February 16, 2016 11:58 ET

Defeating the Challenges in Vendor Model Validation

Interview With John Zhang, Senior Manager of Investment Risk at CPP Investment Board

TORONTO, ON--(Marketwired - February 16, 2016) - Model risk management has become very important across North America. After Letter 11-7 was introduced in the US, many changes were made to model risk practices. There are indications that Canada will have a similar regulation soon, making model risk management, and vendor model validation a top priority for financial institutions in Canada.

John Zhang, Senior Manager of Investment Risk at CPP Investment Board, has recently spoken with GFMI about topics to be discussed at the upcoming Model Risk Management -- Canadian Edition Conference:

What are some of the major challenges in vendor model validation today?

JZ: One of the major challenges in vendor model validation is the full examination of all the models/functions served for the business's purposes within both the framework of regulatory requirements and legal/accounting compliance. It's not a difficult job to validate a single model within the vendor system given market input data. However, it will be a challenge to identify every single model involved in the business application process offered by the vendor system and validate all of them in accordance with necessary regulatory requirements. Another challenge is less detailed and incomplete documentation from vendors. Lastly, a huge problem is the inaccessibility to the source codes, which will make it difficult to justify the hard coded data and extreme case testing.

How are vendor models built, and how much do we know about them?

JZ: Most vendor systems are built to serve, or focused on some special business purpose, such as valuation of derivatives (P&L), market risk (VaR), Counterparty credit risk (PFE), economic capital (EC), etc. A good vendor system usually is designed with more flexibility or compatibility for future development. We start to know the vendor by the scope of functions it can offer. With the complete model validation process, we can have a clear picture of the framework offered by the vendor, and we believe such a model validation process can help model users to better understand the product and make full use of the vendor's functions.

What are the top regulatory expectations around vendor models?

JZ: Most financial institutes are on their way to seeking Basel compliance. For the banks' exposure to the U.S. market, they need to be CCAR compliant as well. As a result of that, the design of the structure/framework of the vendor has to be compliant with such requirements.

What are the key issues in model risk quantification today?

JZ: Model risk quantification is still in its infancy. Actually, it's still far from reaching the level of a fully accepted method and application in the industry due to the following issues: 1. the various types of model involved; 2. the complexity of models; 3. the availability and reliability of market quotes.

What do you think attendees will gain by attending the event?

JZ: Through a series of practitioner's lectures and panel discussions, you will have ample opportunity to actively benchmark with your peers and take home key implementation strategies on model risk management and how to build a concrete model validation framework by maintaining regulatory compliance.

John Zhang is the Senior Manager with Investment Risk at CPP Investment Board. John has over 13 years of financial quantitative modeling experience across banks' capital market derivative pricing models, market risk models, and counterparty credit risk models to alternative investment risk models. Before joining CPPIB in October 2014, John was the Senior Manager with Model Risk Governance and Vetting at BMO for over four years. John headed a small group of quantitative analysts responsible for all model vetting activities related with capital market derivative pricing models (vendor models) and treasury models with BMO. Prior to working with BMO in May 2010, John worked with National Bank of Canada for over six years on a couple of senior quantitative analyst roles from market risk model development to derivative pricing model validation. John received his Master of Quantitative Finance degree in 2003, and Master of Applied Science degree in 2002 -- both from the University of Waterloo.

Join John Zhang at the Model Risk Management Conference, March 7 th - 8 th , 2016 in Toronto, ON, Canada. View the conference agenda to check out John's case study topic. For more information, please contact Jen Jordan, Digital Marketing Coordinator, GFMI at 312.894.6347 or jenjo@global-

About Global Financial Markets Intelligence

GFMI is a specialized provider of content-led conferences for the financial markets. Carefully researched with leading financial market experts, our focused quality events deliver key bottom-line value through targeted presentations, interactive discussions and high-level networking opportunities.

Image Available:

Contact Information