Model Validation Best Practices: Achieving Value-Added

June 20, 2008

Abstract

New York, NY, USA June 20, 2008

Celent outlines a framework geared toward mitigating inherent risks borne by financial institutions’ complex network of models.

The last decade has seen a maturing process for the management of financial risks. One of the drivers of this process has been the increasingly complex dynamics of the global financial markets, which have brought about the necessity for commensurately complex models. The upsurge in usage of models has firmly placed model risk (the risk that models do not perform as intended) on regulators’ and executives’ agendas.

Building a model governance framework to contain this risk has become a prevalent risk management practice. Model validation, essentially a peer review system with the ultimate goal of providing efficient quality control, has been an integral part of these efforts. With a few exceptions, financial institutions have approached model validation as a compliance activity. In a new report, , Celent advises institutions to build a streamlined process that can deliver robust quality control around the entire model lifecycle.

Source: Oliver Wyman

"We have seen quite a bit of activity in this area over the past few years. The focus has mostly been complying with various regulations. The next wave is going to be characterized by achieving a value-added model validation framework that goes beyond compliance," says Umit Kaya, co-author of the report.

The report introduces the concept of a model lifecycle and presents a model governance framework around it derived from first principles. The best practices highlighted in the report outline everything from high-level model governance setups to the actual validation processes that materializes the value-added. Observations on pitfalls from the experiences of many diverse financial institutions are offered throughout the report.

"The recent market turmoil has shown that models are only as good as their weakest links. Model risk is not limited to a particular stage in the model lifecycle. Neither should model validation be," Kaya adds.

The past few years have seen the establishment of model validation units. However, best practices are only employed by a minority, and it would be optimistic to conclude that the state of the market is close to the end game. A significant portion of market participants have approached model validation with the mindset of a one-off exercise or a pure compliance activity. Optimizing returns from the investment in model validation requires more.

The 34-page report contains eight figures and four tables. A table of contents is available online.

 

Members of Celent's Finance, Risk & Compliance research service can download the report electronically by clicking on the icon to the left.  Non-members should contact info@celent.com for more information.

 

Celent is a research and advisory firm dedicated to helping financial institutions formulate comprehensive business and technology strategies. Celent publishes reports identifying trends and best practices in financial services technology and conducts consulting engagements for financial institutions looking to use technology to enhance existing business processes or launch new business strategies. With a team of internationally based analysts, Celent is uniquely positioned to offer strategic advice and market insights on a global basis. Celent is a member of the Oliver Wyman Group, which is a wholly-owned subsidiary of Marsh & McLennan Companies [NYSE: MMC].

Media Contacts

North America
Michele Pace
mpace@celent.com
Tel: +1 212 345 1366

Europe (London)
Chris Williams
cwilliams@celent.com
Tel: +44 (0)782 448 3336

Asia (Tokyo)
Yumi Nagaoka
ynagaoka@celent.com
Tel.: +81 3 3500 3023

Table of Contents

New York, NY, USA June 20, 2008

Model Validation Best Practices:  Achieving Value-Added 

Executive Summary 3
Introduction 6
Market Context and the Case for Best Practice Model Validation 7
Model Validation in the Context of Broader Model Governance 9
Organizational Design 15
Categorization of Models 19
Leading Model Validation Process 23
Lessons for the End Game 31

 

Sign in to download reports and access personalized information