Hedging Valuation Model Validation

Objective

A Fortune 500 life insurer engaged Graeme Group to independently validate the third-party pricing platform used for daily mark-to-market (MTM) valuations of its liability and general account hedges. The platform values a multi-billion dollar derivatives portfolio, with results used in investment accounting, risk reporting, statutory reserving and management reports.

The model had an overall risk rating of High. The rating reflected the size of the positions, the reliance on daily valuations, the range of derivative products and the controls needed for a third-party platform. The pricing routines themselves used established methods.

The client wanted to establish whether the pricing configuration and market data were appropriate, whether valuations could be independently reproduced within tolerance, and whether governance and operating procedures met its model risk requirements.

Scope of Work

The validation covered the derivative positions valued at year-end for liability and general account hedging. Products included cross-currency swaps, interest rate swaps, equity index options and equity index futures.

We reviewed market data sources, trade entry, user access, pricing configuration and methods, daily and month-end valuations, the tool that aggregates the results, and weekly management reporting. Governance testing covered documentation, change management, version control, security, issue escalation and ongoing monitoring.

Services Provided

Input Validation

We checked that market data updates had run successfully at the valuation date and reconciled the inputs to their sources. The review covered money market rates, including SOFR, ESTR, SONIA and TONAR, as well as zero curves, futures prices, swap rates and equity index volatility surfaces. For volatility data uploaded manually, we reviewed the upload procedure and its controls. We also examined how trades were entered and amended, how errors were detected and how changes were recorded, and traced a sample of trades to source records.

Calculation and Output Validation

We built independent Excel/VBA recalculation tools for the main derivative product types, using market-standard pricing methods and independently sourced market data. A documented stratified sampling method selected trades across material exposures, unusual cases and contractual variations.

Each sampled trade was repriced and compared with the platform output against a defined tolerance. We investigated differences to distinguish rounding effects from methodology or input issues. The tools allowed the model owner to refresh inputs and repeat the checks in later reporting periods.

Conceptual Soundness Review

We assessed the pricing methods for each instrument type against market conventions and the contracts being valued. This included linear interpolation of zero rates, Black–Scholes pricing with interpolation across volatility strikes and maturities, and flat forward extrapolation beyond observable tenors. We considered whether simplifications could materially affect valuations.

Governance Assessment

Governance was assessed against the client's model risk framework and policies. We reviewed documentation, the model inventory, hedging and derivatives policies, change management, version control, security, storage, issue escalation and monitoring. We paid particular attention to controls over manual inputs and uploads.

Downstream Output Review

We traced valuations from the platform through the aggregation tool into management reports. We reconciled the transferred values, checked the aggregation logic and reviewed whether the reports accurately represented the underlying positions.

Model Validation Report

We delivered a Model Validation Report for the model owner and second-line model risk function. Findings were classified by severity and type, with the supporting evidence, risk, recommendation, proposed owner and remediation steps.

Deliverables

The engagement resulted in:

• A Model Validation Report concluding Satisfactory with Recommendations, with assessments of inputs, calculations, outputs and governance.
• Reusable Excel/VBA tools for independently recalculating the main derivative product types.
• A sampling workbook recording the selection method and coverage by product type, portfolio and materiality.
• A findings register with recommended actions and owners for the client's issue tracking process.
• Recommendations on manual inputs, version management, reproduction of historical valuations and the model risk rating framework.
• Sign-off by qualified actuarial and quantitative reviewers.

Outcome

• Valuations Reproduced: The sampled MTM valuations were independently reproduced within tolerance.
• Reusable Checking Tools: The model owner received the recalculation tools and sampling workbook for use in future reporting periods.
• Governance Recommendations: Findings on manual inputs, version management and historical valuations included recommended actions and owners.
• Evidence for Audit Review: The report documented the tests, evidence and conclusions against the client's model risk requirements.
• Risk Rating Reviewed: We recommended reducing the Complexity rating from High to Medium because the pricing methods were established and well understood.
• Reporting Checked: The review covered the transfer and aggregation of valuations into management reporting as well as the pricing platform itself.