Independent Validation of a Multi-Billion Dollar Hedging Valuation Model for a Fortune 500 Life Insurer

Objective

Graeme Group was retained by a Fortune 500 life insurer to perform an independent technical validation of the third-party SaaS derivative pricing platform used to produce daily mark-to-market (“MTM”) valuations across the Company’s liability and general account hedging programs. The platform sits at the centre of a multi-billion dollar derivatives book whose valuations feed downstream into investment accounting, risk reporting, statutory reserving, and senior executive dashboards. Any error or instability in the model has the potential to flow directly into the Company’s reported financial position and into the hedge effectiveness conclusions reviewed by senior leadership.

The model carried an overall risk rating of High. That rating was driven not by the analytical complexity of the platform itself — which uses well-established pricing routines — but by the financial materiality of the positions valued, the operational criticality of the daily output, the breadth of derivative product types in scope, and the inherent challenges of governing a third-party SaaS environment that the Company does not directly control.

Against that backdrop, the client wanted independent comfort across three distinct dimensions. First, that the pricing configuration and market data inputs were appropriate, current, and aligned with market convention. Second, that the MTM output produced by the platform could be independently reproduced within tolerance using a credible challenger methodology. Third, that the governance, controls, change management, and operational practices wrapped around the platform were commensurate with its High risk rating and would withstand both internal Model Risk Management (MRM) scrutiny and external audit review.

Scope of Work

The validation covered the full population of derivative instruments valued within the platform as at the year-end anchor date, supporting both liability and general account hedging activity. The book spanned several distinct product types — including cross-currency swaps, interest rate swaps, equity index options, and equity index futures — each with its own pricing mechanics, market data dependencies, and operational nuances. Aggregate notional and market value across the in-scope positions ran well into the multi-billion dollar range.

The engagement deliberately took an end-to-end view, covering not just the pricing engine itself but everything that touches it: market data sourcing, trade booking, user access, the configuration of pricing environments, the methodology applied to each instrument type, the production of daily and month-end MTM output, the downstream aggregation tool that consumes that output, and the weekly executive reporting derived from it. We also reviewed the surrounding governance fabric, including documentation, change management, version control, model security, escalation pathways, and ongoing monitoring controls.

Services Provided

Input Validation

We reviewed every market data source configured within the platform’s pricing environment, confirming that scheduled data pulls had executed successfully as at the anchor date and that the data being consumed reconciled to its underlying source. The review covered money market rates across multiple major currencies (including SOFR, ESTR, SONIA, and TONAR), zero curves, equity index futures prices, bond futures prices, swap rates, and equity index volatility surfaces. For volatility surfaces sourced from a third-party vendor and uploaded manually rather than via API, we walked through the upload process step by step, reviewed the controls wrapped around it, and assessed whether the workflow was commensurate with the criticality of the input. We also examined trade data referencing — how positions are entered, who can amend them, how booking errors are detected, and the audit trail behind every trade-level change — and reconciled a sample of trades back to source records.

Calculation and Output Validation

The core challenge in validating a third-party pricing platform is that the validator does not own the source code. We addressed this by building independent re-calculation tools (“TestWare”) in Excel/VBA for each of the major in-scope product types. Each TestWare module was constructed from first principles using market-standard pricing methodology and was wired to consume independently sourced market data. We then designed and documented a stratified sampling methodology that selected a representative sample of trades, weighted to ensure coverage of material exposures, edge cases, and the full range of contractual variations within each product type.

Each sampled trade was re-priced inside the TestWare and the resulting MTM was compared to the platform output against a defined tolerance. Any breaks were investigated to root cause, distinguishing immaterial rounding differences from genuine methodology or input issues. The TestWare was deliberately built to be reusable: it was structured so that the model owner could refresh inputs and re-run the comparison in subsequent reporting periods without external help, providing a permanent independent challenger capability.

Conceptual Soundness Review

We assessed the appropriateness of the platform’s pricing methodology against established market convention for each instrument type. This included the choice of yield curve construction (linear interpolation on zero rates), the treatment of volatility surfaces (Black–Scholes pricing with linear interpolation across strikes and maturities), and the approach to extrapolating curves and surfaces beyond observable tenors (flat forward extrapolation). For each in-scope product type we considered whether the chosen approach was a reasonable industry-standard method, whether it produced sensible pricing for the actual contractual features in the book, and whether any of the simplifications introduced material risk that should be flagged or remediated.

Governance Assessment

The governance assessment was anchored against the Company’s own model risk management framework and policies. We reviewed model documentation for completeness and currency, entries within the model inventory, hedging programme and derivatives policies, change management practices, version control protocols, model security and storage arrangements, the escalation pathway for issues identified during production use, and the ongoing monitoring controls applied to the platform. Particular attention was paid to the controls around manual touch points — the points at which a human action can affect the integrity of the model output — because these are typically where operational risk concentrates in third-party SaaS environments.

Downstream Output Review

The platform’s MTM output does not stand alone — it feeds a downstream aggregation layer that summarises positions by hedging programme, which in turn feeds the periodic derivative reporting distributed to senior leadership and the investment teams. We traced the data flow end to end, confirming that the values consumed by the aggregation tool reconciled to the platform’s output, that the aggregation logic itself was appropriate, and that the resulting executive reporting accurately represented the underlying positions. This downstream review is critical because errors that originate in the model can be amplified or obscured by the aggregation layer, and senior stakeholders are often consuming the aggregation output rather than the raw model output.

Model Validation Report

The engagement concluded with the delivery of a comprehensive Model Validation Report structured to serve both the Model Owner and the second-line MRM function. Findings were categorised by severity and type, each accompanied by a description of the underlying risk, a recommendation, a proposed owner, and a target remediation pathway. The report was written to be audit-ready, with sufficient narrative for an external reviewer to understand the basis of every conclusion without needing to re-perform the work.

Deliverables

The engagement resulted in:

• A comprehensive Model Validation Report concluding that the model was Satisfactory with Recommendations, with detailed evaluations across inputs, calculations, outputs, and governance.
• Independent re-calculation TestWare in Excel/VBA for the major in-scope derivative product types, each built from first principles and structured for reuse by the model owner in subsequent reporting periods.
• A stratified sampling workbook documenting the methodology, rationale, and selection logic used to derive a representative sample of trades for independent re-pricing, including coverage analysis across product type, book, and materiality.
• A categorised findings register with proposed remedial actions and ownership recommendations, designed to feed directly into the Company’s issue tracking process.
• Targeted recommendations to strengthen controls around manual data inputs, version management of the platform, reproducibility of historical valuations, and refinements to the model risk rating framework where the existing rating overstated genuine risk.
• Sign-off by qualified actuarial and quantitative reviewers.

Outcome

• Valuation Confidence: Independent re-calculation across the in-scope product types confirmed that the platform’s MTM output was accurate within tolerance, giving the client well-evidenced comfort that the valuations feeding downstream regulatory, accounting, and management reporting were reliable.
• Reusable Challenger Toolkit: The TestWare and sampling workbook were transferred to the model owner as a permanent independent challenger capability. The client can now refresh inputs and re-run the comparison in any subsequent reporting period without external assistance, materially reducing the cost of ongoing assurance.
• Governance Roadmap: The validation surfaced a focused set of governance enhancements around manual touch points, version management, and reproducibility of historical valuations. Each was paired with a concrete recommendation and owner, giving the model owner a clear and prioritised remediation roadmap.
• Audit Leverage: The validation report was written to audit-ready standards and aligned to the Company’s model risk management framework, providing evidence that can be lifted directly into internal MRM submissions and external audit walkthroughs without rework.
• Right-Sized Risk Rating: Our review supported a proposed reduction of the model’s Complexity rating from High to Medium on the basis that, while the underlying instruments are complex, the pricing methodology relies on established and well-understood market-standard approaches. This gives the client a more accurate basis for prioritising future model risk effort.
• End-to-End Coverage: By extending the scope of the validation through to the downstream aggregation tool and executive reporting, the client received assurance not just over the pricing engine but over the full chain of data that informs senior decision-making and external disclosure.

This engagement highlights Graeme Group’s ability to deliver rigorous, independent validation of complex third-party derivative pricing platforms. By combining bespoke independent re-calculation tools, defensible sampling methodology, deep familiarity with derivative pricing theory and market convention, and a pragmatic understanding of the operational realities of running a SaaS platform inside a large financial institution, our team challenged a high-risk hedging valuation model on its own terms and delivered the client both immediate comfort and a permanent toolkit for ongoing model risk management.