Model Risk Governance and Review - Associate

JP Morgan

Location: Greater London

Job Type: Full time


About the Group

The Model Risk Governance and Review Group (MRGR) oversees model risk at the firm, conducts independent model reviews, and provides guidance around a model’s appropriate usage. t has a global presence across New York, London and Mumbai, and Paris. The group carries out the reviews of models used across the firm. It assesses and helps mitigate the model risk of complex models used in the context of valuation, risk measurement, the calculation of capital, and more broadly for decision-making purposes. Derivative instruments are widely used in the Bank's businesses as part of the core trading activities or for risk management purposes. The group also focuses on evaluating the usage, deployment and performance of models including the appropriateness of key inputs and consideration of the reliance placed upon all outputs.

About the Team

MRGR CPG (Credit Portfolio Group) covers validation and governance activities for models used in Counterparty Credit Risk (CCR) space. These include a variety of components such as risk factor simulation engines, instrument pricers, correlation models, exposure aggregation and end-usage models such as CVA and FVA (Credit and Funding Valuation Adjustments, or XVA), and PFE (Potential Future Exposure). The distinctive features of the space are its cross-asset nature (with its opportunity to learn about different lines of business and associated models), broadness in terms of usage (e.g. valuation, capital and credit risk management) and relative novelty of the area with active new model development.


  • Act as the first point of contact for questions related to the existing models usage and reviews of new models
  • Carrying out model validation and designing model risk measurement activities, for models used to value and risk manage XVA, compute CCR monitoring metrics and capital requirements.
    • Model reviews: evaluate the conceptual soundness of pricing engines and its suitability for capturing risk; the reasonableness of assumptions and reliability of inputs; the consistency of approaches used across products and asset classes; the completeness of the testing performed to support the methodology choices and the correctness of the implementation; the suitability and comprehensiveness of performance metrics and risk measures associated with the use of the model. The scope of models naturally covers a variety of asset classes and model usages.
    • Model risk measurement: design and implement experiments to measure on-going model performance, the potential impact of model limitations, parameter estimation error or deviations from model assumptions; compare model outputs with empirical evidence and/or outputs from alternative model benchmarks
  • Evaluate model performance on a regular basis and need for re-review of the models
  • Liaise with FO, Quants, Market Risk and Valuation Control Groups to understand usage of models within the business context, assess models’ fit-for-purpose for specific portfolios and netting sets, syndicate the findings of model validation and to ensure that model risk is understood, captured, monitored and managed
  • Identify and highlight any gaps in the model governance and oversight framework, establish good risk and control practices
  • Report on model risk both internally and externally
  • Periodic confirmation of model inventory, and assessment of model risk tiers

Essentialskills, experience, and qualifications:

  • PhD or MS degree in a quantitative areas (Math Finance, Applied Math, Physics, Engineering, Statistics/Econometrics or similar)
  • Solid command of probability theory, stochastic calculus and numerical methods
  • Deep understanding of financial math and derivative pricing
  • Excellent analytical and problem solving abilities
  • Excellent communication skills (written and verbal)
  • Inquisitive nature, ability to ask right questions and escalate issues; risk & control mind-set
  • Knowledge of programming languages (Python/C++)
You’ve got this!