How can financial institutions improve their model risk management?
By Anders Langgaard, Advisor Risk Management, Pre-Sales Risk NEMEA
Model risk management and the future of business
Driven by technological changes and new regulations, banks and financial institutions have come to rely more and more upon models - among other aspects - to manage risks. Analytical models have quickly become the lifeblood of modern financial institutions and integral to their running, helping them to, for example, assess risk, calculate capital and ECL, perform stress-testing and make data-driven business decisions.
But what about the risk management of the risk models themselves? Indeed, these solutions can help to deliver automated decision making and better outcomes for the business but what happens when the risk models are not developed or validated properly or use incorrect data?
The stakes in managing model risk have never been higher. When things go wrong at financial institutions and/or banks, the consequences are severe - the financial crisis of 2007-08 - being a prime example. Back then, modellers had more freedom and worked independently to develop models that they thought were reasonable for the intended purpose - doing so without strict governance or enterprise insight. At most, developers would check that models were conceptually accurate and no other form of validation or assessment would take place.
This problem is also present now. As automation and digital technologies become standard, more complex models are being integrated into processes to deliver more value, but when these models do not undergo the necessary development and validation and are subsequently applied incorrectly, banks and financial institutions expose themselves to unnecessary risks.
Increasing regulatory scrutiny of risk models
As a result, regulators are applying greater levels of scrutiny to the model risk management process; financial firms and banks must not only be able to provide the numbers - but also be able to clearly demonstrate to regulators how the models work, how they are used to generate the numbers, and how they have been validated and approved. Regulators want financial firms and banks to provide a transparent, reliable and robust model management framework with risk reporting at individual model level.
With this considered, to accurately report risk exposure to management and regulators, clear oversight throughout the model life cycle - from development to implementation - needs to be established. Financial firms and banks must take a more end-to-end, holistic approach to the development, implementation, use and validation of models.
The challenges for financial institutions and banks
This is, of course – the biggest challenge. For many financial institutions and banks, the problem is a lack of a model governance framework and workflow process that allows them to effectively manage the model life cycle, implementation and model risk. In many instances, problems arise when financial institutions and banks rely on disconnected and disparate legacy systems which fail to provide a complete and consolidated view of the model development life cycle and model inventory.
This disconnected approach can lead to problems with the development and management of models. Models are then developed in isolation without an appreciation of the overall objective or purpose and as a result, deliver incorrect results which are then acted upon. On the other hand, models might be fundamentally correct and developed in-line with requirements, but are then applied to the wrong environments with different assumptions – again, leading to incorrect results.
However, with systems disconnected and out of date, key issues arise around the functionality they can deliver, auditability, the integrity of data, and whether or not the database technology is fit for the future. Financial institutions and banks need a cost-effective solution that integrates all of their data sources, allows them to support and scale to meet new requirements, includes an audit trail for regulatory reporting and governance, provides a centralised model inventory database for documentation and management, reporting capabilities and streamlines the model risk management life cycle.
Achieving this level of insight and control, however, is entirely possible with a model risk management process that delivers enterprise-wide insight into the entire model life cycle, such as SAS’ Model Risk Management solution.
End-to-end model risk management and governance throughout the life cycle
At SAS, we have developed our Model Risk Management solution to allow you to create a fully integrated model risk life cycle ad govern the entire process.
The SAS Model Risk Management solution delivers transparent, enterprise-level insight into every aspect of the model life cycle, allowing you to keep up to date with the latest model status across all risk categories.
Here are just a few of the crucial – yet fundamental benefits it can provide:
- Centralised model inventory and information
Manage, monitor and analyse your models – regardless of type or source – from one centralised location. SAS Model Risk Management will provide enterprise-wide model inventory information and validation to keep your risk models fresh (through the use of workflows that measure and address model risk to deliver insight) and improve their performance. Information can be easily shared and broken down to deliver granular reporting. Bringing models together in a singular environment allows for more accurate aggregation of risk reporting and analysis of potential business opportunities.
- Auditable model life cycle
Track, review and document model assumptions, as well as classify models and analyse performance. All of this can be done at various levels of the model life cycle stage to ensure complete suitability for assigned purpose and allow for regulators to readily check and validate models.
- Governance by design
Easily tune your model review and validation processes to stay on top of changing risk policies and regulatory requirements. SAS Model Risk Management allows you to modify your reporting environment to facilitate internal and external auditing and reduce manual work for regulatory reporting. Governance tools are included to help manage sign off and testing is managed within a compliant framework.
These are but a few vital functions that the SAS Model Risk Management solution provides financial institutions and banks looking to develop a more holistic approach to model risk management.
To find out how SAS Model Risk Management can help you to improve your model risk management life cycle and governance, click here.
What to read next
The current mantra from regulators in the US and Europe is simple: "Know your models"