), this service dynamically adjusts computing resources according to computational needs, enabling support for portfolios of any size. Results are then made available via feeds which can subsequently be integrated into Client’s downstream reporting system.

While the main onboarding activity is mapping client’s trade files, counterparty hierarchies and legal documents, the chart below summarizes the onboarding activities S&P Global provided:

Once the mapping was completed and tested, S&P Financial Engineers walked the client’s users through the results including sensitivities and the intermediate data that was supplied with the valuation results. The above-mentioned stages were divided into three phases over 6 months, where the first phase of a calculation service was available within 4 weeks of contract signature. To help the client with model validation, the intermediate calculations that contribute to generating the final benchmark measures were retained and made available to the customer. This includes:

1.       Instrument-specific Intermediate Results:

• Swaps’ coupon dates, year fractions, implied forward rates, discount factors, etc.
• Implied Volatilities used to price options, etc.
• Calibrated parameters used for finite-difference pricing or Monte-carlo Simulation, etc.

2.      Parameters of the Simulation Model:

• Calibrated parameters, e.g.: volatilities, mean reversion, correlation, etc.
• Simulated risk factors
• Simulated MTM, exposures, collateral balance, initial margin

3.     S&P Market Data used (IRV)

• The market data used for calculation, e.g. bootstrapped IR curves, swaptions, FX Vols, etc.
• Market data are licensed to be used only within the MRM remit

Along with various benefits of Benchmarking Services, the client could benefit from the S&P Global data universe under IRV services. The S&P Data Universe offers extensive coverage and high-quality datasets that are widely used throughout the industry. It provides multiple daily snapshots and over 10 years of historical data, ensuring comprehensive and reliable information. The data categories include OTC Derivatives Data, CDS Pricing Data, and Pricing & Reference Data, all sourced from multiple market participants and exchanges. 

Key Benefits

Users in the quantitative analysis division were thoroughly impressed by the benchmarking services that S&P Global Market Intelligence provided during the onboarding and production phases. Here are the key benefits associated with this service:

S&P Global Market Intelligence's Benchmarking Services offered a robust solution for enhancing client’s Model Risk Management framework. By providing comprehensive, scalable, and cost-effective solutions, we helped the client achieve regulatory compliance and optimize their risk management processes.